..
ti: ..)....2
13:4.)

515.2...
An; . E 2333...:
Irfluzw .5 £533.43.

:1...
523.5 .53.
.1 :L...
5 .3.
~ 59.6. .mrdflflu
3.7... z.\.|..x 3.5
.. 3 5.12?

3
31‘: 3.1,.» :
V3.4... . .

.1 .
2:; ..
1: 1:: .u ‘.

5..., 5:.
12:. I. 5.
7.113.:

{EL 1.1.»:

73...?
2:1.

{7:22.35 K.

 

: ME, 3..

 

 

SITY LIBRARI

lull lllllllll i

072 ?

 

 

Illllllillllllllllllll

3 1293 0101

This is to certify that the

dissertation entitled

IN VITRO METHODS TO DETERMINE PROTEIN
DEGRADATION OF FEEDS FOR RUMINANTS

presented by

Richard A. Kohn

has been accepted towards fulfillment
of the requirements for

Ph.D. degreein Animal Science

 

 

Wfi/ézp

flajor professor

Date 8/18/93

MSU is an Affirmative A ction/Equal Opportunity Inslilution 0-12771

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.

TO AVOID FINES return on or before due due.

DATE DUE DATE DUE DATE DUE

       

  
  

 

IN VITRO METHODS TO DETERMINE PROTEIN
DEGRADATION OF FEEDS FOR RUMINANTS

By

Richard A. Kohn

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Animal Science

1993

 

ABSTRACT

IN VITRO METHODS TO DETERMINE PROTEIN DEGRADATION OF
FEEDS FOR RUMINANTS

By
Richard A. Kohn

Three laboratory methods for evaluation of feed protein degradation were studied: 1)
in vitro fermentation with 35S used to label microbial protein, 2) in vitro incubation
with cmde extract of enzymes from rumen fluid, and 3) sequential extraction of forage
proteins into 6 fractions postulated to degrade at different, yet uniform first order rates.
In the first method, feeds were incubated in batch culture in bicarbonate-phosphate
media with 35S-cysteine and rumen inocula. At the end of incubation, feed proteins
were separated from free bacteria by sequential centrifugation. Ratio of N to 35$ in
free bacteria was multiplied by 358 in other fractions to estimate microbial N. Two
rinses in HCl solution containing free cysteine, NazSO4 and NaSO3 were required to
remove most non-protein 358. Ratio of N to 35S varied across treatments and lengths
of incubation because feed S was incorporated into microbes. The proposed method
was simpler than previous work, but high variation within treatments persisted. For the
second method, microbial enzymes were dislodged by extraction of rumen solids with
butanol and acetone, and subsequent solubilization of proteases in water. This extract
was incubated with feeds to degrade proteins. The enzyme extract maintained 62% of
the activity in whole rumen fluid, but was more concentrated. Proteolytic activity
appeared to be resistant to self-degradation, but rates of degradation of alfalfa hay and
soybean meal declined with length of incubation. A third method evaluated N
solubility for determination of crude protein degradation. Several forages were
degraded by crude enzyme extract and were subjected to sequential solubilization in: l)

trichloroacetic acid, 2) buffer, 3) acetone, 4) detergent at pH 7, and 5) detergent in IN

H2804. Additionally, 8 forages were degraded in rumen extract followed by sequential
fractionation. Protein degradation of individual fractions was determined by multiple
regression analysis of degradation against fractions, or by direct measurement. These
studies demonstrated that buffer soluble proteins were readily and completely degraded
by rumen extract, but other fractions were degraded incompletely and inconsistently

over time.

 

DEDICATION

He [Professor Jefi‘rey Williams ] wanted to publish her research ndings before she
[Graduate Student Maie El ’Kassby] had a chance to, so he fire her. Professors do
that all the time. They ’re parasites that live on the backs of their graduate students.

--Zolton Ferency, 1922-1993

This dissertation is dedicated to the memory of civil rights activist Zolton Ferency, a
man who taught us by example to be brave, to say and to do what we know is right, to
struggle, and never to compromise our principles. Zolton, as a professor, lawyer and
politician often found himself involved with a corrupt system, but he never became
corrupted by the system.

 

projr
was i

degrc

O'Nc

new;

with 1
every

Stalin

 

 

Sniff;

 

r3363:

diet

5%;

 

grad,

 

 

 

 

ACKNOWLEDGMENTS

I would like to thank my wife Tammy for her support of me regarding this
project, but especially for her support regarding every other aspect of my life. Tammy
was by far the most significant ( P < .0001) discovery I made while working toward my
degree. She reminds me daily of what is truly important.

I would like to express my appreciation to my fellow graduate students, Kitty
O'Neil, Katharine Knowlton, Rick Dado, Steve Mooney and Mary Beth Roe. I could
never begin to describe how much I learned from them.

I also would like to thank Dave Main and Dewey Lon guski for their assistance
with maintaining and optimizing the lab, help on experiments, and for teaching me
everything I'd ever want to know about country music.

I thank Jim Liesman for sharing his broad knowledge of laboratory methods,
statistics, and animal science, and for the interesting philosophical conversations.

I thank my graduate committee members: Dr. Roy Emery, Dr. Kenneth Nadler,
Dr. Steven Rust, and Dr. Michael VandeHaar, and previous members Dr. Charles
Sniffen and Dr. Derek Lamport. All of these members have helped me in different
Ways to improve the quality of my education, and to broaden my understanding of the
research.

I would especially like to thank Dr. Mike Allen, my major professor and
dissertation director, for his interest in my education and research.

Last but not least, I would like to thank the woman with the ruler at the graduate
School , who Will determine in two minutes that this dissertation is acceptable so I can

graduate. I wish only that my committee members were as efficient.

rm

CH4

CHM

 

 

 

TABLE OF CONTENTS

 

 

 

 

 

 

 

 

 

 

 

 

 

INTRODUCTION 1
CHAPTER 1
LITERATURE REVIEW 4
Importance of Ruminal Protein Degradation 4
Protein Metabolism 4
Protein Digestion in Ruminants 5
Ruminant Protein Requirements 5
Effect of Protein Source on Degradation 11
Animal and Microbial Effects on Protein Degradation ............................. 18
Methods to Determine Protein Degradation 21
Measurement of Protein 21
Estimation of Microbial Protein Synthesis 22
In Vivo Methods to Measure Protein Degradation .............................. 28
In Situ Methods to Measure Protein DPgmdafinn 29
In Vitro Methods to Measure Protein Degradation 31
LITERATURE CITED 42
IHAPT ER 2

Storage of fresh and ensiled forages by freezing affects fibre and crude

 

 

protein fractions 57
Abstract 57
1. Introduction 58

 

vi

Cf.

 

2. Experimental

 

2.1. Treatment of samples

 

 

2.2. Analyses

2.3. Statistical

 

2.4. Freeze—drying

 

3. Results and D' ‘

 

3.1. Crude protein fractinne

3.2. Ash, fibre and lignin

 

3.4. Freeze-drying

 

3.6. Relevance of research

 

4. References

 

IHAPTER 3

Observations on determination of protein degradation of feeds by in
vitro batch fermentation with 3SSulfur used to label microbial growth. ............

ABSTRACT

 

INTRODUCTION

 

METHODS

 

Final Method for Microbial Fermentation

 

Development and Validation

 

RESULTS AND DISCUSSION

 

Separation of Incorporated from Non-Incorporated 358 .....................
Addition of reducing agents prior to rinsing. ..............................
Precipitation of cysteine with protein .........................................

Rinsing the pellet of 3SS—cysteine precipitate

 

Solubilization of Incorporated 353
Tissue solubilizer and HCl digestion ..........................................

HCl digestion and chromatography with Dowex ........................

vii

59
59
59
61
61
61
61
62
63
65
66

71
71
72
76
76
78
79
79
79
80
82
92
92
92

 

 

 

CH.A

 

CH4.

 

 

Digestion in sulfuric acid

Direct Measurement of 355 in Microbes ...............................

Ratio of N to 358 in Labeled Microbes

 

REFERENCES
CHAPTER 4

Enrichment of proteolytic activity in preparations from the rumen for in
vitro studies

 

Synopsis

 

Introduction

 

Materials and ‘ ' ‘L ‘

 

 

Validation and improvement of azocascin method

Azocasein degradation

 

Chilling, blending and timing

Incubation in vitro

 

Extraction in buffer, detergent or amtnne

 

Results and 1"

 

Validation and improvement of azocascin method ........................

 

 

 

 

Chilling, blending and rinsing
Incubation in vitro
Extraction in buffer. detergent or Mm-..
REFERENCFQ
CHAPTER 5

An in vitro method to measure protein degradation of feeds using
enzymes extracted from rumen ‘ ‘

 

Synopsis

'L J a'

 

Materials and ‘ ' ‘L ‘

 

 

133
133
133
135

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Solutions and feeds 135
Butanol-acetone extraction 135
Feed protein ‘ 9 ‘ " 135
Reducing agents 138
Cellulase and amylase 138
Proteolytic activity over time 139
Effect of enzyme level 140
Comparison of feeds and variation 140
Results and I" 141
Butanol-acetone extraction 141
Reducing agents 142
Proteolytic activity over time 143
Cellulase and amylase 143
Effect of enzyme level 144
Comparison of feeds and variation 144
Conclusion 147
REFERENCFQ 148
CHAPTER 6
Effect of Plant Maturity and Preservation Method on In Vitro Protein
Degradation of Forages 158
ABSTRACT 158
INTRODUCTION 159
METHODS 159
RESULTS AND DISCUSSION 161
REFERENCES 164
CHAPTER 7
Prediction of Protein Degradation of Forages from Solubility Fractions ........... 178

ix

CHAPTER

PRE

 

 

ABSTRACT 178

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

INTRODUCTION 179
METHODS 180
Forages and Protein Degradation 180
Protein Fractinnatinn 181
Statistics for the First Experiment 133
Solubility Fractions After Degradation 184
RESULTS AND DISCUSSION 185
Fractionation of Forages 185
Prediction of Degradation from Fractions 187
Solubility Fractions After Degradation 189
Comparison to Previous Results 192
Implicatinnc 193
REFERENCES 195
BR 8
EPILOGUE 214
IX

PRELIMINARY EXPERIMENTS AND OBSERVATIONS ............................ 219
Observations on Protein Solubility 219
Measurement of Forage CP Solubility 219
Effect of pH and Eh on CP Solubility 220
Effect of Temperature on Proteolytic Activity of Alfalfa .................... 221
Observations on In Vitro Fermentation 222
Titration of in vitro media 222

Variation in pH and reducing potential for in vitro batch
(‘nltan 223
Effect of Reducing Agents on Microbial Fermentation 224

x

 

 

Effect of carbohydrate source on microbial growth and 358

 

N ratio.

Observations on Crude Rumen Fluid Extract

 

Reducing agents on protein degradation.

 

Centrifugation of rumen extract before use. ........................................

 

Presoaking of alfalfa.

 

Tungstic acid vs. TCA precipitation

xi

225
226
226
229
230
230

 

CHrl

CHAI

 

LIST OF TABLES

{AFTER 2

 

TABLE 1. Effect of freezing on crude protein and fibre fractions of
smooth bromegrass (g 100 g—l M).

TABLE 2. Effect of freezing on crude protein and fibre fractions of
fresh luceme (g 100 g-l DM).

68

69

 

TABLE 3. Effect of freezing on crude protein and fibre fractions of
luceme silage (g 100 g-l

7O

 

{AFTER 3

Table 1. Precipitation of 358 (cpm) with trichloroacetic acid after
incubation of Trypticase with microbial inocula for 10 h with addition of

sulfide or sulfide and acid before rinsing the pellet in water. ........................

Table 2. Precipitation of 358 (cpm) with trichloroacetic acid after
incubation of soybean meal with sterile media for 4 h with addition of

reducing agents before rinsing the pellet in water three times. .......................

Table 3. Precipitation of 35S—cysteine (cpm) with different methods to

Table 4. Method of rinsing the trichloroacetic acid pellet on removal of
3SS—cysteine (cpm) from a feed mixture (starch, cellulose and soybean
meal)

precipitate protein, with and in the absence of soybean meal.1 ......................

 

Table 5. Precipitation of 35S (dpm) from media incubated with rumen
microorganisms for 8 h with Trypticase, or without organisms with
soybean meal.

 

Table 6. Effect of number of times of rinsing a feed sample on removal
of 3SS-cysteine from TCA precipitable material.

 

Table 7. Effect of soaking feed sample during rinsing on precipitation

 

3SS-cysteine.

Table 8. 3SSocysteine precipitated with different feed samples incubated
(38 °C) for 3 h without microbial inocula

 

 

..... 81

81

83

84

86

86

86

87

 

 

Table 9. Effect of rinse solution pH on precipitation of 358- -cysteine

(cpm) after rinsing twice.

89

 

 

Table 10. Effect of adding eth,‘ " ' ‘ ‘ " acid,
dithiolthreotol, or cytyl trimethylammonium bromide to rinse solution
on 3SS-cysteine precipitation (cpm) with starch

Table 11. Effects of timing on precipitation of 358- -cysteine (dpm) with
feed samples 1n pH 6. 8 media.

89

93

 

Table 12. Effects of timing on precipitation of 35S-cysteine (dpm) with
feed samples in 1% HCl.

 

Table 13. Recovery of 35S (cpm) in samples digested in H2804 with
peroxide and recovery per m1 NaOH required to neutralize the H2804 in
the digest.

 

Table 14. Digestion of 35S-cysteine in sulfuric acid and peroxide on
recovery of radioactivity (cpm)

Table 15. Solubilization of 358 from radio—labeled microbes using
different solvents.

 

Table 16. Effect of incubation time with microbial inocula and nitrogen
source on nitrogen and 358 activity insoluble in trichloroacetic acid. ...............

Table 17. Effect of amount and source of nitrogen on nitrogen and 35S
activity insoluble in trichloroacetic acid after incubation with microbial

 

inocula for 18 h.

Table 18. Estimation of microbial nitrogen and undegraded feed protein
after incubation for 20 h with rumen inocula.

 

AFTER 4

Table 1. Carbohydrate composition of media used for incubation of
ruminal organisms in continuous culture.

93

98

98

101

102

104

126

 

Table 2. Treatment of rumen contents on proteolytic activity per m1
rumen fluid (PA) and per mg nitrogen (PA/N).

127

 

 

Table 3. Mean proteolytic activity per ml rumen fluid (PA) and per mg
N 1n rumen fluid (P A/N).

Table 4. Effect of treatment of rumen contents (chilling, rinsing and
blending) on recovery of proteolytic activity per ml rumen fluid (PA)
and per mg N (PA/N) in fractions separated by sequential
centrifugation.

128

129

 

Table 5. Mean proteolytic activity per m1 rumen fluid (PA) and per mg
N in rumen fluid (PA/N) by centrifugation fraction.

130

 

 

xiii

 

 

Table 6. Effect of preincubation in continuous culture at .07/h on
proteolytic activity per ml rumen fluid (PA) and per mg N in rumen fluid

 

131

 

(PA/N) by centrifugation fraction.

Table 7. Effect of solubilizing rumen fluid precipitate in reduced media,
media with detergent, and acetone on proteolytic activity per ml rumen
fluid (PA) and per mg N (PA/N).

 

HAPTER 5

Table 1. Effect of reduction of in vitro media on soybean meal (SBM)

degradation at 16 h in enzyme extracted from the rumen. ...............................

Table 2. Solubility of nitrogen from luceme hay in trichloroacetic acid
as affected by incubation with protease extracted from the rumen (P),
commercial cellulase (C), soaking in pH 4.7 buffer before P, and C
before P.

 

Table 3. Solubility of nitrogen from luceme hay in buffer (pH 6.8) as
affected by incubation with protease extracted from the rumen (P),
commercial cellulase (C), soaking in pH 4.7 buffer before P, and C

 

before P.

Table 4. Solubility of dry matter (DM) from luceme hay in buffer (pH
6.8) as affected by incubation with protease extracted from the rumen
(P), commercial cellulase (C), soaking in pH 4.7 buffer before P, and C
before P.

 

Table 5. Effects of enzyme level and addition of enzyme after 8 h on

 

soybean meal degradation at 8 and 16 h.

Table 6. Average crude protein degradation (percentage of original)

after 0, 6 or 16 h incubation with enzyme extracted from the rumen. .............

[AFTER 6

TABLE 1. Mean forage composition by species and maturity level
)

132

152

153

154

155

156

157

. 167

 

across forage preservation methods (11 = 12 .

TABLE 2. Mean protein degradation at O, 2 and 24 h by forage species
)

and maturity level across forage preservation methods (n :12 . ....................

TABLE 3. Mean protein degradation at 0, 2 and 24 h by forage species
and preservation method across maturities (n = 9)

168

169

 

APTER 7

TABLE 1. Mean percentage protein in each of 6 solubility fractions by
forage species and maturity level across forage preservation methods (11
-12).

.. 198

 

 

TABLE 2. Mean protein percentage in each of 6 solubility fractions by

 

 

 

 

forage species and preservation method across maturities (n = 9). .................... 199

TABLE 3. Prediction of forage protein degradation at 2 h determined

from multiple regression analysis of protein“ 200

TABLE 4. Prediction of forage protein degradation at 24 h determined

from multiple regression analysis of protein“ 201

TABLE 5. Percentage decrease from the initial amount of crude protein

in solubility fractions after 2 h degradation in vitro 202

TABLE 6. Percentage decrease from the initial amount of crude protein

in the respective solubility fractions over time across forages. .......................... 203

TABLE 7. Estimation using the Lucas test of percentage protein

degraded from fractions. 204
APPENDIX

APPENDIX TABLE 1. Effect of straining through cheese cloth.
filtering through Whatrnan # 54, or centrifuging (1500 g) on water
soluble CP, or trichloroacetic acid (T CA) soluble CP in orchard grass

 

hay. 220
APPENDIX TABLE 2. Effect of soaking one tenth bloom freeze-dried
alfalfa on protein solubilized in TCA. 222

 

APPENDIX TABLE 3. Effect of starch and cellulose level (E) and
casein level (P) on pH and Eh of in vitro batch culture. ..................................... 223

APPENDIX TABLE 4. Effect of sodium ascorbate addition on
precipitation of 355 (dpm) with soybean meal fermented with rumen
microorganisms. 224

APPENDD( TABLE 5. Precipitation of 358 after incubation of 358-
cysteine with rumen microorganisms, trypticase and the carbohydrate
sources listed as treatments. 226

APPENDIX TABLE 6. Method of reduction on soybean meal (SBM)
degradation with enzyme extracted from the rumen 227

 

APPENDIX TABLE 7. Effect of the presence of enzyme and reducing
method on protein degradation of alfalfa. 228

 

APPENDIX TABLE 8. Precipitation of CP before and after 16 h
incubation with crude rumen extract using trichloroacetic acid (TCA) or
tungstic acid (W). 232

APPENDIX TABLE 9. Interference of crude rumen extract with
measurement of precipitable protein (mg N of total 8 mg) 1n corn or
soybean meal (SBM) before ‘ 233

 

XV

 

CHAP

CHAP

 

 

 

LIST OF FIGURES

CHAPTER 3

Figure 1. Sulfur Metabolism in the Rumen

75

 

CHAPTER 4

Fig. 1. Optical density of azocascin solution, centrifuged rumen fluid
pellet, and centrifuged rumen fluid supernatant at varying wavelengths
of light.

 

Fig. 2. Treatment of rumen fluid on proteolytic activity per ml rumen ............

Fig. 3. Treatment of rumen fluid on proteolytic activity per mg nitrogen
in enzyme preparation.

CHAPTER 5

 

Figure 1. Effect of incubating rumen enzyme preparation at 38°C on
proteolytic activity on azocascin.

 

Figure 2. Degradation of soybean meal and luceme hay with enzyme
extracted from rumen fluid.

 

CHAPTER 6

Figure 1. Percentage of alfalfa forage protein degraded in early mid and
late maturity forage.

 

Figure 2. Percentage of bromegrass forage protein degraded in early
mid and late maturity forage.

 

Figure 3. Percentage of reed canarygrass forage protein degraded in

 

early mid and late maturity forage.

Figure 4. Percentage of whole plant corn protein degraded in early and
late maturity forage.

 

CHAPTER 7

Figure 1. Scheme for sequential extraction of forage crude protein. ................

123

125

150

151

170

172

174

. 205

 

 

 

Figure 2. Prediction of forage protein degradation at 24 h from
solubility fractions of original material. ' 206

Figure 3. Change in crude protein solubility fractions due to degradation
in ruminal enzyme extract. 208

Figure 4. Lucas test for uniformity of degradation in ruminal enzymes
across forages. 210

Figure 5. Lucas test for uniformity of degradation in ruminal enzymes
across forages. 212

met]

rau'c

indir
film:-

COmI

 

 

INTRODUCTION

In ruminant species optimal protein utilization results when adequate nitrogen is
available to rumen microbes at a utilizable rate, while additional protein remains
undegraded until it is digested and absorbed in the small intestine. Recently introduced
systems to balance cattle rations have emphasized the need to consider ruminal protein
degradation. Though many feed protein sources have been evaluated for protein
degradation, the forage protein, which comprises a major portion of the diet, is highly
variable and the degradation properties are difficult to predict. Therefore, a routine
method to predict forage protein degradation in the rumen is needed to balance cattle
rations for ruminally degraded and undegraded protein.

Rumen degradability of feed protein depends on the rate of degradation of
individual proteins in the rumen, and on the rate of passage of those proteins from the
rumen. Both of these factors may be influenced by protein associations with other feed
components, and by availability of chemical bonds of the proteins to microbial attack.
Separation of feed proteins based on solubility may have practical significance to
rumen protein degradation, because solubility in different media can separate proteins
based on their associations with plant organelles (i.e. cytoplasm, membranes, cell
wall), and solubility may relate to aspects of protein structure which affect degradation.
The proposed hypothesis is that a sequential extraction of crude protein in forages may
provide for protein fractions that degrade at uniform rates across forages. This
information would enable us to routinely predict protein degradation from size of

individual crude protein fractions.

solul

mme

evalu

crude

0n 50]
their c
metho
metho
re‘lUin
occur I
immec

 

The objectives of this work were:
1) to develop methods to separate nitrogen fractions of forages based on
solubility characteristics that may be related to ruminal protein degradation,
2) to develop a laboratory procedure to simulate feed protein degradation in the
rumen, and
3) to evaluate ruminal degradation of the isolated nitrogen fractions and to
evaluate the feasibility of using patterns of nitrogen partitioning to estimate overall
crude protein degradation.
The approach to the first objective of isolating protein fractions in plants based
on solubility was based on previous research on types of proteins found in plants and
their cellular locations as described in Chapter 1. Some preliminary studies to develop
methods to separate and measure these proteins are described in the Appendix. Any
method to routinely predict forage protein degradation based on solubility would
require that forages be preserved before analysis in such a way that changes do not
occur before processing. Since changes in forages during preservation would
immediately result in an artifact for all subsequent measurements, research into how
forages should be preserved before analysis was undertaken as described in Chapter 2.
The second objective was to develop a method to measure protein degradation
in an environment that simulates the rumen, but with the ability to monitor degradation
of individual protein fractions isolated according to the first objective. The first attempt
was to use an in vitro fermentation method that was developed as described in Chapter
3. This method was too labor intensive and variation within treatments was too high to
use the method for the desired purposes. The results are summarized in this dissertation
so that further work might improve development of the method for other purposes.
A second approach to measure protein degradation was to prepare a crude

extract from rumen fluid that would have high proteolytic activity for use with in vitro

 

Slur

11511

{13C [in
degrad
am the
analys-L

after de

how [he
needed r.

3

studies. In order to measure proteolytic activity in rumen fluid preparations, a method
using azocasein was refined and briefly validated for use with rumen fluid preparations
as is described in Chapter 4. This method was then applied to determine gross methods
to concentrate proteolytic activity. The results of these studies, as described in Chapter
4, led to development of a method to extract rumen fluid solids with acetone and
butanol to free the proteases from microbial cells, and then to extract the enzyme from
the acetone-insoluble fraction with water. The final method used was improved as
described in the Appendix and in Chapter 5, and initial validation and studies on uses
and limitations of the method are described in Chapter 5.

The third objective was to measure protein degradation of individual CP
fractions. A wide array of different forages were collected and evaluated for protein
degradation as described in Chapter 6. Cnide protein was fractionated in these forages
and the protein degradation of the fractions was determined from multiple regression
analysis of fractions vs. degradation, and from solubility analysis of the residue left
after degradation of some forages. These results are described in Chapter 7.

Finally, the studies are interrelated in Chapter 8, with particular emphasis on
how the later studies offer insight into the previous ones, and the further research

needed for further development of these methods

 

 

CHAPTER 1: LITERATURE REVIEW

Importance of Ruminal Protein Degradation

Protein Metabolism. Proteins are composed of amino acids which have a
chiral carbon atom with a carboxylic acid group (COOH), an amine group (NH2), a
hydrogen atom (H) and a variable residue (R). The residue differs for each of 20
common amino acids. Amino acids that can be synthesized by the body from common
precursors are "non-essential", and amino acids that cannot be synthesized by the body
from common precursors are termed "essential". "Limiting" amino acids are those
which are absorbed from a given diet in the least concentration relative to requirements.
That is, a decrease in the protein concentration of the diet would result in a decrease in
production or growth due to the limiting amount of these amino acids. Proteins
synthesized by body tissues have specific amino acid sequences. The non-essential
amino acids can be synthesized de novo for the production of these proteins, but
essential amino acids or their precursors must be supplied in the diet. If the body lacks
an essential amino acid, protein synthesis will be arrested until that amino acid is
supplied regardless of how much total protein is provided in the diet.

The amine group of amino acids can be replaced with an oxygen atom that
shares two electrons (double bond) by deamination, thus leaving a keto—acid. The keto—
acids are metabolized for energy. Alternatively, the amine group may be transferred
from an amino acid to a keto-acid by a process called transamination. In animals,
glutamate and aspartate may be synthesized de novo from NH3 and the respective

carbohydrate backbone, and these may be converted to glutarnine and asparagine

 

non

be 11
feed

C005

ahso
mien
abort.
amin.
3mm:
aDim;
this p
inrpo;
diggs
Wail;

{01' p(

blow
in W
being
is beli

5
respectively by amination (Lehninger, 1975). Once a keto-acid is aminated to form an
amino acid, the amine may be transferred to other keto-acids to form the non-essential
amino acids. Proteins and amino acids are continually synthesized and degraded during
normal metabolism. Some essential amino acids are lost in this metabolism, and must
be replaced by the diet.

Protein Digestion in Ruminants. Ruminants have a unique way of digesting
feed protein. Some feed protein is digested by rumen microorganisms, hydrolyzed to
amino acids and eventually to ammonia. This non-protein nitrogen (NPN) is either
consumed by the microorganisms and used to synthesize microbial proteins, or it is
absorbed from the gut as NH3. Some feed protein is not degraded by ruminal
microorganisms. This undegraded feed protein and the microbial protein pass to the
abomasum and small intestine and can be digested by animal enzymes. Peptides and
amino acids may be absorbed from the small intestine and used to meet the animal's
amino acid requirements. Some of this protein is not digested and passes out of the
animal in the feces. Some microbial protein synthesis occurs in the large intestine, but
this protein is not available for absorption. However, microbial growth may be
important for digestion of fiber in the large intestine, and some end-products from this
digestion may be absorbed in the large intestine and used for energy. Therefore,
availability of nitrogen may be important to digestion in the large intestine, especially
for post-gastric fermenters such as the horse. Nitrogen in the blood may be transported
to the large intestine and used for microbial growth.

Absorbed ammonia has one of four possible fates. On high protein diets, most
blood ammonia is transported via the blood to the liver, converted to urea, and excreted
in urine. Some ammonia is excreted directly from the blood by the kidney without first
being converted to urea. Some urea and ammonia are recycled back to the gut by what

is believed to be passive transport. Urea and ammonia diffuse into saliva that enters the

 

 

 

6

gut, and across the rumen and intestinal wall. Microbes are thus able to use this NPN
for synthesis of microbial protein. Finally, ammonia may be used for synthesis of non-
essential amino acids.

Ruminant Protein Requirements. The protein requirements of ruminants are
related to the digestive process. As with non-ruminants, the animal must absorb
essential amino acids from the small intestine. However, ruminal microorganisms are
capable of synthesizing all of the common amino acids de novo. Therefore, ruminants
have the unique ability to convert NPN to protein, and may survive and produce
without consumption of true protein. None-the—less, microbial protein synthesis alone
is not adequate to maintain the high levels of production of modern ruminants (Orskov
et al., 1980; Schwab et al., 1992). Therefore, some feed protein that escapes ruminal

degradation is required in the diet. This protein is referred to as "undegraded intake

 

protein" or "UIP" by a system commonly used to balance dairy cattle diets in the United
States (National Research Council, 1989). The protein which is degraded is referred to
as "degraded intake protein" or "DIP".

Adding sources of protein that are slowly degraded in the rumen has been
shown to increase milk production in some cases (Kung and Huber, 1983; Ciszuk and
Lindberg, 1988; Voss et al., 1988; Hamilton et al., 1992), but not in other cases
(Folman et al., 1981; Robinson and Kennelly, 1988). A requirement of UIP also has
been demonstrated for maximal growth of beef steers (Lindberg and Olsson, 1983;
Hunter and Siebert, 1987 ; Newbold et al., 1987). The amount of amino-acid protein
absorbed in the small intestine was greater for cattle fed diets containing a greater
portion of UIP relative to DIP. In these diets the rumen ammonia levels were lower for
the high UIP diets. In addition, infusion of amino acid protein directly to the small

intestine also increases production (Schwab et al., 1992) and dry matter intake

(Me

pron

optir

pron

intes

IHad.

'______~._—. k

 

 

7

(Meissner and Todtenhofer, 1989), thus demonstrating the importance of undegraded
protein entering the small intestine.

Though providing for protein that escapes ruminal degradation is important to
optimizing production of high-producing ruminants, feeding to increase microbial
protein synthesis may also increase the flow of essential amino acids to the small
intestine. Microbial growth requires an adequate supply of carbohydrate that is
fermented in the rumen, and a rumen environment conducive to microbial growth.
Microbial growth may be limited if carbohydrate degradation is too low to supply the
energy for growth (National Research Council, 1985). Alternatively, on high starch
diets, the pH of the rumen fluid will decrease which will lower microbial growth
(Staples and Lough, 1989). The interactions of these factors result in a limit to the
maximal amount of protein synthesis that can occur. Further research will address
ways to optimize microbial protein synthesis, but additional protein requirements must
be met by provided UIP.

Often times, microbial protein synthesis is limited by the amount and form of
nitrogen available to rumen organisms (Polan, 1988; Rooke and Armstrong, 1989).
Inadequate DIP may result in a decrease in protein entering the small intestine due to a
depression of microbial protein synthesis (Rooke and Armstrong, 1989). If inadequate
N is available to rumen microbes, fiber digestion may also be reduced (McAllan and
Smith, 1983B; McAllan and Griffith, 1987; Adamu et al., 1989; Zorrilla—Rios and
Horn, 1989). When microbes are not available to digest fiber in forages, fiber has a
greater filling effect, and hence feed intake is reduced (Hunter and Siebert, 1987;
Ciszuk and Lindberg, 1988). Therefore, while it is important to supply an adequate
amount of UIP to help meet the needs of the host animal, there is also a requirement for
DIP to provide for microbial growth to synthesize microbial protein, degrade fiber in

the diet, and promote normal rumen function.

 

acid:
1985
short
mixtt
amin.
not re
infusi
synth.
Anus;
0r fist.

fibcrc'

8

While most of the protein needs of rumen microorganisms can be met with
ammonia and carbohydrates, microbial growth is maximized by the inclusion of amino
acids or peptides in the rumen (Mathison and Milligan, 1971; Thomsen, 1985; Polan,
1988). In vitro studies involving incubation of rumen fluid with amino acid mixtures
showed that maximal microbial growth was obtained by supplementation of urea with a
mixture of leucine, methionine and histidine (Fujimaki et al., 1992). Addition of other
amino acids did not result in further increases in microbial growth. Sulfide addition did
not result in as high a microbial growth as either cysteine or methionine. Ruminal
infusion of casein or supplementation with soybean meal increased microbial protein
synthesis in the rumen above that observed from infusion of urea (Rooke and
Armstrong, 1989). Supplementation with soybean meal (McAllan and Griffith, 1987)
or fishmeal (McAllan and Smith, 1983; McAllan and Griffith, 1987) resulted in greater
fiber digestion than supplementation with urea.

Optimal formulation of ruminant diets requires attention to protein and energy
interactions. As was mentioned, the amount of microbial protein synthesis depends in
part on niminally-available carbohydrate to support microbial growth. Therefore, the
amount of DIP required in a ration depends on the level of ruminally-available
carbohydrate. Two carbohydrate sources were compared for effect of energy level on
protein flow to the small intestine (Cecava et al., 1988). A study showed faster
disappearance of soybean meal and sunflower meal from nylon bags suspended in the
rumen when sheep were fed dried grass instead of a barley-based diet with little fiber
(Ganev et al., 1979). This suggests that protein degradation may be affected by
carbohydrate sources in the ration.

Protein requirements for ruminants also are affected by post-absorptive
metabolism of carbohydrates and protein. Some required protein is used for

gluconeogenesis (synthesis of glucose from pyruvate). Because much of the glucose

con.
5an
gluc
rrter
intes
glucr
(Oke
ratio:
lactos
fequir
acids,

consumed by ruminants is fermented to volatile fatty acids in the rumen, considerable
synthesis of glucose is required, and amino acids may be used as precursors for
gluconeogenesis. Intravenous infusion of glucose in lambs resulted in greater nitrogen
retention (Matras and Preston, 1989). A greater amount of starch fermented in the
intestine (due to rumen protection by formaldehyde treatment) resulted in greater blood
glucose and insulin levels and improved efficiency of N utilization by Angus steers
(Oke and Loerch, 1992). It has been suggested that higher undegraded protein in the
ration may result in greater gluconeogenesis which can increase the possibility for
lactose formation and milk production (Veen etal., 1988). This would suggest that the
requirement of UIP in a ration may not be only to supply a few most limiting amino
acids, but also to provide the glucogenic amino acids required for glucose production.

If N available to microorganisms limits microbial growth, then there may be an
advantage to increasing DIP in the rumen. However, once the microbial requirements
have been met, there is no additive effect of additional rumen N (Madsen and
Hvelplund, 1988). The excess DIP is converted to ammonia in the rumen and diffuses
into the blood. In the same way, if the UIP requirement has been met, and no amino
acids are limiting, there is no advantage to further increasing the amount of UIP in the
diet beyond the effect of providing additional energy. Excess amino acids are
transported to the liver, amine groups are transaminated to glutamate or arginine and
eventually urea is synthesized.

Understanding these possible effects and interactions of rumen protein
availability makes it easier to understand why often studies fail to show positive effects
of balancing rations for UIP and DIP. The requirements for each type of protein may
depend on other feeds in the diet, and effects will not be seen unless the level of UIP or
DIP is in fact limiting production. Some authors (Robinson and Kennelly, 1988)

suggest that the requirements for UIP set by the National Research Council (1989) are

Res
diff
UIP
avai

diet,

UPC!
more

that r

mquir

When
Why fr

PTOteir.

 

 

10

too high, and in fact they are 15% higher than recommendations from the Agricultural
Research Council (1984) which are used in Europe. If two diets are compared that
differ slightly in the ruminally unavailable protein, and both diets supply an excess of
UIP, differences in production are unlikely. Because requirements for both ruminally
available and undegraded vary depending on production level and other feeds in the
diet, and because methods to measure protein degradation are crude, the exact
requirements of UIP and DIP are not known, and the exact amount of these protein
types provided in different feeds is not known. Balancing rations for UIP and DIP is
more important for higher levels of production because it is under these circumstances
that ruminal synthesis lags further behind in its ability to provide for the entire protein
requirement.

The only way to be certain to have met the requirements for both UIP and DIP
when balancing a ration is to supply a large excess of both. There are several reasons
why feeding excess protein to high-producing ruminants is undesirable. Sources of
protein are one of the most expensive ingredients in production-animal diets.
Therefore, attempts to reduce amount of protein fed while maintaining production are
important to the profitability of farms. Feeding high levels of protein results in high
levels of blood urea and ammonia. These forms of nitrogen are toxic to tissues and
may result in impaired reproductive performance (Visek, 1984; Ferguson and Chalupa,
1989). High protein diets were also associated with lameness in cattle (Manson and
Leaver, 1988). The synthesis and excretion of urea requires expenditure of energy (7.2
kcal MB per g N) (Tyrrell et al., 1970). Finally, excess N in wastes from animal
production are harmful to the natural water supply (Johnson et al., 1992). Simply
feeding excess protein to ruminants is not adequate so a better understanding of the

amounts of different types of protein required in rations is needed.

 

forag.
dried
prote:
whjcl
Argo

datior
1983)
the ru
of Up
(Gian
PFOIei
Rendr
exam I
imPOr
linked
degfac

 

 

 

11

Effect of Protein Source on Degradation

Feeds differ in the ruminal availability of their protein. legumes and most
forages have high levels of degradable protein compared to rendering by-products and
dried corn grain These feeds differ in the protein composition and interactions of these
proteins with other feed components. Corn grain endosperrn proteins include glutelin
which is cross-linked by disulfide bonds into a network (Duvick, 1961; Paulis, 1981;
Argos et al., 1982) that explains the slow protein degradation (Harvey and Oaks, 1974).
The presence of disulfide bonds has been shown to be associated with reduced degra—
dation of other protein sources as well (Mahadevan et al., 1980; Wallace and Kopecny,
1983). Corn grain storage proteins (zein) are hydrophobic, and also slowly degraded in
the rumen (Harvey and Oaks, 1974). While these proteins may be an important source
of UIP, the lysine content (the first limiting amino acid for many rations) is low
(Gianazza et al., 1977; Esen, 1980). The conformation and amino acid residues of these
proteins would prevent them from coming into contact with proteolytic enzymes.
Rendering by-products are composed of several types of cross-linked proteins. For
example, blood meal contains the protein fibrin which is a cross-linked glycoprotein
important to blood clotting, and meat and bone meal contains collagen, which is cross-
linked by covalent bonding between lysine residues (Lehninger, 1975). Readily-
degraded proteins typically are not cross-linked (Mahadevan et al., 1980).

Proteins of plant and animal sources are frequently heated to reduce their
susceptibility to degradation (Coenen and Trenkle, 1989; Aguilera etal., 1992; Nia and
Ingalls, 1992). Proteins may also be cross-linked by Amadori rearrangement reactions
that link lysine residues with carbohydrates (Wallace and Falconer, 1992). Heating also
crosslinks proteins by formation of isopeptide bonds between the amino residue of
lysine with the carboxyl residue of glutamate and aspartate (Papadopoulos, 1989).

Finally, heating can result in the formation new amino acids that can be cross-linked.

 

Fissun
the am
results
and wi
and cy:
Ironica
soy-b3;
may re
dcgrad.
about t
POSitivt
the Pre
While 5

MOWar

degrad.
n10m
1992).
involve
addif10
Soaps (
PTOVi¢

Protein

degrad
Crude I

 

 

12

Fissure of cystine disulfide bond may give rise to dehydroalanine which can react with
the amino groups of lysine (Papadopoulos, 1989; Sathe etal., 1989). Therefore, heating
results in reduced protein degradation by increased cross-linkage of proteins internally
and with other proteins and with carbohydrates at specific amino acid residues. Lysine
and cysteine are frequently destroyed or made indigestible from these transformations.
Ironically, lysine and methionine are considered the first-limiting amino acids on com-
soy-based rations fed to dairy cattle (Schwab etal., 1992), and destruction of cystein
may require conversion of methionine to cysteine. If one goal of decreasing protein
degradation is to increase the supply of these amino acids, heat treatment may not bring
about the realization of that goal. This may be another reason some studies do not find
positive results from supplementing rations with additional UIP. Mild heat treatment in

the presence of sugars has been applied to reduce the destruction of amino acid protein

 

while still protecting amino acids from microbial degradation (Mosimanyana and
Mowat, 1992; Wallace and Falconer, 1992).

Seed proteins are frequently treated with formaldehyde to reduce ruminal
degradation of both protein and starch (Hunter and Siebert, 1987; Van Ramshorst and
Thomas, 1988; Fluharty and Loerch, 1989; Hamilton et al., 1992; Oke and Loerch,
1992). The formaldehyde treatment also crosslinks protein with starch and may also
involve lysine, but a better understanding of the chemical reactions is needed. In
addition, proteins have also been protected from degradation by coating in calcium
soaps of long-chain fatty acids (Sklan, 1989). This treatment has the dual benefit of
providing increased energy to the small intestine as well as increased undegraded
protein.

Though most research on manipulation of the susceptibility of feeds to
degradation has been with concentrates, forages provide a significant amount of the

crude protein provided to high-producing ruminants. Though most forages appear to be

l3

highly degraded, they differ in their susceptibility to degradation. Frequently-used
tables for balancing dairy cattle rations (National Research Council, 1989) indicate that
proteins of dried grasses are not as readily degraded in the rumen as those of corn
silage, and that corn silage protein is less degraded than alfalfa protein. These values
should be reevaluated with consideration to the techniques used to measure protein
degradation of forages. The disappearance of protein from nylon bags suspended in the
rumen was similar for alfalfa and corn silages, but was lower for orchard grass hay than
for alfalfa hay (Janicki and Stallings, 1989). In a different study, sainfoin crude protein
was degraded to a lessor extent in vitro in continuous culture than alfalfa, cicer
milkvetch or birdsfoot trefoil (Dahlberg et al., 1988), but other forages did not differ
from each other. Despite these effects in vitro, there were no differences when these
forages were examined in situ. Silages of grasses were compared to clover silages
incubated in situ (Vik-Mo, 1989). Initial disappearance of crude protein from the bags
was greater for grasses, but clover silage degraded faster at later points in the
incubation. Switchgrass (a warm season grass) crude protein disappeared from nylon
bags more slowly than that of smooth bromegrass (a cool season grass) (Mullahey et
al., 1992). A different study showed that protein disappeared faster from nylon bags in
situ from the cool season grass species bromegrass and wheatgrass than for the warm-
season grasses big bluestem and switchgrass (Kephart and Boe, 1991). Absolute values
for protein degradation and comparison of crude protein degradation across studies is
not possible due to the large variation in the techniques used to measure and to
calculate protein degradation.

Conservation of forages results in changes in the protein available for
degradation. Frozen grass and hay crude protein (CP) was estimated to have similar
amounts of rumen available CP from in situ data. but silages of the same material were

estimated to have higher amounts of CP available to the rumen (Petit and Tremblay,

1992;
Stalli

than!

prote:
conse
affect
aCCOI

McDr

PrOtei
PTOCe
micro
low p
PFOdu
McD(
0f the
PTOteh

Co'lser

addilic
forage:
1982).

with T84

bags w.

G

14
1992). The same was true for comparison of alfalfa hay and silage (Janicki and
Stallings, 1989). Dried perennial ryegrass provided more CP to the duodenum of sheep
than the same forage that was fed fresh or had been frozen (Beever et al., 1974).

The susceptibility of protein to degradation and digestion, and the amount of
protein degraded in conserved forges may be affected by various factors that affect the
conservation process. Field leaf loss, rate of wilting, anaerobiosis, and heating may all
affect the level of protein degradation of the conserved forages. The wilting process is
accompanied by proteolysis by plant enzymes (Ohshima and McDonald, 197 8;
McDonald, 1982; Petit and Tremblay, 1992). Proteins are degraded to amino acids and
ammonia. Ensiling of forages results in greater solubilization and degradation of plant
proteins by microorganisms (Ohshima and McDonald, 1978; McDonald, 1982). This
process continues until the silage pH is reduced low enough to inactivate plant and
microbial enzymes (McKersie, 1985). The only type of enzymes capable of activity at
low pH, aspartic proteases, are not found in silage (McKersie, 1981). Heat can be
produwd during ensiling or storage of moist hay due to respiration (Ruxton and
McDonald, 1974). While heat production generally has a negative effect on the quality
of the forage, some heat production may decrease the susceptibility of the forage to
protein degradation (Broderick et al., 1993). Generally heat production during
conservation makes the forage protein and carbohydrate indigestible and may reduce
the palatability of the forage (Ruxton and McDonald, 1974; McDonald, 1982). In
addition, clostridial bacteria actively degrade silage and hay proteins when these
forages are not adequately preserved (Ohshima and McDonald, 1978; McDonald,
1982).

Some studies have suggested that advanced plant maturity may be associated
with reduced ruminal degradation of forage protein. Disappearance of CP from nylon

bags was slower for silages made from more mature grasses (timothy, fescue,

br

pl:

prt
brt

{Bit

863:
(Ah
mat
pho
Tron
slou
in p“
C0n(
ferti
on p
Solu
1985
Van:
and

Wu

15

bromegrass) and legumes (red clover, birdsfoot trefoil) (V ik-Mo, 1989). Initial CP
solubility was lower at advanced maturity of alfalfa, red clover, orchard grass, perennial
rye grass, quack grass and timothy, but not different for birdsfoot trefoil and
bromegrass (Hoffman et al., 1992). The rate of disappearance of CP from nylon bags
incubated in situ of the insoluble residue was slower with advancing maturity for most
plant species (Hoffman et al., 1992). Protein disappearance from nylon bags was lower
in mature alfalfa samples compared to early samples (Griffin et al., 1991). Crude
protein disappeared more quickly from nylon bags incubated in situ for early
bromegrass compared to late bromegrass, but bromegrass maturity had no effect on N
retention in non-lactating Holstein cows (Messman et al., 1992A).

The in vitro degradation of CP by protease from S. griseus was affected by
season for pasture samples, with the least degradation occurring in mid-summer
(Abdalla et al., 1988). Several factors could contribute to this variation including plant
maturity, level of regrowth (least when growth is slowest), climate (rain, temperature,
photoperiod), fertilization, or plant species and variety composition. Crude protein
from summer regrowth disappeared from nylon bags suspended in the rumen more
slowly than spring growth of alfalfa (Griffin et al., 1991). The concentration of nitrate
in plants increases during regrowth and periods where soluble carbohydrate
concentrations are lower (Ourry et al., 1989). Though it would seem that nitrogen
fertilization of forages would result in increased uptake of NPN and subsequent effects
on protein degradation, three independent studies showed no differences in CF
solubility or degradation in situ due to level of fertilization (Lindberg, 1988; Coto et al.,
1989; Messman et al., 1992A). Differences in protein degradation among alfalfa
varieties were demonstrated using an in vitro method with rumen enzymes (Broderick
and Buxton, 1991), but were not demonstrated in a different study using the in situ

technique (Kephart and Boe, 1991 ).

A

In
treatment
These pro
However,
protein do
so a grate
carbohydr;
from the n
degradatio
greater 0p;

Pla

{032% P2
lignified sc
Season gas
for in situ d
1989). Gm
bundles, am
green Panic
I\‘3ducti0n of
Panic becaus
digestible m.
by Wise 1ng
FOI’ag

plants. Fresh
nuclfiou'deS,a

(Lflfleton‘ 19

l6

Intentional treatments to reduce forage protein degradation include heat
treatment (Broderick et al., 1993) and treatment with formaldehyde (V ik-Mo, 1989).
These procedures operate using the same principles as discussed for high-protein feeds.
However, there are potentially some disadvantages of treating forages for reduced
protein degradation. The forage protein is associated with high levels of carbohydrate
so a greater bulk must be treated, and there is a greater potential for bonding with
carbohydrate thus reducing its digestibility. Secondly, forages are believed to pass
from the rumen at a slower rate. Therefore, treatments must be decrease protein
degradation rate to a greater extent than is required for concentrates because there is
greater opportunity for ruminal degradation.

Plant anatomy may influence ruminal digestion of carbohydrate and protein in
forages. Parenchymal tissues are more rapidly degraded by rumen microbes than more
lignifred sclerenchymal tissues (Akin, 1979; Grabber and Jung, 1991). The warm—
season grass green panic was compared in situ to the cool-season grass Italian ryegrass
for in situ degradation, as studied by light and electron microscopy (Wilson et al.,
1989). Green panic leaves had a larger cross section, larger and more frequent vascular
bundles, and less but more densely packed mesophyll. Due to the greater rigidity,
green panic leaves were broken down more by chewing than ryegrass leaves. However,
reduction of the chewed leaf particles by microbes was faster in ryegrass than in green
panic because ryegrass epidermis and bundlesheath cells were easily split from the
digestible mesophyll cells, but green panic mesophyll was protected from degradation
by these lignified structures.

Forage nitrogen is found in several different forms within different structures of
plants. Fresh forages contain NPN including nitrate. ammonia, amino acids,
nucleotides, and chlorophyll which comprise 10 to 25% of the total crude protein

(Lyttleton, 197 3). The most abundant true protein is ribulose-bisphosphate

A

l7

carboxylase/oxygenase (RUBP do) which comprises 30% of the total soluble true
protein of temperate forages and 10 to 25% of tropical forages (Lyttleton, 1973). When
isolated RUBP do was readily degraded in vitro (Nugent and Mangan, 1981). This
protein was readily degraded in silage of alfalfa, orchard grass, ryegrass and fescue, but
not in vetch or pearl millet (Messman et al., 1992B). The latter forages contain tannins
which have been shown to bind to RUBP c/o (Jones and Mangan, 1977), and thus could
in turn reduce its susceptibility to degradation . Changing the diet from fresh to dried
forages was shown to decrease in vitro proteolytic activity on RUBP do of rumen fluid.
This suggests that bacterial populations may adjust to the presence of the protein by
changing amount or type of enzymes produced, or by changing the species of
organisms, . (Hazlewood etal., 1983). Other soluble proteins are also generally
enzymatic and have diverse functions (Mangan, 1982). Chloroplast membranes contain
insoluble true proteins that are amphiphilic (Lyttleton, 1973; Askerlund et al., 1989;
Marder and Barber, 1989), These proteins include those involved with electron
transport for photosynthesis (Askerlund et al., 1989; Marder and Barber, 1989; ).
Additional proteins are associated with plant cell wall (Lamport , 1980; Lee et al.,
1990) or are bound to tannins (Albrecht and Muck, 1991) or lignin (Sanderson and
Wedin, 1989A and B).

High molecular mass phenolic compounds (tannins) bind to protein and
carbohydrates to alter the ruminal availability and whole tract digestibility. Tannins
may form cross-links with proteins that are hydrolyzable or condensed (Wong, 1973).
Both forms of tannins would theoretically inhibit protein degradation by preventing
attachment of enzymes, but cross-linkage due to hydrolyzable tannins are reversed by
the acid conditions of the abomasum. Thus these tannins decrease protein degradation
in the rumen but do not affect the overall protein digestion (Jones and Mangan, 1977).

Other authors have correlated the concentration of tannins in sainfoin (Jones and

A

 

18

Mangan, 1977) lotus (Barry and Manley, 1984) and across several species (birdsfoot,
sainfoin, lotus, sericea) (Albrecht and Muck, 1991) with reduced protein degradation in
the rumen and in silage. Tannins have also been shown to be associated with reduced
protein digestion (Hagerman et al., 1992; Hanley et al., 1992). Though much of the
affect of tannins on digestion is likely to be caused by the cross-linkage, tannins may
also be toxic to bacteria (Hartley and Akin, 1989).

Protein complexes with carbohydrates may also inhibit protein degradation.
Malliard product is formed when forages are overheated during ensiling or hay-making,
and this complex makes protein and carbohydrates indigestible (Goering etal., 1972;
Van Soest, 1982). Glycosylated proteins are also found in fresh forages, and these
linkages give the proteins unique solubility (including some TCA soluble proteins) and
slower degradation (Lamport, 1980; Lee et al., 1990). Researchers have reported the
accumulation of glycoproteins in the rumens of steers fed bermuda grass and alfalfa hay
(Windham, 1989), which suggests they are slowly degraded.

Animal and Microbial Effects on Protein Degradation

The percentage protein degraded in the rumen is not completely dependent on
the structural characteristics of the feed. Changes in pH could affect the solubility
and conformation of proteins that may alter availability to peptide bonds for enzymatic
attack (Assoumani et al., 1992). Changes in rumen reducing potential may also affect
rate of protein degradation. One of the most important structural limitations to protein
degradation are disulfide bonds. These bonds are disrupted and formed by non-
enzymatic processes under conditions of high reducing potential. The fraction of bonds
disrupted and formed at a given pH and reducing potential can be determined by
application of the Nerst equation (Segel, 1976). The pH and reducing potential of the
rumen varies to such an extent that as little as 10% and as much as 90% of the available

disulfide bonds are likely to be disrupted on different diets (calculated from Barry et al.

(l

thi

cle

pro

prot
betv
the t
at th

pepti

(Barr
and t
Plant
high

sult‘h;
and a
0min
Prime

Optim

{Ound

20f“

ll

19

(1977) and Segel, (1976)). Some disulfide bonds are not likely to be accessible to the
aqueous environment and will remain intact (Mahadevan et al., 1980). None—the-less,
this aspect of the ruminal environment is likely to influence protein degradation.

Enzymes responsible for protein degradation differ in their mode of action and
cleavage sites on proteins (Barrett, 1985; Wagner, 1985). Proteases may cleave
proteins at specific internal sites (endopeptidases) or at the first, second or third amino
acids from the amino or carboxyl-terminus of the peptide (exopeptidases). Some
proteases have both endo— and exo—peptidase activity. Proteases also cleave proteins
between specific amino acids in the substrate. For example, trypsin cleaves peptides at
the carboxyl group of lysine or arginine, and chymotrypsin and pepsin cleave peptides
at the carboxyl group of phenylalanine, tryptophan or tyrosine. Thermolysin cleaves
peptides at the amine group of 1eucine,isoleucine or valine (Lehninger, 1975).
Proteases are also classified by the active sites on the enzyme into four major categories
(Barrett, 1985). Serine proteases are commonly found in bacteria, plants and animals
and include trypsin and chymotrypsin. Cysteine proteases are also common in bacteria,
plants, and animals and include papain derived from papaya. These enzymes require
high reducing potential and are inhibited by conditions that result in reaction of the
sulfliydryl group of the active cysteine. Aspartic proteases are found primarily in fungi
and animals and include pepsin. These enzymes have a characteristically low pH
optima (pH 2 for pepsin). Metalloproteases require calcium and zinc and are found
primarily in bacteria. The multitude of different enzymes that exist have different
optima for pH, reducing potential, presence of inhibitors, and type of substrate
(Wagner, 1985).

A few studies have been conducted to attempt to characterize the proteases
found in rumen fluid. Most proteolytic activity is precipitated by centrifugation at

20,000 g (Brock et al., 1982; Kopecny and Wallace, 1982). However, gentle agitation

bat
AC

cla
adt

€112

vitr

aCC

link

6an

prof
or a,

20
in phosphate buffer resulted in solubilization of much of this activity from rumen
bacteria (191% of original) without disruption of cells (Kopecny and Wallace, 1982).
Additional activity of similar character was removed by blending and treatment with
detergent (Kopecny and Wallace, 1982). Cysteine proteases were the most dominant
class found in rumen fluid (Brock et al., 1982; Kopecny and Wallace, 1982) , with
additional activity of serine and metallo-proteases (Brock et al., 1982). Rumen
enzymes included a broad range of activities on different synthetic substrates including
both endo— and exoprotease activity (Wallace and Kopecny, 1983). None—the-less, in
vitro (Wallace, 1992) and in vivo (Chen et al., 1987) studies demonstrated an
accumulation of hydrophobic peptides from incomplete protein degradation, which
demonstrates that protease activity in the rumen is limited in the ability to cleave some
linkages.

Protein degradation in the rumen may depend on changes in the rumen
environment that affect the different proteases, and on the changes in the protease
profile found in the rumen. The amount of the different types of proteolytic enzymes,
or the amount of substrate, can affect the fraction of available substrate that can be
degraded according to Michaelis—Menten kinetics (Lehninger, 1975). Different protein
sources resulted in different determinations of Km and Vmax for the same rumen fluid
samples (Broderick and Clayton, 1992). Therefore, some feeds not only may be more
likely to degrade at a faster rate than other feeds, but the susceptibility of some feeds to
enzyme limitation of the rumen fluid may be greater than for other feeds. In other
words, some proteins may degrade more uniformly when fed to different animals and
across diets than other proteins.

The level of protein degradation activity on RUBP do was more than doubled
by changing the ration from hay to fresh forage (Hazlewood et al., 1983). This would

indicate a possibility for adaptation of rumen organisms to the protein source. A

11.7%? fig—"A13 ":7 . ' W

anin
iono‘
nesul
data t
degra

increa

Proleir

Vary dc
There fc

Protein

 

21

different study showed that feeding the resistant protein albumin resulted in increased
numbers of proteolytic bacteria (Butyrivibiro spp.) and increased activity on azocascin,
but the activity on albumen was not increased relative to that of azocascin (Wallace et
al., 1987). Therefore, fluctuations in proteolytic activity may exist, but evidence of
adaptation to specific proteins was not demonstrated.

A lower level of ruminally degraded carbohydrate in the ration was shown to
decrease protein degradation in vivo (Meyer et al., 1986; Cecava et al., 1988) and in
situ (Meyer et al., 1986). However, protein disappeared faster from nylon bags
incubated in rumens of sheep fed grass than those fed barley (Ganev et al., 1979).
Proteolytic activity measured in vitro was higher when rumen fluid was collected from
animals fed high energy diets (Furchtenicht and Broderick, 1987). The addition of the
ionophore monensin, which inhibits the proteolytic organism Streptococcus bovis,
resulted in reduced ammonia production in the rumen (Chen and Russell, 1989). These
data would suggest that proteolytic enzyme activity in the rumen partially limits protein
degradation, because diets that increase this activity, such as those high in starch,
increase the portion of feed protein degraded.

The percentage of protein degraded in the rumen depends upon the rate of
protein degradation and on the rate of passage of the protein from the rumen. The faster
the protein passes, the less time is available for protein degradation. Passage rates will
vary depending on rumen size, diet, intake level, and nature of the feed protein.
Therefore, undue rigidity should not be placed on values of expected UIP for a given
protein source.

Methods to Determine Protein Degradation

Measurement of Protein. Ruminant nutritionists have developed the practice

of measuring nitrogen instead of true protein. Forage protein is about 16% nitrogen, so

the estimation of protein in forages is 6.25 (x) nitrogen content (Association of Official

(H
mt

re;

Fre

8CC
hav
deb:
197t
bind
eitht
With

men
198:
men
Plot.
mea
in Vi

mlCr

22

Agricultural Chemists, 1975). The macro-Kjeldahl method is commonly used to
measure feed nitrogen (Association of Official Agricultural Chemists, 1975). Recent
modifications have made the method faster and reduced the requirements for reagents
(Hach et al., 1985; Schmitter and Ribs, 1989). Comparison of traditional Kjeldahl
method with use of block digestion or peroxide catalysts showed that there was good
repeatability and recovery of nitrogen when using the peroxide method, but some
samples were not adequately digested in block digesters (Watkins et al., 1987).
Frequently protein is measured by spectrophotometry after reaction with folin-phenol
reagent (Lowry et al., 1951). However, several compounds in forages interfere with
accurate measurement of true protein by this method (Askerlund et al., 1989). Attempts
have been made to separate interferences from protein by TCA precipitation before
determination of protein degradation by Lowry's method (Bensadoun and Weinstein,
1976). Attempts to measure forage protein by spectrophotometric determination after
binding with bicinchoninic acid (BCA) were not well correlated with Kjeldahl nitrogen
either before or after precipitation with TCA (Messman and Weiss, 1992). Interference
with forage phenolics or other compounds was suspected as the problem.

Estimation of Microbial Protein Synthesis. Several reviews have described
methods used to measure protein degradation of feeds (Orskov et al., 1980; Lindberg,
1985; Madsen, 1985; Nocek, 1988; Michalet-Doreau and Ouldbah, 1992). These
methods can be confined to one of three categories: 1) in vivo measurement of feed
protein arriving at the small intestine by collection from duodenal cannulas, 2) in situ
measurement of protein disappearance from nylon bags suspended in the rumen, and 3)
in vitro laboratory methods. One common difficulty for methods that involve live
microorganisms is the estimation or separation of microbial protein from undigested

feed protein.

A

E)
S)".
19
ma
ma
prc
(H:

“281

alsc

23

Various procedures have been used to determine the proportion of microbial N
associated with feed in digesta. These procedures include measurement of modified
amino acids or sugars that are found exclusively in microorganisms such as
diaminopimelic acid (DAPA) (Czerkawski and Foulds, 1974), arninoethylphosphoric
acid (AEPA) (Ling and Buttery, 1978), D-alanine (Rooke et al., 1984) or L-glucose
(Rooke et al., 1984). Total RNA or purine concentration also is used with the
assumption that RNA of feed origin is entirely digested (Zinn and Owens, 1986).
Excreted metabolites of nucleic acids have also been used to estimate microbial
synthesis in less intrusive studies ( Lindberg et al., 1989; Lindberg and Jacobsson,
1990; Puchala and Kulasek, 1992). The most problematic aspect of using these internal
markers for estimation of microbial growth is the lack of uniformity of concentration of
marker per amount microbial nitrogen across the microbial ecosystem. For example,
protozoa have no DAPA and have higher concentrations of AEPA titan bacteria
(Henderickx et al., 1972; Ling and Buttery, 1978). Whereas, DAPA is associated with
the cell wall of gram positive bacteria, this concentration will vary with conditions that
result in changes in the type of microbial growth. The RNA content of bacteria may
also vary depending on the microbial population and rate of growth (Matin, 1992).

Stable and radioactive isotopes of nitrogen (Nolan et al., 1972; Varvikko and
Lindberg, 1985; Javorsky et al., 1986; Aharoni et al., 1991), sulfur (S) (Henderickx et
al., 1972; Raab et al., 1983; Spencer et al., 1988; Thomsen, 1985), or phosphorus (P)
(Buckholz and Bergen, 1973; Smith etal., 1978) have been used as microbial markers
but radioisotopes may require costly safety procedures for large animal studies and
mass spectroscopy for 15N may be expensive or inaccessible. Inorganic forms of the
nutrient such as nitrate (N03), phosphate (P04 ) or sulfate (S04) often are placed in the

rumen or in vitro flask and microbial growth is estimated based upon isotope

A

 

 

24

incorporation into microbial matter. Isotopic methods label both bacteria and protozoa
without labeling feed sample.

Several studies have compared the use of different microbial markers on the
estimation of microbial CP. Some authors have reported similar estimates of microbial
CP from estimates made using DAPA vs. RNA (McAllan and Smith, 1983), or DAPA
vs. 358 (Mercer et al., 1980). Since DAPA is to estimate only bacterial protein one
would expect it to underestimate the total microbial CP. However, DAPA predicted
much greater microbial CP than 15N (Varvikko and Lindberg, 1985), or higher and
unrealistic estimates compared to 35S (W hitelaw et al., 1984), and it resulted in an
estimate that was nearly twice as high as from 15N or 358 (Siddons et al., 1979). Other
work did show DAPA to underestimate microbial CP compared to 35S and RNA with
an important diet by marker interaction effect (Ling and Buttery, 197 8; Smith et al.,
1978). Microbial CP estimated using 32F to pass to the duodenum was 85% the level
estimated from RNA (Smith et al., 1978). Because 32? labels the nucleic acids
synthesized by bacteria, the authors concluded that differences in microbial yield
determined with the two markers resulted because some feed RNA (which would not be
32P labeled) was not digested in the rumen.

The assumptions and problems of using different external markers are similar,
but further discussion will focus on use of 35S because this marker was employed in a
study described in this dissertation. Radioactive sulfur often is used owing to its low
energy B-emission which reduces the risk of hazardous radiation exposure. Various
microorganisms bring about the enzymatic reduction of S04 to sulfide (H2S) which is
then converted to the thiol group of cysteine by reaction with pyruvate and ammonia or
alternatively with serine. Methionine is synthesized from the cysteine by rumen
microbes. Therefore, radioactivity associated with protein after incubation of microbes

with inorganic sulfur results from microbial synthesis of amino acids.

 

 

25

A ratio of nitrogen to radioactive sulfur (Nz35S) can be determined for a
standard which is entirely of microbial matter. This ratio can then be multiplied by the
amount of 358 associated with a mixture of dietary and microbial protein. The product
is the estimate of microbial N. The standard of 100% microbial protein may be
obtained by isolating protein from an in vitro rumen incubation that used only non—
protein nitrogen (Henderickx et al., 197 2). Alternatively, microbes may be separated
from dietary protein by differential centrifugation (Walker and Nader, 1975). Another
alternative to evaluating the N2353 ratio is to hydrolyze total protein and isolate total
methionine or cysteine, then quantify percent microbial protein as 35S-methionine or
cysteine/total methionine or cysteine x 100 (Gunn and McMeniman, 1989). This
approach is obviously time consuming, expensive and still requires assumptions about
the sulfur amino acid content of microbial and partially digested protein.

The successful use of 358 as a microbial marker requires the validity of a
several assumptions. Inorganic sulfur must be the only source of sulfur for methionine
or cysteine incorporated into microbial protein. If sulfur amino acids from the feed
protein source itself are used instead of inorganic sulfur, total microbial protein will be
underestimated. In vitro incubation of rumen fluid containing 358 and 14C—labeled
sulfur amino acids as a nitrogen and energy source resulted in less than 1% and 11%
incorporation of free cysteine and methionine respectively directly into microbial
protein (Nader and Walker, 1970). The majority of the microbial sulfur was provided
from the inorganic sulfur pool which resulted in low 14C:35S ratio of microbial sulfur
amino acids. The authors concluded that direct incorporation of sulfur amino acids
from feed sources was negligible.

Results obtained from a similar in vitro procedure indicate that the form and
amount of 358 used is an important consideration (Halverson et al., 1968). Cystine and

methionine added to a sulfate and rumen fluid incubation appeared to be more readily

26

converted to H2S than 358-1abeled 804. Moreover, added H28, cystine and
methionine depressed the incorporation of 35S-labeled 804 into microbial protein
though casein had no consistent effect. The authors attributed the decrease in 353-
labeled 804 incorporation into microbial protein to incorporation of H28 from cystine
and methionine into the protein in place of H28 from 804. These results suggest that
804 may not be an adequate carrier of 358, and that labeled H28 or labeled cysteine are
preferred.

An additional assumption that must be true for accurate measurement of
microbial contamination is that microbial proteins produced from different types of
fermentation have the same N:358 ratio. It is particularly important that the isolated
microbial cells or microbial protein used as a standard have the same N:358 ratio as the
sample of unknown quantity of microbial protein. The N:358 ratio of microbial culture
in NPN media may differ from that of the culture in media with true protein to assay.
Non-protein components of the feed sample might further alter the microbial
metabolism causing a difference in N2358 ratio between standard and samples and
among different samples. Addition of excess energy and microbial growth factors
(vitamins, minerals, isoacids) may diminish effects of non-nitrogen feed components on
microbial action.

The standard used to determine Nz358 ratio alternatively may be obtained from
differential centrifugation rather than culturing microbes in non-protein nitrogen media.
Methods of differential centrifugation may isolate only certain microbes that have
different Nz358 ratios than the overall microbial population. For example, bacteria
bound to fiber and protozoa may not be recovered by such separation and may differ in
sulfur-amino acid content or in their ability to incorporate inorganic sulfur. In addition,
the cell fraction recovered may not have the same S-amino acid content as whole cells.

Bacteria have virtually no S-arnino acids in cell wall (Mathers and Miller, 1980). and

A

_.._. —» 7' f»? hrrigeflf- 1‘ , .7

27

nucleic acids represent 30% of the total N content of microbes though they contain no
sulfur. Bacteria are 52% true protein and 20% RNA and DNA (Gottschalk, 1985).
RNA is 18% N and protein is 16% N (calculated from Lehninger, 1975).

The behavior of inorganic sulfur during fermentation adds complexity to the
use of sulfur as a microbial marker. Elemental sulfur (SO)is insoluble in aqueous
solutions. At pH 7 and Eh -.25 V (possible conditions in the rumen) the ratio of
[80]:[H28] is calculated from the Nerst equation to be .46. Greater reducing potential
and lower pH reduce this ratio exponentially so that if pH is reduced to 5 or Eh is
reduced to -.35 V elemental sulfur is less than 1% of H28 (Segel, 1976). Under
reducing conditions and low pH, H28 gas may be evolved after exposure to air (Walker
and Nader, 1968), and it may be oxidized rapidly to 80. Oxidation of samples to
convert 80 solid to 8042' has been used to adjust samples before analysis. None-the-
less, the volatilization of inorganic sulfur compounds results in additional difficulty in
use of sulfur as a marker, especially for in situ or in vivo studies (Mathers and Miller,
1980). These possibilities necessitate careful technique to maintain a closed
environment, and an excess of inorganic sulfur may be necessary, as well as appropriate
interpretation of results.

Another difficulty related to solubility of sulfur compounds is that as much as
20% of inorganic sulfur was found to associate with protein added to sterile rumen fluid
(Walker and Nader, 1968). In this experiment sulfur was precipitated with barium
chloride (BaCl2) after digestion in nitric acid (HN03) and perchloric acid (HClO4.).
Though the samples were acidified before digestion, some elemental sulfur (80) may
have been associated with the protein especially if oxidation occurred before
acidification . Then 80 would have been converted to sulfate (8042‘) during the
oxidative digestion. In order to eliminate inorganic sulfur associations with protein

other workers hydrolyzed protein in HCl after oxidation with perforrnic acid (H2CO3)

 

_ —____. . wizfi

28

and prior to precipitation of sulfate using BaCl2 (Mathers and Miller, 1980). The
perforrnic acid oxidation converted S0 to 8042'.

In Vivo Methods to Measure Protein Degradation. Determination of protein
degradation in the rumen may be accomplished by collection of digesta samples from
surgically placed cannulas in the abomasum (McAllan and Smith, 1983) or duodenum
(Santos et al., 1984; Van Bruchem et al., 1988; Gasa et al., 1991). Duodenal cannulas
are placed in the proximal region to minimize intestinal absorption before sampling and
to prevent secretions from the pancreatic and bile ducts from contaminating the sample.
Reentrant cannulas are occasionally used, and they can be automated so that total
digesta is diverted outside the intestine from which it is sampled and returned. More
commonly, and with less physical stress and complications, simple "T" cannulas are
used. From these cannulas, samples can be taken during desired periods of time. A
stop-gate mechanism can be inserted to prevent a separation of digesta that results in
the sample being of a different composition than the digesta which passes beyond the
cannula during sampling. None-the—less, it is conceivable that fluid digesta will be
collected in greater quantity than particulate matter, because liquids are more likely to
flow from the cannula after it is opened than are solids. Samples of digesta from simple
cannulas therefore may not be representative of the digesta passing from the rumen.

Use of duodenal cannulas to measure protein degradation requires use of
indigestible markers to determine total digesta flow in order to determine the
proportional sample size. The use of these markers has been reviewed (Faichney,
1975). An indigestible marker must be administered continuously until steady state
conditions are achieved. The passage of digesta (L/d) can be determined as flux of
infusion of marker (g/d) divided by the marker concentration in digesta (g/L). Marker
infusion may be discontinued after steady state conditions are achieved. The passage is

then determined from the decline in marker concentration from the first order equation

 

29

-d[A]/dt = k[A], or more directly from the integral of this equation [A]i = [A]0 (x) e'kt,
whereas d[A]/dt is the fractional change in marker concentration, k is the fractional rate
of passage, and [A] is the concentration of marker initially ([A]0) or at time t ([A]i).
When using simple cannulas that result in unrepresentative sampling of duodenal
contents, a two-marker method is required to correct for the disportionate amount of
fluid phase relative to particulate phase of digesta collected (Faichney, 1980; Faichney,
1992). The currently available markers have various limitations that result in
inaccuracies of the passage estimates. Markers have been shown to be partially
digested and absorbed, to dissociate from the phase they are intended to mark, or to
bind to the gut wall or to other feeds (Warner and Stacy, 1968; Faichney, 1975; Teeter
and Owens, 1983; Cruickshank et al., 1989). These limitations as well as the difficulty
of using markers correctly add to the inaccuracy and impreciseness of in vivo methods.

Estimation of protein degradation in vivo is made by measuring the
concentration of protein in samples collected from the small intestine. This value (g
1 protein lg DM collected) is multiplied by the total passage of digesta through the rumen
, determined by markers (g DM / d). From this value for total protein flow to the small
intestine, microbial protein is subtracted. Microbial protein is determined by using
8 microbial markers with all their limitations as discussed previously. The passage of
undigested feed protein per day is then divided by the total protein intake per day to
determine the fraction of undegraded protein in the diet. In vivo methods are certainly
_ more physiological than other methods but they potentially include great biases and
variation (Nocek, 1988).

In Situ Methods to Measure Protein Degradation. The most common means
to estimate protein degradation of feeds is by submersion of feed samples in mesh bags
. in the rumen of cattle or sheep for different durations in time followed by determination

of residual N. The protein source is often divided into at least 3 fractions for

that

CXIEI

ofth

previ

rum 6!

andth

€Stimal

30

description of its disappearance (Mohamed and Smith, 1977; Nocek, 1988). The "A"
fraction is that which is washed from the bags by rinsing before submersion in the
rumen. This protein is considered degraded in the rumen. The "B" fraction is protein
that is slowly and partially degraded in the rumen, and partially digested and absorbed
in the small intestine. The "C" fraction is protein which remains in the nylon bag for
extended periods of time and which is not digested. The first order rate of degradation
of the B fraction is determined from in situ data using the model described for passage
previously.

The portion of the B fraction of dietary protein calculated to be degraded in the
rumen is determined from the in situ disappearance rate divided by the sum of this rate
and the estimated passage rate (Qtskov and McDonald, 1979; Kristensen et al., 1982).
Often the same passage rates are assumed across diets and feed types, while other
estimates are based on different passage rates for different feeds (Ganev et al., 1979).
Passage rates are likely to vary across feed types and within fractions of the same feed,
1 but determination of the passage rates for the feeds on different diets requires passage
markers which are subjected to the same difficulties as described for in vivo studies.

, Additional calculations include allowance for a lag phase before digestion of sample

, (McDonald, 1981; Dhanoa, 1988), but if lag phase is assumed to affect both digestion
and passage equally it can be excluded from the calculation (Allen and Mertens, 1988).

i In situ N disappearance is affected by bag porosity, particle size of sample, and

sample size (Nocek, 1988). This disappearance is also affected by the diet fed to the

‘ host animal (Murphy and Kennelly, 1987; Caton et al., 1988). While these effects
make it impossible to compare feeds for degradation across studies, protein degradation

1 of a feed is not entirely a function of the feed, but is also affected by the ruminal
conditions, so variation across methods is expected and is interesting in itself. More

troublesome aspects of the in situ method do exist. The pH inside the nylon bag has

fn

sitl

CO!

mmen
Control
animal.
Unifonn

fraCLion .

and how I
degMdatic
962mm,“
ref"(Isifltlptt
0"antenna

31

been shown to be lower than that outside the bags, especially when small pore sizes
were used (Marinucci et al., 1992). The microbial population inside the bags also
differed from that outside the bag. Both protozoa and bacterial populations were lower
inside the bags (Meyer and Mackie, 1986; Marinucci et al., 1992). Analysis of digesta
from nylon bags incubated in vitro showed that proteins escaped the nylon bags before
being digested (Spencer et al., 1988). The microbial attachment to feed incubated in
situ is frequently not measured, though several studies have shown high levels of
contamination (Nocek, 1988).

In Vitro Methods to Measure Protein Degradation. Laboratory methods to
determine protein degradation range from in vitro fermentation with markers of
microbial growth to prediction of degradation from protein solubility. In vitro
techniques are less expensive than in vivo and in situ methods, and they offer the
possibility to analyze and compare several feeds at once. In vitro techniques also allow
the researcher to collect and analyze botlt the residue and the metabolites of microbial
degradation, while more physiological techniques are complicated by fluxes out of the
rumen or out of in situ incubation bags. In vitro methods may ultimately allow for
control of the various non-structural factors that alter degradation of a feed (microbial,
animal. environment) and therefore allow for the possibility to characterize feeds
uniformly for their susceptibility to CP degradation. As with other methods, the
fraction of protein degraded in the rumen requires estimation of rates of passage

The in vitro environment for ruminal fermentation studies can vary substantially
and how this variation occurs and its effects must be understood. The rate of in vitro
degradation of protein depends on the amount of protein in an accessible form and on
the amount and types of enzymes and bacteria present. Non-structural aspects of the
feed sample itself (protein and non-protein components) may affect the type and extent

of fermentation and therefore affect the extent of microbial degradation. For example,

A

32

if a feed sample does not contain adequate ferrnentable carbohydrate or is deficient in
an essential mineral, poor microbial growth may limit protein degradation regardless of
protein structure.

High-producing dairy cattle are fed mixed rations to meet the requirements of
both the animal and its microbial population. This high plane of nutrition results in
extensive fermentation and presumably abundant microbial enzymes in vivo. Therefore
the extent of microbial degradation of protein in vivo is likely to be limited by the
physical nature of the protein itself and by its surrounding feed components, rather than
by the ability of the organisms to grow on the material. If an individual feed is only
one component out of several in a feeding regime, an appropriate in vitro system would
supply other nutrients besides the sample to be tested in order to insure profuse
microbial growth relative to the amount of sample. The resulting protein degradation
then would mainly depend on the structural nature of the feed.

Fermentation in continuous culture systems allows removal of metabolites that
can reduce activity of fermentation (Dahlberg et al., 1988; Fuchigarni et al., 1989).
Incubation in batch culture is more convenient, especially if several treatments are to be
compared. These systems require added buffers to maintain adequately high pH, but
harmful metabolites are not removed. If a system contains an abundance of microbial
nutrients, microbial growth may be limited by accumulation of volatile fatty acids
(VFA) or low reducing potential or pH. Therefore the amount of protein used in an in
vitro system for protein degradation must be low enough so that growth is limited by
ability of the microbes to digest the protein, and not by a reduction in pH or
accumulation of VFA's.

The importance of available protein for microbial growth using the system
described above implies that not only is protein degradability (or lack thereof) likely to

limit microbial growth (and therefore degradation), but also certain amino acids may be

 

limit
requi

prote

deten

Micrt
(Raal
differ
level

(Wall
Walla
mark:

dflsire

M6831

marks:
high 1,
Showe
Protei.
1985)_

and di:

with m

33

limiting. Isoacids of branched-chain amino acids can be added to meet some of these
requirements so that limitations on extent of degradation will be due primarilly to
protein structure and its associations with other feed components and not due to its
amino acid profile.

Most in vitro fermentation methods use markers of microbial growth to
determine microbial contamination of feed sample. Since these markers are difficult to
use and the results are unreliable, attempts have been made to circumvent their use.
Microbial growth has been estimated from turbidity of fluid or from gas production
(Raab et al., 1983). These approaches provide crude estimates with potential for bias as
different fermentation patterns result in different levels of gas production at a given
level of microbial protein synthesis. Proteins have been labeled themselves with 14C
(Wallace et al., 1987; Wallace and Falconer, 1992) or azo-dye (Mahadevan et al., 1979;
Wallace and Kopecny, 1983) for determination of degradation from the release of
marker. This approach is useful in some situations but is more labor-intensive than
desired. The degradation of individual proteins has been followed by separation of
these protein from microbial proteins using electrophoresis (Fahmy et al., 1991;
Messman et al., 1992B; Wallace, 1992).

A possible future approach to measuring protein degradation without microbial
markers, but in the presence of high levels of carbohydrate, is to incubate feeds with
high levels of corn starch and cotton but with low levels of feed protein. Previous work
showed that under these conditions starch was optimally degraded with urea as a
protein source, but that peptides were required for degradation of cellulose (Thomsen,
1985). Therefore, disappearance of starch may be a predictor of ammonia availability
and disappearance of cellulose may be related to amino acid availability of feeds.

Another approach to measure protein degradation has been to incubate feeds

with rumen fluid under anaerobic conditions without addition of a carbohydrate source

34

(Russell et al., 1983). Rumen microbial growth was inhibited this way and therefore
was not required to be quantified (Chen and Russell, 1989). A theoretically similar
technique is to incubate feeds in rumen fluid in the presence of inhibitors of N uptake
by microbes (Broderick, 1987; Neutze et al., 1993). The most serious limitation of
these methods is that enzyme activity is not likely to be in excess of substrate. For this
reason, higher level of substrate in the incubation with rumen fluid and inhibitor
resulted in reduced fractional degradation of CP (Broderick and Clayton, 1992), and
level of activity depended on rumen fluid sample (Furchtenicht and Broderick, 1987).
Ruminal proteolytic enzymes have been shown to be loosely associated with bacterial
cell wall, and can be solubilized while retaining activity by gentle extraction (Kopecny
and Wallace, 1982; Brock et al., 1982). A procedure to compare protein degradation of
feeds from crude extracts of rumen enzymes has been published (Mahadevan et al.,
1987). Using this method, relative differences in liberation of amino acids from
common feeds were realistic, but the fractional CP degradation was much lower than
‘ expected. This would suggest that enzyme was more limiting in this system than would
1‘ be in vivo, or some other aspect of the in vitro system (i.e. lack of microbial
attachment) limited degradation.
1 The difficulties associated with fermentation studies, including standardizing
the fermentation, measurement of microbial growth, and the requirement of access to an
‘[ inocula source (fistulated ruminant), have led to the development of several methods to
imeasure protein degradation using enzyme preparations. Ruminal protease was
enriched from Bacteroides amylophilus and used for studies of degradation of several
:protein sources (Mahadevan et al., 1980). Alternatively, methods have used
commercially-available proteases. Crude protein degradation using five different
commercially-available proteases was compared to that of the in situ method for several

concentrate feeds (Poos-Floyd et al., 1985). The solubility of CP was determined by

filter
and l
diffe
data.
feeck
degrz
1983
both

Tetra
prote-
1992]

lheol
exist i
Dossit
exist i

llunch

en 2m

35

filtering after degradation with: Streptomyces griseus protease, papain, bromelain, ficin
and Aspergillus oryzae protease. Though absolute degradation with proteases was
different than observed in situ, all enzymatic degradations were correlated to in situ
data, suggesting the possibility for using enzymes to detect relative differences among
feeds. Use of S. griseus enzyme has become popular for prediction of protein
degradation of feeds but the optimal pH f0r the enzyme is 8.0 (Krishnamoorthy et al.,
1983), which suggests a type of activity not found in the rumen. It has been used at
both normal rumen pH with reduced activity (Chamberlain and Thomas, 1979;
Terramoccia et al., 1992) and at the higher pH (Krishnamoorthy et al., 1983). A neutral
protease was recently introduced to measure protein degradation (Assoumani et al.,
1992).

The use of commercial enzymes has not provided consistent results.
Theoretically, there may be a problem that only one of the many types of proteases that
exist in the rumen is present in a purified enzyme. More serious, however, is the
possibility that the type of protease activity present in the purified enzyme does not
exist in the rumen. A comparison of protein degradation by crude extract from the
rumen and by S. griseus protease showed very different patterns of degradation across
several feeds (Mahadevan et al., 1987). In addition, proteases may not be the only
enzymes required for degradation of plant proteins. Structural limitations from
carbohydrates could prevent degradation by protease enzymes. Crude protein
degradation was lower when incubated with 8. griseus protease compared to in situ
methods, which suggests a synergism of protease with amylase and cellulase
(Terramoccia et al., 1992). This difference was especially apparent for forages (J anicki
and Stallings, 1989). Addition of amylase and beta-glucosidase increased protein

degradation with neutral protease (Assoumani et al., 1992).

 

 

36

Various researchers have attempted to characterize feed protein based upon
solubility properties in aqueous solutions such as saline, buffers, or autoclaved rumen
fluid (Wohlt etal., 1973; Pichard and Van Soest, 1977; Crawford et al., 1978; Crooker
et al., 1978; Mahadevan et al., 1980; Krishnamoorthy et al., 1982; Nocek et al., 1983;
Stern and Satter, 1984). Nitrogen solubility varies greatly for different feedstuffs. For
example, while nitrogen of brewer's grains was only 3% soluble in borate-phosphate
buffer, oat nitrogen was 55.4% soluble (Krishnamoorthy et al., 1982). Buffer soluble
nitrogen usually is comprised mostly of non-protein nitrogen (NPN), such as ammonia,
urea, nitrates, amino acids and small peptides (Pichard and Van Soest, 1977;
Krishnamoorthy et al., 1982). Nucleic acid nitrogen is the major NPN fraction not
soluble in neutral buffer, but generally it is low in quantity, and is underestimated by
nitrogen analysis using the Kjeldahl procedure (Lyttleton, 1973). In addition to NPN,
some true protein is soluble to varying degrees among feeds.

Protein solubility can be influenced by various factors associated with the
solvent or extraction procedure. Subtle changes in chemical composition of solvent can
have pronounced effects on nitrogen solubility. For example, substitution of
ammonium chloride with sodium chloride in Burrough's mineral mixture (BMM)
resulted in increased nitrogen solubility for several concentrates (Crooker et al., 197 8).
It is difficult to draw meaningful conclusions, however, as to the effect of solvent
composition on nitrogen solubilization because substantial interaction among
feedstuffs, methodologies and solvents exists. For example, in one study, nitrogen
solubility was found to be higher in BM or .15 M sodium chloride (NaCl) than in
autoclaved rumen fluid (ARF) (Crawford et al., 197 8), but other authors have reported
the opposite to be true (Waldo and Goering, 1979), and still others found no significant
difference among the three solvents (Lamport, 1980). It is noteworthy that similar

concentrate feeds were analyzed in each of the above studies.

nit

ion:

37

Comparisons of ARF, NaCl, BMM and hot water revealed that variance among
feeds was most pronounced for ARF, suggesting that this method may best separate
different feeds according to solubility of protein (Waldo and Goering, 1979). However,
while homogeneous variance within samples was detected in the above experiment
(Waldo and Goering, 1979), the use of ARF may not be repeatable from laboratory to
laboratory, or from time to time, as ARF composition varies (Jancarik and Proksova,
1970). Solvents often are evaluated according to their similarity to nitrogen solubilized
in ARF (Krishnamoorthy et al., 1982). Ironically, solubility in BMM was found to be
more highly correlated to nitrogen disappearance from nylon bags than was solubility in
ARF (Craward et al., 1978). As with ARF, BMM has a short shelf life and contains
nitrogen unless it is modified (Krishnamoorthy et al., 1982).

Purified proteins are insoluble in isoelectn'c solutions, or those in which the
ionic concentration is such that the protein's positively charged molecules are equaled
by its negatively charged molecules (Lehninger, 1975). Ionic concentration of NaCl
was demonstrated to effect solubility of soybean meal proteins (Smith et al., 1959).
Only 45% of soybean meal nitrogen was soluble at .1 ionic strength, while 98% was
soluble in distilled water, and about 80% was soluble in .5 ionic strength NaCl
(interpreted from graph) (Smith et al., 1959). These results were not confirmed,
however, when seven other concentrates were compared for solubility in BMM, NaCl
and modified BMM, where NaCl was substituted for ammonium chloride, with ionic
strengths of .11, .15 and .19 (Crooker et al., 1978). The narrow number and range of
ionic strengths studied may be responsible for this finding. Ionic strength of rumen
fluid was calculated as .15 (Salobir et al., 1970).

The pH of a solvent has been shown to influence nitrogen solubility of
concentrate feed protein (Smith et al., 1959; Wohlt et al., 1973; Waldo and Goering,

1979). Eighty-five percent of soybean meal protein was soluble in aqueous HCl (pH

 

38

2.0) and water (pH 7.2), while less than 10% was soluble at pH 4.0 (Smith et al., 1959).
Significant differences in solubility of isolated soy protein and casein also were
observed when pH of autoclaved rumen fluid and BM were varied from pH 5.5 to pH
6.5 (W ohlt et al., 197 3). Therefore, the effect of pH on protein solubility is an
important consideration when choosing an appropriate solvent. Ironically, solubility
often is measured in neutral pH solutions, even though rumen fluid is acidic to varying
degrees. Whereas sodium chloride has almost no buffering capacity, and bicarbonate-
phosphate buffer has an unstable pH, a borate phosphate buffer has been suggested as
an appropriate substitute to measure nitrogen solubility without variation due to
fluctuation in pH (Krishnamoorthy et al., 1982).

Solubility of feed nitrogen as a function of time is important in relating data to
the dynamic system of the rumen. Feeds vary in the rate of solubilization, and these
rates have been interpreted as single (Mehrez and Qrskov, 1977) and multi-exponential
curves (Waldo and Goering, 1979). Whereas plant enzymes may be active,
solubilization over extended periods of time may be due to degradation of protein
instead of simple solubility.

Buffer-soluble CP appears to be more readily degraded in the rumen than
insoluble nitrogen. There is a close association between buffer-soluble CP and nitrogen
disappearance from nylon bags suspended in the rumen, especially when feeds are
incubated in situ for short periods of time, such as 1 h (Nocek et al., 1979; Stern and
Satter, 1984). In addition, the amount of buffer soluble nitrogen in a feed was
correlated strongly to increases in rumen ammonia concentration after feeding (Jones
and Mangan, 1977; Waghom, 1991). However, other authors found no consistent
relationship between nitrogen solubility in various aqueous solvents and ammonia
concentration (Crooker et al., 1978). Nitrogen which is available to microbes is

converted to ammonia is used for microbial synthesis of true protein. Soluble NPN is

 

 

 

39

rapidly utilized by rumen bacteria, and solubility in mineral buffer of pure proteins was
correlated strongly to degradation rate in rumen fluid (Hamilton et al., 1992).

Based on these findings, one may suspect that buffer-soluble nitrogen predicts
immediate availability of nitrogen to rumen microorganisms for a given feed, provided
that soluble nitrogen is primarily NPN or is easily—degradable true protein. This
assumption would appear to be true for most of the feeds that have been analyzed.
However, rate of degradation of soluble true protein varies among sources, and a few
feeds have been found to release some soluble protein that is not rapidly degraded
(Nocek et al., 1983; Mahadevan et al., 1980). For example, when soluble and insoluble
feed fractions from dehydrated alfalfa and corn gluten feed were subjected to
Streptomyces griseus protease, amino acid release rates were only half as great for
soluble fractions as for insoluble fractions of the same feeds (Nocek et al., 1983).
Therefore, one may wish to determine the amount of soluble true protein that comprises
the soluble nitrogen fraction for some feeds. This task is easily accomplished by
protein precipitation from solution with such reagents as trichloroacetic acid (T CA)
(Bensadoun and Weinstein, 1976).

The rate of degradation of soluble protein has not been determined for many
feeds. This observation especially is true for forages. The major soluble protein of
plant leaf material, ribulose bisphosphate carboxylase/oxygenase (RUBP-Clo) or
fraction I protein, has been shown to be readily degradable in vitro and in vivo when
derived from fresh forage (Nugent and Mangan, 1981; Hazlewood et al., 1983).
However, it may bind to tannins or other forage components or be altered during forage
conservation causing it to become less readily available in other circumstances.

Some authors have reported a poor correlation between nitrogen solubility in
buffer with in vivo protein degradation across several feeds (Stern and Satter, 1984).

This result is not surprising given that the amount of soluble feed protein appears to

 

 

40

have no relationship to the degradation rate of the remaining insoluble protein fraction
of a given feed (Mahadevan et al., 1980; Nocek et al., 1983). Therefore, one can
conclude that estimation of nitrogen solubility in neutral buffers may predict amount of
nitrogen immediately available to bacteria, but it may not be accurate for all feeds as
discussed above, and nitrogen solubility in aqueous solutions is not sufficient to
determine protein degradation rate or protein degradability.

Crude protein solubility in detergents has also been studied briefly. The amount
of pH-7—detergent insoluble hay protein was negatively correlated (R = -.83) to in situ
degradation, but the correlation for silages was lower (R = -.51). Crude protein
insoluble in pH 7 detergent was shown to be slowly degraded in the rumen (Lindberg,
1988). Heat treatment has been shown to increase the detergent-insoluble crude protein
as well as protect forages from ruminal degradation (Weiss et al., 1989). This fraction
also increased with advancing maturity, and was more concentrated in stems than
leaves of alfalfa (Sanderson and Wedin, 1989). Detergent (pH 7) insoluble protein was
also shown to be higher in pasture samples that had higher concentrations of S. griseus-
protease-resistant protein (Abdalla et al., 1988).

Crude protein that is insoluble in detergent in 1 N H2804 (ADCP) has been
associated with indigestibility of protein and carbohydrate. Severe heat damage renders
protein indigestible by non-enzymatic browning (Maillard) , and the product of this
reaction is insoluble in sulfuric acid-detergent solution (Weiss et al., 1986A; Weiss et
al., 1986B; Van Soest, 1965). The amount of ADCP and pepsin-insoluble CP were
correlated across a large range of heat-damaged alfalfa, but with an R2 of only .66
(Goering, 1972), though it was well correlated with apparent in vivo indigestibility of
forages (Goering et al., 1972). Some ADCP is digestible so it is not a uniform fraction
in forages or concentrates (Britton et al., 1986; Weiss et al., 1986; Waters et al., 1992;

Broderick et al., 1993). Theoretically, Maillard product from short peptides and

 

 

41

saccharides could also be filterable yet indigestible, and so added error may be

associated with use of ADCP for estimation of the indigestibility from heat damage.

42

LITERATURE CITED

Abdalla, H. 0., D. G. Fox, and R. R. Seaney. 1988. Variation in protein and fiber
fractions in pasture during the grazing season. J. Anim. Sci. 66: 2663.

Adamu, A. M., J. R. Russell, A. D. McGilliard, and A. Trenkle. 1989. Effects of added
dietary urea on the utilization of maize stover silage by growing beef cattle. Anim. Feed
Sci. Technol. 22: 227.

Agricultural Research Council. 1984. Report of the Protein Group of the Agricultural
Research Council Working Party on the Nutrient Requirements of Ruminants.
Commonw. Agric. Bur., Slough, Engl.

Aguilera, J. F., M. Bustos, and E. Molina. 1992. The degradability of legume seed
meals in the rumen - effect of heat treatment. Anim. Feed Sci. Technol. 36: 101.

Aharoni, Y. , H. Tagari, and R. C. Boston. 1991. A new approach to the quantitative
estimation of nitrogen metabolic pathways in the rumen. Br. J. Nutr. 66: 407.

Akin, D. E. 1979. Microscopic evaluation of forage digestion by rumen
microorganisms - a review. J. Anim. Sci. 48: 701.

Albrecht, K. A. and R. E. Muck. 1991. Proteolysis in ensiled forage legumes of varying
tannin concentration. Page 76 in US Dairy Forage Research Center, 1991 Research
Summaries, Madison, WI.

Allen, M. S. and D. R. Mertens. 1988. Evaluating constraints on fiber digestion by
rumen microbes. J. Nutr. 118: 261.

Argos, P., K. Pedersen, M. D. Marks, and B. A. Larkins. 1982. A structural model for
maize zein proteins. J. Biol. Chem. 257: 9984.

Askerlund, P, C. Larsson, and S. W. Widell. 1989. Cytochromes of plant plasma
membranes. Characterization by absorbance difference spectrophotometry and redox
titration. Phyiologia Plantarum 76: 123.

Association of Official Agricultural Chemists. 1975. Offical Methods of Analysis. 12th
edition. A. O. A. C., Washington, D. C.

- Assoumani, M. B., F. Vedeau, L. Jacquot, and C. J. Sniffen. 1992. Refinement of an

enzymatic method for estimating the theoretical degradability of proteins in feedstuffs

g for ruminants. Anim. Feed Sci. Technol. 39: 357.

.. )7

I‘T‘IFH

43

Barrett, A. J. 1985. Chapter 1: The classes of proteolytic enzymes. Plant Proteolytic
Enzymes. 1' 1

Barry, T. N. and T. W. Manley. 1984. The role of condensed tannins in the nutritional
value of Lotus pendunculatus for sheep. Br. J. Nutr. 51: 493.

Barry, T. N., A. Thompson, and D. G. Armstrong. 1977. Rumen fermentation studies
on two contrasting diets. 1. Some characteristics of the in vivo fermentation, with
special reference to the composition of the gas phase, oxidation/reduction state and
volatile fatty acid proportions. J. Agric. Sci., Camb. 89: 183.

Beever, D. E., S. B. Cammell, and A. Wallace. 1974. The digestion of fresh, frozen and
dried perennial ryegrass. Proc. Nutr. Soc. 33: 73A.

Bensadoun, A. and Weinstein. 1976. Assay of proteins in the presence of interfering
materials. Anal. Biochem. 70: 241

Britton, R. A., T. J. Klopfenstein, R. Cleale, F. Goedekin, and V. Wilkerson. 1986.
Methods of estimating heat damage in protein sources. Proc. Dist. Feed Conf. 41: 67.

Brock, F. M., C. W. Forsberg, and J. G. Buchanan-Smith. 1982. Proteolytic activity of
rumen microorganisms and effects of proteinase inhibitors. Appl. Environ. Microbiol.
44: 561.

Broderick, G. A. 1987. Determination of protein degradation rates using a rumen in
vitro system containing inhibitors of microbial nitrogen metabolism. Br. J. Nutr. 58:
463.

Broderick, G. A. and D. R. Buxton. 1991. Genetic variation in alfalfa for ruminal
protein degradability. Can. J. Plant Sci. 71: 755.

Broderick, G. A. and M. K. Clayton. 1992. Rumen protein degradation rates estimated
by non- linear regression analysis of Michaelis-Menten in vitro data. Br. J. N utr. 67:
‘ 27.

5 Broderick, G. A., J. H. Yang, and R. G. Koegel. 1993. Effect of steam heating alfalfa
hay on utilization by lactating dairy cows. J. Dairy Sci. 76: 165.

Buckhoz, H. F. andW. G. Bergen. 1973. Microbial phospholipid synthesisasamarker
f for microbialprotein synthesisin the rumen. Appl. Microbial. 25: 504.

‘ Caton, J. 8., A. S. Freeman, and M. L. Galyean. 1988. Influence of protein

. supplementation on forage intake,‘ in situ forage disappearance, ruminal fermentation

‘: 22d digesta passage rates in steers grazing dormant blue grama rangeland. J. Anim. Sci.
1 226

‘ Cecava, M. J., N. R. Merchen, L. L. Berger, and G. C. Fahey, Jr. 1988. Effects of
dietary energy level and protein source on site of digestion and duodenal nitrogen and
. amino acid flows in steers. J. Anim. Sci. 66: 961

 

44

Chamberlain, D. G. and P. C. Thomas. 1979. Prospective laboratory methods for
estimating the susceptibility of feed proteins to microbial breakdown in the rumen.
Proc. Nutr. Soc. 38: 138A.

Chen, G. and J. B. Russell. 1989. More monensin-sensitive, ammonia-producing
bacteria from the rumen. Appl. Environ. Microbiol. 55: 1052.

Chen, G., C. J. Sniffen, and J. B. Russell. 1987. Concentration and estimated flow of
peptides from the rumen of dairy cattle: Effects of protein quantity, protein solubility,
and feeding frequency. J. Dairy Sci. 70: 983.

Ciszuk P. and J. E. Lindberg. 1988. Responses in feed intake, digestibility and nitrogen
retention in lactating dairy goats fed increasing amounts of urea and fish meal. Acta
Agric. Scand. 38: 381.

Coenen, D. J. and A. Trenkle. 1989. Comparisons of expeller processed-and solvent-
extracted soybean meals as protein supplements for cattle. J. Anim. Sci. 67: 565.

Coto, G., R. S. Herrera, M. Perez, J. Alert, and S. Poppe. 1989. A study on the soluble
and insoluble nitrogen fractions in star grass fertilized with high nitrogen doses. Cuban,
J. Agric. Sci. 23: 115.

Crawford, R. J. Jr., W. H. Hoover, C. J. Sniffen, and B. A. Crooker. 1978. Degradation
of feedstuff nitrogen in the rumen vs. nitrogen solubility in three solvents. J. Anim. Sci.
46: 1768.

Crooker, B. A., C. J. Sniffen, W. H. Hoover, and L. L. Johnson. 1978. Solvents for
soluble nitrogen measurements in feedstuffs. J. Dairy Sci. 4: 437.

Cruickshank, G. J., D. P. Poppi, and A. R. Sykes. 1989. Theoretical considerations in
the estimation of rumen fractional outflow rate from various sampling sites in the
digestive tract. Br. J. Nutr. 62: 229.

Czerkawski, J. W. and C. Foulds. 1974. Methods for determining 2-6-diaminopimelic
acid and 2-aminoethylphosphoric acid in gut contents. J. Sci. Food Agric. 25: 45.

Dahlberg, E. M., M. D, Stern, and F. R. Ehle. 1988. Effects of forage source on ruminal
microbial nitrogen metabolism and carbohydrate digestion in continuous culture. J.
Anim. Sci. 66: 2071.

Dhanoa, M. S. 1988. Research note on the analysis of dacron bag data for low
degradability feeds. Gr. Forage Sci. 43: 441.

Duvick, D. N. 1961. Protein granules of maize endosperm cells. Cereal Chem. 38: 374.

Esen, A. 1980. Estimation of protein quality and quantity in corn (Zea mays L.) by
assaying protein in two solubility fractions. J. Agric. Food Chem. 28: 529.

Fahmy, W. G., J. Aufrere, D. Graviou, C. Demarquilly, and K. El-Shazly. 1991.
Comparison between the mechanism of protein degradation of 2 cereals by enzymatic
and in situ methods, using gel electrophoresis. Anim. Feed Sci. Technol. 35: 115.

:l'Tl warn

KEITH

 

45

Faichney, G. J. 1975. The use of markers to partition digestion within the gastro-
intestinal tract of ruminants. Pages 277-291 in Digestion and Metabolism in the
Ruminant. ed. 1. W. McDonald and A.C.I Warner, University of New England,
Arrnidale.

Faichney, G. J. 1980. The use of markers to measure digesta flow from the stomach of
sheep fed once daily. J. Agric. Sci., Camb. 94: 313.

Faichney, G. J. 1992. Application of the double-marker method for measuring digesta
kinetics to rumen sampling in sheep following a dose of the markers or the end of their
continuous infusion. Aust. J. Agric. Res. 43: 277.

Ferguson, J. D. and W. Chalupa. 1989. Symposium: interactions of nutrition and
reproduction. Impact of protein nutrition on reproduction in dairy cows. J. Dairy Sci.
: 746.

Fluharty, F. L. and S. C. Loerch. 1989. Chemical treatment of ground corn to limit
ruminal digestion. Can. J. Anim. Sci. 69: 173.

Folman, Y., H. Neumark, M. Kaim, and W. Kaufmann. 1981. Performance, rumen and
blood metabolites in high-yielding cows fed varying protein percents and protected
soybean. J. Dairy Sci. 64: 759.

Fuchigami, M., T. Senshu, and M. Horiguchi. 1989. A simple continuous culture
system for rumen microbial digestion study and effects of defaunation and dilution
rates. J. Dairy Sci. 72: 3070.

Fujimaki, T., Y. Kobayashi, M. Wakita, and S. Hoshino. 1992. Amino acid
supplements -— a least combination that increases microbial yields of washed cell
suspension from goat rumen. J. Anim. Physiol. Anim. Nutr. -- Zeitschrift Fur
Tierphysiologie Tieremahrung und Futtermittelkunde 67: 41. ‘

Furchtenicht, J. E. and G. A. Broderick. 1987. Effect of inoculum preparation and
gietary energy on microbial numbers and rumen protein degradation activity. J. Dairy
ci. 70: 1404.

Furchtenicht, J. E. and G. A. Broderick. 1987. Effect of inoculum preparation and
gietary energy on microbial numbers and rumen protein degradation activity. J. Dairy
ci. 70: 1404.

Ganev, G., E. R. Orskov, and R. Smart. 1979. The effect of roughage or concentrate
feeding and rumen retention time on total degradation of protein in the rumen. J. Agric.
801., Camb. 93: 651.

Gasa, J., K. Holtenius, J. D. Sutton, M. S. Dhanoa, and D. J. Napper. 1991. Rumen fill
and digesta kinetics in lactating friesian cows given 2 levels of concentrates with 2
types of grass silage ad—lib. Br. J. Nutr. 66: 381.

Gianazza, E., V. Viglienghi, P. G. Righetti, F. Salamini, and C. Soave. 1977. Amino
acrd composition of zein molecular components. Phytochem. 16: 315.

46

Goering, H. K. 1972. A laboratory assessment on the frequency of over-heating in
commercial dehydrated alfalfa samples. J. Anim. Sci. 43: 869.

Goering, H. K., C. H. Gordon, R. W. Hemken, D. R. Waldo, P. J. Van Soest, and L. W.
Smith. 1972. Analytical estimates of nitrogen digestibility in heat damaged forages. J.
Dairy Sci. 55: 1275.

Gottschalk, G. 1986. Bacterial Metabolism. 2nd ed. Springer-Verlag, New York.

Grabber, J. H. and G. A. Jung. 1991. In viu'o disappearance of carbohydrates, phenolic
acids, and lignin from parenchyma and sclerenchyma cell walls isolated from
cocksfoot. J. Sci. Fd. Agric. 57: 315.

Griffin, T. 8., 0. B. Hesterrnan, and S. R. Rust. 1991. Alfalfa maturity and cultivar
effects on chemical and in situ estimates of protein degradability. Page 23 in American
Forage and Grassland Council, Interpretive Summaries. Amway Grand Plaza Hotel,
Grand Rapids, MI.

Gunn, K. J. and N. P. McMeniman. 1989. Incorporation of 358 into rumen microbial
protein. Proc Nutr Soc (Aust) 14: 141.

Hach, C. H., S. V. Brayton, and A. B. Kopelove. 1985. A powerful kjeldahl nitrogen
method using peroxymonosulfuric acid. J. Agric. Food Chem. 33: 1117.

Hagerrnan, A. E., C. T. Robbins, Y. Weerasuriya, T. C. Wilson, and C. Mcarthur.
1992. Tannin chemistry in relation to digestion. J. Range Manag. 45: 57.

Halverson, A. W., G. D. Williams, and G. D. Paulson. 1968. Aspects of sulfate
utilization by the microorganisms of the ovine rumen. J. Nutr. 95: 363.

Hamilton, B. A., J. R. Ashes, and A. W. Carmichael. 1992. Effect of formaldehyde-
treatedasgnflower meal on the milk production of grazing dairy cows. Austr. J. Agric.
Res. 4 : 79.

Hanley, T. A., C. T. Robbins, A. E. Hagerman, and C. Mcarthur. 1992. Predicting
digestible protein and digestible dry matter in tannin-containing forages consumed by
ruminants. Ecology 73:537.

Hartley, R. D. and D. E. Akin. 1989. Effect of forage cell wall phenolic acids and
derivatives on rumen micorflora. J. Sci. Food Agric. 49: 405.

Harvey, B. M. R. and A. Oaks. 1974. The hydrolysis of endosperm protein in Zea mays.
Plant Physiol. 53: 453.

Hazlewood, G. P.,C. G. Orpin, Y. Greenwood, M. E. Black. 1983. Isolation of
proteolytic rumen bacteria by use of selective medium containing leaf fraction 1 protein
(rubulosebisphosphate carboxylase). Appl. Environ. Microbiol. 45: 17 80.

Henderickx, H. K., D. I. Demeyer, C. J. Van Nevel. 1972. Problems in estimating
microbial protein synthesis in the rumen. Pages 57-68 in Tracer Studies on Non-Protein
Nitrogen for Ruminants. Inter. Atomic Energy Agency. Vienna.

47

Hoffman, P. C., S. J. Sievert, R. D. Shaver, and D. K. Combs. 1992. In situ CP
degradation kinetics of perennial forages. J. Dairy Sci. 75(Suppl. 1):210.

Hunter, R. A. and B. D. Siebert. 1987. The effect of supplements of rumen-degradable
protein and formaldehyde-treated casein on the intake of low—nitrogen roughages by
Bos taurus and Bos indicus steers at diferent stages of maturity. Aust. J. Agric. Res, 38:
209.

Jancarik, A. and M. Proksova. 1970. The breakdown of protein in the rumen in relation
to the physical and chemical characters of the rumen juice. Pages 304-306 in Nutr.
Proc. 8th Intern. Congr. Prague. Excerpta Medica, Amsterdam.

Janicki, F. J., and C. C. Stallings. 1988. Degradation of crude protein in forages
determined by 1n vitro and' in situ procedures. J. Dairy Sci. 71: 2440

Javorsky, P., I. Havassy, E. Rybosova, M. Kralova, K. Horsky, and K. Kosta. 1986.
Synthesis of bacterial nitrogen substances in the sheep rumen from nitrogen-15 labelled
urea. J. Anim. Physiol. Anim. Nutr. 5: 281.

Johnson, D. E., G. M. Ward, and J. Torrent. 1992. The environmental impact of bovine
somatotropin use in dairy cattle. J. Environ. Quality 21: 157.

Jones, W. T. and J. L. Mangan. 1977. Complexes of the condensed tannins of sainfoin
(Onobrychis viciifolia Scop) with fraction 1 leaf protein and with sub-maxillary
mucoprotein, and their reversal by polyethylene glycol and pH. J. Sci. Food Agric. 28:
126.

Kephart, K. D. and A. Boe. 1991.Rumen degradable protien of Rumen degradable
protein of cool- and warm—season forage grasses. Page 24 in Amer. Forage and
Grassland Council, Interpretive Summaries. Amway Grand Plaza Hotel, Grand Rapids,
MI.

3 Kopecny, J. and R. J. Wallace. 1982. Cellular location and some properties of
proteolytic enzymes of rumen bacteria. Appl. Environ. Microbiol. 43: 1026.

Krishnamoorthy, U, T. V. Muscato, C. J. Sniffené andP. J. Van Soest 1982. Nitrogen
fractions 1n selected feedstuffs. J. Dairy Sci. 65:21.7

Krishnamoorthy, U., C. J. Sniffen, M. D. Stern, and P. J. Van Soest. 1983. Evaluation
of a mathematical model of rumen digestion and an in vitro simulation of rumen
proteolysis to estimate the rumen-undegraded nitrogen content of feedstuffs. Br. J.
Nutr. 50: 555.

Kristensen, E. S., P. D. Moller, and T. Hvelplund. 1982. Estimation of the effective
protein degradability 1n the rumen of cows using the nylon bag technique combined
with the outflow rate. Acta Agric. Scand. 32: 1.23

Kung, L. Jr. and J. T. Huber. 1983. Performance of high producing cows in early
12actation fed protein of varying amounts, sources, and degradability. J. Dairy Sci. 66:
27

48

Larnport, D.T.A. 1980. Structure and function of plant glycoproteins. Pages 501—541 in
The Biochemistry of Plants, vol. 3, Academic Press Inc., New York.

Lee, K. B., D. Loganathan, Z. M. Merchant, and R. J. Linhardt. 1990. Carbohydrate
analysis of glycoproteins: a review. Appl. Biochem. Biotech. 23: 53.

Lehninger, A. L. 1975. Biochemistry. The Molecular Basis of Cell Structure and
Function. 2nd edition. Worth Publishers, Inc. New York.

Lindberg, J. E. 1985. Estimation of rumen degradability of feed proteins with the in
sacco technique and various in vitro methods. A review. Acta Agric. Scand. Suppl. 25:
64.

Lindberg, J. E. 1988. Influence of cutting time and N fertilization of the nutritive value
of timothy: 2. Estimates of rumen degradability of nitrogenous compounds. Swedish J.
Agric. Res. 18: 85.

Lindberg, J. E., H. Bristav, and A. R. Manyenga. 1989. Excretion of purines in the
urine of sheep in relation to duodenal flow of microbial protein. Swedish J. Agric. 19:
45.

Lindberg, J. E. and K. G. J acobsson. 1990. Nitrogen and purine metabolism at varying
energy and protein supplies in sheep sustained on intragastric infusion. Br. J. Nutr. 64:
359.

Lindberg, J. E. and I. Olsson. 1983. Live weight gain in intensively reared bulls fed
rations with low and high degradable protein. Pages 243-246 in IVth Int Symp. Protein
metabolism and nutrition, Clermont—Ferrand (France) Sept. 5-9. NRA Publ. 1983. II
(les colloques de INRA no 16), Clermont, France.

1 Ling, J. R. and Buttery, P. J. 1978. The simultaneous use of ribonucleic acid, 358, 2,6,-
‘ diaminopimelic acid and 2-aminoethylphosphonic acid as markers of microbial
nitrogen entering the duoddenum of sheep. Br. J. Nutr. 39: 165.

Lowry, O. H., N. O. Rosenbrough, A. L. Fan, and R. J. Randall. 1951. Protein
measurement with the Folin-phenol reagent. J. Biol. Chem. 193: 265.

Lyttleton, J. W. 1973. Proteins and nucleic acids. Pages 63-103 in Chemistry and
Biochemistry of Herbage. ed. G. W. Butter and R. W. Baily. Academic Press, New
York, NY.

Madsen, J. 1985. Protein evaluation for ruminants. Acta Agric. Scand. Suppl. 25: 9.

Madsen, J. and T. Hvelplund. 1988. The influence of different protein supply and
feeding level on pH, ammonia concentration and microbial protein synthesis in the
‘ rumen of cows. Acta Agric. Scand. 38: 115.

Mahadevan, 8., J. D. Erfle, and F. D. Sauer. 1979. A colorimetric method for the
determination of proteolytic degradation of feed proteins by rumen microorganisms. J.
Anim. Sci. 48: 947.

 

49

Mahadevan, 8., J. D. Erfle, and F. D. Sauer. 1980. Degradation of soluble and insoluble
proteins by Bacteroides amylophilus protease and by rumen microorganisms. J. Anim.
Sci. 50: 723.

Mahadevan, 8., F. D. Sauer, and J. D. Erfle. 1987. Preparation of protease from mixed
rumen microorganism and its use for the in vitro determination of the degradability of
true protein in feedstuffs. Can. J. Anim. Sci. 67: 55.

Mangan, J. L. 1982. 2.1 The nitrogenous constituents of fresh forages. Pages 25-39 in
Forage Protein in Ruminant Anim. Prod., ed. Thomson, D. J. et al., Occasional
Publication No. 6, Br. Soc. Anim. Prod., London.

Manson, F. J. and J. D. Leaver. 1988. The influence of dietary protein intake and of
hoof trimming on lameness in dairy cattle. Anim. Prod. 47: 191.

Marder, J. B. and J. Barber. 1989. The molecular anatomy and function of thylakoid
proteins. Plant, Cell and Environment 12: 595.

Marinucci, M. T., B. A. Dehority, and S. C. Loerch. 1992. In vitro and in vivo studies
of factors affecting digestion of feeds in synthetic fiber bags. J. Anim. Sci. 70: 296.

Mathers, J. C. and E. L Miller. 1980 A simple procedure using 358 incorporation for
the measurement of microbial and undegraded food plotein in ruminant digesta. Br. J.
Nutr. 43. 503.

Mathison, G. W. and L. P. Milligan. 1971. Nitrogen metabolism in sheep. Br. J. Nutr.
25: 351.

Matin, A. 1992. Physiology, molecular biology and applications of the bacterial
? starvation response. Journal of Applied Bacteriology 73: S49.

1 Matras, J. and R. L. Preston. 1989. The role of glucose infusion on the metabolism of
nitrogen in ruminants. J. Anim. Sci. 67: 1642.

McAllan, A. B. and E. S. Griffith. 1987. The effects of different sources of nitrogen
supplementation on the digestion of fibre components in the rumen of steers. Anim.
Feed Sci. Technol. 17:65.

McAllan, A. B. and R. H. Smith. 1983A. Estimation of flows of organic matter and
nitrogen components in postruminal digesta and effects of level of dietary intake and
physical form of protein supplement on such estimates. Br. J. Nutr. 49: 119.

McAllan, A. B. and R. H. Smith. 1983B. Influence of different dietary nitrogen sources
on fibre digestion in the rumen. Pages 255-257 in IVth Int. Symp. Protein Metabolism
and Nutrition, Clermont-Ferrand (France) Sept. 5-9. INRA Publ. 1983. II (les colloques
de INRA no 16)

McDonald, 1. 1981. A revised model for the estimation of protein degradability in the
rumen. J. Agric. Sci., Camb. 96: 251.

50

McDonald, P. 1982. The effect of conservation processes on the nitrogenous
components of forages. Pages 41-49 in Forage Protein in Ruminant Animal Prod.
Occasional Publication No. 6 Br. Soc. of Anim. Prod.

McKersie, B. D. 1981. Proteinases and peptidases of alfalfa herbage. Can. J. Plant Sci.
61: 53.
McKersie, B. D. 1985. Effect of pH on proteolysis in ensiled legume forage. Agron. J.
77: 80.

Mehrez, A. Z. and E. R. Qrskov. 1977. A study of the artificial fibre bag technique for
determining the digestibility of feeds 1n the rumen. J. Agric. Sci., Camb. 88: 64 5.

Meissner, H. H. and U. Todtenhofer. 1989. Influence of casein and glucose or starch
supplementation in the rumen or abomasum on utilization of Eragrostis curvula hay by
sheep. S. Afr. J. Anim. Sci. 19: 43.

Mercer, J. R., S. A. Allen, and E. L. Miller. 1980. Rumen bacterial protein synthesis
and the proportion of dietary protein escaping degradation in the rumen of sheep. Br. J.
Nutr. 43: 421.

Messman, M. A. and W. P. Weiss. 1992. Extraction and colorimetric determination of
forage protein. J. Dairy Sci. 75(Suppl. 1):l75.

Messman, M. A., W. P. Weiss, and D. O. Erickson. 1992A. Effects of nitrogen
fertilization and maturity of bromegrass on nitrogen and amino acid utilization by cows.
J. Anim. Sci. 70: 566.

Messman, M A., W. P. Weiss, and M. E. Koch. 1992B. Electrophoresis as a tool to
ascertain effects of preservation method on forage protein composition. Pages 35 to 39
\in Ohio Agric. Experiment Station Research Highlights, Wooster, OH.

Meyer, J. H. F. andR. I. Mackie. 1986. Microbiologicalevaluation of the 1ntrarum1nal
in sacculus d1gest1ontechn1que Appl. Environ. Mic1obiol. 51. 622

Meyer, J. H. F. S. I. van der Walt, and H. M. Schwartz. 1986. The influence of diet and
protozoa] numbers on the breakdown and synthesis of protein in the rumen of sheep. J.
Anim. Sci. 62: 509.

Michalet-Doreau, B. and M. Y. Ouldbah. 1992. In vitro and in sacco methods for the
estimation of dietary nitrogen degradability in the rumen - a review. Anim. Feed Sci.
Technol. 40: 57.

Mohamed, O. E. and R. H. Smith. 1977. Measurement of protein degradation in the
rumen. Proc. Nutr. Soc. Soc. 36:152A.

Mosimanyana, B. M. andD. N. Mowat. 1992. Rumen protection of heat- treated
soybean proteins. Can. J. Anim. Sci. 72:

Mullahey, J. J, S. S. Waller, K. J. Moore, L. E. Moser, andT. J. Klopfenstein.1992.1n
iiéu ruminal protein degradation of switchgrass and smooth bromegrass. Agron. J. 84:
3.

51

Murphy, J. J. and J. J. Kennelly. 1987. Effect of protein concentration and protein
source on the degradability of dry matter and protein in situ. J. Dairy Sci. 70: 1841.

Nader, C. J. and D. J. Walker. 1970. Metabolic fate of cysteine and methionine in
rumen digesta. Appl. Microbiol. 20: 677.

National Research Council. 1985. Ruminant Nitrogen Usage. Subcommittee on
Nitrogen Usage in Ruminants Committee on Animal Nutrition Board on Agriculture.
National Academy Press. Washington, D. C.

National Research Council. 1989. Nutrient Requirements of Dairy Cattle. 6th rev. ed.
Natl. Acad. Sci., Washington, DC.

Neutze, S. A., R. L. Smith, and W. A. Forbes. 1993. Application of an inhibitor in vitro
method for estimating rumen degradation of feed protein. Anim. Feed Sci. Technol.
40: 251.

Newbold, J. R., P. C. Gamsworthy, P. J. Buttery, D. J. A. Cole, and W. Haresign.
1987. Protein nutrition of growing cattle: Food intake and growth responses to rumen
degradable protein and undegradable protein. Anim. Prod. 45: 383.

Nia, S. A. M. and J. R. Ingalls. 1992. Effect of heating on canola meal protein
degradation in the rumen and digestion in the lower gastrointestinal tract of steers. Can.
J. Anim. Sci. Sci.72: 83.

Nocek, J. E., K. A. Cummins, and C. E. Polan. 1979. Ruminal disappearance of crude
groteisn7 and dry matter in feeds and combined effects 1n formulated rations. J. Dairy Sci.
2

Nocek, J. E., J. H. Herbein, and C. E. Polan. 1983. Total amino acid release rates of
soluble and insoluble protein fractions and concentrate feedstuffs by Streptomyces
griseus. J. Dairy Sci. 66: 1663.

Nolan, J. V., W. B. Norton, and R. A. Leng. 1972. Dynamic aspects of nitrogen
metabolism in sheep. Pages 13-24 in Tracer Studies on Non-Protein Nitrogen for
Ruminants. Inter. Atomic Energy Agency. Vienna.

Nugent, J. H., andJ. L. Mangan. 1981. Characteristics of the rumen proteolysis of
fraction 1 (188) leaf protein from luceme (Medicago sativa L). Br. J. Nutr. 46: 39.

L Ohshirna, M. and P. McDonald. 1978. A review of the changes in nitrogenous
. compounds of herbage during ensilage. J. Food Agric. 29:497.

Oke, B. O. and S. C. Loerch. 1992. Effects of energy and amino acid supply to the
small intestine on amino acid metabolism. J Nutr. Biochem. 3. 63.

@rskov, E. R. M. Hughes- J-ones, and I McDonald. 1980. Degradability of protein
supplements and utilization of undegraded protein by high— producing dairy cows. Pages
{5- -98 1n Recent Advances 1n Animal Nutrition. ed. W. Haresign, Butterworths,

ondon.

52

firskov, E. R., and I. McDonald. 1979. The estimation of protein degradability in the
rumen from incubation measurements weighted according to rate of passage. J. Agric.
Sci., Camb. 92: 499.

Ourry, A., B. GonzaleLand J. Boucaud. 1989. Osmoregulation and role of nitrate
during regrowth after cutting of ryegrass (Lolium perenne). Physiologia Plantarum 76:
177.

Papadopoulos, M. C. 1989. Effect of processing on high-protein feedstuffs: a review.
Biol. Wastes 29: 123.

Paulis, J. W. 1981. Disulfide structures of zein proteins from corn endosperm. Cereal
Chem. 58: 542.

Petit, H. V. and G. F. Tremblay. 1992. In situ degradability of fresh grass and grass
conserved under different harvesting methods. J. Dairy Sci. 75: 774.

Pichard, G. and P. J. Van Soest. 1977. Protein solubility of ruminant feeds. Page 91 in
Cornell Nutr. Conf. Feed Manufacturers, Syracuse, NY.

Polan, C. E. 1988. Update: dietary protein and microbial protein contribution. J. of
Nutr. 118: 242.

Poos-Floyd, M., T. Klopfenstein, and R. A. Britton. 1985. Evaluation of laboratory
techniques for predicting ruminal protein degradation. J. Dairy Sci. 68: 829.

Puchala. R. and G. W. Kulasek. 1992. Estimation of microbial protein flow from the
rumen of sheep using microbial nucleic acid and urinary excretion of purine derivatives.
Canadian J. Anim. Sci. 72:821.

Raab, L, B. Cafantaris, T. Jilg, and K. H. Menke. 1983. Rumen protein degradation and
biosynthesis l. a new nethod for determination of protein degradation in rumen fluid in
vitro. Br. J. Nutr. 50: 569.

Robinson, P. H., and J. J. Kennelly. 1988. Influence of intake of rumen undegradable
protein on milk production of late lactation holstein cows. J. Dairy Sci. 71: 2135.

Rooke, J. A. and D. G. Armstrong. 1989. The importance of the form of nitrogen on
microbial protein synthesis in the rumen of cattle receiving grass silage and continuous
intrarumen infusions of sucrose. Br. J. Nutr. 61: 113.

Rooke, J. A., H. A. Greife, and D. G. Armstrong. 1984. The effect of in sacco rumen
incubation of a grass silage upon the total and D-amino acid composition of the residual
silage dry matter. J. Agric. Sci., Camb. 102: 695.

Russell, J. B., C. J. Sniffen, and P. J. Van Soest. 1983. Effect of carbohydrate limitation
on degradation and utilization of casein by mixed rumen bacteria. J. Dairy Sci. 66: 763.

Ruxton, I. B. and P. McDonald. 1974. The influence of oxygen on ensilage. 1.
Laboratory studies. J. Sci. Food Agric. 12: 706.

 

53

Salobir, K., A. Pen, and O. Muck. 1970. The importance of solubility of proteins in
feed for the evaluation of their propriety for ruminants. Page 329 in Proc. 1st Yugoslav
Intern. Conf. Anim. Prod.

Sanderson, M. A. and W. F. Wedin. 1989A. Nitrogen concentrations in cell wall and
lignocellulose of smooth brome grass. Grass Forage Sci. 44: 151.

Sanderson, M. A. and W. F. Wedin. 1989B. Nitrogen in the detergent fibre fractions of
temperate legumes and grasses. Grass Forage Sci. 44: 159.

Santos, K. A., M. D. Stern, and L. D. Satter. 1984. Protein degradation in the rumen
and amino acid absorption in the small intestine of lactating dairy cattle fed various
protein sources. J. Anim. Sci. 58: 244.

Sathe, S. K., A. C. Mason, C. M. Weaver.
1989. Thermal aggregation of soybean (glycine max L.) sulfur-rich protein. J. Food
Sci. 54: 319.

Schmitter, B. M. and T. Rihs. 1989. Evaluation of a macrocombustion method for total
nitrogen determination in feedstuffs. J. Agric. Food Chem. 37: 992.

Schwab, C. G., c. K. Bozak, N. L. Whitehouse, and M.M.A. Mesbah. 1992. Amino acid
limitation and flow to the duodenum at four stages of lactation. 1. Sequence of lysine
and methionine limitation. J. Dairy Sci. 75: 3486.

Segel, I. H. 1976. Biochemical Calculations. 2nd ed., John Wiley and Sons, Inc., New
York.

Siddons, R. C., D. E. Beever, J. V. Nolan, A. B. McAllan, and J. C. Macrae. 1979.
:Estimation of microbial protein in duadenal digesta. Ann. Rech. Vet. 10: 286.

Sklan, D. 1989. In vitro and in vivo rumen protection of proteins coated with calcium
soaps of long-chain fatty acids. J. Agric. Sci., Camb. 112: 79.

Smith, C. R. J r., F. R. Earle, I. A. Wolff, and Q. Jones. 1959. Comparisons of solubility
characteristics of selected seed proteins. J. Agric. Food Chem. 7: 133.

Smith, R. H., A. B. McAllan, D. Hewitt, and P. E. Lewis. 1978. Estimation of amounts
of microbial and dietary nitrogen compounds entering the duodenum of cattle. J. Agric.
Sci., Camb. 90: 557.

Spencer, D., T. J. V. Higgins, M. Freer, H. Dove, and J. B. Coombe. 1988. Monitoring
the fate of dietary proteins in rumen fluid using gel electrophoresis. Br. J. Nutr. 60: 241.

Staples, C. R. and D. S. Lough. 1989. Efficacy of supplemental dietary neutralizing
agents for lactating dairy cows. A review. Anim. Feed Sci. Technol. 23: 277.

Stern, M. D. and L. D. Satter. 1984. Evaluation of nitrogen solubility and the dacron
bag technique as methods for estimating protein degradation in the rumen. J. Anim. Sci.
58: 7 14.

54

Teeter, R. G. and F. N. Owens. 1983. Characteristics of water soluble markers for
measuring rumen liquid volume and dilution rate. J. Anim. Sci. 56: 717

Terramoccia, S., S. Puppo, L. Rizzi, and F. Martillotti. 1992. Comparison between in-
sacco and in vitro protein rumen degradability. Annales de Zootechnie 41: 20.

Thomsen, K. V. 1985. The specific nitrogen requirements of rumen microorganisms.
Acta Agric. Scand. Suppl. 25: 125.

Tyrrell, HE, P. W. Moe, and W. P. Flatt. 1970. Influence of excess protein intake on
energy metabolism of the dairy cow. Pages 68-71 in Publ. 16, European Assoc. Anim.
Prod., London.

Van Bruchem, J., A. K. Kies, R. Bremmers, M. W. Boxch, H. Boer, and P. W. M. Van
Adrichem. 1988. Digestion of alfalfa and grass silages in sheep. 2. Digestion of protein
in reticulorumen and intestine. Neth. J. Agric. Sci. 36: 365.

Van Ramshorst, H. and P. C. Thomas. 1988. Digestion in sheep of diets containing
barley chemically treated to reduce its ruminal degradability. J. Sci. Food Agric. 42: 1.

Van Soest, P. J. 1965. Use of detergents in analysis of fibrous feeds. III. Study of
effects of heating and drying on yield of fiber and lignin in forages. J. Assoc. Offic.
Anal. Chem. 48: 785.

Van Soest, P. J. 1982. Chapter 4. Fecal composition, mathematics of digestion
balances and markers. Page 39 in Nutritional Ecology of the Ruminant, Cornell
University Press, Ithaca, NY.

Varvikko, T. and J. E. Lindberg. 1985. Estimation of microbial nitrogen in nylon-bag
residues by feed 15-N dilution. Br. J. Nutr. 54: 473.

.Veen, W. A. G, J. Veling, and Y. T. Bakker. 1988. The influence of slowly and rapidly
degradable concentrate protein on nitrogen balance, rumen fermentation pattern and
blood parameters in dairy cattle. Nether. J. of Agric. Sci. 36. 127.

Vik-Mo, L. 1989. Degradability of forages in sacco 2. Silages of grasses and red clover
at two cutting times, with formic acid and without additive. Acta Agric. Scand. 39: 53

Visek, W. J. 1984. Ammonia: its effects on biological systems, metabolic hormones,
and reproduction. J. Dairy Sci. 67: 481.

Voss V. L, D. Stehr, L. D. Satter, andG. A. Broderick. 1988. Feeding lactating dairy
cows proteins resistant to ruminal degradation. J. Dairy Sci. 71: 2428.

"Waghom, G. C. 1991. Electronegativity and redox potential of rumen digesta in situ in
,cows eating fresh luceme. N. Z. J. Agric. Res. 34: 359

‘Wagner, F. W. 1985. Chapter2. Assessment of methodology for the purification,
characterization, and measurement of proteases. Pages 17- 39 1n Plant Proteolyuc
Enzymes. 1: l7.

 

55

Waldo, D. R. and H. K. Goering. 1979. Insolubility of proteins in ruminant feeds by
four methods. J. Anim. Sci. 49: 1560.

Walker, D. J. and C. J. Nader. 1968. Method for measuring microbial growth in rumen
contents. Appl. Microbiol. 16: 1124.

Walker, D. J. and C. J. Nader. 1975. Measurement in vivo of rumen microbial protein
synthesis. Aust. J. Agric. Res. 26: 689.

wallace, R. J. 1992. Gel filtration studies of peptide metabolism by rumen
microorganisms. J. Sci. Fd. Agric. 58: 177.

Wallace, R. J ., G. A. Broderick, and M. A. Brammall. 1987. Protein degradation by
ruminal microorganisms from sheep fed dietary supplements of urea, casein, or
albumin. Appl. and Environ. Microbiol. 53: 751.

Wallace, R. J. and M. L. Falconer. 1992. In vitro studies of conditions required to
protect protein from ruminal degradation by heating in the presence of sugars. Anim.
Feed Sci. Technol. 37: 129.

Wallace, R. J. and J. Kopecny. 1983. Breakdown of diazotized proteins and synthetic
substrates by rumen bacterial proteases. Appl. Environ. Microbiol. 45: 212.

Warner, A. C. I. and B. D. Stacy. 1968. The fate of water in the rumen 1. A critical
appraisal of the use of soluble markers. Br. J. Nutr. 22: 369.

Waters, C. J, M. A. Kitcherside, and A. J. F. Webster. 1992. Problems associated with
estimating the digestibility of undergraded dietary nitrogen from acid- -detergent
insoluble nitrogen. Anim. Feed Sci. Technol. 39: 279.

‘ Watkins, K. L., T. L. Veum, and G. F. Krause. 1987. Total nitrogen determination of
various sample types: a comparison of the Hach, Kjeltec, and Kjeldahl methods. J.
Assoc. Offic. Anal. Chem. 70: 410.

Weiss, W. P., H. R. Conrad, andW. L. Shockey. 1986A. Amino acid profiles of heat-
damaged grasses. J. Dairy Sci. 69: 1824.

Weiss, W. P. ,H. R. Conrad, andW. L. Shockey. 1986B. Digestibility of n1trogen 1n
heat-damaged alfalfa. J. Dairy Sci. 69: 2658.

, Weiss, W. P, D. O. Erickson, G. M. Erickson, and G. R. Fisher. 1989. Barley distillers
grains as a protein supplement for dairy cows. J. Dairy Sci. 72: 980.

Whitelaw, F. G., J. M. Eadie, L. A. Bruce, and W. J. Shand. 1984. Microbial protein
synthesis in cattle given roughage concentrate and all concentrate diets: the use of 2,6-
diaminopimelic acid, 2-aminoethylphosphonic acid and 35S as markers. Br. J. Nutr.
52: 249.

Wilson, J. R., D. E. Akin, M. N. McLeod, and D. J. Minson. 1989. Particle size
reduction of the leaves of a tropical and a temperate grass by cattle. II. Relation of
anatomical structure to the process of leaf breakdown through chewing and digestion.
Grass Forage Sci. 44: 65.

56

Windham, W. R. 1989. Potential characterization of a protein-carbohydrate complex
from the rumen of steers fed high quality forages. J. Agric. Food Chem. 37: 912

Wohlt, J. E., C. J. Sniffen, and W. H. Hoover. 1973. Measurement of protein solubility
in common feedstuffs. J. Dairy Sci. 56: 1052.

Wong, E. 1973. Plant Phenolics. Pages 265-322 in Chemistry and Biochemistry of
Herbage. ed. G. W. Butler and R. W. Bailey. Academic Press, New York, NY.

Zinn, R. A. and F. N. Owens. 1986. A rapid procedure for purine measurement and its
use for estimating net ruminal protein synthesis. Can. J. Anim. Sci. 66: 157.

Zorrilla-Rios, J. and G. W. Horn. 1989. Effect of ammoniation and energy
supplementation on the utilization of wheat straw by sheep. Anim. Feed Sci. Technol.
22: 305.

 

 

 

ef
ad
ov
ga
aff

Prc

57

Reprinted with permission from J. Sci. Food Agric.
(Kohn. R. A. and M. S. Allen. 1992. 58:215-220.

CHAPTER 2: STORAGE OF FRESH AND ENSILED FORAGFS BY
FREEZING AFFECTS FIBRE AND CRUDE PROTEIN FRACTIONS.I

Abstract
The effect of freezing on fibre and crude protein fractions of forages was determined.
Fresh and ensiled luceme and fresh bromegrass were processed immediately after
collection or stored at -25 °C for l d, 1 mo, 6 mo or 12 mo before drying at 55 °C.
Samples were frozen quickly by submersion in liquid nitrogen or slowly at -25 °C.
Samples which were not frozen were processed immediately or after 1 h delay at room
temperature. All treatments were replicated (n=3). Samples were analyzed for crude
protein (CP), trichloroacetic acid soluble CP (TSCP), phosphate buffer soluble CP
(BSCP), neutral detergent insoluble CP (NDCP), acid detergent insoluble CP (ADCP),
neutral detergent fibre (NDF), acid detergent fibre (ADF), acid detergent sulfuric acid
lignin (lignin) and ash. Freezing decreased BSCP and increased NDCP by more than
40% for bromegrass. Freezing also changed NDF, ADF, lignin, ash, CP and ADCP in
different ways depending on forage type and length of time frozen. No significant
effects were observed for method of freezing or a l h delay in processing. An
additional experiment showed that freeze-drying resulted in less insoluble protein than
oven-drying. Prior freezing of forages appeared inconsistently to change the extent of
gaseous loss during drying, and resulted in precipitation of protein. These changes also
affected fibre estimates. Forages should not be frozen and thawed before analysis of

protein or fibre fractions.

 

'AbsmwesenwdameanwfingofmeAmdmnDaindweeAssodmimu-fllum
l990.Raleigh,NC,USA.

Frb

Ru
H0
fib

58

Key words
Fibre, NDF, ADF, protein fractions, freezing, freeze-drying, forage analysis.

1. Introduction
Ruminants normally consume fresh forages and silages without prior freezing.
However, freezing is a common method of storing forage samples prior to analysis of
fibre and nitrogen fractions. Neutral detergent fibre (NDF) is related to forage intake
by ruminants (Van Soest 1965a; Mertens 1973) and is used to measure fibre
requirements of dairy cattle (Mertens 1987). Acid detergent fibre (ADF) is negatively
related to digestible energy content of ruminant feeds (Van Soest 1965a). Lignin is an
indicator of feed quality and is associated with decreased digestibility of feeds
(Woodman and Stewart 1932; Jung 1989).

Crude protein (CP) soluble in trichloroacetic acid (TCA), neutral buffers, or
neutral detergents generally is more rapidly converted to ammonia by ruminal microbes
than insoluble protein (Mahadevan etal 1980; Nocek etal 1983; Van Soest and Sniffen
1984; Lindberg 1988). The rate of protein degradation may, in turn, affect protein
utilization by ruminants (National Research Council, 1985). Crude protein which is
insoluble in acid detergent (ADCP) is a measure of heat damage to silage or hay that
renders the carbohydrate and protein indigestible (Van Soest 1965b; Weiss etal 1986).
Freezing of vegetative material and foods is associated with precipitation of protein
(Levitt 1962; Macrae 1970; Powrie 1973; Fennema 1973).

This experiment investigated the effect of freezing for 1 d to 12 mo on changes
in protein and fibre fractions. In addition, short-term changes (within 1 hour at room
temperature) were evaluated to determine if accurate assessment of fresh and ensiled

forages is possible. A mechanism for changes that occur during freezing is proposed.

2L1.

Ines
cutti
harv
lfloo:
HNJS

andt

follor
dean:
ofun
bags,

Hqukfi

Specfl
fiBezi

61D01

Subs;
Ifli7r
NaCH)
buffer1
(BSCP.

59

2. Experimental
2.1. Treatment of samples
Fresh smooth bromegrass (Bromus inermis Leyss.) in vegetative stage of second
cutting, and fresh luceme (Medicago sativa) in early bloom of second cutting were
harvested from university plots on different days in June 1989. Ensiled luceme (early
bloom, second cutting) was collected in June 1989 from a university tower silo after 10
mo storage as silage. Samples were chopped to 3 cm length with a paper cutter, mixed
and divided into 30 sub-samples (40 g 12) which were randomly assigned to 10
treatments. Protein solubility fractions were determined for each forage 40 min
following harvest or collection from the silo (t=0 h, n=3), 1 h following the first
determination (t=1 h, n=3), and following freezing by two methods for several lengths
of time (n=24). The samples which were frozen were placed in doubled Whirl-PakTM
bags. Half of these bags (n=12) were frozen immediately (t = 0 h) by submersion in
liquid nitrogen for 15 min. These and the remaining bags (n=12) were then placed in a
p freezer at -25 °C. Several workers processed samples but each was assigned to a
, specific task (e. g. cutting, sampling, freezing, analyzing). Samples (n=3) from each
freezing treatment were thawed for 15 min i1 at 25° C before analysis after 1 d, 1 mo,
, 6 mo or 12 mo of frozen storage.
2.2. Analyses
‘ Sub-samples of fresh or thawed forages (2 g DM basis) were homogenized in 100 ml of
~ pH 7.0 saline—phosphate buffer (4.7 g 1-1 NazHPO4, 2.3 g 1-1 NaH2P04, 4.9 g 1-1
' NaCl) for 2 minutes, and strained through cheese cloth after 15 min at 25° C. This
buffer is similar in pH and osmotic activity to rumen fluid. Buffer soluble crude protein
(BSCP) was determined by centrifuging (1500 g, 15 min) 30 m1 of filtrate and

analyzing the supernatant for nitrogen. Trichloroacetic acid soluble crude protein

 

 

 

(ISCP) we
15 sec, wa
nitrogen ]
Kjeldahl p

emulsion 1

ADCP. T
for 6 mo a
drying at
DM basis
fractions 1
1970) wit
Z-ethoxyt
sulfite (R
nimgent
analysis 1
NDCP at
Cellulose
1982) as

C
E
Grand I
carried 0

qualitati'

60

(TSCP) was determined by adding 10 ml of 24% (w/v) TCA to 30 ml filtrate, vortexing
15 sec, waiting 15 min, centrifuging (1500 g, 15 min), and analyzing the supernatant for
nitrogen. Nitrogen content of liquid fractions was determined using a modified
Kjeldahl procedure (Watkins en! 1987) with the addition of 1 ml 5% Sigma anti-foam
emulsion A and a boiling chip to reduce foaming.

After sampling for determination of the soluble protein concentration, the
remaining sample was dried at 55 °C in a forced air oven for 48 h, ground with a Wiley
mill (1 mm screen), and analyzed for DM. ash, NDF, ADF, lignin, CP, NDCP, and
ADCP. The first 6 treatments were randomized for these analyses while samples frozen
for 6 mo and 1 yr were each analyzed separately. Dry matter was determined after
drying at 100 °C in a forced air oven for 16 hours, and all analyses are expressed on a
DM basis unless otherwise indicated. Samples were ashed at 500 °C for 5 h. Fibre
fractions (NDF, ADF, lignin) were determined sequentially (Goering and Van Soest
1970) with modification of NDF procedure by the substitution of triethylene glycol for
2-ethoxyethanol (18 g kg' 1), and the omission of decahydronaptlralene and sodium
sulfite (Robertson and Van Soest 1977). Crude protein was determined as Kjeldahl
nitrogen times 6.25 (Watkins M 1987). Residues from non-sequential detergent
analysis were collected on filter paper and nitrogen content determined to calculate
NDCP and ADCP expressed as a percentage of sample DM.

Cellulose and hemicellulose estimates were based on detergent analysis (Van Soest,
1982) as follows:

Cellulose = ADF - lignin

Hemicellulose = NDF - ADF - NDCP + ADCP.

Ground orchardgrass hay was used as a standard for each analysis and all analyses were
carried out by the same person. Effect of freezing before drying was also examined

qualitatively with light and electron microscopy.

 

Liihmh
Data were
DM and a1
ash basis,
samples. -'
(e0hhm
were detei
Method 0‘
standards
of the star
141%“
An additi
drying (11
C. Addit
compare
describet
buffer us

Models 1
3.1. Cu

Ematest
bmmegr
NDCP t
indicatit

freezing

61

2.3. Statistical
Data were analyzed by the General Linear Models procedure of SAS (1985) on both a
DM and an ash basis by forage type using orthogonal contrasts. For calculations on an
ash basis, values were divided by percentage ash. Fresh were compared to frozen
samples. To evaluate immediate changes after harvesting, the first treatment of fresh
(t=0 h) was compared to the second treatment (t=1 h). Duration of freezing effects
were determined (1 (1 vs. 1, 6 and 12 mo; 1 mo vs. 6 and 12 mo; and 6 mo vs. 12 mo).
Method of freezing (fast vs. slow) was compared within each time period. Dry
standards did not vary over time for any analyses except BSCP; therefore mean values
of the standards were used as covariates for statistics on BSCP.
2.4. Freeze-drying
An additional experiment compared drying at 55 °C without freezing (n=6) to freeze-
drying (n=2) for NDCP levels of each forage. Freeze-dried samples were frozen at -25°
C. Additionally, bromegrass analyzed for NDCP directly after freeze-drying (n=2) was
compared to freeze-dried sample that had been resolubilized in phosphate buffer (as
1 described previously) for 2 h (n=2), and sample resolubilized in reduced phosphate
buffer using 3 mg ml‘1 NazS (n=2). The data were analyzed using SAS General Linear
Models procedure for unbalanced non-orthogonal contrasts.
3. Results and Discussion

3.1. Crude protein fractions

Freezing affected solubility of CP in different solvents (Tables 1-3) with the
greatest changes in forage composition observed for protein fractions of smooth
bromegrass (Table 1). There was a 41% reduction in BSCP and a 96% increase in

NDCP by freezing for l (1 compared to fresh samples. The change in ADCP was slight,

indicating the protein which became insoluble in buffer and neutral detergent due to

freezing was not made insoluble in acid. There was a subsequent decline in NDCP

 

when freer
similar to 1
however, I
solubility t
in CP and
CP percen
unfrozen 1
3.2. Ash,
Ash perce
For exam
for l d (T
is intuitiv
temperan
1'espiratm
would be
darnage 1
frozen sa
lucemec
Percenta
randomi
1
(Tables
for 24 h
NDCP;
LikGWis
attributg

62

when freezing for longer duration. Changes in BSCP and NDCP of fresh luceme were
similar to those of bromegrass, but less dramatic (Table 2). In contrast to fresh forages,
rowever, there was a slight decline in NDCP due to freezing for ensiled luceme. Buffer
solubility of luceme silage declined when stored for 6 mo or longer (Table 3). Changes
m CP and TSCP for all forages were small but occasionally significant. The decline in
SP percentage for fresh luceme appears to result from a greater respiration of the
infrozen material, as CP did not change relative to ash.

3.2. Ash, fibre and lignin

Ash percentage of the fresh forages appeared to vary depending on duration of freezing.
For example, ash percentage of fresh luceme decreased from 7.2 to 6.9 due to freezing
for 1 d (Table 2). Ash is that material which remains after heating at 500° C for 5 h. It
.s intuitively unlikely that this material could be lost from the feed as gas at much lower
emperatures. Changes in the percentage ash are therefore most likely due to

respiratory losses of other feed components. The most likely time for such losses

would be during the thawing or 55 °C drying periods. Freezing or thawing may

‘iamage catabolic enzymes and therefore result in decreased respiration during drying of
frozen. samples. The increase in ash from 6 to 12 mo of freezing of bromegrass and
*uceme contradicts this explanation. For some treatments, the variation in ash
)ercentage could be due to randomization error as the last 4 treatments were not
'andomized with the first 6 treatments.

Fibre fractions of fresh forages also changed due to freezing for 24 h or longer
Tables 1-3). There was an 11% increase in NDF content of bromegrass after freezing
'or 24 h (Table 1 and Fig 2), and 73% of this increase was attributed to an increase in
tlDCP; however, estimates of hemicellulose and lignin also increased significantly.
hikewise, freezing increased NDF of fresh luceme (Fig 2), and this increase was also

.ttributed to increased NDCP, hemicellulose and lignin. The slight increase in NDF

 

due to free:
estimate in
Sm
-25 °C W61
treatments
rates of in
(contrast 3
coincident
significan
analyzing
3.3. Visu
Frozen sa
freeze dri
Particles
differenc
Simply ft
in surfac
more 310
Howeve
34. Fl‘t
Oven (11
Percent;
(3.1% 0
that We
Freem-

Silage (

63

due to freezing of luceme silage, was attributed to an increase in the hemicellulose
estimate in spite of decreased NDCP and lignin.

Small differences due to freezing rapidly in liquid nitrogen vs. more slowly at
-25 °C were noted for ADCP and TSCP for bromegrass when comparing the 2
treatments within the 24 h time period. Because differences that resulted from different
rates of freezing were infrequent, only contrasts from the 24 h time period are shown
(contrast 3 in Tables 1—3). Statistical significance of these contrasts may be
coincidental given the large number of contrasts made, and there is little practical
significance due to the small magnitude of the differences. There was no effect of
analyzing immediately after harvest compared to a l h delay at room temperature.
3.3. Visual differences
Frozen samples were darker in color and browned compared to samples dried fresh or
freeze dried for the fresh forages but did not appear to differ for the ensiled luceme.
Particles from these samples appeared thicker using light microscopy, but these
differences may have resulted from deposits inside or outside of vascular tissue or
simply from a color change. Scanning electron microscopy did not reveal any changes
1 in surface morphology. Bromegrass samples that had been frozen and thawed filtered
more slowly after refluxing in neutral detergent than samples freeze dried or not frozen.
However, all samples filtered adequately before cooling.
3.4. Freeze-drying
Oven dried and freeze-dried treatments were significantly different in NDCP
percentage for all forages (P<0.05). Freeze-drying decreased NDCP of bromegrass
(3.1% of DM ) compared to oven drying without prior freezing (4.9%). Fresh luceme
_ that was freeze—dried had lower NDCP (2.7 %) than oven dried luceme (4.0 %).
Freeze—dried luceme silage had slightly lower NDCP (2.6 %) compared to oven dried

silage (3.1%). These results are similar to a previous study on freeze-dried and oven

 

dried fora:
resolubiliz
freeze-drit
content of
dried we
study did
protein pr
3.5. Poss
Results 01
precipitat
Freeze-d1
liquid wa
formatior
increased
carbohyd
vacuoles
cOIIlplex
Sumere 5
increase(
may be t
or mace
iIlsolublt
importer
DOtentiaj.

64

dried forages (Abdalla c431 1988). In addition, freeze-dried bromegrass that was
resolubilized in saline-phosphate buffer before analysis had higher NDCP (5.6%) than
freeze-dried bromegrass analyzed directly without resolubilization (3.1%). The NDCP
content of bromegrass resolubilized in reduced buffer (3.1%) did not differ from freeze-
dried sample analyzed directly (3.1%). Standard errors of the treatment means for this
study did not differ significantly and averaged 0.17%. These results suggest that forage
protein precipitates in the presence of liquid water after being frozen.

3.5. Possible mechanism

Results of these experiments suggest NDCP and other insoluble compounds are
precipitated during oven-drying, especially if the samples are previously stored frozen.
Freeze-drying appears to inhibit the reaction by preventing transport of reactants via
liquid water in frozen and dry samples. High reducing potential seems to inhibit
formation of NDCP. Given these characteristics, a possible mechanism for the
increased NDCP after freezing and thawing may include the polymerization of proteins,
carbohydrates and phenolic compounds. Peroxidase enzymes may be released from
vacuoles that are ruptured by water crystals during freezing. These enzymes may
complex proteins and carbohydrates into acid-stable and acid-labile complexes (Van
Sumere M 1975; Whitrnore 197 8). The acid stable complexes would result in
increased ADCP while the acid labile complexes would only increase the NDCP, and
may be broken down in the abomasum under acidic stomach conditions. The enzymes
or reactants may become less active over time, thus resulting in decreased formation of
insoluble material with longer duration of frozen storage. The reaction may be less
important to ensiled materials where enzymes have been damaged and reducing

potential is higher.

 

3.6. Role
Freezing l
affect fibr
in forage 1
temperatu
degradabi
Lignin an
NDCP an
work has
compared
Mo 1989]
between 1

frozen ant

65

3.6. Relevance of research

Freezing has previously been shown to change protein solubility, but was found not to
affect fibre fractions of forages (Mme et al. 1975). More recently, however, changes
in forage fibre composition have been demonstrated for short-term storage at several
temperatures (O'Neil and Allen 1990). Carbohydrate and protein digestibility and
degradability may also differ due to changes in chemical composition as observed here.
Lignin and ADCP are negatively related to forage digestibility, and greater portions of
NDCP and less BSCP may result in less readily degradable protein in forages. Other
work has shown faster ruminal degradation of DM and CP from freeze-dried feeds
compared to oven-dried feeds for samples submerged in the rumen in dacron bags (V ik-
Mo 1989), and in vivo differences in digestion of perennial ryegrass were noted
between fresh and frozen samples (Beever 9:31 1974). Fresh forages should not be

frozen and thawed before analysis of fibre and nitrogen fractions.

 

Abdalla 1
fresh for.

Beever l
perennia

Fennem:
offend:
Inc, Nev

Natio
Wash

4. References

Abdalla H 0, Fox D G, Van Soest P J 1988 An evaluation of methods for preserving
fresh forage samples before protein fraction determinations. LAW 66 2646-2649.

Beever D E, Cammell S B, Wallace A 1974 The digestion of fresh, frozen and dried
perennial ryegrass. W 33 73A-74A.

Fennema O R 1973 Nature of the freezing process. In. W
W ed Fennema O R, Powrie W D, Marth E H. Marcel Dekker
Inc, New York, pp 150-239.

Goering H K, Van Soest P J 1970 Forage fiber analyses (apparatus, reagents,
procedures, and some applications). Agricultural Handbook Number 379. Agricultural
Research Service USDA.

Jung n G 1989 Forage lignins and their effects on fiber digestibility. Agmnl 81 33-33.

Levitt J 1962 A sulfhydryl-disulfide hypothesis of frost injury and resistance in plants. .I
W 3 355-391.

Lindberg J E 1988 Influence of cutting time and N fertilization on the nutritive value of
timotléy8 2. .sgstimates of rumen degradability of nitrogenous compounds. W
Res 1 5-

MacRae J C 1970 Changes ln chemical corrapositsign of freeze-stored herbage. New
ZealandloumalangrieulluraLReseamb 1 45

MacRae J C 1975 Changes in the biochemical composition of herbage upon freezing
and thawing. W 84 125-131.

Mahadevan S, Erfle J D, Sauer F D 1980 Degradation of soluble and insoluble
proteins by Bacteroides amylophilus protease and by rumen microorganisms. .LAnim
Sci 50 723-728.

Mertens D R 1973 Application of theoretical mathematical models to cell wall
digestion and forage intake in ruminants. Ph.D. Dissertation Cornell University, Ithaca,
NY.

Mertens D R 1987 Predicting intake and digestibility using mathematical models of
ruminal function. .LAninLSm'. 64 1548-1558.

National Research Council 1985 W National Academy Press,
Washington, D. C.

 

Nocek l E.
insoluble r
Sgi66 166

O'Neil K1
temperatlt

Powrie W
freeze-pm
Fennema

Robertsor
! . S .

SAS User

Van Soes
by min.

mm

Van Soc:
of heatin
48 785-7

Van Soc
7594.

Van Soe

Canine

Van Sur
Phenolit
Sumere

Vik~Mc
and fro:

Watkin
sample

Weiss ‘
damagt
Whitm
241-24
Wood]

Organi
feedin

 

67

Nocek J E, Herbein J H, Polan C E 1983 Total amino acid release rates of soluble and
insoluble protein fractions and concentrate feedstuffs by Streptomyces griseus. 1D mg
Sci 66 1663- 1667.

O'Neil K A, Allen, M S 1990 Forage fiber values can be affected by storage time and
temperature before drying. JD airy Sci Suppl 7312.8

Powrie W D 1973 Characteristics of food phytosystems and their behavior duringr
freeze- -preservationIn: Lw —T m r P rv i n fF n Livin M ed
Fennema O R, Powrie W D, Marth E H. Marcel Dekker Inc, New York, pp 352- 386.

Robertson J B, Van Soest P J 1977 Dietary fiber estimation in concentrate feedstuffs. J

Anim Sci 1Snppl 45 254.
SAS User's Guide. 1985 SAS Inst. Inc. Carry, NC.

Van Soest P J 1965a Symposium on factors influencing the voluntary intake of herbage
by ruminants: Voluntary intake in relation to chemical composition and digestibility ,1
mm; 24 834-843.

Van Soest P J 1965b Use of detergents in analysis of fibrous feeds. IH. Study of effects
of heating and dryng on yield of fiber and lignin in forages. J Asscc Qffic Anal thm
48 7 85-790.

Van Soest P J 1982 Chapter 6: Analytical systems for evaluation of feeds. In: Micrcbia!
Ecclcgy cf thc Ruminant, ed Van Soest, P J. O and E Books, Inc. Corvallis, Oregon, pp
75-94.

Van Soest P J, Sniffen C J 1984 Nitrogen fractions in NDF and ADF. Distillcrs Fccg
anfcrcncc, Cincinnati OH 39 73-82.

Van Sumere C F, Albrecht J, Dedonder A, de Poorer H, P6 I 1975 Plant proteins and
phenolics. In: The thmistg and Bicchcmistry cf Plant Prctcins, ed Harbome J B, Van
Sumere C F. Academic Press Inc, London, pp 210-264.

Vik- Mo L 1989 Degradability of forages 1n sacco. 1. Grass crops and silages after oven
and freeze drying. Acta Agrjc Scand 39 43- 52.

Watkins K L, Veum T L, Krause G F 1987 Total nitrogen determination of various
sample types: A comparison of the Hach, Kjeltec, and Kjeldahl methods. 1A sscc Qffic
AaLChm 70 410- 412.

Weiss W P, Conrad H R, Shockey W L 1986 Digestibility of nitrogen in heat—
damaged alfalfa. J Dairy Sci 69 2658-2670.

EVhitmore F W 1978 Lignin-protein complex catalyzed by peroxidase. Plant Sci Q11 13
41—245.

, Woodman H E, Stewart J 1932 The mechanism of cellulose digestion in the ruminant
organism. The action of cellulose-splitting bacteria on the fiber of certain typical
feeding stuffs. ,1 Agric Sci 22 527-547.

manages» mane.

 

5..me r mayo" on mama—i o: 3.8 3083 E5 38 35:25 on «885 :83» «3% Am So «L 02:.

 

 

 

 

EB new? menu. EFL?
u u E :5 one Baal BEE—en
helm». z n z n IE
9% no.» u; N: a.» a: e: we... we... N3 N5 .3 5
new .5 Bu :5 6... :8 Ga Eu :6 Sb 6... .8 2m we
do? 5 ..u ..o C 5 i S C C C .8 2m rue
wmQum PM 96 ab Pm Pc arm PM u.» wb Pm .3 farm {warm
28E 9— 9.» ch PM uh uh uh uh flu SN .3 Turn rwk
>001” _.~ 90 EN Hb Pm Pm fin fin Pm ab .3 Farah warub
22a do.» do.» So one 2b 25 o: 2.: 3a are .3 C in
.53 a: 8.» 3... 8e are 8e Ba u; u; 3d .2 _ rub
Readies 2.» NS Sb NZ u: N; Sq Sn Ba Ba .3 re 5
8:38." 22 N3 8.» 3e 23 N3 web N3 N3 2... B we we
was. we we 3 3 u... u... we we .2 we be 5.3 rub
a: 5 So 3 o.» 9e 9e 3 9e 3 .5 .8 we --

 

£8398 anon am an gases" 825m. cam—.583 8:55 ~88 9.8 25388 on a 3 Base. 3am Qnabuv. axon? CZ 305 5
~89 <m. mason“ 8 an: <9 nu; 85v.“ 8 ~ g <9 r a. B :5 @255“ .3 _ :5 Se. a BE 5 Be 533% uv a Se. 5 :5 9.83m“ 3

mason 332v. <m. act—v. 395 we a. 85:58:, 8:555 3 £53 982 925.8.“ o: E. 8: Sam 955$. %88: Bea—v. 9
mean 3.3%... $88: act—a on bu on E madman. ".3 Susan. «02% @385 u inane: x 9N9 51039.88? 85 852m 9.. .

55838 95.2 «2ch Ow. .225: 8.2%.: 5852a Gm. 553 8888. 58.56 9v. .225— Gnanwna 38. 3.66
884%.? no?

099.93» 39.0.

 

 

69

..,.»me N. M500. 0. 0.008.... 0.. 0800 08.0... .50 .030 08000.5 00 0.00.. 000080 8 .8 m. 6.5.

 

 

 

: 3N: EPLWE‘
u .. u .. .0 .30 E 9.5 .881 IE... 0%..
l20 .0. z .u z .., z .u
9.. .3 N... .3 No.0 3... 3.. 03 .3 .3 mm... 33 ..N
0... 5.. .3 .3 3.. .3 .3 .3 .3 .3 .3 ..s . 20
do... 3 0... 0.. 3 3 3 0.0 3 3 3 ..o 0.3 9.3
$0.... 3 0.0 M... 3 3 0.. 3 a... 3 3 .00 3... 3...
2.00... ...o .... .3 .3 .3 .3 .3 .3 ..... .3 .8 .... ....
.60..A .3 Z E ..o 3 .3 ..0 ..0 ..m .3 .8 20 3
2...“. ..3... 3.. ..3 ..m... ..3 ..m... .3... .30 ..m... .6... .00 . .
>9“... .0... 0.3 0.... 00... 0... 0.... 03 m... 0... B. .00 20 .
55.80%... .3 5.0 .w. .3 .3 a... .3 .3 ..3 ..3 .00 3 3
85.08 03 .3 8.0 N... 0.... ..3 0.... 8.0 8... N... 3.. 20 20
0...... q 0 3 3 3 3 3 3 3 3 3 .0. .43 .
8.. . N 3 a... 3 a. 3 3 a... a. 3 ..N 3 --

 

3.00080 0.8. o. 0.0 .8280... .0008. 5.8.0.00... 80.8.... ..8... 00.0 0......30 0.. 0 0Q 80:0. 000.0 %Aobmv. 0x00... UK 0.30.... 5
$0... <0. 0.000... N. .no <0. .n... 00.8.“ 8 . 0 .5. 0. 0. 5 Bo 805%. .3 . :8 <0. 0 000 5 Bo $000.5“ 8 0 <0. 5 So 3005.... 8
.88.. 8.00;. <0. 3.02... 50...... N. 0. 0%....008... 000.808 00 0030 0.80... 8.00.30 0.. 0.. 00.. 000.0 QAobmv. 0$30.. 80.0... .0
.550 3.8%... 0330.. 0.02... 0. .8 on .0 $020.. 5a. 30.8.. m0800 08.0... n 3.8%.. x 0.8. 0.08000800000 00.0 30.020 Ow.
5.80.5.0 0.5.0. 8.00.0 Ow. 0.200.... 00.050... 0.50.5.0 0... 5.0.0 00.080... 0.00.0000 n... .2005: 00.05%... 08.0. 050.0
00.0.m0... 00.0.

 

 

70

.5me w. Q80. 0. 0800.8 0.. 0800 08.0... 8.0 .....8 080.800 0. .8080 00.80 8 .oo 8-. U7...

 

 

 

E .0800. ...0an mmgm. $80.88.“
u .. n... ... .0.. 3.... .N ...0 0.02... 06>...
l0 8. 2 .u z .u 2 m
02.. ..3. ..0... ...... ...... ...3 ...... ..3... ..30 ..3.. ...N .00 ....
000 N.... N... N... N3 NN.0 N... NN.. N... N3 N... N. 20 20
0.00.... .03 .0.. .0 .03 .... .00 .00 .0.. .3 ..0 .N3 .3 .3
000.... ...3 ...3 ...3 .30 .33 .3... ...3 ..3 ..... ...0 .00 .. ..
200... ..N 0.. 0.. N0 3 N. 3 3 N.. N.. .... . .
>00... ... .0 ... ... ... .3 ... 3 ..0 .0 ..0 3.3 3.3
2.0... 030 000 0... ...0 03 0.... 003 03 003 .03 .NN .3... ..
>00... N30 N33 N3 N3. N3.N N33 N30 N30 N3. 030 .N0 0... 20
0.008080. .00 .00 .0... .00 .0.. .0.. ...0 .00 .0.. .0.. ..3 ..0 .
800.000 ...0 ...0 ...0 .... ...N ..... N0.N .... ..3 ...3 ..3 3.3 20
00.... 0.0 03 3.. 3.. 0.0 ..N 0.. 3.0 3... 00 .0. .3... .3
00.. 3 3 3 3 3 ... ... ... ... 0.0 ..N 20 --

 

N8.00.008 0.8. 0. 0.0 .8080... 800.00. $80500... 00080.0 ..8... 00... 8.00800 0.. 0 08 80.8. 00.0.0 @08me 0x00... UK :00...” C
.80.. <0. ..88... 8 We <0. .n... 00...... w. . 0 <0. .. 0. .N ...0 3000.8. .3 . ...0 <0. 0 ....0 .N ...0 0800.8. 3. 0 <0. 5 ...0 3800.8. 0.
.88.. 80.0... <0. 0.02... $0.0... NA ... 880.88.. 00080.0 00 80<0 080... 800.800 0.. .... 00.. 000.0 $qu.90.. 0.88.. 80.0..< ...
.580 ....880... 0.088.. 0.02.... 0. .3 on ... 0808.. .0.... 80.8.. 80800 08.0... H 0.880.. x 0.8. 080008800000 00.0 00.0.0.0 O...

.2880»... 0008. 00.8.0 0... .2008. 00.080... 800.030 Ow. 58.0 00.080... 80080.0 O... .2005... 00.080... ...08. 8.8.0
00.080... 088.

 

DEGM

A batch f
feeds ust
microbia
or incorr
limiting
and free
protein ‘
contami
pellet tc
bacteria
finsing
NaSOg
compa
incoqy
tWice 1
differe
incor;
lower

degra

71

CHAPTER 3: OBSERVATIONS ON DETERMINATION OF PROTEIN
DEGRADATION OF FEEDS BY IN VITRO BATCH FERMENTATION WITH

35SULFUR USED TO LABEL MICROBIAL GROWTH.

ABSTRACT
A batch fermentation method was developed to measure crude protein degradation of
feeds using 358-cysteine to label microbial protein. Previous attempts to use 358 as
microbial marker involved inefficient or expensive methods to isolate incorporated 358,
or incorrect estimation of N: 358 in microbes. Feeds were incubated (38 °C) in N-
limiting media under C02 with rumen microbial inocula. Following fermentation, feed
and free bacteria were separated by differential centrifugation. Undegraded feed
protein was determined as crude protein of feed pellet corrected for microbial
contamination. The Nz358 of microbial pellet was multiplied by 358 content of feed
pellet to obtain microbial N. Several methods of rinsing non-incorporated 358 from
bacterial and feed pellets were compared with best results obtained by immediately
rinsing at least twice in 1% HCl (30 ml) with 1 g/L of each: cysteine.HCl, Na2804, and
NaSO3. Several methods of solubilizing microbial pellet to determine 358 were
compared with adequate results obtained by solubilizing in water. Feed sulfur was
incorporated into microbial protein, thereby increasing Nz358 of bacteria by more than
twice that of bacteria grown on NH4. Since feeds differ in S available to microbes,
differential centrifugation of each treatment was required for determination of marker
incorporation. Using a higher level (3 vs. 6 vs. 11 mg N) of soybean meal resulted in a
lower percentage degradation at 20 h. Fraction of feed protein remaining after 20 h
degradation was greater (P<.05; n=3) for corn (.40) and meat and bone meal (.32) than

 

for soybe
standard
publishe
treatmen
(Keywo
protein)
Abbrev
CP = or
DTI‘ =

nitroger

The rat
for the
1989).
by fen
al., 19:
than it
severa
finely:
Physir

incub

meas

feet
PTOd'

72
for soybean meal (.12), alfalfa hay (.11) and NH4CO3 (.01). Average within treatment
standard deviation was .079. The proposed method is simpler than previously
published methods using 358 as microbial marker, however high variation within
treatments persists.
(Keywords: in vitro protein degradation, 35Sulfur incorporation, rumen microbial
protein)
Abbreviation Key:

CP = crude protein, CTAB = cytyl trimethylammonium bromide, DM = dry matter,

 

DTT = dithiolthreotol, EDTA = ctlry‘ " ' ‘ ‘ aacetic acid, NPN = non-protein

nitrogen, TCA = trichloroacetic acid.

INTRODUCTION
The rate and extent of protein degradation in the rumen is an important consideration
for the optimal feeding of high-producing ruminants (National Research Council,
1989). Researchers have developed methods to measure protein degradation of feeds
by fermentation with live microbes from the rumen (Henderickx et al., 1972; Raab et
al., 1983; Thomsen, 1985; Spencer et al., 1988). In vitro techniques are less expensive
than in vivo and in situ methods and they offer the potential to analyze and to compare
‘ several feeds at once. In vitro techniques also allow the researcher to collect and to
l analyze both the residue and metabolites of microbial degradation, while more
physiological techniques are complicated by fluxes out of the rumen or out of in situ
l incubation bags.

Methods that involve fermentation with live microbes offer an opportunity to
measure protein degradation in an environment similar to that of the rumen. However,
these methods require estimation or separation of microbial protein associated with the
feedstuff. The extent of protein degradation at a point in time is equal to the sum

production of non-protein nitrogen (NPN) and microbial protein. Various procedures

 

have beer
digesta.

(Walker
Miller, 1
sulfide (l
pyruvate
synthesi
protein
associat
nitrogen
entirely
associat
0f mien
Pr
compar
fennent
detenni
An add
methoc
after a

[hill 83}

reQuin
‘0 be n

micro]

73
have been used to determine the proportion of microbial N associated with feed in
. digesta.
‘ Radioactive sulfur (358) often is used to label microbial protein in digesta
‘ (Walker and Nader, 1968; Beever, et al., 1974; Walker and Nader, 1975; Mathers and
V Miller, 1980). Microorganisms bring about the enzymatic reduction of sulfate ($04) to
, sulfide (HzS) which is then converted to the thiol group of cysteine by reaction with
. pyruvate and ammonia or alternatively with serine (Lehninger, 1975). Methionine is
synthesized from cysteine, and these amino acids are incorporated into microbial
protein. Therefore, incubation of microbes with inorganic sulfur results in radioactivity
associated with microbial protein (Emery, et al., 1957). See Figure l. A ratio of
nitrogen to radioactive sulfur (N: 358) can be determined for a standard which is
entirely of microbial matter. This ratio can then be multiplied by the amount of 358
associated with a mixture of dietary and microbial protein. The product is the estimate
of microbial N.
Previous methods to label microbial protein have been too labor-intensive to
compare a large number of feeds at once. Residual feed protein from microbial
. fermentation is digested in nitric or perchloric acid for solubilization of 358 before
, determination of radioactivity (Walker and Nader, 1968; Mathers and Miller, 1980).
An additional fraction is digested separately for N determination using the Kjeldahl
. method. Previously, as much as 20% of the radioactivity has been associated with feed
; after a few hours incubation in sterile media (Walker and Nader, 1968). This indicates
4 that separation of incorporated and non-incorporated 35$ is inadequate.
Determination of the ratio of Nz358 also has been labor intensive and has

I required acceptance of several assumptions. This ratio is determined on what is hoped

to be microbial protein that is not contaminated with feed protein. The standard

microbial protein may be obtained by isolating protein from an in vitro rumen

 

incubatior
be separat
1975). In
the micro
Ar
microbial
feeds W01
accompli
that have
incorpor.
al., 1972
compare
'1
method 1
methods
Separatit
determir
degmda
sources
not enti
80 that :

BXperie

74
incubation that used only NPN (Henderickx et al., 1972). Alternatively, microbes may
be separated from dietary protein by differential centrifugation (Walker and Nader,
1975). In either case, one assumes that the protein isolated has the same N23SS ratio as
the microbial protein associated with the feed sample.

An early goal of the present study was to develop a system to measure
microbial growth using an abundance 35S-cysteine, so that contribution of sulfur from
feeds would be negligible compared to that included in the system. If this goal were
accomplished, the Nz3SS ratio might be constant across all treatments and standards
that have only soluble protein. (NPN). Cysteine was desired as a marker because direct
incorporation of feed amino acids may be favored over inorganic sulfur (Henderickx et
al., 1972), thus lowering the incorporation of 358 of microbes incubated with feeds
compared to those incubated with blanks.

The objective of this study was to simplify an in vitro microbial fermentation
method to measure protein degradation of feeds. This required simplification of
methods to solubilize microbial protein for detection of radioactivity, a more thorough
separation of incorporated and non-incorporated radioactivity, and simple but accurate
determination of Nz358. The method was evaluated by determining fractional
degradation of protein supplied at different levels, comparison of several protein
sources and the variation associated with those measurements. These objectives were
not entirely met, but the results of several experiments are presented in this dissertation

so that anyone wishing to continue this work will have the benefit of the these

experiences.

 

 

cl

 

 

75

SO3—
sulfate

802'

sulfite

S(s) <—> H28

elemental sulfide
sulfur

pyruvate + NH3

NH3—CH-COOH —> NH3-CH—COOH

|
CH2-SH CH2—CH2—S-CH3

cysteine methionine
Microbial Protein

Figure 1. Sulfur Metabolism in the Rumen.

 

 

 

 

 

Final Metht
Freeze-drier
mm screen I
drying in a‘
a modified
In V
media is bl
bicarbonat
digest of 0
make the 1
concentra‘
microbial
carbohyd
addition .
microbia
protein, 5
at .2 m1]
dairy co
microbi
Ilnmedi
Cyseir
dithiolt
E
tWice t

“final-

 

76

METHODS

Final Method for Microbial Fermentation

Freeze-dried forages and dry concentrates were ground through a 5 mm and then a 1
mm screen of a Wiley mill. Dry matter (DM) was determined by weighing hot after
drying in a forced air oven for 8 h at 100 °C. Crude protein (CP) was determined using
a modified Kjeldahl method (Hach et al., 1985).

. In vitro media was adapted from that of Goering and Van Soest (1970). This
media is buffered with NaHCO3 and includes macro- and micro-minerals. Ammonium
bicarbonate was replaced with an equi-molar concentration of NaHCO3, and trypsic
digest of casein (trypticase, Sigma Chemical Co., Saint Louis, MO) was excluded to
make the media low in N. Magnesium Sulfate was replaced with an equi-molar
concentration of MgC12.6H20 to prevent a decrease in specific activity of 358 in
microbial protein due to dilution with cold sulfur from added sulfate. Soluble
carbohydrates were added in the quantities shown (Chapter 4, Table 1). Vitamin
addition was adapted from Russell et al. (1983). In order to decrease affects on
microbial growth, and hence protein degradation, due to amino acid profile of added
protein, isoacids (valerate, isovalerate, isobutyrate and 2-methyl-butyrate) were added
at .2 ml/L of media. This rich media was used to simulate that of the high-producing
dairy cow fed a mixed diet, and to remove constraints to protein degradation (i.e. low
microbial growth) that are unrelated to protein structure of the feed to be tested.
Immediately before use, media was purged with C02 until the pH stabilized at 6.8.
Cysteine.HCl (.05 g, except where indicated otherwise) and 353-1abe1ed L-cysteine in
dithiolthreotol (DTT, Aldrich Chem. Co., Milwaukee, WI) was mixed into the media.

Rumen fluid was collected 2 to 4 h after feeding from a Holstein dry cow fed
twice daily a diet based on corn silage and alfalfa hay. Rumen fluid was blended for 1

min and strained through 4 layers of cheese cloth followed by glass wool. Rumen fluid

 

 

 

 

 

was incub
method (it
Feet
tubes. Tu
25 m1 of 1
incubated
At the en
with Ant
temporar
In l

min. Th
Trichlon
supernat
consider
(20.000
consider
supema
combin.
of the f.
in the ri
Sample
time in
labeled
incorp.

I

Suspt‘n

77

was incubated in continuous culture at a dilution rate of .07/h for 24 h according to the
method described elsewhere, except media was as described previously.

Feed samples containing 4 to 6 mg N were placed in 50 ml plastic centrifuge
tubes. Tubes were sealed with black rubber stoppers connected to a source of CO2, and
25 ml of labeled media and 8 ml of rumen inocula were added. Samples were
incubated at 38 °C while in equilibrium with 1 atm C02 for varying amounts of time.
At the end of the incubation, degradation was arrested by addition of 2 ml of 6 N HCl
with Antifoam Emulsion A (Sigma Chemical Co., Saint Louis, MO) and sample storage
temporarily at 4 °C.

In the final procedure, digested feed samples were centrifuged at 400 g for 30
min. The supernatant was decanted and centrifuged again at 20,000 g for 30 min.
Trichloroacetic acid (TCA, 10 ml per 30 ml sample) was added to this second
supernatant which was centrifuged a third time (1500 g, 20 min). The first pellet was
considered residual feed protein, bound bacteria and protozoa. The second pellet
(20,000 g) was considered to be composed only of free bacteria. The third pellet was
considered to be composed of soluble feed and released microbial protein. The final
supernatant was discarded as degraded protein. The first and third pellets were
combined and rinsed with 1% (vol/vol) HCl solution (30 ml) containing 1 g/L of each
of the following: cysteineHCl, NaSOg, and Na2804. Sample was allowed to soak
in the rinse solution for l h after which time TCA (10 ml, 240 g/L) was added and
samples were centrifuged at 1500 g. The pellet was rinsed and precipitated a second
time in the same manner. Three replicates of feed were also incubated for 3 h in
labeled media with no inocula to determine amount of activity that precipitates without
incorporation in microbial CP.

Rinsed pellets were suspended in H20 (10 m1), and an aliquot (.2 ml) of this was

suspended in scintillation cocktail (BioFluorTM, New England Nuclear, Dupont Co.,

 

Boston, l
Searle A1
standard
samples :
sample tl

remaindt

pellet).
Develop
Several
protein 1
were co
was add
should 1
Cellulo.
h witho
precipit
several
methoc
Resuln
in plac
Variant
disinte
quencl
detem
sol/lit:

 

78

Boston, MA). Radioactivity was determined in a scintillation counter (Isocap 300,
Searle Analytic Inc.). Quench was determined from channel ratios determined on
standard curves with differing amounts of soybean meal added. Radioactivity in
samples incubated without microbial inocula was subtracted from the respective feed
sample that had been inoculated and incubated. Crude protein was determined on the
remainder of the sample as Kjeldahl N times 6.25. Undegraded feed CP was estimated
as: Feed pellet CP - (353 in feed pellet (x) CP in bacterial pellet/ 35$ in bacterial
pellet).
Development and Validation
Several experiments were conducted to determine the best method to separate microbial
protein from the original label. Most of these experiments (unless indicated otherwise)
were conducted as those involving microbial fermentation except no microbial inocula
was added. Incubation with labeled media without the addition of microbial inocula
should result in no precipitation of 358 because there should be no microbial growth.
Cellulose (160 mg), corn starch (200 mg) and soybean meal (100 mg) were incubated 4
h without addition of rumen inocula. Following incubation, samples were directly

precipitated with TCA (rather than centrifuging out the bacterial pellet). Because

l several experiments were conducted with variations to the methods used (as the final
method was not yet developed), these methods are described in more detail in the
Results and Discussion section of this paper. Generally, replicate values are presented
in place of means because the sample size was small for these experiments, and the
variance was not uniform across treatments. Radioactivity is presented as
disintegrations per minute unless quench was not determined (assuming uniformity of
quenching across treatments) in which case counts per min are given. Quench was

' determined by channel ratios adjusted to a standard curve when quench was added from

soybean meal, or by internal standards when indicated.

 

Sev
detection

cysteine

experimt
digest ot
TCA pn
peptide
Smgd
Results
amount
bacteri:
ccntrifr
previor
solven
one ha
meat 2
run. I

fenne

Sepa

non.
addi‘

Dtec

 

79

Several experiments were conducted to compare methods to solubilize 358 for
detection in the supernatant. These methods included comparing recovery of 35S
cysteine from solutions as well as detection of activity in labeled microbial samples.

Radio-labeled microbial protein was isolated from feed protein in two different
experiments. In the first experiment, microbes were grown in the presence of trypsic
digest of casein (Sigma Chemical Co.), or NH4Cl. The ratio 35$:N was determined for
TCA precipitate of this degradation after incubation for 18 h with different amounts of
peptide or ammonia N added initially (0, 1, 3, or 5 mg), and with incubations with 0 or
5 mg of peptide or ammonia N for different periods of time (0, 3, 9, 18, or 48 h).
Results were analyzed for consistency of 353 to ratio over time and with different
amounts of substrate. In the second experiment, 358 to N ratio was determined on the
bacterial pellet from degradation of feeds. The free bacteria were isolated by
centrifugation at 20,000 g after removal of feed that precipitates at 400 g, as described
previously. Feeds used and their respective percentage CP content on 21 DM basis were:
solvent extracted soybean (Glycine max) meal -- 48.0, alfalfa (Medicago sativa) hay at
one half bloom —- 19.1, dried and ground corn (Zea mays L.) grain -- 10.8, and dried
meat and bone meal —- 55.2. Three replicates of each feed were incubated in the same
run. Protein degradation was determined by the final method for microbial

fermentation described previously for the 5 feed samples.
RESULTS AND DISCUSSION

Separation of Incorporated from Non-Incorporated 35$

1 ' n f r u in n ri r rin in Rumen bacteria may synthesize
non-protein sulfur compounds such as sulfide (82'). This experiment was to see if
addition of Na2S is necessary to remove sulfide to prevent its association with

precipitable protein. Trypticase (61 mg), Starch (200 mg) and cellulose (160 mg) were

 

digested i
1i)6 (1me
of N328}.
incubatic
cocktail t
than whr
would bt
would b

other su‘:

 

 

80

digested 10 h with rumen inocula and media containing cysteine (1 g/L) and 35S (8.2 x
106 dpm) to produce labeled microbial protein and sulfide. Treatments were addition
of Na2S.9H20 (40 mg), NazS.9H20 and HCl (1 ml 6 N) or no addition at end of
incubation and before rinsing once in water. Pellet was suspended in scintillation
cocktail (10 ml). Total radioactivity (cpm) in pellet was greater for NazS treatments
than when none was added as indicated in Table 1. One would expect that H28 gas
would be removed from the system by addition of acid and that more radioactivity
would be remOved by addition of Na2S. However, the reducing agent may have made
other sulfur compounds more likely to bind when rinsed in distilled water.

A second experiment was conducted at the same time to determine the effect of
additions prior to rinsing when no fermentation has occurred (only cysteine in the
system). Cellulose (160 mg), starch (200 mg) and soybean meal (100 mg) were
incubated 4 h without addition of rumen inocula. Treatments were addition of NazS
(40 mg), ‘2-mercaptoethanol (5 m1), DTT (50 mg), or no addition prior to TCA
precipitation and rinsing three times with water. Precipitation of 35S (dpm) was much
greater for all treatments than for the previous experiment, but there were no
differences among treatments. The greater precipitation of 35S in this experiment
compared to the previous one may be explained by greater binding to soybean meal
than trypticase. Total radioactivity in pellet is given in Table 2.

Pmcipitatign of cysteine with protein, This experiment was conducted to
compare different methods of precipitating protein for their effects on precipitation of
35S from cysteine. Trichloroacetic acid, tungstic acid, perchloric acid and copper
sulfate precipitation were compared, each at two concentrations of precipitating agent
(5 or 10 ml of reagent to 30 ml buffer). Each treatment was centrifuged (1500 x g; 15
min) or filtered (Whatman #1), and was carried out with cellulose (160 mg) and starch

(200 mg) and no protein or with these carbohydrates and soybean meal (100 mg).

 

 

 

Mel

Cor

81

Media (30ml) with 358-cysteine (1.6 x 106 dpm) was added to centrifuge tubes, and
precipitating reagent added immediately. There was no added cold cysteine.
Concentrations of reagents were 24% TCA, 20% perchloric acid, .05 M copper sulfate
and 1 N sodium hydroxide, and 10% sodium tungstate and 1.07 N sulfuric acid.

Table 1. Precipitation of 358 (cpm) with trichloroacetic acid after incubation of
Trypticase with microbial inocula for 10 h with addition of sulfide or sulfide and acid
before rinsing the pellet in water.

 

 

Treatment Rep 1 Rep 2 Rep 3
Na28 25,960 27,362 17,077 .
Na2S + HCl 48,492 34,368 26,181
No addition 12,977 14,491 16,032

 

Table 2. Precipitation of 358 (cpm) with trichloroacetic acid after incubation of
soybean meal with sterile media for 4 h with addition of reducing agents before rinsing
the pellet in water three times.

 

 

Treatment Rep 1 Rep 2 Rep 3
NaZS 95,143 86,896 101,271
2-mercaptoethanol 94,468 100,169

D'IT 84,245 108,337 103,942

No addition 109,500 34,215 89,564

 

Tungstic :
pellet wa
in Table i

4 filter p
4 filter p
presence
did not :

precipit

determi
to deter
of mic]
cystein
and 1
and 1t
(10 m'.
15 mi
Shake
TCA

suSpe

rinsi
ante
Was

dlfft

82

Tungstic acid and copper sulfate precipitations used 2.5 or 5 ml of each reagent. Total
pellet was suspended in scintillation cocktail (10 ml), and radioactivity (cpm) is given
in Table 3.

Tungstic acid preparations would not filter through Whatman # 1 so Whatman #
4 filter paper was used. Copper sulfate preparations filtered poorly through Whatman #
4 filter paper as well. Cysteine associated with the precipitate in all cases. The
presence of soybean meal appeared to increase precipitation of cysteine. These results
did not suggest much improvement would be attained by changing from the TCA
precipitation procedure.

Rinsing the pellet of LES—cysteine precipitate. This experiment was to
determine the best method to rinse non-protein 35S-cysteine from TCA precipitate, and
to determine if an adequate amount of 35S could be removed for accurate determination
of microbial sulfur. Rinse solutions were: A. water, B. cysteine HCl (1.3 mg/ml), or C.
cysteine.HCl and Na2$ (1.3 mg/ml). Samples were washed once in solutions A and B
and 1, 2 or 3 times in solution c. In vitro media (30 ml) with 353-cysteine (4.1 x 106)
and 10 mg or 0.5 mg unlabeled cysteine was added to soybean meal (100 mg). TCA
(10 ml) was added after a 15 min delay. Samples were shaken and centrifuged (1500 g;
15 min), and supernatant discarded. Rinse solution (30 ml) was added and samples
shaken. After 15 min delay (30 min delay for one treatment with rinse C one time),
TCA was again added and centrifugation and decanting repeated. Total pellet was
suspended in scintillation cocktail (10 ml), and radioactivity (cpm) is shown (Table 4).

Rinsing in cysteine solution was more effective at removing 35S-cysteine than
rinsing in water. There was no apparent improvement by adding Na28. Rinsing twice
appeared better than once and three times was better than twice. However, this effect
was only apparent at higher level of initial unlabeled cysteine. Considering the 20 fold

difference in initial cold cysteine concentration with little difference in precipitation of

 

Table 3.
protein, ‘

Ira-1|

1313M
Water.

 

83

Table 3. Precipitation of 35S-cysteine (cpm) with different methods to precipitate
protein, with and in the absence of soybean meal.1

 

 

Reagent SBM Conc. Method rep 1 rep 2
Trichloroacetic acid
0 10 ml centrifuge 23,062 32,706
filter 8,625 9,403
5 m1 centrifuge 28,218 —-—
filter 11,874 10,552
100 mg 10 m1 centrifuge 70,629 49,630
filter 35,041 72,682
5 ml centrifuge 23,818 94,512
filter 31,396 30,904

Tungstic acid
10 ml centrifuge 38,897 45,840

filter 36,095 25,806
5 m1 centrifuge 32,120 38,343
filter 41,017 24,625

100 mg 10 m1 centrifuge 50,076 44,462
filter (#54) 50,366 45,648
5 m1 centrifuge 103,501 86,995
filter (#54) 70,695
Perchloric acid
10 ml centrifuge 43,364 38,550

filter 39,716 28,184
5 ml centrifuge 43,951 47,077
filter 80,730 66,210
100 mg 10 ml centrifuge 30,984 32,531
filter 99,133 82,333
5 ml centrifuge 51,304 49,821
filter
Copper sulfate
0 10 ml centrifuge 73,897
filter 23,376 10,864
5 m1 centrifuge 58,059 44,936
filter 21,641 15,269
100 mg 10 ml centrifuge 28,204 26,635
filter (#54) 23,496
5 ml centrifuge 23,31 l 40,529

 

1SBM = soybean meal, conc. = concentration of reagent (5 or 10 m1 used with 30 ml
water.

 

 

Tablt
(cpm

Initi.‘

lOn

84

Table 4. Method of rinsing the trichloroacetic acid pellet on removal of 35S—cysteine
(cpm) from a feed mixture (starch, cellulose and soybean meal).

 

 

Initial Cys Rinse Times Delay r9) 1 rep 2

10 mg Water 1 15 min 8038 6407
Cys 1 15 min 5843 2991
Cys+Na2S 1 15 min 6925 4479
Cys+Na2S 2 15 min 3813 3623
Cys+Na2S 3 15 min 3117 3172
Cys+Na2$ 1 30 min 2605 6965

0.5 mg Water 1 15 min 7975 5895
Cys l 15 min 3191
Cys+Na2S 1 15 min 3661 4079
Cys+Na2S 2 15 min 2731 7875
Cys+Na2$ 3 15 min 6261 8468

Cys+Na2S l 30 min 8798

 

35S t

cystei
theon
There

effect

and c
conta
mg) i
incub
TCA
33ml)
deter

show

orgat
solut
incul

from

85

358, either a constant amount of 35S precipitates or a constant percentage of total
cysteine. Decreasing the unlabeled cysteine in the ill vitro media would therefore,
theoretically increase the ratio of labeled protein to interference from non-protein 358.
Therefore, if low levels of unlabeled cysteine are used in the media, 35S should be
effective in labeling microbial S.

Another experiment compared rinsing 1 time with a solution containing Na28
and cysteine, compared to water or no rinsing. Samples were incubated in media
containing 35S-cysteine (5.0 x 105 dpm) for 8 h with starch (200 mg), cellulose (160
mg) and trypticase (62 mg) or soybean meal (100 mg). Samples with trypticase were
incubated with live rumen organisms while soybean meal was not inoculated. After
TCA precipitation and rinsing in 30 ml solution and TCA precipitation repeated,
samples were digested in H2804 with perchloric acid (Hach et al., 1985), quench was
determined on external standards with H2304, and total pellet radioactivity (dpm) is
shown in Table 5.

These results confirmed the previous experiment for incubation without live
organisms. Rinsing in H20 reduced dpm compared to not rinsing, and rinsing in a
solution containing cysteine and Na2S reduced dpm compared to rinsing in H2O for
incubations with soybean meal and trypticase. Percentage activity precipitated ranged
from 1.5 to 5.3.

An experiment compared rinsing l, 2 or 3 times in solution (30 ml) containing 1
g/L of each: Na2SO4, NaSO3, Na2S and cysteine-HG]. Soybean meal, starch and
cellulose were incubated for 3 h in media containing cysteine (.2 g/L) and 3SS-cysteine

(3.0 x 105 dpm), and without rumen bacteria. After TCA precipitation, addition of

. rinse and repeated precipitation in TCA, the pellet was digested in H2SO4 with

peroxide. Total radioactivity precipitated is given Table 6.

 

Table 5.
for 8 h v

’

Treattnt

 

Trypticz
No 1
H2(
Cys

Soybea
No

86

Table 5. Precipitation of 35S (dpm) from media incubated with rumen microorganisms
for 8 h with Trypticase, or without organisms with soybean meal.

 

 

Treatment Rep 1 RELZ W
Trypticase
No rinse 24,340 18,085 26,383
H20 rinse 12,298 18,383 18,426
Cys-Na2S 16,936 14,255 13,702
Soybean meal
No rinse 16,128 20,723 26,553
H20 rinse 15,191 19,021 17,787
Cys-Na2S 7,277 7,745 9,787

 

Fl‘otal activity in each tube was 5.0 x lOme.

Table 6. Effect of number of times of rinsing a feed sample on removal of 35S-cysteine
from TCA precipitable material.1

 

 

Rinse Rep 1 Rep 2 Rep 3

1 time 7,106 12,681 7,489
2 times 1 4,809 4,979 5,617
3 times 4,383 4,383 4,043

 

1Total activity in each tube was 3.0 x 105 dpm.

Table 7. Effect of soaking feed sample during rinsing on precipitation 35S-cysteine.1

 

 

Soak Rep 1 Rep 2 R933

0 h 1477 3191 407 8

2 h 1950 2246 2128
20 h 2 187 2009 2305

 

1Rinse solution was a .1 M sodium phosphate buffer with 1 g/L of each cysteine.HCl,
Na28, Na2$04, NaSO3. Total activity per tube was 3.6 x 103 dpm.

sul

in

 

 

87

Table 8. 355-cysteine precipitated with different feed samples incubated (38 °C) for 3
h without microbial inoculal

 

 

Treatment rep 1 rep 2 rpp 3
Soybean meal 4,213 4,766 5,191
Bromegrass 13,361 13,787 15,702
Starch 2,085 1,872 2,979
Cellulose 4,340 4,043 5,021
flbean—acid rinsed 1,191 1,957 2,383

 

1Each sample was rinsed 3 times in pH 7 phosphate buffer (.1 M), with 1% KBr and 1
g/l each of the following: cysteine-HCl, Na2SO4, NaSO3 and Na2S, or in pH 1.5 dilute
sulfuric acid. Total activity placed in each tube was 5.2 x 106.

 

Additional rinses reduced the precipitation of 355, but the difference between 2
and 3 rinses was not great. Another consistent trend is becoming apparent. The
reduction in precipitated 358 due to rinsing procedures is accompanied by a reduction
in variation among samples within a treatment.

An experiment was conducted to determine the effect of allowing the sample to
soak in rinse before adding TCA. Each sample was rinsed twice and the pellet was
allowed to soak in the rinse for 0, 2 or 20 h before addition of TCA for the second rinse.
The pellet was made radioactive as described in the previous experiment except 3.6 x
105 dpm 358 was added. The rinse solution was the same as that above except it was
made in a .1 M phosphate buffer (pH 7).

There appeared to be a reduction in 358 precipitation due to soaking for 2 h as
compared to 0 h. However, there was no effect of soaking for 20 h compared to 2 11.

Another experiment compared substrates for their ability to precipitate 35S.

Soybean meal (100 mg), smooth bromegrass hay (300 mg), starch (200 mg), or

 

 

cellulos
microbi
in the in
pH 7 ph
cysteine
in dilute
Bach trt
113A fo

cysteine
precipit
through
to reduc

better u
(pH 1.5
contain
Sllpema
cenm‘fu
and the
measun

Supema

CYSteint

That is
the Star

88

cellulose (160 mg) were incubated with 358-cysteine (5.2 x 106 dpm) without
microbial inocula for 3 h at 38 °C. Amounts of substrate were used that would be used
in the in vitro procedure for protein degradation (4 to 6 mg N). Samples were rinsed in
pH 7 phosphate buffer (.1 M), with 1% KBr and 1 g/l each of the following:
cysteineHCl, Na2804, NaSO3 and Na28. One u'eatment of soybean meal was rinsed
in dilute sulfuric acid (pH 1.5) instead of phosphate buffer and with the other reagents.
Each treatment was rinsed three times as before with a 16 h delay before addition of
TCA for the final rinse. Total radioactivity (dpm) is shown in Table 8.

These results show that different substrates result in different amounts of 35S-
cysteine precipitated. This should emphasize the importance of minimizing the
precipitation of non-microbial 358 because simply subtracting a constant for all feeds
throughout the degradation may result in systematic bias. Rinsing in acid also appeared
to reduce the precipitation compared to neutral buffer.

A rapid method to measure 358 precipitation was used in several experiments to
better understand the association of 358-cysteine with the pellet. Dilute sulfuric acid
(pH 1.5. 30 ml) containing 358—cysteine (6.8 x 103 dpm) was added to centrifuge tubes
containing starch (200 mg). The samples were centrifuged (15 min; 1500 x g) and
supernatant discarded. Rinse solution (20 ml) was added to the samples followed by
centrifugation. This was repeated twice. The final suspension (before centrifugation)
and the final supernatant were sampled (.2 ml) for radioactivity. Precipitated 35S was
measured as the difference in radioactivity between the final suspension and the final
supernatant.

The first use of this technique showed that starch pellet that already had 358-
cysteine associated with it did not result in additional precipitation of 35S-cysteine.
That is it appeared that the starch was saturated with radioactive cysteine. However,
the starch

 

Table

Treatr
pH 1
pH 7

M

Table
trimetl
starch.
Treatm
EDTA
DTT
CTAB

Control

may hat

OCCUITBI

shown 1
have 8hr

c)lStine i

 

89

Table 9. Effect of rinse solution pH on precipitation of 35S—cysteine (cpm) after rinsing
twice.

 

 

 

Treatment rep 1 rep 2
pH 1 3 10
pH 7 295 250
pH 10.6 80 25
Table 10. Effect of adding -11 J ‘ " ' ‘ ‘ " acid, dithiolthreotol, or cytyl

 

trimethylammonium bromide to rinse solution on 35S-cysteine precipitation (cpm) with
starch.

 

 

Treatment Rep 1 Rep 2 Rep 3
EDTA 164 88 157
DTT 247 239 162
CTAB 205 178 245
Control 218 242 258

 

may have been reduced after the first precipitation so that no appreciable oxidation
occurred the second time.

The effect of pH was investigated for its affect on the association with starch as
shown in dpm below. The association was inhibited by acid compared to pH 7. Others
have shown than the oxidation of cysteine to cystine occurs fastest at neutral pH, and
cystine is insoluble. Low pH rinse solution resulted in the lowest precipitation of 35S.

The pellet from pH 7 treatment was then rinsed 3 times in dilute sulfuric acid
(pH 1.5) to see if the acid rinse could reverse the binding. There was no decrease in
binding due to rinsing in dilute acid. Rinsing an additional time in pH 0.5 sulfuric acid
also did not reduce the association of 35$ with starch. Rinsing again by soaking for 2 h

in pH 0.5 solution also did not remove 35S from the pellet. However, nearly all 358

 

 

 

was s
(repli

fOI'IIII

ethyl
(CTr
stare
The '
and I
2 ml
were
the i

scin

tend
no t

solu

beft

botl
witl
eitl.
cen

tin:

 

90

was solubilized by rinsing in 1% HBr solution brought to pH 7 by phosphate buffer

(replicated twice). The Br' ions may have been able to substitute for bound cysteine
formed in this particular case. Further exploration of use of KBr may be desired, but
HBr vapors are hazardous and so use of both acid rinse and HBr may not be desired.

An experiment was conducted to determine the effect of adding

 

etlr,‘ " ' ‘ ‘ ancetic acid (EDTA), DTT, or cytyl trimethylammonium bromide
(CTAB) to rinsing the pellet of 353-cysteine. Samples included .17 g cellulose, .2 g
starch and .062 g soybean meal and were incubated for 4 h without microorganisms.
The Control rinse contained 10 ml/L HCl and 1 g/L of each: cysteine-HCl, Na2804, l
and NaSO3. Treatments were addition of: EDTA at 3 mM final concentration, D'I'I‘ at l
2 mM final concentration, and CTAB at 8 mM final concentration. The treatments
were for the first rinse only, and a second rinse with control rinse solution was used for
the second rinse. Samples were dissolved in 40% H2SO4 and diluted before
scintillation counting. Total cpm are shown in Table 10 for each of three replicates.
The results show a high variation in precipitation of 353. There was a

tendency for the EDTA treatments to be lower in precipitable 35$ than the control with
no effect of the other treatments. The EDTA did not dissolve in the acidic rinse
solution.

An experiment was conducted to determine whether there is precipitable 353
before the incubation, and if so whether it could be precipitated before the experiment.
A year old source of 35S-cysteine was used for the experiment. The media and rinse
both contained 1% HCl to prevent oxidation of cysteine. Samples were rinsed twice
with the solution used previously. The treatments included centrifugation of the media
either in the presence or absence of starch (.2 g) and cellulose (.16 g). After
centrifugation the supernatant was added to starch, cellulose and soybean meal and

rinsed twice as previously. An additional treatment was prepared the same way but

01

de

of

di;

ad
Ti

Wt

dir

in'

inc

 

 

91

contained 5 times more cold cysteine in the media. The control treatment was not

centrifuged to remove precipitable material before adding to feed sample and rinsed.
Each treatment was replicated 3 times. There were no significant differences among
treatments and 0.8 % of the total radioactivity precipitated ill each case. The
experiment was repeated the next day using the same isotope. This time the precipitate
ranged from 0.45 % of the total for samples subjected to centrifugation before rinsing,
compared to 1.5% for samples not subjected to centrifugation before rinsing. Addition
of 5 times more cold cysteine also reduced the precipitation of 35$ to 0.18% of the
total. The results suggest that some precipitable 358 can be removed by centrifugation
before incubation.

If some ferrnentations result in greater precipitation of non-microbial 35S than
others, then bias is introduced by this precipitation. An experiment was conducted to
determine the effect on precipitation of 35S due to the reduction procedure and length
of incubation. Media was degassed by heating to boiling and cooling under C02. In
retrospect, this was not a very efficient way to degas media compared to using a bubble
disperser. Media was reduced with cysteine-HCI (1 g/L ) and 35S-cysteine (2.3 x 106
dpm/tube) was added. Feed (.1 g soybean meal, .2 g starch and .16 g cellulose) was
added to centrifuge tubes and 30 ml media added to the feed and incubated at 38 °C.
Timing of reduction, addition of 35S, addition to the feed and precipitation in TCA
were varied as indicated in Table 11. Each sample was rinsed twice in 1% HCl with 1
g/l each of the following: cysteine-HCI, Na2SO4, NaSO3, and radioactivity determined
directly on a sample (.2 ml) of a suspension of the final pellet. At the same time, some
samples were incubated with 1% HCl in place of in vitro media, these results are shown

in Table 12.
The most important effect to be determined from this experiment is that

incubation without organisms results in precipitation of 358 over time. The longer the

 

rn

at

92
incubation, the more likely is the precipitation. It could not be prevented by reducing
conditions or even acid. Most researchers have used 35SO4 to label microbes. This
compound may be less likely to adhere to feeds. However, it is readily converted to SZ‘
which, like cysteine, is quickly bound to feed samples (Walker and Nader, 1968).
Solubilization of Incorporated 358

Tissue solubilizer and Hg] digestion. Commercial tissue solubilizer (BTS-450,
Beckman Co., Fullerton, CA) and 6 N HCl were used separately to solubilize the pellet
from TCA precipitate. Due to the high amount of carbohydrate in the samples both
approaches left a black viscous residue that resulted in high (>70%) quench. Sample
size was reduced from entire pellet to .2 ml resolubilized pellet (resolubilized in 10 m1
H20). This procedure resulted in an apparently adequate digestion in BTS-450 with
low quench, as will be discussed later.

Hg :1 digestion and chromatography with Dowex, The purpose of these
experiments was to determine if 35S-protein could be determined by HCl digestion
followed by separation using chromatography. Orchard grass hay (.25 g) was digested
for 24 h with 6 N HCl with 2—mercaptoethanol (15 ml) at 110 °C. After digestion 35S-
cysteine was added to digest. 35S—cysteine in 1 N HCl (n=2) or 6 N HCl digest (n=2)
was put on Dowex anion exchange columns (.2 ml). The columns were rinsed with .01
N HCl and cysteine eluted with .5 N NH4OH. Only about 20% of the 353 was
recovered from the columns when cysteine was put on the columns directly. Only
about 2% of the 35S from the digest was recovered from the columns. Moreover,

recovery increased only 20% for fractions of NH4OH over fractions of HCl.

 

 

 

 

Tab]
in pl

WISE

Oh
Oh
121
121

93

Table 11. Effects of timing on precipitation of 35S—cysteine (dpm) with feed samples
in pH 6.8 media.

 

 

Media Feed Precip-
reduced 353 added added itated Rep 1 Rep 2 Rep 3
+incubated

0 h 0 h 12 h 15 h 13,950 15,550 14,400

0 h 12 h 12 h 15 h 20,700 19,100 14,300

12 h 12 h 12 h 15 h 24,350 14,600 19,600

12 h 12 h 12 h 63 h 39,450 25,700 - 27,600

0 h 12 h 0 h 15 h 7,850 9,050 4,000

0 h 12 h 10 h 12 h 4,050 5,050 3,600 1
0 h 12 h 10 h 15 h 9,400 7,100 6,350 l
0 h 12 h 10 h 36 h 24,600 20,500 22,950

0 h 12 h 10 h 60 h 17,350 32,950 4,300

 

1Each sample was rinsed 3 times in 1% HCl with 1 g/l each of the following:
cysteine-HCl, Na2SO4, NaSO3 and Na2S. Total activity placed in each tube was 2.3 x
106

Table 12. Effects of timing on precipitation of 35S-cysteine (dpm) with feed samples
in 1% HCl.

 

 

Media Feed Precip-

reduced 353 added added itated Rep 1 Rep 2 Rep 3
+incubated

0 h 12 h 10 h 12 h 31,750 8,250 9,600

0 h 12 h 10 h 15 h 9,550 30,000 12,550

0 h 12 h 10 h 36 h 21,700 18,050 20,950

0 h 12 h 10 h 60 h 13,250 22,250 18,950

 

1Each sample was rinsed 3 times in 1% BC] with 1 g/l each of the following:
cysteine-HCI, Na2SO4, NaSO3 and Na2S. Total activity placed in each tube was 2.3 x
106

 

 

5.6‘
the

dirt
Six
in 1
COT
firs
we

sec

W3

0X

ha

lal
tht

94

A second attempt was made to elute 358-cysteine from anion exchange columns
(n=2). This time a higher concentration of NH40H (50%) was used. The first HCl
fraction from 358.cysteine in 1 N HCl contained 48% of the original 353. The first
Nmon fraction (3 ml) contained 15% of the original 358, with an additional 6.2%
recovered in other fractions. When the 358-cysteine was added with acid digest (n=2),
5.6% eluted with acid wash and 5.2% eluted in the first NH40H fraction and 3.5% in
the remaining two fractions.

The attempt to elute cysteine was repeated by putting 35S-cysteine in l N HCl
directly on the columns (n=2). This time .1 N HCl was used as rinse instead of .01 N.
Sixty-two percent of the 358 activity was recovered that was recovered was recovered
in the elute. The first HCl rinse fraction contained 22.3% and the first NH40H fraction
contained 26.9%. The experiment was repeated and similar results were obtained. The
first NH40H fraction was added to .1 N HCl and put on new columns. Two peaks
were again found except that this time they were in the second fraction of HCl and
second fraction of NH40H. The BC] peak was 3.8 times higher than the NH40H peak.

It is possible that the cysteine was oxidized to cystine or degraded once the pH
was neutralized with NH4OH. Cystine is insoluble which could explain why low
recoveries were found. Addition of reducing agents and an anaerobic environment, or
oxidation to cysteic acid, may have improved the recovery. The cysteine may not have
had time or surface area enough to bind to the column thus explaining the recovery in
the HCl fraction.

The attempt to use ncr digestion and chromatography to solubilize 358 in the
pellet was discontinued for several reasons. The procedure was found to be more
laborious than necessary. Much of the 358 in cysteine appeared to bind irreversibly to

the column. The 358 in cysteine did not elute in a single fraction. There was

95

considerable interference from other compounds in the HCl digest with elution on the
column.

W. The first experiment was conducted to determine if
digestion in concentrated sulfuric acid with H202 could be used for determination of
35S in TCA precipitate. This would allow digestion for radioactivity to be
simultaneous to digestion for ninogen determination by Nessler's reaction. However, N
and S differ in that N is non-gaseous in acid even when boiling at high temperatures.
0n the other hand S can be expected to evaporate with sulfuric acid. This experiment
measured the amount of 358 present after digestion in H2804. The amount of H2804
remaining was determined by titration with 8 N NaOH. An equal amount of 358 was
added to each of 5 flasks. Orchard grass (.25 g) was added to flasks 1 to 3, and H2804
(4 ml) was added to each additional flask. The first 5 flasks were digested for varying
amounts of time to digest sample and allow some evaporation. The flasks were titrated
with TKN solution as indicator, then were filled to 100 ml with H20. Samples (.2 ml)
for 358 activity were taken.

Samples that were digested resulted in lower acidity and lower cpm than the
undigested sample. This suggests that 35S disappeared with the reduction in acid.
However, the sample that was not digested had a higher ratio of cpm/acidity, suggesting
that more cpm were lost from samples that were digested than acidity was lost.
Therefore, on the basis of one sample, this may not be an appropriate way to digest
samples to determine 358 activity. On the other hand, the cpm per ml NaOH required
to neutralize acid was constant for the other samples. Perhaps, activity lost in these
samples was from the sulfide contamination rather than cysteine. Further work with
more replication is required to reevaluate these preliminary results. Addition of 358-

cysteine to 1 ml 20% H2804 resulted in 94% efficiency of scintillation counting

 

 

 

{I'll-Ir}! l?

 

96

compared to the same radioactivity in 20 m1 aqueous environment. This indicates that
the quenching from acid is not a serious problem to overcome.

A different approach to estimating H2804 remaining after digestion was
investigated. The concentrated acid was sampled after digestion for radioactivity. The
flask was then filled to 100 ml with H20 and sampled again for 353 and N. The acid
remaining after digestion was calculated as follows:

Conc. H2804 = (100 ml x dpm/ml of diluted acid) / (dpm/ml of cone. acid)
Theamount of acid could then be used to calculate the percentage reduction in N from
sampling the concentrated acid for radioactivity. The sampling of concentrated acid
was highly variable with a conventional repeater pipette, but was improved
substantially by using a positive displacement pipette. Sampling of concentrated
H2804 was still variable because of the viscosity of the material.

Alternatively, the amount of acid remaining after digestion was determined by
weight. This approach assumes that only acid remains after the digestion process. An
experiment was conducted to see if 358 from free cysteine was lost proportionally to
H2804 during digestion. Four treatments were compared: 1) no digestion of H2804,
2) boiled in H2804 for 10 min., 3) boiled in H2804 for 10 min and 10 m1 H202 added,
and 4) boiled with in H2804 with 20 ml H20 and l Kemwipe with H202 added for
complete digestion. The replicate weights of the residues and cpm in 0.2 ml of digest
diluted to 100 ml with H20 are presented. The initial weight of the acid varied from
7.27 to 7.39 with all but one sample between 7.36 and 7.39.

Weight of the residual acid appeared to increase due to digestion for treatment 2
(Table 14). The digestion decreased weight of residual for treatment 4, presumably to
evaporation of sulfuric acid. Digestion (treatment 2-4) decreased the recovery of cpm.
There was a greater decrease in cpm than in weight. This could be due to an incorrect

measurement of acid or to a greater loss of 358 than of total S. In either case, the

97

digestion in sulfuric acid resulted in a loss of 35S that could not be easily accounted for.
Use of weight to determine evaporation of H2804 would be further complicated if feed
samples contain ash that increases weight of digested samples. With concentrate feeds,
sample size was small relative to the amount of acid, which would decrease this error.
Additional experiments demonstrated that digestion in cold concentrated
H2804 resulted in high quenching (> 50%) even if diluted to 20% acid for scintillation
counting. The acid accounted for 4% of the quenching, with the remainder due to the
coloration of acid digested samples. Quenching could be reduced by the addition of
H202 for decolorizing, but addition of .3 ml 50% H202 to .2 ml 20% H2804 resulted
in 20 % quenching from H202. The quenching (as determined by internal standards)
from color and H202 was reduced by incubation in the oven for 30 min at 55 °C. and
was reduced more by incubation at 100 °C. However, this treatment resulted in an
unexplained loss of 358 activity. The final method of solubilizing the microbial pellet
was. to suspend sample in 6 m1 H20 and add 4 ml concentrated H2804. The solutions
are rapidly mixed to prevent coloration of the sample. This mixture is diluted 50%
before scintillation counting to reduce acid quenching and facilitate dissolution in
cocktail. All microbial protein was dissolved in this solution as was purified starch and

cellulose. However, some feed protein remained undissolved.

 

98

Table 13. Recovery of 358 (cpm) in samples digested in H2SO4 with peroxide and
recovery per ml NaOH required to neutralize the H2SO4 in the digest.

 

 

NaOH (m1) cpm cRm/NaOH
16.0 1735 108
17.2 1828 106
14.5 1671 105
18.6 2025 109
17.2 1904 110
21.0* 2771 131

 

*Last sample was not digested.

Table 14. Digestion of 35S-cysteine in sulfuric acid and peroxide on recovery of
radioactivity (cpm).

 

 

 

Treatment Residual mass cpm cpm/mass
Not digested 7 .27 g 4803 661
7.36 4899 666
Boiled in H2SO4 7.73 g 4000 517
7.76 4256 548
H2S04 + H202 7.33 g 4409 602
7.53 4605 612
Digested Kemwipe 5.77 g 2529 438

6.71 3032 452

 

 

99

WW. An experiment was conducted to
determine if the efficiency of measuring 353 radioactivity in microbial protein directly
without digestion. Treatments were: 1) radio-labeled microbes suspended in water and
suspended in scintillation cocktail, 2) radio-labeled microbes suspended in 20%
(vol/vol) H2804, and 3) radio-labeled microbes solubilized in BTS-450. There were 3
replicates per treatment. Microbes were by incubating with rumen inocula at 38 °C for
17 h with trypticase (3 mg/ml) and carbohydrates in media. The microbes were
precipitated with TCA and rinsed twice using the final method determined previously.
Quench was determined by internal standard.

Though treatment with 20% H2804 appeared to decrease recovery of 358 from
microbes compared to solubilizing in water or BTS-450, treatments appeared equal for
effect of quenching determined on internal standards. These results suggest that use of
BTS-450 is not necessary for detection of 358 activity. However, perhaps none of the
treatments solubilized all of the activity in microbial samples. These treatments should
be compared to samples treated by complete acid digestion to ensure complete
recovery. Secondly, a problem with use of 358 is the high variability within treatments.
A larger number of samples per treatment is required to evaluate effects of these
solubilizing treatments on variability.

Ratio or N to 35s in Labeled Microbes

Effects of length of incubation on microbial growth are demonstrated in Table 16.
Microbial inocula was incubated for 24 h in continuous culture before addition to
centrifuge tubes with media and carbohydrates. This inocula was already radio-labeled
when added to the tubes. Microbial growth appeared to cease within 3 h of incubation
for both ammonia and trypticase. Though 5 mg of N was added to each tube, microbial
pellet only contained 3.5 mg N. Perhaps some microbial N was solubilized in TCA at

 

100

the end of the incubation. The 35S to N ratio increased over time for NH4 but
remained fairly constant for trypticase.

The preincubation media had trypticase in it. This media thus supplied additional sulfur
to growing microbes which diluted the incorporation of 358. Additional growth on
NH4 resulted in an increase in N to 358 ratio because less sulfur was available in the
media. This ratio continued to increase up to 48 h even though additional microbial
growth plateaued at 3 h. It is possible that more 35S precipitated with the pellet
without incorporation, but this did not occur for the samples incubated in trypticase, so
this is an unlikely explanation. Alternatively, microbial protein may have been
degraded by microbes and resynthesized resulting in more microbial 35S incorporated
over time. In any system in which the amount of S in the incubation is different from
that of the preincubation, there may be a change in 35S:N over the length of the
incubation, thus requiring that this ratio be determined for each time point to which it is

to be applied.

Table 15. Solubilization of 35S from radio-labeled microbes using different solvents.

 

 

Treatment dpm* EfficielgLl
Solubilized in water 74835 0.867
Solubilized in 20% H2SO4 69285 0.856
Solubilized in BTS-450 76298 0.853

SBM (n=3) 1576 0.019

 

*Significant effect determined by ANOVA, P < .05.

1Fraction of internal standard measured in spiked samples.

 

 

T:

ni

 

 

a"D

Table 16. Effect of incubation time with microbial inocula and nitrogen source on
nitrogen and 355 activity insoluble in trichloroacetic acid.

 

 

Time (h) mg N dpm dpm/N
NH4 (5 mg N) u u a...
0 1.6 955 617
3 3.0 1915 639
9 3.4 2776 816
18 3.5 3281 954
48 3.2 3583 l 134
SEM (n=3) .1 119 65
Trypticase (5 mg N) **
0 2.4 995 423
3 4.4 1564 356
9 3.9 1384 353
18 3.9 1627 414
48 3.3 1519 462
SBM (n=3) .5 164 9
Blank HI IHI
0 1.6 826 509
3 1.9 1080 57 1
9 1.9 1 121 594
18 1.7 1263 727
48 1.7 1337 768
SEM (n=3) .1 41 27
P, N sources .0001 .0001 .0001

 

*Differences among treatments within a row and N sources, P < .05. ** P < .01.

Table 17. Effect of amount and source of nitrogen on nitrogen and 35S activity
insoluble in trichloroacetic acid after incubation with microbial inocula for 18 h.

 

 

Added N (m g) %N dpm dpm/N
NH4 u u an

0 1.7 1263 727

l 2.2 1341 628

3 2.8 2460 865

5 3.4 3281 954

SEM (n=3) .1 205 95
Trypticase sul- lul- u

0 1.7 1263 727

l 2.3 1479 634

3 3.2 1557 490

5 3.9 1627 414

SEM (n=3) .1 46 16
P, N source .0001 .0001 .001

 

”Differences among treatments within a row and N sources, P < .01.

 

 

103
The effects of altering the amount of nitrogen incubated with microbes for 18 h are
shown in Table 17. The greater the amount of NH4 supplied to the media, the greater
the ratio of 35S to N. This effect is due to the increased incorporation of 35S from the
media with a higher specific activity than that used in the preincubation as described for
the time effects. Incubation with greater amounts of trypticase resulted in lower 35SIN
ratio due to the dilution of 35S by S in the trypticase. Again the ratio of 35S to N in the
microbes depended on the contribution of S from the nitrogen source.

Undegraded CP in different feeds after incubation with ruminal microbes for 20 h
as described for the final method to determine protein degradation of feeds (Methods
section) is shown in Table 18. Corn grain and meat and bone meal were less degraded
than alfalfa hay and soybean meal. These trends are expected (National Research
Council, 1989). The ammonia was shown to be completely degraded as is expected.
This sample served as a control to evaluate the ability to measure protein degradation
and microbial growth. The undegraded protein is presented as the fraction of the initial
5 mg N from feeds incubated with rumen microbes. The amount of undegraded feed
CP was much less than the amount of bound microbial protein estimated from 35$.
Often when concentrates are evaluated for protein degradation, there is no correction
for microbial contamination of the microbial protein associated with feed protein. This
correction was shown to be very important to estimation of protein degradation.

Data in Table 18 also shows that the level of soybean meal incubated with
enzyme affects the fraction degraded. This may be an artifact of the method, because a
negative value was obtained at the lower level. Possible differences due to sample size
would indicate that first order kinetics do not apply. Further research is needed to

determine if it is possible to eliminate this variation due to sample size and to determine

a fractional degradation.

 

 

104

Table 18. Estimation of microbial nitrogen and undegraded feed protein after
incubation for 20 h with rumen inocula.

 

 

 

Flee Microbial Pellet
Undgr. Bound

Feed (fraction) m.o. N N (mg) dpm/N
Corn grain _ .40a 2.7c .33c 5489":
MBM 323.!) 2.60 .336 475b,c
Alfalfa hay .1 lead 3.31).c .51b 444c
NH4C03 .01d.c 3.11).c .54b 951a
Soybean meal

3 mg N -.0t5e 4.213 .320 604b

5 mg N .1204! 2.7‘3 .53b 537b,c

11 mg N .231),c 6.13 .88a 497bac
SBM (n=3) .047 .42 .043 43

 

arbchifferent superscripts within a column indicate differences between means, P <

.05.

lUndgr. = Undegraded feed protein (fraction remaining of 5 mg N used at start), Bound
m.o. N = microbial nitrogen associated with feed pellets, dpm/N = disintegrations per
minute per mg N of microbial pellet (20,000 g), MBM = dried meat and bone meal.

105

Estimation of ”SN varied for different feeds due to the incorporation of feed
sulfur into microbial protein. This result means that this ratio must be determined for
each. feed analyzed, because this ratio is important to the calculation of microbial
protein associated with undegraded feed. Since the amount of feed sulfur available for
incorporation into microbes will change over time and with sample size, 3SSIN must be
determined for every ueatrnent. Not only does this add to the labor intensity of the
method, but the variation associated with determination of ”SM is also transferred to
the final measurements.

The goal of incorporating 35S into microbial protein in the presence of large
amounts of cysteine was not attained. There is always some precipitation of
radioactivity with feed protein even when no incorporation has occurred. This
precipitation was very high relative to the specific activity of 353 in microbial protein
when large amounts of cysteine were used in the incubation. It was possible to
effectively label microbial protein, but low amounts of sulfur were required in the
media, so that a higher percentage of the 358 was incorporated into microbial protein,
and a lower percentage of non-incorporated 35S was precipitated with feed and
microbes. Using media low in sulfur resulted in a change in 35S:N for different
treatments due to the incorporation of feed sulfur in place of 35S into microbial protein.

This work exposes many of the limitations to using 353 to label microbial growth
for the determination of protein degradation of feeds. The final method presented here
was simplified from previous efforts, but it is still too complicated for analysis of a
large number of feeds at once. The variation, even within a run as shown in this paper,
was too high to detect any but huge differences between feeds. This variation was due
to the fact that calculation of protein degradation was dependent on six different
measurements all of which had variation associated with them. Each sample required

estimation of: l) 353 activity in feed incubated with no microbial inocula, 2) 353 in

 

106

microbial pellet, 3) 35S in feed pellet, 4) N in feed pellet, 5) N in microbial pellet, and
6) initial N of feed sample. Some of this variation might be reduced in the future by
lowering the precipitation of 35S with feeds and by developing more consistent
methods to solubilize the residual feed for determination of 353. Use of sulfate instead
of cysteine may result in less precipitation of non-incorporated 358, but sulfate must be
converted to cysteine before it can be incorporated into microbial protein. The
microbes may preferentially use feed sulfur over sulfate, thus decreasing the
incorporation of 35S04.

A method is still needed that uses microbial fermentation to measure protein
degradation of a large number of feeds, and with the ability to analyze the end-products.
The available methods all have many of the limitations shown in this paper. The work
described in this paper is limited by the lack of replication, and methods used to rinse
and solubilize pellet may have been inadequate for some experiments. Further
development of the techniques described here will require the iterative approach used
so far. Development of better means to solubilize microbial protein for detection of
radioactivity requires a better means to remove non-microbial 35S from samples.
Better means to determine 35S2N require better rinsing and solubilizing of microbial
protein. The description of the results described in this paper is intended to make it
easier to continue development of these methods, and to avoid the pitfalls of the present

work.

...l

N

b.)

P

E"

O\

>‘

9°

107

REFERENCES

. Beever, D. E., D. G. Harrison, D. J. Thomson, S. B. Cammell, and D. F. Osbourn.

1974. A method for the estimation of dietary and microbial protein in duodenal

digesta of ruminants. Br. J. Nutr. 32: 99.

. Emery, R. S., C. K. Smith, and C. F. Huffman. 1957. Utilization of inorganic sulfate

by lumen microorganisms. I. Incorporation of inorganic sulfate into amino acids.

Appl. Microbiol. 5: 360.

. Goering, H. K. and P. J. Van Soest. 1970. Forage Fiber Analyses (Apparatus,

Reagents, Procedures, and Some Applications). Agric. Handbook N0. 379. ARS-
USDA, Washington, DC.

Hach, C. H., S. V. Brayton, and A. B. Kopelove. 1985. A powerful Kjeldahl
nitrogen method using peroxymonosulfuric acid. J. Agric. Food Chem. 33: 1117.
Henderickx, H. K., D. I. Demeyer, C. J. Van Nevel. 1972. Problems in estimating
microbial protein synthesis in the rumen. Page 57 in Tracer Studies on Non-Protein

Nitrogen for Ruminants. Inter. Atomic Energy Agency. Vienna.

. Lehninger, A. L. 1975. Biochemistry. The Molecular Basis of Cell Structure and

Function. 2nd ed. Worth Publishers, Inc. New York, NY.

Mathers, J. C. and E. L. Miller. 1980. A simple procedure using 35S incorporation
for the measurement of microbial and undegraded food protein in ruminant digesta.
Br. J. Nutr. 43: 503.

National Research Council. 1989. Nutrient Requirements of Dairy Cattle. 6th rev.

ed. Natl. Acad. Sci., Washington, DC.

 

 

108

9. Raab, L, B. Cafantaris, T., and K. H. Menke. 1983. Rumen protein degradation and
biosynthesis 1. A new method for determination of protein degradation in rumen
fluid in vitro. Br. J. Nutr. 50: 569.

10. Russell, J. B., C. J. Sniffen, and P. J. Van Soest. 1983. Effect of carbohydrate
limitation on degradation and utilization of casein by mixed rumen bacteria J.
Dairy Sci. 66: 763.

11. Spencer, D., T. J. V. Higgins, M. Freer, H. Dove, and J. B. Coombe. 1988.
Monitoring the fate of dietary proteins in rumen fluid using gel electrophoresis. Br.
J. Nutr. 60: 241.

12. Thomsen, K. V. 1985. The specific nitrogen requirements of rumen
microorganisms. Acta Agric. Scand. Suppl. 25: 125.

13. Walker, D. J. and C. J. Nader. 1968. Method for measuring microbial growth in
rumen contents. Appl. Microbiol. 16: 1124.

14. Walker, D. J. and C. J. Nader. 1975. Measurement in vivo of rumen microbial

protein synthesis. Aust. J. Agric. Res. 26: 689.

 

 

109

CHAPTER 4: EN RICHMENT 0F PROTEOLYTIC ACTIVITY IN
PREPARATIONS FROM THE RUMEN FOR IN VIIRO STUDIES

Swords
Three experiments compared treatment of rumen contents on recovery of proteolytic
activity per ml rumen fluid (PA) and per mg nitrogen (N). Proteolytic activity was
determined by degradation of azocascin for 2 h, and N was determined by Kjeldahl
method. The first experiment compared effects of chilling, blending, rinsing,
fractionating by centrifugation at 400 x g and 20,000 x g, and their interactions. The
second experiment compared rumen fluid from a f'lstulated cow to the same rumen fluid
that was incubated in continuous culture (dilution rate = .07Ih) in media with high sugar
content. The third experiment compared effects of extracting centrifugation pellets of
rumen fluid with buffered media, media with detergent, and acetone. For all
experiments, most activity was found in the centrifugation pellets. Rinsing of rumen
solids with cold buffer increased PA and PAIN. Chilling tended to increase PA
slightly, but had no effect on PAIN. Blending rumen contents did not affect PA, but
decreased PAIN. Incubation in continuous culture did not affect PAIN. Mixing rumen
fluid precipitate with buffered media solubilized 18% of PA. Treatment with detergent
or acetone solubilized 51 and 48 % of PA respectively. Activity per mg N was 3.9
times greater in supernatant than in original sample after treatment with detergent or
acetone. Proteolytic activity was enriched most by treatment with detergent or acetone,
and additional recovery and enrichment were provided by rinsing of rumen solids.

Protein degradation in vitro: Proteolytic activity

110

Introduction

Various researchers have developed methods to measure protein degradation of feeds
by incubation in vitro with enzymes from rumen organisms. One approach has been to
incubate feeds with rumen fluid in the presence of microbial growth inhibitors
(Broderick, 1987; Neutze er al. 1993). Another approach has been to extract enzymes
from ruminal contents (Mahadevan er al. 1987, Kohn and Allen, 1993). These methods
are limited by interference of non-enzymatic proteins in rumen fluid preparation.
Contaminating proteins may compete with feed protein for enzyme sites thereby
reducing protein degradation of feeds . In addition, contaminants may interfere with the
ability to measure degraded or undegraded protein of the feed. Since the contaminant
may degrade at a slower rate when in the presence of feed protein due to competition
for enzyme, subtraction of values from blank (no feed) samples may not be adequate.
Ideally, these methods require that enzyme preparations from the lumen are
representative of whole rumen contents, yet do not contain additional proteins.

Previously, proteolytic activity in rumen fluid preparations has been increased
by chilling rumen contents before straining through cheese cloth, or by rinsing rumen
solids in cold buffer, and blending of rumen contents has had inconsistent effects (Craig
et al., 1984; Furchtenicht and Broderick, 1987). These treatments may increase
proteolytic activity (PA), and still have a negative impact on proteolytic activity per
unit protein or nitrogen (N).

Sequential centrifugation at 500 x g and 20,000 x g divides rumen fluid into a)
soluble contents, b) free bacteria, and c) protozoa, feed and bound bacteria (Kopecny
and Wallace, 1982). Understanding the enzyme partitioning in these fractions may
elucidate possibilities for enrichment of PA. Sacchrolytic rumen bacteria are primarily
responsible for protein degradation (I-Iungate, 1966). These organisms have high

reproduction rates (Lin et al., 1985), so incubation in continuous culture at high dilution

lll

rate and with high availability of saccharides and starch may be another method to
increase PA or PAIN of rumen contents. Extraction of enzymes by solubilization in
water (Wallace and Kopecny, 1983), and by treatment with detergent (Kopecny and
Wallace, 1982) or acetone (Mahadevan et al., 1987) has been successful at solubilizing
proteolytic activity, but these methods have not been compared to each other within a
single study. Interaction effects of many of these means of enrichment need elucidation
to determine which combinations are optimal for recovery and concentration of PA.
Azocasein is used to measure proteolytic activity of rumen contents
(Mahadevan et al. 1979; Wallace and Kopecny, 1983). Enzymatic degradation of
azocasein results in the solubilization of azo dye in trichloroacetic acid (TCA), but this
could result from either cleavage of dye from protein or from actual protein
degradation. The azo dye may itself be degraded from the activity of rumen contents
thus causing an underplediction of activity (Mahadevan et al., 1979). Usually
absorption of light at 440 nm wavelength is used to detect azo dye in alkaline solution
(Charney and Tomarelli, 1947), but interference from absorption of rumen fluid may
complicate this measurement. Rumen fluid has been centrifuged to remove interfering
pigments before analysis of azo casein degradation (Mahadevan et al., 1979), but if
some proteolytic activity remains soluble, the total activity may be underestimated.
The objectives of the present study were to: 1) validate and improve
measurement of PA in rumen contents using azocascin, and 2) determine effects of
chilling, blending, rinsing, fractionation by centrifugation, incubation in vitro, and

extraction of rumen contents on proteolytic activity in vitro.

MATERIALS AND METHODS
Validation and improvement of azocascin method
Color and nitrogen liberation from azocascin during degradation were compared to

ensure that solubilization of dye is due to protein hydrolysis (which releases N) and not

112

simply cleavage of dye. Azocasein was solubilized in sodium phosphate buffer (.1 M,
pH 7) and nitrogen and optical density at 440 nm (OD) were measured for standards
not subjected to TCA precipitation. Centrifuge tubes containing 2 ml phosphate buffer,
16 mg azocascin and 1 m1 rumen inocula were incubated at 38°C for 0, 1, 2, 4, or 8 h.
At the end of the incubation, TCA (1 ml) was added to precipitate undigested protein.
After centrifugation, OD. and N were measured on 1 ml aliquots. Samples which did
not contain azocasein (blanks) were also incubated and azocascin added immediately
before addition of TCA. These blanks were used to measure N and absorbance not due
to azocasein solubilization, and they were subtracted from the digested azocascin
samples.

The optimal wavelength to measure light absorption in solutions with azocascin
was determined. Strained rumen fluid was centrifuged at 20,000 x g for 30 min. The
pellet was resolubilized in an equal volume of water as the initial volume of rumen
fluid. Resolubilized rumen fluid pellet, the rumen fluid supernatant, and azocascin
solution (.2 mg azocascin I ml water) were each combined with KOH at a final
concentration of IN. The absorbance was then determined at wavelengths ranging from
260 to 800 nm at intervals of 20 nm.

Effect of adding microbial growth inhibitor to azocascin degradation was
determined. Azocasein degradation was determined as described in the final method
except media was not reduced and sodium hydrazine (final concentration 2 mM) was
added to the inhibited treatment. The inhibitor was chosen because it has previously
been shown to inhibit microbial growth at the concentration used without affecting
enzymatic activity of proteases (Broderick, 1987). Each treatment was replicated three
times with separate batches of rumen fluid.

Two experiments compared timing between steps on absorbance of degraded

azocascin samples. In the first experiment, KOH was added immediately after

 

113

centrifugation of final preparation, or after soaking in TCA overnight. Each treatment
was replicated three times. In a second experiment, six samples were read for
absorbance immediately or after soaking in KOH solution overnight. A precipitate
appeared to form in the bottom of tubes soaked in KOH, so the optical density of these

samples was measured both after remixing or centrifugation.

Azocasein degradation: final method
In vitro media used in these experiments except where counter indicated was modified
from Goering and Van Soest (1970) by replacing NH4HCO3 with an equimolar
concentration of NaHC03. The media was diffused with CO2 gas before use until the
pH stabilized at 6.8, and was reduced with .6 g/L of each cysteine-HCI and Na28.
Azocasein (8 mg/ml) was solubilized in the in vitro media. Ruminal enzyme source (.5
m1) and media with azocascin (1.5 ml) were added to 15 m1 centrifuge tubes. Tubes
were incubated under C02 at 38°C for 2 h. The degradation was then arrested and
undigested protein precipitated with addition of ml TCA (240 gfkg water). Blank
samples were determined by incubating enzyme preparations without azocascin
solution, then adding azocasein immediately before addition of TCA. After addition of
TCA, samples were centrifuged (1500 x g, 20 min). An aliquot of supernatant (1 ml)
was transferred to another centrifuge tube and 2.5 N KOH (1 ml) was added. Optical
density (O.D.) was determined at 480 nm wavelength with a spectrophotometer
(Beckman DU 460, 1 cm path length). Optical density of the blank from each sample
was subtracted from that of the digested azocascin for that sample. This value was
compared to a standard curve derived from dilute azocascin solution in l N KOH. Each

step in the procedure was completed without delay to avoid loss of color over time due

to TCA or N aOH solutions.

 

 

 

 

114

Chilling, blending and rinsing

Rumen contents were collected from a fistulated Holstein dry cow fed twice daily a
ration based on maize (Zea mays L.) silage and luceme (Medicago sativa) hay. Rumen
contents were collected two to four hours after feeding and divided into 4 parts. Two
parts were chilled for 4 h and two were processed immediately. One part from each
initial treatment (chilled vs. not chilled) was blended and one was not. A114 parts were
strained through 4 layers of cheese cloth followed by glass wool. Cold (4°C), reduced
in vitro media as described for use in azocasein degradation was poured over the
filtrand, and the rinse collected and kept separate from the strained fluid that was
initially removed. Each treatment was thus divided into two additional fractions
(strained fluid and the rinse of contents). These eight samples were centrifuged
sequentially at 400 x g for 15 min and 20,000 x g for 30 min and each fraction
(supernatant, pellet l and pellet 2) analyzed for N and 2-H azocascin degradation. The
experiment was replicated 6 times using different rumen fluid samples from the same
cow on different days.

Results were analyzed in a split-plot design blocked by rumen fluid sample
(run). Chilling and blending treatments were nested within rumen fluid sample.
Rumen fluid and the rinse of contents were nested within chilling and blending
treatments. Centrifugation fractions (400 g pellet, 20,000 g pellet, supernatant) were
nested within rinse and fluid. The mean of PAIN determined in fractions is not
proportional to PAIN determined on the whole of a treatment. Therefore, three separate
ANOVA's were computed. First, total PAIN for each replicate (n = 6) of chilling and
blending treatments was determined by finding the sum of PA in the six fractions (2
rinse x 3 centrifugation fractions), and dividing that value by the sum of N in the six
fractions. Chilling, blending and their interaction effects were determined after

blocking by run using these total PAIN values. In a second ANOVA, PAIN was

 

 

 

 

115

determined for each replicate used in comparison of rumen fluid to rinse of solids.
Replicate values were calculated as the sum of PA from the three centrifugation
fractions divided by the sum of N. Comparison of rumen fluid to rinse and the
interactions of this contrast with chilling and blending were determined using the
calculated replicates, after blocking for run, chilling, blending and chilling x blending
interaction effects on the PAIN calculated the same way. In a third AN OVA, effects of
centrifugation fractions and all their interactions with chilling, blending and rinsing
were determined on the PAIN determined on the individual rumen fluid fractions.
Effects of chilling, blending, rinsing and their interactions and effect of run on PAIN
determined on fractions were removed from error. Interactive effects were elucidated
by computing the previously described contrasts for each rinsing and centrifugation
fraction rather than using the sum of fractions. For this and all subsequent experiments,

statistics were performed using JMP (SAS Institute Inc. 1991).

Incubation in vitro
Cheese cloth filtrate of rumen contents that was blended but not chilled (500 ml) as
described for the previous experiment was used. In vitro media as described for
azocascin degradation was reduced with .6 g/l Na2S.(H20)9. Carbohydrate (Table l)
and tryptic digest of casein (1.5 g/L) were added to select for proteolytic organisms
based on recommendations of Pichard (1977). Concentrations of carbohydrate chosen
to select for proteolytic organisms. The incubation was carried out in a one liter flask
with side arm, under diffusion of C02 and in a 38° C water bath. Media with
carbohydrates was added at constant rate of .07/h using a peristaltic pump to perfuse
media through glass tubing to the bottom of the flask. Effluent was collected in a
separate flask. The contents of the fermentation flask were continually stirred with a

magnetic bar. Incubation in continuous culture was continued for 24 h.

 

 

 

116

Both the incubate and the initial rumen fluid were subjected to sequential
centrifugation (400 x g and 20,000 x g, 30 min), and the proteolytic activity on
azocascin and N content of each fraction was determined. The experiment was
repeated on six different days (run). Comparison of initial rumen fluid with infusate
was made after blocking by run for each centrifugation fraction and for the total of the
three fractions.

Extraction in bufl'er, detergent or acetone
Different solvents were compared for extraction of enzyme from rumen contents.
Rumen fluid was collected and prepared as described for the previous experiment.
Strained rumen fluid (50 ml per sample) was cooled to 4 °C and then centrifuged at
20,000 x g for 30 min. These centrifugation pellets were resolubilized in 25 m1 reduced
media (previously described), reduced media with detergent (deoxycholate) or acetone.
Each sample with solvent was blended for 30 sec in a tissue homogenizer (Sorvall
Omni-mixer). Samples treated with acetone were centrifuged again (20,000 x g, 30
min) and acetone soluble fraction was discarded. This pellet was blended with 25 ml
reduced media. After resolubilization, all treatments were centrifuged (20,000 x g, 30

min).

RESULTS AND DISCUSSION
Validation and improvement of azocascin method
The ratio of azo dye to N solubilized was constant over time and the same as for the
sample not treated with TCA. This indicates that azo dye liberation is the same as N
liberation and represents true protein degradation, not merely the cleaving of the azo
dye from the protein. If azo dye was degraded by ruminal enzyme as has been recorded
for long-term degradation (Mahadevan, 1979), the ratio of dye to N solubilized would

decrease over time or differ from the undigested sample. The solubilization of azo dye

 

117

was linear (constant mg/ml/h) for up to 8 h. Since the results are positive, azocascin
degradation was used in several other experiments.

The absorbance of light due to rumen fluid pellet and supernatant, and to
azocascin is shown in figure 1. The wavelength chosen to measure azocasein
absorption for the remaining experiments was 480 nm because, though it was not the
wavelength of peak light absorption for azocascin (450 nm), it resulted in less
interference from light absorption of rumen fluid than the lower wavelength.

There was no effect of adding hydrazine on azocascin degradation (P > .1). The
mean degradation was 2.28 mg azocascin hydrolyzed/h/ml rumen fluid and the standard
error of the mean was .24. Therefore hydrazine was not added in subsequent
experiments.

Soaking in TCA overnight resulted in lower (P < .01) OD. than not soaking (.44
vs. 1.1 mg/h). Soaking in KOH overnight did not affect CD. but centrifugation at this
step did result in lower 0D. (1.2 vs 1.4 mg/h, P < .01). Because it was determined that
solubilized dye faded if left in TCA supernatant addition of KOH was completed
without delay and samples were mixed by gentle shaking before analysis (azocascin

fragments that are soluble in TCA but precipitable in KOH were considered degraded.

Chilling, blending and rinsing

Chilling of rumen contents before straining tended to increase PA recovered slightly,
but blending had no effect (Tables 2 to 3). Rinsing of rumen contents increased total
recovery of PA by 42% (Tables 2 to 3). Chilling increased PA in rumen fluid pellets,
but did not affect PA in rinse (Figure 2). This effect explains most of the interaction
with chilling and rinsing, and shows that rinsing rumen solids in cold buffer is nearly as
good as chilling before straining at recovering enzyme activity. The rinse of rumen
solids had higher PAIN than fluid (Tables 2 to 3), indicating that rinsing not only helps

in recovery of PA but that it can concentrate that activity. Chilling had no overall effect

118

on PAIN (Table 2), but chilling increased PAIN in the rinse of solids with no impact on
the fluid (Fig. 3). Blending decreased PAIN (Tables 2 to 3) and therefore is not
recommended as a treatment to concentrate proteolytic activity. There was also a
significant chilling by blending interaction on PAIN in the rinse fraction (Table 2).

Proteolytic activity differed for centrifugation fractions with most activity
recovered in pellets (Tables 2 to 3). Sequential cenuifugation separates rumen fluid
proteins into three fractions. The supematant contains free soluble protein. The 20,
000 x g pellet (pellet 1) contains free bacterial protein and the 400 x g pellet (pellet 2)
contains protozoa, insoluble feed protein and bound bacteria. Most proteolytic activity
was associated with the feed/protozoa and bacterial fractions and little enzyme appears
to have been soluble. These results confirm earlier work that showed most proteolytic
activity was insoluble (Kopecny and Wallace, 1982). Chilling increased PA in pellets
but not supernatant (Tables 4 to 5). Therefore, it did not appear to be such a harsh
treatment as to dislodge the enzymes from the bacteria. Much of the activity in the
rinse fraction was precipitated at 400 x g, while little was located in the bacterial
fraction of the rinse (Tables 4 to 5). This indicates that rinsing may not actually
dislodge free bacteria, but rather dislodge protozoa or feed with bound bacteria.
Proteolytic activity per unit N was greater in the 400 x g pellet and supernatant of the
rinse than for the strained fluid (Tables 4 to 5), thus confirming this hypothesis. Most
of this increase was found in the rinse from chilled samples (Figure 3).

The present work, as well as previous studies, demonstrates that rinsing of
rumen solids is effective at increasing both PA and PAIN. However, none of the effects
on concentration of proteolytic enzyme relative to crude protein would be adequate to
ensure excess enzyme for extended periods of incubation. Rinse of rumen solids would
be important to increase the percentage of activity recovered, and potentially this

recovery could impact how representative the recovered enzyme would be. Based on

119

these results, it was clear that both centrifugation pellets would be important to
recovery of enzyme, yet since the supernatant represented a small portion of the total

activity it could be removed.

Incubation in vitro
Results of incubation of rumen fluid in continuous culture are shown in Table 6.
Preincubation resulted in a build up of brown sludge in the bottom of the flask,
demonstrating that solids were not washed out of the flask. This fraction was
precipitated into the 500 x g pellet which contained a surprisingly high level of PA.
However, PAIN was lower in preincubated contents than for the original for this pellet.
The 20,000 x g pellet had lower PA in preincubate than rumen fluid. This would most
likely be explained by a dilution of microorganisms in continuous culture owing to the
high dilution rate relative to microbial growth. The PAIN was not different for the free
bacterial fraction indicating that the free bacteria of preincubate did not have greater
capacity to degrade protein than those bacteria in the original fluid. The PA in the
supernatant was also lower in preincubate than the original rumen contents, and PAIN
tended (P =.08) to be lower in the preincubate for this fraction. Preincubation had no
effect on the total PA or PAIN recovered in all three fractions. It was hypothesized that
the high concentration of starch and maltose and the high dilution rate would select for
proteolytic organisms. In addition, the dilution would remove non-microbial protein
from the fluid. It appears that both hypotheses were false.

Extraction in bufl’er, detergent or acetone

Total PA was higher when treated with deoxycholate, than for the original pellet, but
other treatments did not increase total PA (Table 7). Detergent may have altered
protein conformation to increase digestion. Since the rumen does not contain detergent,

it may not be a desired method to increase PA for in vitro studies. Mixing rumen fluid

 

 

120

precipitate with buffered media solubilized 18% of PA. Treatment with detergent or
acetone solubilized 51 and 48 % of PA respectively. Activity per mg N increased 3.9
times in supernatant compared to original sample after treatment with detergent or
acetone. These results demonstrate the potential of developing an in vitro system with
enzyme extracted from pelleted rumen fluid. Gentle extraction with buffer may be
advantageous because only cell wall proteases would be extracted and these are thought
to be primarily responsible for degradation in vivo (Kopecny and Wallace, 1983). On
the other hand, much more activity was recovered by extraction with acetone, so this
enzyme is potentially more representative of whole rumen contents. Previous work has
used enzyme recovered from acetone extraction for in vitro studies (Mahadevan, 1987).
Treatment with acetone or detergent seemed more effective than preincubating,
chilling, rinsing or blending of rumen contents at increasing proteolytic activity per unit

protein of rumen preparations.

 

 

 

121

REFERENCES

Broderick, G. A. (1987). Determination of protein degradation rates using a rumen in

vitro system containing inhibitors of microbial nitrogen metabolism. Br. J Nutr.
58: 463- 475.

Chamey, J. & Tomarelli, R. M. (1947). A colorimetric method for the determination of
the proteolytic activity of duodenal juice. J. Biol. Chem. 171: 501-505.

Craig, W. M., Hong B. J., Broderick G. A., & Bula, R. J. (1984). In vitro inoculum
enriched with particle-associated microorganisms for determining rates of fiber
digestion and protein degradation. J. Dairy Sci. 67: 2902

Furchtenicht, J. E. & Broderick, G. A. (1987). Effect of inoculum preparation and
dietary energy on microbial numbers and rumen protein degradation activity. J.
Dairy Sci. 70: 1404-1410.

Goering, H. K. & Van Soest, P. J. (1970). Forage fiber analyses (apparatus, reagents,

procedures, and some applications). Agric. Handbook No. 379. USDA-ARS.
Washington, DC.

Hungate, R. E. (1966). Chapter II: The rumen bacteria. The Rumen and Its Microbes.
New York: Academic Press.

Kohn, R. A. 1993. In vitro methods to determine protein degradation of feeds for
ruminants. Ph.D. Thesis, Michigan State University, East Lansing, MI.

Kohn, R. A. & M. S. Allen. (1993) An in vitro method to measure protein degradation
of feeds using enzymes extracted from rumen contents. J. Dairy Sci. 76: Suppl. 1

Kopecny, J. & Wallace, R. J. (1982). Cellular location and some properties of
proteolytic enzymes of rumen bacteria. Appl. Environ. Microbiol. 43: 1026-1033.

Lin, K. W., Patterson, J. A. & Ladisch, R. R. (1985) Anaerobic Fermentation: microbes
from ruminants. In: Enzyme and Microbial Technology 7: 98-107.

Mahadevan, S., Erfle J. D., & Sauer, F. D. (1979). A colorimetric method for the
determination of proteolytic degradation of feed proteins by rumen
microorganisms. J. Anim. Sci. 48: 947—953.

Mahadevan, S., Sauer F. D., & Erfle, J. D. (1987). Preparation of protease from mixed
rumen microorganism and its use for the in vitro determination of the
degradability of true protein in feedstuffs. Can. J . Anim. Sci. 67: 55.

Neutze, S. A., Smith R. L., and Forbes, W. A. (1993). Application of an inhibitor in
vitro method for estimating rumen degradation of feed protein. Animal Feed
Science and Technology 40: 251-265.

Pichard, G. R. (1977). In vitro rumen fermentation and the assessment of the nutritive
value of forages. Ph.D. Diss., Cornell University, Ithaca, NY

 

 

122

SAS Institute Inc. (1991). Statistical Visualization for the Macintosh. JMP ver. 2.02.
Cary, NC: SAS Institute Inc.

Wallace, R. J. & Kopecny, J. (1983). Breakdown of diazotized proteins and synthetic
substrates by rumen bacterial proteases.

 

123

 

Optical Density

4..

2-1

 

 

 

600 700 800 900
Wavelength

 

Fig. 1. Optical density of azocascin solution ( o ), centrifuged rumen fluid pellet ( D ),

and centrifuged rumen fluid supernatant ( 9 ) at varying wavelengths of light.

 

 

 

 

124

 

    

Strained Rlnee Strained Rlnee Strained Rlnee Strained Rlnee
Rumen oi Rumen of Rumen oi Rumen of
Fluid Solid. Fluid Sollde Fluid Sollde Fluid Solide

 

Blended Not Blended Blended Not Blended

 

 

Chilled Not Chilled

Azocasein degradation (mg degraded] hl ml)

Fig. 2. Treatment of rumen fluid on proteolytic activity per m1 rumen fluid (E5
supernatant, I 20,000 x g pellet, I 400 x g pellet). Ruminal contents were chilled for
4 h or analyzed immediately, blended for l min or not, and analyzed directly after
straining or solids were rinsed with an equal volume as rumen fluid of in vitro media
and the rinse analyzed. The mean of three replicates are shown. Chilling and blending
effects were not significant, strained > rinsed, 400 x g pellet > 20,000 x g pellet >
supernatant, chilling x centrifugation and rinse x centrifugation interactions were

significant, P = .01.

 

125

 

 

Strained Rinse Strained Rinse Strained Rinse Strained Rinse
Rumen 0f Rumen of Rumen of Rumen of
Fluid Solids Fluid Solids Fluid Solids Fluid Solids

Blended Not Blended Blended Not Blended

 

 

Chilled Not Chilled

’z‘6_
a:
E
35
\
8
.04
re
a
«1:3
'0
a)
£2
5
:1
tr:
'0
E
m0
o
'o
.E
o
en
re
0
o
N
<

Fig. 3. Treatment of rumen fluid on proteolytic activity per mg nitrogen in enzyme
preparation. Ruminal contents were chilled for 4 h or analyzed immediately, blended
for 1 min or not, and analyzed directly after straining or solids were rinsed with an
equal volume as rumen fluid of in vitro media and the rinse analyzed. The mean of

three replicates are shown. Chilling and blending effects were not significant, rinse >

strained, P = .05.

126

Table I. Carbohydrate composition of media used for incubation of ruminal organism

in continuous culture.

 

 

Carbohydrate Amount (glL)
Starch" 6.7 gIL
Cellulose" 5.3 g/L
Pectin 0.85 g/L
Maltose 0.5 gIL
Malic acid 0.2 g/L
Sucrose 0.2 g/L
Xylan 0.2 g/L
Citric acid 0.12 g/L
Cellobiose 0.1 g/L
Glucose 0.1 g/L
Succinic acid 0.08 E2

*Added separately in 2 parts at 12 h intervals.

127

Table 2. Treatment of rumen contents on proteolytic activity per ml rumen fluid (PA)
and per mg nitrogen (PAW).

 

 

Effect df PA (P) PAIN (P)
Run (rep) 5 .0001 .0001
Chilling (C) l .09 NS
Blending (B) 1 NS .01

C x B l .06 NS
Error 1 15 NS NS
Rinse vs. Fluid (R) 1 .0001 .03
R x C 1 .08 .03

R x B 1 NS NS
R x C x B 1 .07 .05
Error 2 20 NS NS
Centrifugation fraction (F) 2 .0001 NS

F x C 2 .01 NS

F x R 2 .001 .01

F x C x B 2 .06 .07

F x C x R 2 .0001 .08

F x B 2 NS NS

F x B x R 2 NS NS

F x C x B x R 2 NS NS

   

Errorl(df)=erep(5)+Bxrep(5)+Cxerep(5). Error2(df)=erep(S)+R
xerep(5)+Rxerep (5)+RxCxB xrep (5). Error3=allotherinteractions
with rep. Error 1 was tested against Error 2 and Error 2 against Error 3. Effects rep, C,
BandCwaeretested againstError 1. EffectsR,RxC,RxBandeCwaere
tested against pooled Errors 1 and 2. Effect F and its interactions were tested against
pooled Errors 1 to 3. NS = not significant at P> .1.

Table 3. Mean proteolytic activity per ml rumen fluid (PA) and per mg N in rumen

 

128

 

fluid (PAW) .
Main Effects PA PAIN
Chilling
Chilled 6.60 4.20
Not chilled 6.06 4. 2 6
Standard error of mean (11 = 12) .22 .1 1
Blending **
Blended 6.42 4.02
Not blended 6.30 4.50
Standard error of mean (n = 12) .22 .1 1
Fluid or rinse ** *
Rumen fluid 3.66 4.08
Rinse of solids 2.64 4.74
Standard error of mean (n = 24) .13 .20
Centrifugation pellet **
400 x g pellet 1.44 4.2
20,000 x g pellet .96 4.86
Supernatant .78 4.38
Standard error of mean (n = 48) .05 .03

 

Total activity and N recovered (sum of fluid and rinse for 3 centrifugation fractions) is
shown for chilled, not chilled, blended and unblended treatments. Total activity and N
of rumen fluid or rinse (sum of 3 centrifugation fractions) is shown for rumen fluid and
rinse of solids. Mean activity in centrifugation fractions (for rumen fluid and rinse) is

also indicated. *Indicates significant main effect at P = 05. **Indicates significant

main effect at P = .01.

 

129
Table 4. Effect of treatment of rumen contents (chilling, rinsing and blending) on
recovery of proteolytic activity per ml rumen fluid (PA) and per mg N (PA/N) in

fractions separated by sequential centrifugation.

 

 

 

PA (P) PAIN (P)
Effect Pellet l Pellet 2 Supern. Pellet l Pellet 2 Supern.
Chilling .001 .01 NS NS NS NS
Blending NS .08 NS .001 .01 NS
Chilling x blending NS NS .01 .001 NS .05
Rinse .01 .0001 .08 .01 NS .01
Rinse x chilling .0001 .01 .06 .0001 NS NS
Rinse x blending .05 NS NS .05 NS NS
R x C x B NS NS .08 .07 NS .07

 

Pellet 1 = 400 x g pellet, Pellet 2 = 20,000 x g pellet, Supern. = supernatant). Data

analyzed in a split plot design according to Tab. 1 and description in text.

 

130

Table 5. Mean proteolytic activity per ml rumen fluid (PA) and per mg N in rumen fluid
(PAflV) by centrifugation fraction.

 

 

 

 

PA PA/N
Effect pellet 1 pellet 2 supem. pellet 1 pellet 2 supem.
Chilling ** **
Chilled 3.18 2.10 1.38 3.96 4.86 3.84

Not chilled 2.70 1.50 1.86 4.02 4.86 4.2

SEM (n = 12) .08 .09 .15 .09 .22 .31
Blending *
Blended 2.94 1.8 1.68 3.78 4.38 4.02

Unblended 2.88 1.98 1.50 4.26 5.34 4.02

SEM (n = 12) .08 .09 .15 .09 .22 .31
Fluid or rinse ** ** *
Fluid 1.5 1.26 0.96 3.84 4.98 3.36
Rinse 1.38 .6 .66 4.56 4.68 5.46
SEM (n = 24) .035 .044 .10 .14 .26 .52

 

Pellet 1 = 400 x g pellet, Pellet 2 = 20,000 x g pellet, Supern. = supernatant. Total
activity and N recovered (sum of fluid and rinse) is shown for chilled, not chilled,
blended and not blended treatments. Mean activity in for rumen fluid and rinse is also

given. * Indicates significant main effect at P = .05. **Indicates significant main

effect at P = .01.

13 1
Table 6. Efi'ect of preincubation in continuous culture at dilution rate of .07/h on
proteolytic activity per ml rumen fluid (PA) and per mg N in rumen fluid (PAW) by

cenmficgation fraction.

 

PA (mglhlml) PAIN (mg/h/mg N)
Treatment Pellet 1 Pellet 2 Supern. Total Pellet l Pellet 2 Supern. Total
Effect * it * ¢
Preincubated 1.92 .54 .42 2.88 3.66 4.62 3.30 4.32
Rumen fluid 1.26 1.02 1.56 3.84 5.10 4.14 5.16 4.56
SEM (n=6) .14 .07 .25 .39 .36 .42 .61 .36
Pellet l = 400 x g pellet, Pellet 2 = 20,000 x g pellet, Supern. = supernatant. "' Indicates

significant main effect at P = .05. ”Indicates significant main effect at P = .01.

132

Table 7. Efl'ect of solubilizing rumen fluid precipitate in reduced media, media with
detergent, and acetone on proteolytic activity per ml rumen fluid (PA) and per mg N
(PAW).

 

 

 

 

PA (mg/hlml) PAIN (mghlmg N)
Solvent Pellet Supern. Total Pellet Supern. Total
Media 1.26ti .28c 1.5b 3.783 7.32b 4.1413
Detergent 1.02b 1.083 2.1a 3.63 15.12a 5.943
Acetone .72c 0.66b 1.36b 2.46b 15.183 4.08ID
Untreated 4- - 1.36b -- - 3.9b
SBM .031 .029 .064 .124 1.164 .237

After treatment in acetone, samples were centrifuged and pellet re-suspended in media.
This and other treatments were centrifuged at 20,000 x g before determination of PA on
azocasein and N. Different superscripts within a column indicate significant

differences among treatments (P=.01). SEM = standard error of mean, 11 = 4.

 

133

CHAPTER 5: AN IN VITRO METHOD TO MEASURE PROTEIN
DEGRADATION OF FEEDS USING ENZYMES EXTRACTED FROM RUMEN
CONTENTS.

Synapsis
A method was developed to measure protein degradation of feeds using proteolytic

enzymes extracted from rumen contents. Strained rumen fluid (1.01) was centrifuged

 

and the pellet was extracted with butanol and acetone. Enzymes were then solubilized
from the residue in water, and frozen at ~25°C for later use. Feed protein was degraded

by incubating feeds with enzyme at 38°C. At the end of degradation, trichloroacetic

 

acid (15 m1; 240 gIl) was added to precipitate undigested protein. The final enzyme
preparation maintained 62% of the proteolytic activity on azocasein of the original
rumen fluid (plus rinse of solids). When enzyme was incubated at 38°C, activity on
azocascin decreased slightly for 12 hours and then increased. There was no difference
between 3, 5 or 10 m1 of enzyme on degradation of soya bean meal or luceme hay
protein at 6 or 16 h. This suggests that enzyme was supplied in excess for the duration
of the incubation. Soya bean meal protein degraded at .15/h for the first 2 h and slowed
to .01/h from 8 to 24 h. Similarly, luceme hay protein degraded at .06/h for the first 2 h
and then slowed to .01/h. The proposed method is simple with low variation within
treatment, and it uses enzymes from the rumen in concentrations that are maintained in
excess of substrate for extended periods of incubation.

Introduction
Current systems to balance dairy cattle rations emphasize the need to consider ruminal

protein degradation of feeds (V érité, 1979; Agricultural Research Council, 1984;

 

134

Madsen, 1985; National Research Council, 1989). Methods to evaluate ruminal protein
degradation have limitations which restrict their use as has been reviewed (Nocek,
1988; Michalet—Doreau and Ouldbah, 1992). In vivo methods are time consuming and
expensive, and the lack of adequate microbial and passage markers results in problems
of accuracy and precision. In situ methods are the most commonly used procedures to
measure protein degradation. Feeds are suspended in bags placed in the rumen of
fistulated cows or sheep for differing durations of time. The residue left in the bag after
in situ degradation is analyzed for nitrogen (N). Microbial contamination in the bag is
occasionally determined by using a microbial marker such as nucleic acids or diamino
pimelic acid (DAPA) which are associated with microbial matter but assumed to be
digested from the original feed. In situ methods actually measure protein which does
not pass out of the nylon bags in the rumen. Once a protein is solubilized or even
reduced adequately in particle size to leave the bag it is considered degraded. Much of
the protein of feeds leaves nylon bags before being completely degraded when
incubated in vitro (Spencer et al. 1988).

Various authors have developed methods to measure protein degradation that
use proteolytic enzymes. Two of the most commonly used enzymes are ficin (Poos-
Floyd et al. 1985) and endo-protease from S. teptomyces griseus (Krishnamoorthy et al.
1983). Ficin is a cysteine protease (Barrett, 1985) as are many of the proteases found
in the rumen (Brock et al. 1982). However, some authors have criticized the use of
plant-derived proteases to estimate degradation of protein sources by microbial
proteases. On the other hand, S. griseus protease demonstrates maximal activity at pH
8 (Krishnamoorthy et al. 1983) which is not physiological. A newer method uses
neutral protease of bacterial origin (Assoumani et al. 1992). Other in vitro methods use
enzymes from ruminal organisms. While one approach is to incubate feeds with

microbes in the presence of drugs that inhibit microbial growth (Broderick, 1987;

135

Neutze et al. 1993), another is to extract enzymes from ruminal organisms (Mahadevan
et al. 1987).

These enzymatic methods have two major limitations. The enzyme activity
declines over time as the proteolytic enzymes degrade themselves (Broderick, 1987;
Mahadevan et al. 1987). Therefore, these methods are only appropriate for short-term
degradation (1-2 b). However, feeds may be comprised of multiple protein fractions
that degrade at different rates. The total feed protein may initially degrade fast and then
slow due to disappearance of the fast fraction. Because of this, methods are required
that measure protein degradation over extended periods of time. A second limitation is
that first order degradation kinetics do not apply to these systems because the level of
enzyme activity limits the degradation velocity (Mahadevan et al. 1987; Broderick and
Clayton, 1992). Though these methods are used to detect some relative differences
among feeds, they have not been adequate to determine fractional degradation rates
because rate of degradation under enzyme-limiting conditions depends on feed sample
size, types of enzymes and level of activity.

The objective of this work was to develop and validate a method to measure
protein degradation using enzymes extracted from rumen contents, and to maintain
activity over time, so that enzyme activity would be supplied in excess for extended

periods of incubation.

MATERIALS AND METHODS
Solutions and feeds

Buffered media used in these experiments (except where indicated otherwise) was
modified from Goering and Van Soest (1970) by replacing NH4HCO3 with an
equimolar concentration of NaHCO3. The media was diffused with C02 gas before

use until pH stabilized at 6.8. At this time, media was reduced by addition of 1.2 gIl of

136

each Na28.(1120)9 and L-cysteine.HCl.I-120 (Sigma Chemical Co., St. Louis, MO,
USA). The media was maintained under 1 atm C02 to prevent oxidation or an increase
in pH due to loss of C02.

. Feeds used in development of procedures and their respective percentage crude
protein content on a dry matter basis were: solvent extracted soya bean (Glycine max)
meal - 48.0, luceme (Medicago sativa) hay at one half bloom -- 19.1, dried and ground
maize (Zea mays L.) grain -- 10.8, and dried meat and bone meal -- 55.2. Each feed
was ground through a 1 mm screen of a Wiley mill.

Butanol-acetone extraction
The method used to extract enzyme from rumen contents was modified from that of
Mahadevan et al. (1987) based on observations recorded elsewhere [Appendix] (Kohn,
1993). Rumen fluid (5 l) was collected from 2 fistulated cows 24 h after feeding. The
cows were fed a diet based on luceme hay and maize silage twice daily. Rumen fluid
was blended for 1 min and strained through cheese cloth, though results of another
study showed that rumen contents should not be blended as blending decreased
proteolytic activity per mg N [Chapter 4] (Kohn and Allen, Submitted). The solids
were twice rinsed with reduced in vitro media (1.25 l), and this was added to the
strained rumen fluid. This mixture was stored at 38°C under C02 until it could be
processed (0-5 h).

The following extraction procedure was repeated up to six times within a day.
Rumen fluid (1.0 l) was centrifuged for 30 min at 4°C and 13,000 x g. It was decanted
with care taken to keep loose sediment with pellet. Water was added to pellets to bring
the total volume up to 250 ml. Equal volume cold (4°C) butanol was added to pellets
and mixed for 1 h using a magnetic stir bar over an ice-water bath. After an hour, this
mixture was added intermittently for 45 min to acetone (1.01) which was also stirred

with a magnetic bar over an ice-water bath. The acetone was stirred for an additional

 

137

15 min and then filtered through Whatman # 54 filter paper in a Buchner funnel with
suction. The filtrand was rinsed with two 500 m1 aliquots of cold acetone and dried in
the funnel. Each pellet was then stored overnight in a refrigerated desiccator. The
following day, pellets were each resolubilized in 150 ml distilled water in centrifuge

bottles. The bottles were shaken vigorously to wet the sample, and were placed in a

shaking ice-water bath for one hour. The fluid was centrifuged for 30 min at 26,000 x g

and 4°C, and decanted while keeping loose sediment with supernatant. The loose
sediment was removed later by centrifugation before using the rumen extract. The
supernatant contained enzyme activity and was frozen at -25°C. After decanting and
saving the supernatant, an additional 100 m1 of water (4°C) was added to the pellet, the

bottle shaken again for 1 h and centrifugation and decanting repeated.

Feed protein degradation
Feed samples containing 8 mg nitrogen were weighed into 50 ml plastic centrifuge
tubes. Tubes were placed under C02 before adding in vitro media (25 m1). Reduced
media was added because previous work showed that proteins may precipitate in the
presence non—reduced media, and this could alter degradation (Kohn and Allen, 1992).
Enzyme preparation was thawed, centrifuged (30 min, 26,000 x g), and decanted with
exclusion of loose sediment before adding supernatant to tubes (6 ml) and incubating at
38°C under C02. Trichloroacetic acid ( 15 ml, 24%) was added to end the degradation
and precipitate undigested protein. Samples were centrifuged at 1500 x g for 20 min
and the supernatant discarded. A modified Kjeldahl method (Hach et al. 1985) was

used to determine nitrogen content in the pellet, and undegraded protein was

determined as nitrogen times 6.25.

 

 

 

138

Reducing agents

Proteolytic activity was determined on azocasein as described [Chapter 4]
(Kohn and Allen, Submitted) except for addition of reducing agents. A control with no
reduction was compared to two levels of cysteine.HCl, cystine, sodium sulphide (.6 g/l
or 1.2 g/l) and dithiothreitol (2 or 4 mM). The experiment was repeated on two
different days with two replicates of each treatment on each day.

Effect of reduction on soya bean meal degradation was determined. Soya bean
meal was degraded for 16 h with extracted enzyme preparation as described previously.
Treatments included: 1. media that was neutralized with HCl but not saturated with
C02 (aerobic), 2. media that was equilibrated with C02 but not reduced (anaerobic),
3. media that was reduced with 1.2 g/l sodium sulphide and incubated with constant
perfusion of 1 atm C02 (open), and 4. media was reduced with 1.2 g/l sodium sulphide
and maintained in stoppered centrifuge tubes under C02 so that sulphide could not
escape (closed). Results were compared using the Student t test. All statistics in these

studies were performed with J MP (SAS Inst. 1992).

Cellulase and amylase
One microliter alpha-amylase (T ermamyl, Novo Nordisk Bioindustn'als, Inc., Danbury,
CT, USA) was added to maize grain samples incubated with enzyme as described. This
amount of amylase was estimated to be adequate to maximize fractional degradation
rate of starch. Commercially-produced cellulase from Trichiderma virde (Sigma
Chemical Co., St. Louis, MO, USA) was added to samples containing protease
preparation. There were no differences in protein degradation with cellulase at 16 h
because the cellulase was minimally active at pH 6.8, and was rapidly degraded by

protease. Thereafter, a two-step procedure was used for degradation with cellulase.

139

. Degradation with cellulase was adapted from Jones and Hayward (1973). A
citrate-phosphate buffer was prepared (.133 M Citric acid, .047 M dibasic sodium
phosphate and .065 M sodium chloride), equilibrated with 1 atm C02, and adjusted to
pH 4.6 with sodium hydroxide or hydrochloric acid. In the first step, luceme hay was
incubated with 10 ml citrate-phosphate buffer with cellulase from T. virde (50
rug/sample; Sigma Chemical Co., St. Louis, MO). In the second step, 15 ml
bicarbonate buffer was added to samples to raise the pH to 6.75 (+I- .01). Bicarbonate
buffer (.12 M sodium bicarbonate, .21 ml micro-mineral solution (Goering and Van
Soest, 1970), 1.0 mM magnesium chloride) was prepared by equilibrating with 1 atm
gaseous C02, reducing with 2 g/l of each cysteine.HCl.H20 and Na28.9H20 and then
adding 5 N KOH (9 mill). Enzyme preparation extracted from rumen contents (6 ml)
was added to neutralized samples. Several variations were compared to determine the
effect of cellulase pretreatment on protein degradation by extracted enzyme. Effect of
cellulase pretreatment for 4 or 16 h was determined with or without degradation with
rumen enzyme for 4 or 16 h respectively. These were compared to samples with no
pretreatment or that were soaked in pH 4.6 buffer without cellulase for 4 or 16 h before
degradation with rumen enzyme. Nitrogen was determined after precipitation with
TCA as described, and dry matter and nitrogen were determined on additional samples
that were filtered with suction through Whatman # 54 filter paper after degradation
without addition of TCA. Each treatment was replicated 3 times. Results were
analyzed statistically using AN OVA to compare treatment effects by time and time

effects by treatment for each response variable.

Proteolytic activity over time
The enzyme preparation (.25 mll tube) was dispensed to 15 ml centrifuge tubes.
Reduced in vitro media (.75 ml) was added under C02, and samples were stoppered

and frozen at -25°C. Samples were removed from the freezer and incubated at 38°C for

140

0, l, 2, 3, 4, 8, 12, 16, or 24 h before addition of reduced azocascin solution (12 mg, 1
m1). Samples were removed from the freezer at different times and all treatments were
combined with azocascin at the same time. Two-hour azocascin degradation was
completed twice using different enzyme preparations with two replicates per run as
described [Chapter 4] (Kohn and Allen, Submitted).

Efl'ect of enzyme level

Soya bean meal and luceme hay were subjected to incubation with 0, l, 3, 5 or 10 ml of
enzyme preparation. Protein degradation at 16 h was determined as previously
described. Three replicates of each treatment were used.

A different experiment was to test whether enzyme activity remained in excess
for the duration of the experiment. Soya bean meal was incubated for 0, 8 and 16 h
with 5 and 10 m1 enzyme. An additional treatment was soya bean meal incubated with
5m] enzyme for 8 h, and then 5 ml more enzyme added, followed by further incubation
for 8 h. Feed protein degradation was determined as described previously except
enzyme preparation was not centrifuged before use because importance of this step had

not been determined. The experiment was repeated twice.

Comparison of feeds and variation of method

Degradation of soya bean meal and luceme hay was determined at 0, 2, 4, 8, 16, 24, 32
and 48 h using the method described for protein degradation of feeds. The experiment
was repeated twice with rumen fluid from different days. Rate of protein degradation
was determined as the natural log of the fraction degraded in a time interval divided by
that time interval.

Four feeds described previously (soya bean meal, luceme hay, maize grain and
meat and bone meal) were compared for protein degradation using the final method
described previously. Each feed was digested for 6 and 16 h. The experiment was

141

repeated three times using rumen fluid extracted on different days in order to evaluate
day to dayvariation. Each treatment was replicated three times each day in order to

evaluate within day and day x treatment variation.

RESULTS AND DISCUSSION
Butanol-acetone extraction
After extraction with butanol and acetone, 62 % (SD. = 10, n = 5) of the rumen fluid
proteolytic activity on azocascin was solubilized in water. Activity was either
recovered in the first (75 %, SD. = 5, n = 17) or second extraction with water (25 %,
SD. = 5, n = 17), with little recovery in the third extraction. Therefore, the standard
procedure was to use only two extractions with water. A second centrifugation at
26,000 x g on the extracted rumen enzyme reduced TCA precipitable nitrogen in the
enzyme preparation below detectable levels without any reduction in proteolytic
activity. This enzyme preparation therefore enabled determination of feed protein
degradation without interference in measurement from proteins in the enzyme
preparation.
A different study showed that 50% of proteolytic activity could be recovered in
one extraction with buffer after extracting rumen fluid solids with acetone [Chapter 4]
(Kohn and Allen, submitted). Extraction in acetone is somewhat easier than butanol-
acetone but does not remove pigments from the final solution, and therefore would
interfere with colorimeuic measurement of protein. However, it could be sufficient for
determinations based on N. A different method enabled recovery of 30% of proteolytic
activity from rumen fluid after extraction in butanol-acetone (Mahadevan et al. 1987).
One reason the method used to extract protease in the present study resulted in higher
recovery of activity compared to the previous method is because previously enzyme

preparation was filtered which results in a loss of some activity. Contamination from

142

protein in the enzyme preparation was a limitation of the previous work. This problem
was eliminated with the additional centrifugation step, and final separation with TCA.
The proposed method of extraction is also less labor intensive than the previous
method.

Reducing agents

Since some effects of addition of chemical reducing agents on degradation of
azocascin were inconsistent across runs, each run was analyzed separately. In both
runs, addition of cysteine.HCLH20 increased proteolytic activity. Comparison of level
of cysteine using student t test showed that 1.2 mg/ml resulted in more protein
degradation than .6 mg/ml in both runs. Sodium sulphide addition resulted in a slight
increase in PA over control, which was significant only for the lower level in one run.
Sulfide (H28) is easily lost from the system by diffusion into the C02 head space of
tubes. This may explain its inconsistent behavior. Dithiothreitol (DTT) and cystine
had no effect on protein degradation in either run. Cystine was added as a negative
control in that it is not a reducing agent but is in fact the oxidized form of cysteine.HCl.

The dramatic results shown in Table 1 indicate that the enzyme extracted from
the rumen clearly benefits from addition of reducing agents. Simply excluding oxygen
from the media did not adequately reduce it. Additional experiments demonstrated the
importance of reducing agents on luceme hay degradation, and there was no difference
between sulphide, cysteine or a combination thereof on degradation of soya bean meal
protein [Appendix] (Kohn, 1993). Based on these results, 1.2 gll of each reducing
agent (Na28 and cysteine.HCL) were added to media after equilibrating with 1 atm
C02, and media was maintained under a 1 atm C02 environment with care taken to

limit the exchange of gases to prevent evolution of H28.

 

 

143

Proteolytic activity over time
The 2-h degradation of azocascin (proteolytic activity) decreased when rumen fluid was
first incubated without azocascin. This decrease was at a rate of .049Ih for 0 to 12 h
(Figure 1). Surprisingly, the proteolytic activity at 24 h was much greater than the
initial level. When this experiment was repeated to verify these results, proteolytic
activity was variable over time and degradation in activity was not significant. Again
degradation at 24 h was as high as in the data shown. It appears that the enzymes may
be auto-activated by incubation in vitro (Lehninger, 1975). The slow rate of
degradation of enzyme (or no degradation) and the possibility for auto-activation is
promising in that enzyme preparations may be viable for degradation of feeds over

extended periods of time.

Cellulase and amylase

Addition of amylase had no effect on protein degradation of dried maize grain at 16 h.
The volume of maize grain was visibly reduced in all treatments, and the residue after
degradation appeared to be a gummy substance. Enzymes extracted from rumen
contents using a similar method demonstrated the presence of amylolytic activity
(Mahadevan et al. 1987). Previous work has demonstrated that addition of amylase
increases protein degradation due to treatment with commercially available protease
(Assoumani et al. 1992). In the present study, there may have been no effect of
amylase addition because starch was adequately removed by extracted enzyme that it
did not interfere with protein degradation.

Cellulase pretreatment decreased protein degradation of luceme hay with
protease slightly at 4 h, and increased it dramatically at 16 h (Table 2). The decrease at
4 h may have been caused by an incorrect estimate of feed residual protein due to

contamination by cellulase. Blank samples containing cellulase and enzyme but no

 

 

 

144

feed were incubated for 4 or 16 h, and the residual protein subtracted from those with
feed. The contaminating protein may have degraded more slowly with feed samples
than with blanks due to the additional protein occupying sites on the enzymes. By 16 h,
cellulase protein was not detectable in the blanks. The dramatic increase in protein
degradation after 16 h in protease when preheating with cellulase indicates that fiber
may interfere with protein degradation especially during long-term degradation of
forages. There was a slight degradation of protease over time when incubated only with
cellulase. This could be due to proteolytic activity of cellulase, contaminants in the
enzyme preparation, or the plant's own proteolytic enzymes. The fact that pretreatment
in pH 4.7 buffer before proteolytic degradation does not increase protein degradation
over protease alone suggests that the increased activity with cellulase is due to the
cellulase and not to presoaking.

Nitrogen solubility in buffer (pH 6.8) determined by filtering was greater than
nitrogen solubility in TCA across treatments (Tables 2 to 3). This shows that T CA
precipitation would be expected to provide a lesser value for protein degradation than in
vitro or in situ methods that rely on solubility. Correlation between solubility of
nitrogen in TCA and buffer was .73 and between nitrogen solubility in TCA and DM
solubility was .66. The highest correlation was between DM and nitrogen solubility in
buffer (.85). Dry matter (DM) degradation was greatest for samples pretreated with
cellulase followed by protease degradation compared to just cellulase 0r protease (Table
4). Though each enzyme preparation resulted in significant DM disappearance, the
combination was most effective. These results suggest that a substantial amount of
protein in forages may be protected from degradation by fiber. Addition of cellulase

may be important to the measurement of the long-term protein degradation of forages.

 

 

 

 

 

145

Efl'ect of enzyme level

There was no difference in degradation of soya bean meal or luceme hay at 6 or
16 h when 3, 5 or 10 ml of enzyme preparation was used. There was a reduction in
protein degradation when 1 ml enzyme was used compared to 3 ml. Subsequent
experiments using this enzyme preparation used 6 ml of enzyme in order to use at least
twice the amount needed to supply enzyme in excess.

Table 5 shows the results of adding additional enzyme at 8 h to soya bean meal
that had already been degraded for 8 h. There was a slight decrease in protein
degradation by adding half the enzyme after 8 h degradation compared to adding all the
enzyme initially. There was no difference between adding 5 ml or 10 ml of enzyme for
degradation at either 8 or 16 b. If enzyme activity decreases overtime and is limiting to
degradation of feeds, we would expect that addition of 5 ml enzyme after 8 h
degradation would result in greater degradation at 16 h than addition of all enzyme at
the start. Since this increase was not observed, it appears that enzyme activity was in
excess for the duration of the incubation. The lower degradation at 6 and 16 h as
compared to other experiments in this paper appears to have resulted from an
underestimation of crude protein in blanks, since blank nitrogen was much higher
because the enzyme preparation was not centrifuged before use. Another experiment
demonstrated the importance of centrifuging the enzyme preparation before use on

measurement of protein degradation [Appendix] (Kohn, 1993).

Comparison of feeds and variation of method
Protein degradation of soya bean meal and luceme hay averaged .15/h and .06/h
respectively for the first 2 h of incubation (Figure 2). These rates slowed to .01/h from
8 to 16 h. These results emphasize the importance of long-term degradation of feeds to
determine protein degradation. Prediction of protein degradation at extended periods of

time from initial rates may overestimate protein degradation.

146

The results of degrading four protein sources using enzyme are shown in table
6. There was no difference in enzyme preparation obtained on different days for either
point in time. However, there was a significant enzyme x run interaction at 16 h (P =
.02). When effect of enzyme was determined by ANOVA on each feed, the enzyme
effect was only significant for dried maize grain at 16 h (P = .04). Average degradation
of maize grain at 16 h was 20.3, 23.8 and 28.8 % for each the three runs. Since maize
grain was lowest in percentage protein the greatest quantity was used on a dry matter
basis (since an equivalent amount of N was used for each feed). This may have resulted
in the lower .degradation of maize than the other feeds and increased variation from run
to run for maize. Overall, the low variation within and across runs is promising.

The actual degradation of each feed protein was less than expected, especially at
16 h. The fact that these values may be lower than has been determined for similar
feeds in situ (National Research Council, 1989) is not surprising because the in situ
procedure would overestimate protein degradation as particles leave the bag prior to
complete digestion. The degradation in the present study may also be lower than that
observed for feeds degraded with rurrrinal enzyme in the presence of inhibitors
(Broderick, 1987). The inhibitor system uses only short-term degradation with
extrapolation out to longer time points. The short-term (2 h) degradation of luceme hay
and soya bean meal was similar in the present study to that measured using the inhibitor
system (Broderick, 1987). These same feeds were also degraded in vitro in a system
employing sulphur-35 (358) to measure microbial growth [Chapter 3] (Kohn, 1993).
The long-term degradation using the 353 system was higher for all feeds than in the
present work. The system using 358 or other markers may also overestimate protein
degradation because the free bacteria fraction separated with sequential centrifugation
may be contaminated with feed protein from chloroplasts or mitochondria (Pichard,

1977). The fact that cellulase was not present in the enzyme preparation may explain

147

additional decrease for luceme. Other enzymes may also have been selected against or
inactivated by extraction, or enzyme associations with microbes, may influence the
degradation of some proteins. Enzymes that are ATP-dependent would be inactive in
extracted preparations, and such proteases have been identified in E. coli (Matin, 1992).

CONCLUSION
Extraction of enzymes from rumen fluid provides a means to concentrate enzyme
activity for degradation of feeds in vitro. This enzyme preparation seems to maintain
its activity for extended incubations (> 16 h), but degradation of feeds slows markedly
with longer incubation. Interaction of cellulase with protein degradation indicates that
protein degradation may require more than the presence of protease for degradation
Further research is needed to characterize this enzyme activity in the crude extract on
different types of feed proteins, and other structural components that may interfere

with protein degradation.

 

 

 

 

148

REFERENCES

Agricultural Research Council (1984). The Nutrient Requirements of Ruminant
Livestock, Suppl. 1. Slough: Commonwealth Agriculture Bureaux.

Assoumani, M. B., Vedeau F., Jacquot L., and Sniffen, C. J. (1992). Refinement of an
enzymatic method for estimating the theoretical degradability of proteins in
feedstuffs for ruminants. Anim. Feed Sci. Technol. 39: 3-4 357-368.

Barrett, A. J. (1985). Chapter 1: The classes of proteolytic enzymes. Plant Proteolytic
Enzymes. 1: 1-16.

Brock, F. M., Forsberg C. W., and Buchanan-Smith, J. G. (1982). Proteolytic activity of
rumen microorganisms and effects of proteinase inhibitors. Appl. Environ.
Microbiol. 44: $61-$69.

Broderick, G. A. (1987). Determination of protein degradation rates using a rumen in
vitro system containing inhibitors of microbial nitrogen metabolism. Br. J. Nutr.
58: 463-475.

Broderick, G. A. and Clayton, M. K. (1992). Rumen protein degradation rates estimated
by non- linear regression analysis of Michaelis—Menten in vitro data. Br. J. Nutr.
67: 27 -42.

Goering, H. K. and Van Soest, P. J. (1970). Forage fiber analyses (apparatus, reagents,
procedures, and some applications). Agric. Handbook No. 379. USDA-ARS,

Washington, DC.

Hach, C. H., Brayton S. V., and Kopelove, A. B. (1985). A powerful Kjeldahl nitrogen
method using peroxymonosulfuric acid. J. Agric. Food Chem. 33: 1117 .

Jones, D. I. H. and Hayward, M. V. (1973). A cellulase digestion technique for
predicting the dry matter digestibility of grasses. J. Sci. Food Agric. 24: 1419-
1426.

Kohn, R. A. (1993). In vitro methods to determine protein degradation of feeds for
ruminants. Ph.D. Diss., Michigan State University, East Lansing, MI.

Kohn, R. A. and Allen, M. S. (1992). Storage of fresh and ensiled forages by freezing
affects fibre and crude protein fractions. J. Sci. Food Agric. 58: 215-220.

Kohn, R. A. & M. S. Allen. Enrichment of proteolytic activity in preparations from the
rumen for in vitro studies. Br. J. Nutr. Submitted.

Krishnamoorthy, U., Sniffen C. J., Stern M. D., and Van Soest, P. J. (1983). Evaluation
of a mathematical model of rumen digestion and an in vitro simulation of rumen
proteolysis to estimate the rumen-undegraded nitrogen content of feedstuffs. Br. J.
Nutr. 50: SSS-568.

Lehninger, A. L. (1975). Biochemistry. The Molecular Basis of Cell Structure and
Function. 2nd edition. Worth Publishers, Inc. New York.

 

 

 

149

Madsen, J. (1985). Protein evaluation for ruminants. Acta Agric. Scand. Suppl. 25: 9-

Mahadevan, S., Sauer F. D., and Erfle, J. D. (1987). Preparation of protease from mixed
rumen microorganism and its use for the in vitro determination of the
degradability of true protein in feedstuffs. Can. .1. Anim. Sci. 67: 55.

Matin, A. (1992). Physiology, molecular biology and applications of the bacterial
starvation response. Journal of Applied Bacteriology 73: S49—S57.

Michalet-Doreau, B. and Ouldbah, M. Y. (1992). In vitro and in sacco methods for the
estimation of dietary nitrogen degradability in the rumen - a review. Animal Feed

Science and Technology 40: 57-86.

National Research Council (1989). Nutrient Requirements of Dairy Cattle. 6th ed.,
Washington, DC: National Academy Press.

Neutze, S. A., Smith R. L., and Forbes, W. A. (1993). Application of an inhibitor in
vitro method for estimating rumen degradation of feed protein. Animal Feed
Science and Technology 40: 251-265.

Pichard, G. R. (1976). In vitro rumen fermentation and the assessment of the nutritive
value of forages. Ph.D. Thesis, Cornell University, Ithaca, NY.

Poos-Floyd, M., Klopfenstein T., and Britten, R. A. (1985). Evaluation of laboratory
techniques for predicting ruminal protein degradation. J. Dairy Sci. 68: 829-839.

Spencer, D., Higgins T. J. V., Freer M., Dove H., and Coombe, J. B. (1988).
Monitoring the fate of dietary proteins in rumen fluid using gel electrophoresis.
Br. J. Nutr. 60: 241—247.

Vérité, R. (1979). Appreciation of the nitrogen value of feeds for ruminants. Pages 87-
96in Standardization of Pages Analytical Methodology for Feeds. Eds. W. J.

Pigden, C. C. Balch and M. Graham.

 

 

.nfl

  
    

 

 

E
.2
an
55.
OD
865
.E'E'
5932
O\
000
at;
Q
a
0 I I I I I
0 5 10 15 20 25

Hours pre-incubated

Figure 1. Effect of incubating rumen enzyme preparation at 38°C on proteolytic

activity on azocascin. Error bars indicate standard error of the mean for points

indicated; 11 = 2.

 

 

 

 

A [
§ 8'0
"U
'8 1
£3 60.
00
d.)
"U
:1
s 40.
.E
8.
g 20.
0 l I 1 I I
0 10 20 30 40 50

Time (h)

Figure. 2. Degradation of soya bean meal (—- El 4) and luceme hay (-— O --) with

enzyme extracted from rumen fluid. Standard error of the mean for soya bean meal =

1.9 and for luceme hay = 2.6; n = 2.

 

 

152

Table 1. Efi‘ect of reduction of in vitro media on soya bean meal (SBM) degradation at

16 h in enzyme extracted from the rumen.

 

 

Media SBM degradation (%)
Aerobic 19.4a
Anaerobic (C02) 21.5a
Reduced (open) 48.0b
Reduced (closed) 45.4b

 

Media was reduced with 1.2 g/l Nazs and maintained under diffusion of C02 (open) or

stoppered under C02 (closed). aDifferent superscripts indicate differences among

means, it = 2, P < .05).

 

153

Table 2. Solubility of nitrogen from lucerne hay in trichloroacetic acid as afi‘ected by
incubation with protease extracted from the rumen (P), commercial cellulose (C),

soaking in pH 4. 7 bufi‘er before P, and C before P.

 

 

 

Time effect (P)
Treatment 0h 4h 16h 0v.4h 4v.16h
Protease (P) 27.8 37.3til 42.4at .0005 .01
Cellulase (C) 27.8 275') 32.913 NS .004
Buffer, P 27.8 37.03 40.94b .003 .09
C, P 27.3 31.2e 51.2e NS .002

SEM .89 1.09 2.47 -- ..-

 

SBM = standard error of mean, 11 = 3. Different superscripts within a column indicate

differences among means (P < .05).

 

154

Table 3. Solubility of nitrogen from luceme hay in bufi‘er (pH 6. 8) as affected by
incubation with protease extracted from the rumen (P), commercial cellulose (C),

soaking in pH 4. 7 bufier before P, and C before P.

 

 

Time effect (P)
Treatment 0h 4h 16h 0v. 4h 4v.16h
Protease (P) 38.9 45.5a 52.221?!) .004 .008
Cellulase (C) 38.9 37.2b 43.43 NS NS
Buffer, P 38.9 46.13 seat .0002 .001
C, P 38.9 58.90 629‘) .0001 .08

SEM ' .59 .55 2.12 m ---

SEM = standard error of mean, 11 = 3. Different superscripts within a column indicate

differences among means (P < .05).

 

155

Table 4. Solubility of dry matter (DM ) from luceme hay in bufier (pH 6. 8) as affected
by incubation with protease extracted from the rumen (P), commercial cellulase (C),

soaking in pH 4. 7 buffer before P, and C before P.

 

 

 

Time effect (P)
Treatment 0h 4h 16h 0v. 4h 4v. 16h
Protease (P) 24.7 29.44 35.2a .0001 .0001
Cellulase (C) 24.7 35.5b 42.4b .0003 .003
Buffer, P 24.7 29.9a 33.1a .0002 .002
c, P 24.7 46.6c 53.1c .0001 .0001

SEM .099 .384 .61 --- ---

SEM = standard error of mean, it = 3. Different superscripts within a column indicate

differences among means (P < .05).

156

Table 5. Efl‘ects of enzyme level and addition of enzyme after 8 h on soya bean meal
degradation at 8 and 16 h.

 

 

 

 

Enzyme added (ml) Protein degraded
0 h 8 h 0 h 8 h 16 h
10 0 1.3 29.7 39.5‘1
5 0 8.1 26.2 39.8a
5 s -- 36.49
SEM -- 2.1 2.2 .64

 

SEM = standard error of the mean, 11: 2. Different superscripts within a column

indicate differences among means (P < .05).

 

157

Table 6. Average crude protein degradation (percentage of original) after 0, 6 or 16 h

incubation with enzyme extracted from the rumen.

 

 

 

Degradation (%)

. Feed 0 h 6 h 16 h
Soya bean meal 5_7a 37.9a 47_3a
Luceme hay 24.1c 38.2a 41.7b
Meat and bone meal 15.0b 36.68 424*)
Maize grain (dried, ground) 13.6b 24.8b 24.3c
Standard error 1.006 1.073 .870

of mean (n = 9)

Different superscripts within a column indicate differences among means (P < .05).

 

 

——__._ 3.5. Lira-v 1r : -' :—£»‘-f."-'." ”gang... _ , - .-_-_ —- --

158

CHAPTER 6: EFFECT OF PLANT MATURITY AND PRESERVATION
METHOD ON IN VITRO PROTEIN DEGRADATION OF F ORAGES

ABSTRACT
The influence of maturity and method of conservation on protein degradation was
determined on four different forage species. Alfalfa, smooth bromegrass and reed
canary grass were harvested at three maturity levels and whole plant corn was harvested
at two maturity levels. Samples of each forage were freeze dried, or prewilted and
ensiled in mini-silos at two DM contents. Additional samples of all forages except corn
were field dried to hay. Ground sample was incubated for O, 2 and 24 h with crude
enzyme extract from rumen contents. Degraded protein as a percentage of total crude
protein was determined as the protein soluble in trichloroacetic acid (80 g/L) after
degradation. Increased maturity resulted in lower protein degradation for alfalfa,
bromegrass and canarygrass. Though ensiling increased protein degradation compared
to dried forage, there was no consistent effect of silage DM content. Freeze dried
material generally had less degraded protein than hay, but bromegrass protein
degradation at 24 h was lower in hay than freeze dried samples. Protein degradation of

forages was highly variable and depended on plant maturity and conservation method.
(Key words: protein degradation, forages, maturity, hay, silage)

Abbreviation key: TCA = Trichloroacetic acid

 

 

159

INTRODUCTTON
Currently used systems to balance dairy cattle rations emphasize the need to consider
ruminal protein degradation (1, 20). Forages generally have low crude protein content
compared to protein supplements. However, since forages are the main ingredient of
most dairy rations, they supply a significant portion of the total crude protein content of
the diet. There has been little research on the protein degradation properties of forages.
Since various methods to determine protein degradation have limitations (21), the few
results that are available require confirmation using an alternative means to measure
protein degradation. As plants mature, different proteins may be required to meet the
needs of changing plant metabolism and structure. Conservation of forages may also
result in changing profiles and tertiary structure of plant proteins. These changes may
alter the proteolytic degradation of forage protein. The objective of this study was to
determine effects of plant maturity and forage preservation method on in vitro protein

degradation of four forage species.

METHODS

Alfalfa (Medicago sativa L.) was harvested in bud stage (Early), one tenth bloom
(Mid), and full bloom (Late). Smooth bromegrass (Bromus inermis Leyss) and reed
canarygrass (Phalaris arundinacea L) were harvested in boot (Early), early
reproductive (Mid) and seeded (Late) stages. Whole plant corn (Zea mays L.) was
harvested at half milk line (Early) and black layer formation (Late). Perennial forages
were harvested in the first cutting of 1989 from randomized university plots that were
established the previous spring. Corn was harvested from a different university plot.

After harvesting withva hand sickle, each forage was mixed and then divided
into parts to provide three replicates of each treatment. Each forage was preserved four

different ways (three ways for corn) at each maturity level. Forage samples were frozen

 

_" "' “FR" *4.“ 2.134e-15: - t

160

at -25°C, and then freeze dried with care taken to avoid thawing of samples, or they

were prewilted for silage or hay (except for com). Hay samples were dried outdoors on

pavement to prevent leaf loss in the field. Alfalfa hay at mid maturity was moved
indoors and drying completed in front of a forced air heater (30°C) to prevent damage
due to out-door precipitation. Samples of each forage were prewilted on pavement to
30-35% and 40-48% dry matter content before ensiling in mini silos (2.5 L) made from
polyethylene tubing. Forages were coarsely chopped in a 24-knife hammermill, and
packed into mini silos to a density of .8 kg/L. The moist silage from late corn was
made by direct addition of water after chopping but before ensiling. The drier silage
from early corn was made by prewiltin g indoors in front of a forced air heater (30°C)
before chopping. After ensiling for at least 60 d, silage samples were frozen, and freeze
dried. Crude dry matter content was determined from the difference in mass of original
and freeze dried samples. Dried samples were ground through a 5 mm and then a 1 mm
screen of a Wiley mill.

Dry matter content of ground and dried forages was determined by drying at
100°C for 16 h (6), and this was multiplied by crude dry matter for final dry matter
determination. All other analyses were determined on freeze dried and ground samples,
and are expressed on a dry matter basis. Samples of each forage were ashed at 500°C
for 5 h (6). Fiber fractions (NDF, ADF, lignin were determined sequentially (6)with
modification of the NDF procedure by the substitution of triethylene glycol for 2-
ethoxyethanol (18 g/kg) and the omission of decahydronaphthalene and sodium sulfite

(24). Crude protein was determined as Kjeldahl nitrogen (8) x 6.25. Crude protein

_~;.‘.A. n .

degradation was determined after 0, 2 and 24 h incubation with enzymes extracted from

rumen contents (11). Degraded crude protein was considered to be that soluble in

trichloroacetic acid (TCA) at a final concentration of 80 g/L.

 

 

 

161

Statistics were performed using JMP (26). Each forage species was analyzed
separately due to significant interaction effects of species with maturity and
preservation method. Maturity, preservation method and the interaction effects were
determined using AN OVA at each length of incubation. Further elucidation of main
effects was made using the following specific contrasts: early vs. mid maturity, mid
vs. late maturity, freeze dried and field dried vs. ensiled forages, freeze dried vs. field

dried, and the wetter silage vs. the drier.
RESULTS AND DISCUSSION

Forage analysis results are presented for each forage species and maturity level (Table
1). Effects of maturity were as expected for each component shown (20). Crude
protein, fiber, lignin and ash were slightly higher in silage and hay compared to freeze
dried samples (P < .05, data not shown). These effects would have resulted from
respiration of other feed components in samples during wilting or ensiling (7, 23).
Alfalfa and bromegrass that were prewilted to a higher DM content before ensiling
resulted in greater ash content compared to silage preserved at lower DM (P < .05).
This indicates that lower silage DM resulted in reduced respiration, as would be
expected from previous research (25).

Alfalfa crude protein was degraded more than that of other forages regardless of
length of incubation with enzyme (Table 2). Cool season grasses (bromegrass and
canarygrass) had higher protein degradation than corn. Previous research on protein
degradation of forages in situ confirms these relative differences (4, 9, 10 12, 18). A
previous production trial showed that milk production increased more due to
supplementation with unde graded protein when dairy cattle were fed diets based on
alfalfa silage instead of corn silage (30). This effect was likely due to the deficiency of
undegraded protein on the alfalfa silage diet.

 

 

162

Advanced maturity resulted in lower degradation of CP for alfalfa and
bromegrass samples regardless of length of incubation (Table 2). The overall main
effect of maturity on canarygrass CP degradation at 24h was significant (P < .05),
though individual contrasts were not. There was no maturity effect on corn CP
degradation. In previous research, advanced maturity was associated with lower
disappearance of CP from nylon bags incubated in situ (9, 16, 29). However, maturity
effects were not demonstrated with grass and legume silage studied in vivo (13).

Ensiled forages were consistently higher than dried forages in CP degradation
(Table 3), as has been demonstrated previously (5, 14, 22). Silage DM affected grass
CP that was soluble in TCA before incubation with enzyme in vitro (0 h), but there was
no effect of silage DM on CP degradation after 2 or 24 h incubation with enzyme. The
initial increased CP degradation in drier silages of bromegrass may have resulted from
proteolysis during the wilting process (15). However, drier silages of canarygrass and
corn had lower initial CP degradation. Previous studies conducted in vivo have
demonstrated both increased (19) and decreased (13) protein degradation due to
prewilting. Heat production may bind protein to reduce the susceptibility to
degradation (3), but the present study was conducted with mini silos in which heat was
quickly dissipated. Hay was higher in initially degraded protein than freeze dried
forages. This effect was probably due to degradation during wilting from plant enzyme
activity (17). Bromegrass CP degradation was lower at 24 h for hay than for freeze
dried forage. Previous work showed that in vivo protein degradation was reduced for
dried compared to fresh forages (2, 28).

Maturity x preservation method interaction effects were significant (P < .05) for
TCA soluble protein before enzymatic degradation of alfalfa, bromegrass and com.
This interaction was significant for alfalfa and canarygrass after 2 h degradation, but

was only significant for alfalfa after 24 h degradation. Silage protein degradation

 

 

fi— — ' ‘35» a Iwm‘ _1 54". "It'd: . %— _.

163

consistently declined with increased maturity of alfalfa, but hay CP degradation was
higher in mid-maturity alfalfa than in early or late (Figure 1). The mid maturity alfalfa
was harvested in a wet period, and was dried indoors at a slower rate than other forages.
Forages increase uptake of nitrogen after rainfall (27), and proteolysis may have had
greater opportunity to proceed during the slower drying. Maturity did not affect TCA
soluble CP in freeze dried bromegrass that was not incubated with enzyme, though it
affected other bromegrass samples (Figure 2). Freeze dried mid maturity canarygrass
was higher in CP degradation at 2 h compared to early and late‘canarygrass. This
contradicted other preserved canarygrass which tended to decline in degraded protein
with increased maturity (Figure 3). Early freeze dried corn was lower than late freeze
dried corn in TCA soluble protein before degradation with enzyme, but ensiled corn did

not differ (Figure 4).

 

 

 

 

164

REFERENCES

 

. Agricultural Research Council. 1984. Report of the Protein Group of the
Agricultural Research Council Working Party on the Nutrient Requirements of
Ruminants. Commonw. Agric. Bur., Slough, Engl.

. Beever, D. B., S. B. Cammell, and A. Wallace. 1974. The digestion of fresh, frozen
and dried perennial ryegrass. Proc. Nutr. Soc. 33: 73A.

. Broderick, G. A., J. H. Yang, and R. G. Koegel. 1993. Effect of steam heating
alfalfa hay on utilization by lactating dairy cows, J. Dairy Sci. 76: 165.

. Brown, W. F. and W. D. Pitrnan. 1991. Concentration and degradation of nitrogen
and fibre fractions in selected tr0pical grasses and legumes. Tropical Grasslands
25:305.

. Filmer, D. G. 1982. Assessment of protein degradability of forage. Page 1 in Forage
Protein in Ruminant Animal Prod. Occasional Publication No.6 Br. Soc. Anim.
Prod.

. Goering, H. K. and P. J. Van Soest. 1970. Forage Fiber Analyses (Apparatus,
Reagents, Procedures, and Some Applications). Agric. Handbook No. 379. ARS-
USDA, Washington, DC.

. Greenhi11,W. L. 1959. The respiration drift of harvested pasture plants during
drying. J. Sci. Food Agric. 10: 495.

. Hach, C. H., S. V. Brayton, and A. B. Kopelove. 1985. A powerful Kjeldahl
nitrogen method using peroxymonosulfuric acid. J. Agric. Food Chem. 33: 1117.

. Hoffman, P. C., S. J. Sievert, R. D. Shaver, and D. K. Combs. 1992. In situ CP
degradation kinetics of perennial forages. J. Dairy Sci. 75(Suppl. 1):210.

10. Janicki, F. J ., and C. C. Stallings. 1988 Degradation of crude protein in forages

determined by in vitro and in situ procedures. J. Dairy Sci. 71:2440.

11. Kohn, R. A. and M. S. Allen. 1993. An in vitro method to measure protein

degradation of feeds for ruminants using enzymes extracted from rumen contents. J.
Dairy Sci. 76: Suppl.

12. Kwakkel, R. P. ,J. Van Bruchem, G. Hof, andH. Boer. 1986. The 1n sacco

degradation of crude protein and cell wall constituents in grass, alfalfa and maize
silages. Neth. J. Agric. Sci. 34:116.

13. Makoni, N. F., J. A. Shelford, and L. J. Fisher. 1989. Effect of wilting and maturity

(3)8 1grass/legume silage on protein degradation in the rumen. J. Dairy Sci. 67 :Suppl.

 

 

165

14. McDonald, P. 1982. The effect of conservation processes on the nitrogenous

15.

16.

17.

18.

19.

20.
21.
22.
23.
24.
25.
26.

27.

28.

components of forages. Page 41 in Forage Protein in Ruminant Animal Prod.
Occasional Publication No. 6 Br. Soc. of Anim. Prod.

McKersie, B. D. 1981. Proteinases and peptidases of alfalfa herbage. Can. J. Plant
Sci. 61: 53.

Messman, M. A., W. P. Weiss, and D. O. Erickson. 1992. Effects of nitrogen
fertilization and maturity of bromegrass on nitrogen and amino acid utilization by
cows. J. Anim. Sci. 70: 566.

Messman, M. A., W. P. Weiss, and M. E. Koch. 1992. Electrophoresis as a tool to
ascertain effects of preservation method on forage protein composition. Page 35 in
Ohio Agric. Experiment Station Research Highlights, Wooster, OH.

Michalet-Doreau, B. and M. Y. Ould-bah. 1992. Influence of hay making on in
situ nitrogen degradability of forages in cows. J. Dairy Sci. 75: 782.

Narasimhalu, P., E. Teller, M. Vanbelle, M. Foulon, and F. Dasnoy. 1989.
Apparent digestibility of nitrogen in rumen and whole tract of friesian cattle fed
direct-cut and wilted grass silages. J. Dairy Sci. 72: 2055.

National Research Council. 1989. Nutrient Requirements of Dairy Cattle. 6th rev.
ed. Natl. Acad. Sci., Washington, DC.

Nocek, J. E. 1988. In situ and other methods to estimate ruminal protein and energy
digestibility: a review. J. Dairy Sci. 71: 2051.

Petit, H. V. and G. F. Tremblay. 1992. In situ degradability of fresh grass and
grass conserved under different harvesting methods. J. Dairy Sci. 75: 774.

Rees, D.V.H. 1982. A discussion of sources of dry matter loss during the process of
haymaking. J. Agric. Eng. Res. 27: 469.

Robertson, J. B. and P. J. Van Soest. 1977. Dietary fibre estimation in concentrate
feedstuffs. J. Anim. Sci. 45: Suppl. 254.

Ruxton, I. B. and P. McDonald. 1974. The influence of oxygen on ensilage. 1.
Laboratory studies. J. Sci. Food Agric. 12: 706.

SAS Institute Inc. 1991. Statistical Visualization for the Macintosh. JMP ver. 2.2.
SAS Institute Inc., Cary, NC.

Thom, E. R., G. w. Sheath, and A. M. Bryant. 1989. Seasonal variations in total
nonstructural carbohydrate and major element levels in perennial ryegrass and
paspalum in a mixed pasture. New Zealand J. Agric. Res. 32: 157.

Thomson, D. J ., E. Pfeffer, and D. G. Armstrong. 1969. Effects of drying grass on
srtes of its digestion in sheep. Proc. Nutr. Soc. 28: 26A.

 

 

 

 

 

166

29. Vik-Mo, L. 1989. Degradability of forages in sacco 2. Silages of grasses and red
clover at two cutting times, with formic acid and without additive. Acta Agric.
Scand. 39: 53.

30. Voss, V. L., D. Stehr, L. D. Satter, and G. A. Broderick. 1988. Feeding lactating
dairy cows proteins resistant to ruminal degradation. J. Dairy Sci. 71: 2428.

 

 

167

TABLE 1. Mean forage composition by species and maturity level across forage
preservation methods (11 = 12).

 

Item CP NDF ADF Lignin Ash

 

 

%
Alfalfa
Early 27.0 35.7 25.9 5.1 9.6
Mid 20.1 48.7 35.2 8.0 8.5
Late 16.5 52.6 38.1 9.0 9.0
SEM .22 .48 .42 .11 . 10
Bromegrass
Early 19.4 53.7 29.0 3.1 8.6
Mid 14.3 63.7 35.9 4.7 7.6
Late 12.3 66.1 37.7 5.6 7.7
SEM .17 .27 . 19 .06 .07
Canarygrass
Early 20.8 53.9 28.5 2.7 8.1
Mid. 16.6 . 64.6 34.3 4.1 7.8
Late 13.9 65.5 35.5 4.8 7.4
SEM .22 .41 .28 .09 .14
Corn ’
Early 7.3 46.6 25.5 2.7 3.7
Late. 7.4 42.3 22.3 2.5 3.4
SEM .13 .74 .47 .08 .08

 

lSequential extraction was used for determination of NDF, ADF and sulfuric acid

lignin.

 

 

168

TABLE 2. Mean protein degradation at 0, 2 and 24 h by forage species and maturity
level across forage preservation methods.1

 

 

 

Forage 0 h 2 h 24 h
%
Alfalfa
Early 39.3 55.4 63.7
Mid 36.6 49.8 57.5
Late 30.8 42.6 51.6
P, early vs. mid?- .05 .0001 .0001
P, mid vs. late .0001 .0001 .0001
Bromegrass
Early 25.6 39.9 47.8
Mid 21.3 31.6 39.7
Late 17.6 25.3 36.5
P, early vs. mid .0001 .0001 .0001
P, mid vs. late .0001 .0003 .05
Canarygrass
Early 28.6 39.5 48.3
Mid 27.3 39.0 47.0
Late 28.6 37.5 45.3
P, early vs. mid NS NS NS
P, mid vs. late NS NS NS
Corn
Early 25.4 26.6 36.2
Late 24.7 26.6 34.9
P, early vs. late NS NS NS

 

1Values expressed as percentage CP soluble in TCA after incubation with ruminal
enzymes for time indicated. For perennials n = 12, com 11 = 9.

2Significance of contrast, NS indicates P > .1.

 

169

TABLE 3. Mean protein degradation at 0, 2 and 24 h by forage species and
preservation method across maturities.1

 

 

 

Forage 0 h 2 h 24 h
%
Alfalfa
Freeze dried 19.8 39.1 52.7
Hay - 24.2 42.6 56.5
Silage (30% DM) 49.2 57.3 59.9
Silage (38% DM) 49.1 58.0 61.3
P, dried vs. ensiled2 .0001 .0001 .0001
P, freeze dried vs. hay .01 .002 .001
P, wet vs. dry silage NS NS NS
Bromegrass
Freeze dried 10.0 25.6 39.1
Hay 15.0 27.7 34.8
Silage (33% DM) 29.7 36.5 45.2
Silage (44% DM) 31.4 39.3 46.3
P, dried vs. ensiled .0001 , .0001 .0001
P, freeze dried vs. hay .0001 NS .03
P, wet vs. dry silage .01 NS NS
Canarygrass
Freeze dried 12.8 27.6 40.4
Hay 20.9 34.2 41.9
Silage (34% DM) 42.8 47.3 53.6
Silage (43% DM) 36.1 45.5 51.5
P, dried vs. ensiled .0001 .0001 .0001
P, freeze dried vs. hay _ .01 .0001 NS
P, wet vs. dry silage .05 NS NS
Corn
Freeze dried 11.8 21.7 29.0
Silage (31% DM) 32.5 30.8 39.4
Silage (38% DM) 30.8 27.4 38.3
P, dried vs. ensiled .0001 .001 .002
.05 NS NS

P, wet vs. dry silage

 

1Values expressed as CP soluble in TCA after incubation with ruminal enzymes for
time indicated. For perennials n = 9; corn n = 6.

2Significance of contrast, NS indicates P > .1.

170

Figure 1. Percentage of alfalfa forage protein degraded in early mid and late maturity
forage. I Indicates freeze dried sample, I hay, 5 low DM silage, and
D high DM silage. A: Percentage soluble at 0 h, B: 2 h and C: 24 h.

 

 

 

 

)

171
(A

 

1'11

 

 

Early

 
 
   

 

 

 

m w
a a
L L
M m M m M
y y
H H
a a
E E
4.. 0 5 0. 4.. 0 5 0 4.. 0. 5 0 5 0 4.. 0 .... 0
7 6 4 3 1 7 6 4 3 1 7 6 4 3 1
Aoev 5395 25.5 ecumewon 36v 5889 25.5 coeaLweQ 36v £32m 32.5 33..on

 

 

 

172
Figure 2. Percentage of bromegrass forage protein degraded in early mid and late
maturity forage. I Indicates freeze dried sample, 3 hay, E3 low DM silage, and
U high DM silage. A: Percentage soluble at 0 h, B: 2 h and C: 24 h.

 

 

 

173

)

A

(

 

5 0 5 0 5 0
7 6 4 3 1

$3 589:. 35.5 @223on

a, and

 

Late

Mid

Early

Late

Mid

Early

 

 

5 0 5
7 6 4

30
15-
0

Aoev 5395 25.5 mega—wen—

)

C

(

 

 

0 5 0
3 1

45-

?ev 5395 35.5 movemen—

Late

Mid

Early

174

Figure 3. Percentage of reed canarygrass forage protein degraded in early mid and late
maturity forage. I Indicates freeze dried sample, . hay, '3 low DM silage, and

U high DM silage. A: Percentage soluble at 0 h, B: 2 h and C: 24 h.

 

 

   

175
A)

(

Late
Late

Mid
B)

(

Early

  

 

 
 

r

 

 

5
0
5

50 050 5
1 31

7
6
4

5 0 5 0
7 6 4 3

$3 £39:— 339 Boe.—us: $3 589:. 25.5 02:29: 33 589:. 25.6 evasion

176
Figure 4. Percentage of whole plant corn protein degraded in early and late maturity
forage. I Indicates freeze dried sample, 1!?! low DM silage, and 0 high DM
silage. A: Percentage soluble at 0 h, B: 2 h and C: 24 h.

 

Degraded crude proteln (%) Degraded crude protein (95)

Degraded crude protein (%)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

178

CHAPTER 7: PREDICTION OF PROTEIN DEGRADATION OF FORAGES
FROM SOLUBILITY FRACTIONS

ABSTRACT
Two experiments were conducted to determine the relationship between ruminal
digestion of forage protein and fractionation based on solubility. The first experiment
used 42 forages (each replicated 3 times) including different species, maturities, and
means of conservation. Crude protein was fractionated into 6 parts for each forage:
trichloroacetic acid soluble, buffer soluble, acetone soluble, neutral detergent soluble,
acid detergent soluble and insoluble. Crude protein degradation at 2 and 24 h
(determined previously) was regressed against solubility fractions. Protein
fractionation patterns differed among forages due to effects of species, maturity and
means of conservation. Multiple regression analysis resulted in prediction of crude
protein degradation at 2 and 24 h with R2 of .88 and .81 respectively. Higher levels of
buffer and neutral detergent soluble fractions were associated with higher protein
degradation, while higher levels of acetone soluble, acid detergent soluble, and
insoluble crude protein were associated with lower degradation. In the second
experiment, eight forages (each replicated twice) were digested with ruminal enzyme
for 0, 2, 6 and 24 h, and then were extracted as described. Solubility in TCA and
acetone increased during degradation, while buffer soluble, acid detergent soluble and
insoluble CP decreased during degradation. In both experiments, buffer soluble CP was
the only uniform fraction, while other fractions contained proteins that degraded at

diverse rates. Solubility of CP is related to degradation properties, but further research

 

 

179

is needed to isolate uniform fractions for use in models that predict CP degradation
rates.
(Key words: protein degradation, protein solubility, forages)

Abbreviation key: TCA = Trichloroacetic acid.

INTRODUCTION
The rate and degree of ruminal protein degradation of feeds is an important
consideration for formulating dairy cattle rations (22). Forages provide much of the
protein consumed by dairy cattle, and the degradation of forage protein varies
depending on plant species, maturity, conservation method (13) and season (1). Since
forages are an important variant, a routine method of analysis or prediction of forage
protein degradation is required to balance dairy cattle rations accurately.

Rumen degradation of forage protein depends on rate of degradation of
individual proteins in the rumen, and on rate of passage of those proteins from the
rumen. Both of these factors may be influenced by protein associations with other feed
components, and by availability of chemical bonds of the proteins to enzymatic attack.
Separation of feed proteins based on solubility may have practical significance to
ruminal protein degradation, since ultimately, the solubility of a protein determines the
microenvironment, or feed componerits, with which it associates (i.e. aqueous, lipid,
etc.), and this association may alter the availability of peptide bonds, or alter the
passage rate of the protein. Additionally, solubility may relate to other aspects of
protein structure which affect protein degradation.

Various researchers have attempted to quantify ruminally degraded protein
based on solubility properties in saline solutions, buffers or detergents. Buffer solutions
dissolve most non-protein nitrogen and varying amounts of true protein (16). In
general, buffer-soluble true protein has been shown to be more rapidly degraded by

bacterial protease and by ruminal microorganisms than insoluble protein, but the rate of

 

 

 

180

degradation varies among protein sources and some soluble proteins are not rapidly
degraded_(l7, 23).

Insoluble forage crude protein includes hydrophobic compounds such as
chlorophyll that contain non-protein nitrogen, and are selectively extracted in acetone
(31). In addition, membranes contain amphiphilic true proteins that may be protected
from degradation by the lipid bilayer (16). These proteins can be extracted in detergent
solutions. Detergent solution at pH 7 extracts cell contents and leaves CP associated
with plant cell wall (28), which is degraded slowly in the rumen (15). Acid detergent
insoluble crude protein is a measure of heat damage that renders protein indigestible
(32, 29).

The objectives of the present study were: 1) to compare changes in protein
solubility fractions across forage species, maturities and preservation methods, 2) to
determine rates of degradation of individual protein fractions, and 3) to use this
information to evaluate the possibilities for predicting forage protein degradation from
solubility fractions.

METHODS
Forages and Protein Degradation
Forages used in this study were described elsewhere (13). These forages included
alfalfa (Medicago sativa L), smooth bromegrass (Bromus inerrnis Leyss), reed
canarygrass (Phalaris arundinacea L), and whole plant corn (Zea mays L) at different
stages of maturity. Samples of each forage were freeze dried immediately after
harvesting, and others were preserved as hay and silage at high and low DM content.
Three field replicates were taken for each of 42 forage types.

Protein degradation was previously determined for each replicate forage by
incubation with a crude extract of ruminal enzyme for 2 and 24 h (13). Enzymes were

extracted from rumen fluid by solubilization in water after releasing them from

 

181

microbes by extraction in butanol and acetone. After incubation with rumen fluid
extract, undegraded protein was determined as that which precipitated in trichloroacetic
acid (I CA) at a final concentration of 80 g/L (12).
Protein Fractionation
Each forage replicate was subjected to sequential extraction of crude protein according
to the scheme depicted in Figure 1. Forage samples containing 16 mg of N were
weighed into 50 ml centrifuge tubes, which were then randomized before extraction.
Tubes were placed under C02 and 30 ml reduced media (12) was added. The media
was a pH 6.8 bicarbonate buffer with minerals and reducing agent ( NazS.9H20).
Tubes were purged with C02 while adding media, and then were capped, shaken and
stored for 2 h at 4 °C. Reduced media was used to prevent precipitation of forage
protein due to enzymatic reactions that were demonstrated previously in bromegrass
(14). Preliminary research (11) indicated that storage at 4°C resulted in solubilization
of crude protein without significant degradation of that protein by plant enzymes.
Storage at low temperature resulted in reduced NPN formation over time compared to
storage at 38°C, and was comparable to samples stored with enzyme inhibitor
(phenymethylsulfonyl fluoride). After 2 h in buffered media, samples were centrifuged
at 4°C and 1500 g, for 20 minutes. The supernatant was removed to different
centrifuge tubes, 15 ml of 240 g/L TCA was added, and these tubes were centlifuged at
1500 g, for 20 minutes. The supernatant from this extraction was presumed to be
soluble NPN (ammonia, nitrate, amino acids, peptides) or "fraction 1", and pellet was
buffer soluble true protein (fraction 2) (18).

The first pellet containing most of the original forage was extracted with 80%
acetone (20% water, vol/vol) for 2 h. Acetone (40 ml) was added and the samples
shaken vigorously three times while stored at 25°C. These samples were centrifuged

without chilling at 1500 x g for 20 min. The supernatant was removed to different

 

 

 

182

centrifuge tubes and the acetone evaporated over N2 in a sand bath at 40°C. This
fraction was presumed to be acetone soluble NPN (fraction 3) which is composed of
strongly hydrophobic pigments and membrane-associated compounds (i.e. chlorophyll)
(7 ).

The residue was next extracted with pH-7 detergent. The detergent solution was
adapted from Goering and Van Soest (9) with modification of the solution by
substitution of triethylene glycol for 2-ethoxyethanol (18 g/kg) and the omission of
decahydronaphthalene and sodium sulfite (25). Additionally EDTA was replaced with
sodium oxalate (25 g/L) in order to remove nitrogen from the detergent solution, yet
still provide a calcium chelator to solublize pectins. This detergent solution (40 ml)
was added to the residual forage samples and incubated for l h at 95°C. It was cooled
to 50°C and centrifuged at 1500 x g for 20 min. The supernatant was removed and
presumed to be cell-wall associated crude protein (fraction 4).

The pellet was extracted again using a similar method except for replacement of
pH-7 detergent with sulfuric acid and detergent solution. Triton-X 100 (2 mlIL, Sigma
Chemical Co., Saint Louis, MO) was added to 1 N sulfuric acid. This solution (40 ml)
was added to the residue of forage samples, which were incubated for 1 h at 95°C
before centrifugation. When the detergent cooled before and during centrifugation, a
film formed at the top of the supernatant and adhered to the sides of the tube. This film
was included with the supernatant fraction. The supernatant was presumed to be cell-
wall-associated crude protein (fraction 5) and the pellet was presumed to be bound
protein (fraction 6) (32, 29).

Each of these protein fractions was subjected to a modified Kjeldahl analysis for
nitrogen (10). Crude protein in each fraction was determined as N x 6.25. This value

was expressed as a fraction of the total crude protein for the replicate (determined

 

183

previously). The percent yield was determined as 100 times the total of all six
fractions.
Statistics for the First Experiment

Statistics were performed using JMP (26). For comparisons of forages for
solubility fractions, each forage species was analyzed separately due to significant
interaction effects of species with maturity and preservation method. Maturity,
preservation method and the interaction effects were determined using AN OVA at each
length of incubation. Further elucidation of main effects was made using the following
Specific contrasts: early vs. mid maturity, mid vs. late maturity, freeze dried and field
dried vs. ensiled forages, freeze dried vs. field dried, and wet vs. dry silage.

The portion of each protein fraction that was degraded at 2 and 24 h was
estimated from results of multiple regression. Several models were compared to
determine rates of degradation for protein fractions that were theoretically possible.

The model determined to be most appropriate a priori was as follows:
6
Y = 2 fi (X 3
i=1 .

The portion of protein degraded (I’CA soluble) at 2 or 24 h is represented by Y, Xi
represents the pool size of each solubility fraction (i) expressed as a fraction of total
crude protein (determined previously), and fi represents the fractional portion of cnrde
protein that is degraded at 2 or 24 h for each solubility fraction (1 to 6). The values of
fi were determined as the slopes from multiple regression of degraded protein at 2 or
24 h by the initial solubility fractions. The intercept was fixed at zero for the regression
because this is theoretically required, and allowing it to vary altered the prediction of
intercepts. Adding this fixed value inflates the value of R2 (because it is adding a point
at the low end of the curve), therefore R2 was determined separately from plots of

observed vs. predicted degradation when using the previously calculated slopes.

‘l

 

 

184

Alternative models fixed negative slopes at 0 to determine other slopes more accurately,
or fractions were pooled where separation of fractions was not believed to be adequate.

Rate of protein degradation (k) was determined on the undegraded residue (undg
= 1 - degraded) as follows: k = [1n (undg at time t + At) - 1n (undg at time t )] IAt,
where t + At represents the duration of degradation (2 or 24 h) and t represents the
previous determination (0 or 2 h), and At is the time interval involved in the
measurement (2 or 22 h).
Solubility Fractions After Degradation

Eight forages were subjected to a second experiment where crude protein was
fractionated after degradation in enzyme for 0, 2, 6, or 24 h. Forages used were a
subset of those in the previous experiment: early alfalfa that was freeze dried, mid-
maturity alfalfa hay, late alfalfa silage, early bromegrass silage, late bromegrass hay,
mid-maturity canarygrass that was freeze dried, mid-maturity canarygrass silage, and
late corn silage. Forages were incubated with ruminal enzyme as described previously,
and after incubation with enzyme crude protein was fractionated using a modified
procedure to that described previously. Since extraction required twice the amount of
sample as was used in the degradation method, two samples of each replicate were
incubated in different centrifuge tubes, and these were combined during the extraction.
At the end of the incubation with enzyme, samples were cooled to 4°C and centrifuged
at 1500 x g for 20 min. The supernatants were decanted to clean tubes and the pellets
were combined for subsequent extractions carried out as described previously. Buffer
soluble true protein was precipitated from the supernatant by addition of TCA.
Nitrogen was determined on combined pellets. Crude protein soluble in TCA was
determined indirectly by subtracting the total of all other fractions from the total CP
which was determined previously. This was done because enzyme was TCA soluble so

that this fraction could not be determined directly. All analyses were determined in 6

 

185

runs with the same rumen extract (2 replicates of 3 lengths of incubation were
determined separately).

Percentage of each protein fraction that was degraded was determined as 100
times the crude protein in the fraction at a point in time divided by the initial crude
protein in that fraction. Effects of forage type and length of incubation were evaluated
by analysis of variance for protein degraded within each fraction. Since interaction
effects were significant, forages were compared for percentage protein degraded in each
fraction at each length of incubation . Dried were compared to ensiled forages by
student t test. Noting differences, similar comparisons were made among ensiled and
dried forages.

The Lucas test of nutritional uniformity was applied to estimate percentage
protein degraded in each fraction and variation across forages (30). This procedure is
based on the principle that the apparent degradation of a nutrient (Dax) is a function of
the true degradation of the nutrient (Dtx), metabolic contribution of the nutrient (Mx),
and amount of the nutrient in the diet (Xi) as follows: Dax = Dtx - Mx I Xi.
Multiplying through by Xi gives the equation for a line: Dax(Xi) = Dtx(Xi) - Mx. If
Dax(Xi) is plotted on the abscissa and Xi on the origin, the slope of the line is the true
degradation and the negative intercept is the metabolic contribution of the nutrient. The
R2 of the line indicates the uniformity of degradation of the fraction across diets. In the
present experiment, amount of each fraction in the forage sample was plotted on the
origin and this times the percentage degradation of the fraction was plotted on the
abscissa. Separate plots were drawn for each length of incubation

RESULTS AND DISCUSSION
Fractionation of Forages
Results of protein fractionation are presented in Tables 1 and 2. Most of the nitrogen

(N) was soluble in TCA, neutral detergent or was insoluble. Corn also contained a

 

186

large amount of acetone soluble N (Table 1). While the acetone soluble fraction was
dark green for other forages, it was bright yellow for corn. Corn proteins (especially
zein) may degrade to hydrophobic peptides that resist ruminal degradation. It was
presumed to be composed of chlorophyll or other pigments and may be indigestible
NPN because compounds would have to be highly non-polar to dissolve in acetone.
Such compounds are likely to contain long hydrocarbon chains but not oxygen. It
would therefore be difficult for bacteria to use these compounds in the anaerobic
environment of the rumen.

Ensiling of feeds resulted in several important changes in protein fractionation
(Table 2). Crude protein soluble in TCA increased for all forages due to ensiling. This
result would have occurred due to the degradation of forage protein to peptides, amino
acids or ammonia which is expected during ensiling (19). Acetone soluble N also
increased for silages compared to dried forages. Several speculations are plausible to
explain this effect. It may have resulted from an inadequate extraction of pigments in
the dried forages, but ensiling may release these pigments from the plant structure.
Alternatively, this fraction may contain hydrophobic peptides which were insoluble in
buffer and therefore not extracted by TCA. Buffer soluble true protein was greatly
reduced in ensiled compared to dried forages. Previous work has demonstrated the
rapid degradation of soluble protein in alfalfa (6). Neutral detergent soluble protein
(fraction 4) was also reduced considerably by ensiling, and acid detergent soluble
(fraction 5) and insoluble (fraction 6) protein were reduced less dramatically. The fact
that nearly all protein fractions were reduced in ensiled compared to dried forages

demonstrates that even buffer insoluble CP is susceptible to protein degradation during

ensiling.

 

 

 

187

Prediction of Degradation from Fractions

The prediction of protein degradation at 2 h from multiple regression analysis is
shown in Table 3. Including all fractions in the model resulted in an R2 of .88 (model
1). The intercept was highly significant. perhaps due to multiple pools within some
fractions. If the size of each fraction (xi) were 0, one would expect the degraded
protein at 2 or 24 h to also be 0. This was the grounds for forcing the intercept through
0 for model 2. The objective was to estimate the percentage protein degraded for each
fraction, not necessarily provide for the best fit line. The slopes presented in model 2
should be better estimates of percentage protein degraded at 2 h than those from model
1 because the intercept is not allowed to bias the prediction. About 100% of buffer
soluble N (fractions 1 and 2) is predicted to be degraded at 2 h. Since TCA solubility is
the measurement of degradation in these studies this estimate is expected. The fact that
100% of buffer soluble protein is degraded is consistent with results comparing ensiled
and dried forages. Fraction 4 was the next most degraded at 2 or 24 h, again
confirming results from comparison of ensiled and dried forages. Acid detergent
soluble and insoluble crude protein were not predicted to be degraded at 2 h.

The estimate of acetone soluble crude protein degradation was negative. If the
difference in amount of acetone soluble crude protein resulted from differing efficiency
of extraction on different feed types, as has been postulated, then the higher extraction
in acetone would be associated with a lower extraction of another fraction. It is likely
that pigments not removed by acetone would be removed by detergent solution, so an
underestimation of fraction 3 could result in an overestimation of fraction 4. Since
fraction 4 is degraded, and 3 may not be, the overestimation of degradation due to high
levels of fraction 4 would be balanced by a negative slope associated with fraction 3.

Further experimentation with the model was based on three considerations.

First, most of the crude protein was found in fractions 1, 4 and 6. Second, the estimates

188

for the smaller fractions 3 and 5 (Tables 3 and 4) were negative, though we can not
expect negative degradation of crude protein. Third, coefficients of variation were high
for fractions 3 and 5. Models 3 and 4 were designed to minimize the contribution from
fractions 3 and 5 to the slopes of the other fractions by either pooling fractions (model
3) or forcing the slopes of fractions 3 and 5 to 0 (model 4). The results are consistent
with previous results. Fractions 1 and 2 again were nearly completely degraded at 2 h.
Fractions 3 and 4 together (model 3) were predicted to degrade more slowly than
fraction 4 by itself (model 4). This demonstrates the importance of removing fraction 3
from fraction 4, and if this were done more thoroughly, the degradation of fraction 4
may have been higher.

Crude protein degradation at 24 h was estimated for each fraction using the
models described for degradation at 2 h (Table 4). The strong relationship between
solubility and degradation at 24 h (Figure 2) indicates the potential for analyzing
forages routinely based on solubility. These results are consistent with those for
degradation at 2 h. Model 4 estimated fractions 1 and 2 to be nearly completely
degraded, fraction 4 to be 60% degraded, fraction 6 to be 11% degraded, with fractions
3 and 5 set to 0% degradation. The rate of degradation of fraction 4 is estimated from
the percentage degraded at 2 h (I' able 3, model 4) as 13.5%Ih, while it is estimated as
3.8 %Ih over the course of 24 h (Table 4, model 4). Likewise, the rate of degradation of
fraction 6 is estimated as 1.7%Ih vs. .5%Ih over 2 and 24 h respectively. The estimates
for rate of degradation declines from fast to slow over time. This change in degradation
rate indicates the probability of multiple fractions within fraction 4 and possibly 6
which degrade at different rates. Further work may be warranted to investigate further

separation, perhaps exploiting better removal of acetone soluble fraction.

189

Solubility Fractions After Degradation

Change in protein fractions due to degradation with ruminal enzyme is shown
(Figure 3). Crude protein soluble in TCA and acetone increased over time in dried
forages. Though acetone soluble crude protein also increased over time for ensiled
forages, TCA soluble crude protein actually decreased. Amount of neutral detergent
soluble crude protein did not change due to incubation with enzyme. All other fractions
decreased when incubated with enzyme.

In the first experiment, TCA soluble crude protein increased due to protein
degradation of all forages including silage. The difference in the present experiment
may relate to the measurement of TCA soluble protein. When crude protein
degradation was measured in forages in the previous experiment, TCA was added
directly to in vitro media after incubation with ruminal enzyme. In the present
experiment, TCA was added to the supernatant from the degraded forage only after
centrifugation of the residue from degradation. Thus crude protein that was insoluble in
media was not available to be solubilized in TCA. On the other hand, if TCA is added
directly to feed proteins, especially when hydrophobic crude protein (pigments or
peptides) have been released due to digestion, the TCA may solubilize some of these
compounds. However, if TCA is added only to the buffer soluble crude protein, the
strongly hydrophobic compounds have already been removed, and therefore are not
candidates for TCA solubility.

Hydmphobic compounds would be solubilized in acetone. This solution was
chosen to solubilize pigments and other NPN compounds that would otherwise be
contained in the detergent soluble fraction. Observation that the neutral detergent
soluble fraction does not decrease due to digestion, while acetone soluble CP increases,
suggests that this CP may be extracted from another fraction. It may contain

hydrophobic or glycated peptides. In fact it may relate to the soluble protein identified

190

in other experiments (apparently peptides) that appears to accumulate due to ruminal
digestion (5). A better understanding of acetone soluble crude protein is needed to
evaluate whether or not it should be included in the degraded, undegraded, or
indigestible fraction, especially considering the fact that after degradation it comprises
15 to 20 % of the total crude protein (Figure 3).

Buffer soluble crude protein (fractions 1 and 2) was nearly completely degraded
by 2 h (Figure 3, Table 5). This observation is consistent with the discussion of the
previous experiment. However, the fact that fraction 4 was apparently not affected by
degradation contradicts the changes from dried to ensiled feeds, and the predictions
from regression described in the previous experiment. A possible explanation may be
that though protein from this fraction was degraded, crude protein from other fractions
may have been made soluble in neutral detergent by degradation, thus negating the
previous effect. This protein may have been supplied from the acid detergent soluble or
insoluble fractions, which did show disappearance in this experiment.

Though protein fractions appear to be degraded by enzyme at the initial stages,
this degradation is slowed or stopped at the later time points (Figure 3 and Table 6).
Buffer soluble protein is the only fraction that was nearly completely degraded. The
enzymeused in these experiments appears to maintain its activity on azocascin for
extended incubations, and addition of fresh enzyme after 8 h incubation did not increase
soybean meal protein degradation rate (13). Therefore, the slowing rate of degradation
is likely to be a function of the forage and not only the method of analysis. Plant
structures associated with proteins may inhibit degradation of some forage protein
fractions using this system. Forages varied in the percentage protein degraded at each
length of incubation for most fractions. These effects are shown for 2 h degradation

(Table 5). Much of the variation was between dried forages and silage. Since much of

191

the readily degraded protein was already lost from the silage samples, the remaining
fraction would degrade at a slower rate.

The Lucas test is an efficient means to estimate the percentage protein degraded
within a fraction across forages, and the variation associated with that estimate (30).
Fraction 2 protein was shown to be entirely degraded within 2 h across all forages
(Figure 4A, Table 7). The small but significant negative intercept probably was due to
contamination of protein from other fractions, or to a constant amount of undegraded
crude protein associated with this fraction for all forages. During the extraction,
insoluble suspensions were observed after centrifugation for buffer soluble protein.
These suspensions were precipitated by addition of TCA. Most likely these particles
caused this negative intercept.

Other fractions appeared less uniform for degradation of protein when analyzed
using the Lucas test. The percentage degradation was significant at 24 h for fraction 4
(Figure 4B, Table 7 ). None-the-less there was considerable variation across forages
and a low percentage degradation. Degradation of acid detergent soluble and insoluble
fractions appeared to be greater than that of neutral detergent soluble, and were more
consistent across forages (Figure 5, Table 7). The accuracy of these estimates of
percentage degradation is questionable where the intercept is negative. When the Lucas
test is applied to animal digestibility data, the negative intercept refers to metabolic
protein. In the present system, this CP could be contributed from other fractions.
However, it is more likely to be due to non-uniform fractions. The most important
differences in protein degradation and in amount of CP in a fraction are due to effects
of ensiling. The ensiled forages have lower amounts of some fractions, and degradation
of those fractions is slower because the more readily degraded portion was lost during

ensiling. Therefore, with certain fractions, lower amounts of the fraction may be

 

—1—_—v—Tn_—r '-~ ; ' .2: .q._..=:-.:>_.-_ —.;:«.r§:_-“_~j “...

 

 

192

associated with slower degradation. This would result in an artificially lower intercept,
and an inflated slope for the line (Figure 5).
Comparison to Previous Results

Previous efforts to predict protein degradation of forages from solubility have
provided mixed results. Buffer soluble protein has been shown to be rapidly degraded
by proteolytic enzymes (6), but since it comprises a small amount of the total CP, a
relatively high amount of buffer soluble protein does not necessarily prescribe a high
degradability of the forage protein (24). Likewise, neutral detergent insoluble protein
was shown to degrade at a slow rate for some forages degraded in situ (15), but this
fraction also represented a small portion of the total CP. Acid detergent insoluble crude
protein is used as an indicator of heat damage that renders the protein indigestible (8,
29), but some studies have shown a degree of degradation of this fraction (4, 33).

- Other researchers have investigated the possibility of using gel electrophoresis
to isolate forage proteins for the prediction of protein degradation (20). This work
demonstrated that even though the exact same protein may be isolated from different
forages (i.e. ribulose-bis-phosphate-carboxylase/oxygenase), its rate of degradation may
vary from forage to forage (20), perhaps due to protein associations with other forage
components. For example, tannins have been implicated in the slowing of degradation
of forage protein (3), and cellulase treatment increased degradation of forage protein in
vitro (13). Therefore, measuring the amount of a specific protein fraction may not be as
important to predicting overall degradation as measuring the degree of association of
forage proteins with tannins, fiber, or membranes.

The methods used to fractionate protein and determine degradation may affect
the results. In the present study, forages were freeze dried and resolubilized in reduced
media. This drying process was chosen because it was determined to decrease changes

in solubility and degradation compared to freezing and thawing [Chapter 2], (14) or

 

 

 

 

 

 

 

193

oven drying (2), however, results may have differed if fresh feeds were analyzed. The
present study also utilized a sequential extraction so that independent fractions would
be obtained. Nitrogen was removed from detergent solutions to make sequential
extraction possible. Solubility was determined by centrifugation ratherthan filtration,
thus potentially decreasing the overall solubility. The method of measuring protein
degradation also differed from some previous work in that TCA insoluble protein was
regarded as the undegraded protein and microbial contamination was eliminated by
using extracted enzyme. Other methods may over-estimate the undegraded crude
protein due to microbial attachment to feeds (21), or underestimate it because protein is
considered degraded as soon as it is reduced in particle size enough to leave the nylon
bag in the rumen, or filter through filter paper used for in vitro methods (27). On the
other hand, the method used in the present study may be limited by the enzyme profile
that was extracted and the possible importance of live microbes to the degradation.
Implications

The’actual amount of degraded and undegraded protein supplied by a forage
depends on both degradation and passage rates. The protein fractions isolated in this
study may in fact pass from the rumen at different rates owing to their solubility and
associations with certain plant tissues. Further research could ultimately lead to the
isolation-of protein fractions that degrade and pass at constant and uniform rates, and
would enable accurate prediction of protein degradation in vivo.

This is the first time all of the CP of forages was extracted sequentially to
determine the possibility of predicting protein degradation from a complete array of
solubility fractions. The prediction of protein degradation at 2 h or 24 h indicates a
potential for use of solubility to predict degradation. Buffer soluble protein was
degraded rapidly in all forages, and acetone soluble CP appeared not to be degraded.

Other fractions degraded to differing extents depending on the forage, and much of this

 

 

 

 

194

inconsistency appears to have resulted from the presence of multiple pools within
certain fractions. Prediction of protein degradation from solubility fractions can be
improved by isolating more homogenous fractions with respect to protein composition
and interactions of the proteins with other feed components. Further research is needed

to fractionate proteins into pools that degrade completely at the same rate.

 

 

10.

ll.

12.

13.

14.

15.

195

REFERENCES

Abdalla,H. O., D. G. Fox, andR. R. Seaney. 1988. Variationinprotein and fiber
fractions 1n pasture during the grazing season. J. Anim. Sci. 66: 2663.

Abdalla, H. O., D. G. Fox, and P. J. Van Soest. 1988. An evaluation of methods
for preserving fresh forage samples before protein fraction determinations. J.
Anim. Sci. 66: 2646.

Barry, T. N. and T. W. Manley. 1984. The role of condensed tannins in the
nutritional value of Lotus pendunculatus for sheep. Br. J. Nutr. 51: 493.

Broderick, G. A., J. H. Yang, and R. G. Koegel. 1993. Effect of steam heating
alfalfa hay on utilization by lactating dairy cows. J. Dairy Sci. 76. 165.

Chen, G. H. J. Strobe], J. B. Russell, andC. J. Sniffen. 1987. Effect of
hydrophobicity on utilization of peptides by ruminal bacteria 1n vitro. Appl.
Environ. Microbiol. 53: 2021.

Chemey, D. J. R., J. J. Volenec, and J. H. Cherney. 1992. Protein solubility and
degradation in vitro as influenced by buffer and maturity of alfalfa. Anim. Feed
Sci. Technol. 37:9.

Dietz, K. J. 1989. Leaf and chloroplast development in relation to nutrient
availability. J. Plant Physiol. 134: S44.

Goering, H. K, C. H. Gordon, R. W. Hemken, D. R. Waldo, P. J. Van Soest, and
L. W. Smith. 1972. Analytical estimates of nitrogen digestibility 1n heat damaged
forages. J. Dairy Sci. 55: 1275.

Goering, H. K. andP. J. Van Soest. 1970. Forage Fiber Analyses (Ap Mrspara
Reagents, Procedures, and Some Applications). Agric. Handbook No. 379. ARS-
USDA, Washington, DC.

Hach, C. H., S. V. Brayton, and A. B. Kopelove. 1985. A powerful kjeldahl
nitrogen method using peroxymonosulfuric acid. J. Agric. Food Chem. 33: 1117.

Kohn, R. A. 1993. In vitro methods to determine forage protein degradation in the
rumen. Ph. D. Diss., Michigan State Univ., East Lansing.

Kohn, R. A. and M. S. Allen. An 1n vitro method to measure protein degradation
of feeds using enzymes extracted from rumen contents. Br. J. Nutr. Submitted.

Kohn, R. A. and M. S. Allen. Effect of plant maturity and preservation method on
in vitro protein degradation of forages. J. Dairy Sci. submitted.

Kohn, R. A. and M. S. Allen. 1992. Storage of fresh and ensiled forages by
freezing affects fibre and crude protein fractions. J. Sci. Food Agric. 58: 215.

Lindberg, J. E. 1988. Influence of cutting time and N fertilization of the nutritive
value of timothy. 2. Estimates of rumen degradability of nitrogenous
compounds" Swedish J. Agric. Res. 18:85.

 

 

16.

17

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

 

 

196

Lyttleton, J. W. 1973. Proteins and nucleic acids. Pages 63-103 in Chemistry and
Biochemistry of Herbage. ed. G. W. Butter and R. W. Baily. Academic Press,
New York, NY.

Mahadevan, S., J. D. Erfle, and F. D. Sauer. 1980. Degradation of soluble and
insoluble proteins by Bacteroides amylophilus protease and by rumen
microorganisms. J. Anim. Sci. 50:723.

Marals, J. P. and T. K. Evenwell. 1983. The use of trichloroacetic acid as
precipitant for the determination of "true protein" in animal feeds. South African
J. of Anim. Sci. 13: 138.

McKersie, B. D. 1981. Proteinases and peptidases of alfalfa herbage. Can. J. Plant
Sci. 61: 53.

Messman, M. A., W. P. Weiss, and M. E. Koch. 1992. Electrophoresis as a tool to
ascertain effects of preservation method on forage protein composition. Pages 35
to 39 in Ohio Agric. Experiment Station Research Highlights, Wooster, OH.

Michalet—Doreau, B. and M. Y. Ouldbah. 1992. In vitro and in sacco methods for
the estimation of dietary nitrogen degradability in the rumen - a review. Anim.
Feed Sci. Technol. 40: 57.

National Research Council. 1989. Nutrient Requirements of Dairy Cattle. 6th rev.
ed. Natl. Acad. Sci., Washington, DC.

Nocek, J. B., J. H. Herbein, and C. E. Polan. 1983. Total amino acid release rates
of soluble and insoluble protein fractions and concentrate feedstuffs by
Streptomyces griseus. J. Dairy Sci. 66: 1663.

Pion, R., C. Genest, G. Bayle, and P. Thivend. 1983. Assessment of protein
degradability in concentrates using an enzymatic method. Pages 207-210 in IV'th

Int. Symp. Protein Metabolism and Nutrition, Clermont-Ferrand (France), Sept. 5—
9, 1983. Inst. Natl. Rech Agron. Publ.

Robertson, J. B. and P. J. Van Soest. 1977. Dietary fibre estimation in concentrate
feedstuffs. J. Anim. Sci. 45: Suppl.254

SAS Institute Inc. 1991. Statistical Visualization for the Macintosh. SAS
Institute Inc., Cary, NC.

Spencer, D., T. J. V. Higgins, M. Freer, H. Dove, and J. B. Coombe. 1988.
Monitoring the fate of dietary proteins in rumen fluid using gel electrophoresis.
Br. J. Nutr. 60: 241.

Van Soest, P. J. 1964. Symposium on nutrition and forage and pastures: new
chemical procedures for evaluating forages. J. Anim. Sci. 23: 838.

Van Soest, P. J. 1965. Use of detergents in analysis of fibrous feeds. 111. Study of
effects of heating and drying on yield of fiber and lignin in forages. J. Assoc.
Offic. Anal. Chem. 48:785.

 

 

 

 

‘l

 

30.

31.

32.

33.

197

Van Soest, P. J. 1982. Chapter 4. Fecal composition, mathematics of digestion
balances and markers. Pages 39-57 in Nutritional Ecology of the Ruminant,
Cornell University Press, Ithaca, NY.

Waghom, G. C., I. D. Shelton, and V. J. Thomas. 1989. Particle breakdown and
rumen digestion of fresh ryegrass (Lolium perenne L.) and luceme (Medicago-
sativa L.) fed to cows during a restricted feeding period. Br. J. Nutr. 61: 409.

Weiss, W. P., H. R. Conrad, and W. L. Shockey. 1986. Amino acid profiles of
heat-damaged grasses. J. Dairy Sci. 69: 1824.

Weiss, W. P., H. R. Conrad, and W. L. Shockey. 1986. Digestibility of . nitrogen in
heat-damaged alfalfa. J. Dairy Sci. 69: 2658.

 

198

TABLE 1. Mean percentage protein in each of 6 solubility fractions by forage
species and maturity level across forage preservation methods (11 =12).l

 

 

 

 

 

 

 

Crude protein fraction
Forage 1 2 3 4 5 6 % Yld
% of CP
Alfalfa
Early 39.3 3.9 5.8 22.5 3.6 18.9 93.9
Mid 36.6 4.1 5.1 20.7 4.7 24.9 96.1
Late 30.8 4.8 4.6 23.6 5.2 29.0 98.1
P, early vs, mid2 .05 NS NS .01 .001 .0001 NS
P, mid vs. late .0001 .05 NS .0001 .1 .0001 NS
Bromegrass
Early 25.6 6.3 3.9 23.9 6.4 28.3 94.4
Mid 21.3 4.0 3.7 22.6 8.3 38.0 97.8
Late 17.6 3.1 4.2 21.9 8.5 40.5 - 96.0
P, early vs. mid .0001 .0001 NS .1 .0001 .0001 .01
P, mid vs. late .0001 .01 .01 NS NS .01 NS
Canarygrass '
Early 28.6 5.3 3.4 23.8 7.0 25.4 93.5
Mid 27.3 5.2 4.0 20.2 7.4 26.0 91.4
Late 28.6 4.1 3.6 21.7 7.0 27.7 91.2
P, early vs. mid NS NS .05 .01 NS NS NS
P, mid vs. late NS .1 NS NS NS .05 NS
Corn
Early 25.4 2.2 17.7 20.2 10.2 21.2 96.9
Late 24.7 2.6 21.5 21.0 8.4 19.5 97.8
P, early vs. late NS NS .001 NS NS .05 NS

 

1Crude protein fractions were extracted sequentially and separated by centrifugation as
follows: 1) soluble in trichloroacetic acid, 2) soluble 1n reduced med1a(b1carbonate), 3)
soluble in 80% acetone, 4) soluble in pH-7 detergent at 95°C, 5) soluble 1n sulfuric ac1d

and detergent at 95°C, and 6) insoluble.

2Significance of contrast, NS indicates P > .1.

199

TABLE 2. Mean protein percentage in each of 6 solubility fractions by forage species
and preservation method across maturities (n = 9).

 

 

 

 

 

Crude protein fraction
Forage 1 2 3 4 5 6 % Yld
% of CP
Alfalfa
Freeze dried 19.8 7.6 3.4 28.4 4.0 31.9 95.1
Hay 24.2 6.1 2.8 24.7 5.5 . 30.8 94.1
Silage (30% DM) 49.2 2.2 7.6 18.5 3.8 16.7 96.8
Silage (38% DM) 49.1 1.2 6.8 17.3 4.8 17.7 98.1
p, dried vs. ensiledz .0001 .001 .0001 .0001 .05 .(XJOI .05
P, freeze dried vs. .01 .01 NS .0001 .0001 NS NS
hay
P, silage 1 vs. 2 NS .05 NS .05 _.01 NS NS
Bromegrass
Freeze dried 10.0 9.0 2.6 31.8 7.3 35.5 96.1
Hay 15.0 6.4 3.0 24.1 8.8 40.0 97.3
Silage (33% DM) 29.7 1.4 5.9 17.6 7.6 30.0 92.1
Silage (44% DM) 31.4 1.1 4.2 17.7 7.2 37.0 98.7
P, dried vs. ensiled .0001 .0001 .0001 .0001 .05 .0001 NS
P, freeze dried vs. .0001 .0001 .l .0001 .001 .0001 NS
hay
P, silage 1 vs. 2 .01 NS .0001 NS NS .0001 .0001
Canarygrass , -
Freeze dried 12.8 6.9 3.3 33.0 9.4 25.8 91.1
Hay 20.9 7.1 2.6 20.0 6.2 34.0 90.9
Silage (34% DM) 42.8 2.0 4.6 18.3 6.3 22.3 96.1
Silage (43% DM) 36.1 3.4 4.2 16.4 6.6 23.4 90.1
P, dried vs. ensiled .0001 .0001 .0001 .0001 .001 .0001 NS
P. freeze dried vs. .01 NS .05 .0001 .0001 .0001 NS
hay
P, silage 1 vs. 2 .05 1 NS NS NS NS .05
Corn
Freeze dried 11.8 5.0 18.1 22.4 11.8 24.1 93.3
Silage (31% DM) 32.5 1.2 ' 20.2 19.2 8.2 18.3 99.5
Silage (38% DM) 30.8 1.1 20.5 20.2 8.1 18.5 99.2
P. dried vs. ensiled .0001 .0001 .05 .001 .01 .0001 .01
P, silage 1 vs. 2 .05 NS NS NS NS NS NS

 

1Crude protein fractions were extracted sequentially and separated by centrifugation as

follows: 1) soluble in trichloroacetic acid, 2) soluble in reduced media (bicarbonate), 3)
soluble in 80% acetone, 4) soluble in pH-7 detergent at 95°C, 5) soluble in sulfuric acid
and detergent at 95°C, and 6) insoluble.

2Significance of contrast, NS indicates P > .1.

200

TABLE 3. Prediction of forage protein degradation at 2 h determined from multiple
regression analysis of protein fractions.1

 

 

Fraction Estimate S.E. P < '
Model 1 (R2 = .88)
Intercept . 58.6 7.4 .0001
1 43.5 8.2 .0001
2 28.2 22.7 NS
3 -82.4 10.6 .0001
4 -31.4 11.1 .005
5 -103.6 21.3 .0001
6 -525 8.6 .0001
Model 2 (R2 = .82)
1 105.0 3.1 .0001
2 117.8 24.1 .0001
3 -31.0 10.3 .01
4 29.8 9.7 .01
5 -23.4 22.9 NS
6 2.6 6.1 NS
Model 3 (R2 = .76)
1 + 2 103.1 3.3 .0001
3 ,t 4 1.5 5.5 NS
5 + 6 12.7 4.1 .05
Model 4 (R2 = .78)
1+2 996 29 .mmr
4 23.7 7.1 .001
6 3.3 5.3 NS

 

 

 

1Crude protein fractions were extracted sequentially and separated by centrifugation as
follows: 1) soluble in trichloroacetic acid, 2) soluble 1n reduced 1n vrtro med1a
(bicarbonate), 3) soluble in 80% acetone, 4) soluble 1n neutral detergent at 95 C, 5)

soluble in acid detergent at 95°C, and 6) insoluble.

2Probabi1ity that estimate at 0, NS indicates P > .1.

201

TABLE 4. Prediction of forage protein degradation at 24 h determined from multiple
regression analysis of protein fractions.1

 

 

Fraction Estimate (%) S.E. P <
Model 1 (R2 = .81)
Intercept 62.1 8.1 .0001
1 32.7 9.0 .0005
2 3.4 24.8 NS
3 -73.2 11.6 .0001
4 5.7 12.1 NS
5 -138.3 23.2 .0001
6 42.4 9.3 .0001
Model 2' (R2 = .72)
1 98.8 3.3 .0001
2 98.5 26.1 .0002
3 -18.8 11.1 .1
4 70.6 10.5 .0001
5 -53.4 24.8 .05
6 15.9 6.6 .05
Model 3 (R2 = .54)
1 + 2 96.4 3.9 .0001
3+4 237 66 .mms
5 + 6 25.5 5.0 .0001
Model 4 (R2 = .67)
1 + 2 94.2 3.1 .0001
4 60.2 7.5 .0001
6 11.4 5.6 .05

 

1Crude protein fractions were extracted sequentially and separated by centrifugation as

follows: 1) soluble in trichloroacetic acid, 2) soluble in reduced media (bicarbonate), 3)
soluble in 80% acetone, 4) soluble in pH-7 detergent at 95°C, 5) soluble in sulfuric acid
and detergent at 95°C, and 6) insoluble.

2Probability that estimate a: 0, NS indicates p > .1.

202

TABLE 5. Percentage decrease from the initial amount of crude protein in solubility
fractions after 2 h degradation in vitro. Negative values indicate percentage increase in

crude protein in that fraction.

 

 

Me 1 2 3 4 5 6
Alfalfa, early, freeze dried -46.6 83.8 -345.5 6.3 32.6 39.1
Alfalfa, mid, hay -24.0 81.8 -239.0 1.6 11.7 26.8
Bromegrass, late, hay -58.9 72.7 -l83.1 -14.2 1.6 34.5
Canarygrass, mid, freeze dried -43.2 74.9 -269.7 .4 3.0 35.9
Alfalfa, late, silage (%DM) 3.2 -19.4 -7l.5 8.1 6.0 -.7
Bromegrass, early, silage 1.6 ~3.9 -134.7 -4.8 ~17.8 31.9
Corn, late, silage (%DM) 6.7 .3 -21.1 -4.7 3.2 12.3
Canarygrass, mid, silage (%DM) 4.2 -31.6 ~100.7 -7.1 -28.4 32.2
SEM, n = 2 5.1 8.3 25.0 4.6 7.4 6.9
[3’ dried vs. silagez .0001 .0001 .0001 NS .01 .05
P, within dried .l .01 .07 .01 .05 NS
P, within silage NS NS .06 NS 'Ns NS

 

1Crude protein fractions were extracted sequentially and separated by centrifugation as

follows: 1) soluble in trichloroacetic acid, 2) soluble in reduced media (bicarbonate), 3)
soluble in 80% acetone, 4) soluble in pH-7 detergent at 95°C, 5) soluble in sulfuric acid
and detergent at 95°C, and 6) insoluble.

2Significance of contrast, NS indicates P > .1.

203

TABLE 6. Percentage decrease from the initial amount of crude protein in the
respective solublllty fractrons over time across forages. Negative values indicate

percentage increase in crude protein in that fraction.1

0 Hi“ :~- -.~ 2mm nan—a a; .'_o‘ - ,

 

 

Time 1 2 3 4 5 6
Dried forages
2 h -43.2 78.3 -259.3 -1.5 12.2 34.1
6 h _ -46.4 83.3 -469.8 -3.3 54.7 34.6
24 h ~38.5 80.9 -539.2 -.1 37.6 ' 36.1
SEM, n = 8 4.5 1.5 24.1 1.5 3.2 3.3
Ensiled forages I l
2 h 3.9 -13.7 -82.0 -21 ' -9.3 18.9
6 h 11.7 29.1 -155.4 -3.5 17.2 4.9
24 h 7.9 2.5 -230.8 -8.7 30.4 8.0
SEM, n = 8 1.4 5.3 12.8 3.3 5.3 4.0

 

1Crude protein fractions were extracted sequentially and separated by centrifugation as

follows: 1) soluble in trichloroacetic acid, 2) soluble in reduced in vitro media
(bicarbonate), 3) soluble in 80% acetone, 4) soluble 1n neutral detergent at 95°C, 5)
soluble in acid detergent at 95°C, and 6) insoluble.

 

 

204

TABLE 7. Estimation using the Lucas test of percentage protein degraded from
fractions. Slope of the line times 100 equals percentage of fraction degraded. The
negative intercept results from metabolic protein, contamination from other protein

fractions, or inappropriateness of the model.1

 

 

Estimate g 2 4 5 6

2 h degradation
Intercept, % -1.9"'* -2.8 -2.7 -4.2"'
Slope x 100 102.3” 11.6 25.6 48.8”
R2 .99 .17 .06 .77

6 h degradation
Intercept -1.1** -3.4 -6.3 -4.2
Slope x 100 97.5“ 12.6 86.7" 42.8”
R2 .99 .1 1 .26 .54

24 h degradation

. Intercept -l.7** -5.7* -4.5 -9.1**

Slope x 100 102.5M 24.3* 70.4" 70.3"
R2 .96 .31 .47 .83

 

*Probability that estimate at 0, P < .05, or ** P < .01.

1Crude protein fractions were extracted sequentially and separated by centrifugation as

follows: 1) soluble in trichloroacetic acid, 2) soluble in reduced media (bicarbonate), 3)
soluble in 80% acetone, 4) soluble in pH-7 detergent at 95°C, 5) soluble in sulfuric acid
and detergent at 95°C, and 6) insoluble.

205

 

 

 

In vitro media
(bicarbonate buffer)
soluble
insoluble Add trichloroacetic acid
soluble
80% Acetone insoluble E ii I' l
W
soluble
insoluble \ > E . 3
V
Detergent, pH 7.0
soluble
insoluble > E . I

Detergent, 1 N H2804

soluble

insoluble > E . 5

Fraction

Figure 1. Scheme for sequential extraction of forage crude protein.

206

Figure 2. Prediction of forage protein degradation at 24 h from solubility fractions of
original material. A: model included 6 fractions and intercept R2 = .81. B: model
included fractions 1 + 2, 4 and 6, R2 = .67.

nsof

Predicted. protein degradation at 24 h (%)

Predicted protein degradation at 24 h (%)

207

A. Model 1

70-

60:

50-

40-

30-

 

20-

10'

 

 

0 10 20 30 40 50 60 70 80
Measured protein degradation at 24 h (%)

O

B. Model 4

70-
60"
50‘

40'

 

30-

20-

10‘

 

O 10 20 30 40 50 60 70 80
Measured protein degradation at 24 h (%)

O

i}

 

208

Figure 3. Change in crude protein solubility fractions due to degradation in ruminal
enzyme extract. A: mean of 4 dried forages. B: mean of 4 ensiled forages. -- . ---

Soluble in trichloroacetic acid, —- °-—- soluble in reduced media, -—l-- soluble in
acetone, -- I —-- soluble in pH 7 detergent, -- Am soluble in l N H2804 with

detergent, -- Am insoluble CP.

 

 

 

209

A. Dried Forages

 

 

 

 

 

 

.5

to

 

 

nw
2

50.
4o
30.
10.
0

EU .58 cc .5 5:25 5 £32.. 830

Time (h)

B. Ensiled Forages

 

 

 

 

15

IO

 

 

50.
4o
30.
0
10.
0

Co .33 .e .5 =88... ... £38.. ..ch

Time (h)

210

Figure 4. Lucas test for uniformity of degradation in ruminal enzymes across forages.
0 Indicates ensiled forages, I hay. Slope of the line is true degradation at time
indicated. Intercept is "metabolic" protein. A: Buffer soluble CP (fraction 2) degraded
at 2 h, y = 1.02x - 1.9, R2 = .99. B: Neutral detergent soluble CP (fraction 4) degraded
at 24 h, y = .24x - 5.7, R2 = .31.

 

 

 

211

72

1'0

Initial Fraction 2 CP (% of total)

l6

l4

 

 

 

- — _ - _ . 0
mol 8 6 4 2 O 7..
2425 cc see
N :2“an Soc m0 m0 304

 

15

 

 

- — _ - _ _ 0
93:5 mo 6ch
V £030me 89¢ EU MO mmOsH

Initial Fraction 4 CP (% of total)

212

Figure 5. Lucas test for uniformity of degradation in ruminal enzymes across forages.

0 Indicates ensiled forages, I hay. Slope of the line is true degradation at 24 h. Intercept is
"metabolic" protein. A: Acid detergent soluble CP (fraction 5), y = .70x - 4.5, R2 = .47. B:
Insoluble CP (fraction 6), y = .70x - 9.1, R2 = .83.

 

213

 

 

 

 

 

 

_1().
'2‘ o
.2
8A 5-i I
Era
ge
we 0- ° °
€361
‘4-4
0
a ‘5 I I I I r 1
3 0 3 6 9 12 15 18
Initial Fraction 5 CP (% of total)
B

20-
\O
t: 15-
.2
8A 10"
Era
gzé 5‘
i3 0-
05°
“5" -5"
a '10 I I I I I r I I
.3 0510152025 303540

Initial Fraction 6 CP (% of total)

 

 

214

CHAPTER 8: EPILOGUE

The theory that high-producing ruminants respond to balancing rations for ruminally
undegraded and degraded forms of protein is well established. However, the practice of
balancing such rations is in its infantile stage. Previous research described in Chapter 1
has established that the amount of protein optimally degraded in the rumen varies
depending production and nutritional factors. The amount of protein that is degraded in
the rumen depends on animal and microbial factors, and on the protein source. The
current methods to balance rations for degraded and undegraded protein do not begin to
address this level of complexity. None-the-less, they do offer an opportunity to
formulate and evaluate rations that generally meet the needs of ruminants.

Research is needed to improve the estimation of protein requirements of cattle
in order to reduce the amount of protein fed, and reduce costs, animal health and
environmental concerns associated with feeding excess protein. Much of this needed
research depends on better methods of chemical analysis of feeds. There is tremendous
variation in protein degradation of forages (Chapters 1 and 6), including variation
within a feed type, due to chemical or structural associations, supply of microbial
growth or inhibition factors, or effectors of protease enzymes. Chemical analyses need
to be deve10ped that can measure these modifiers of protein degradation in feeds. All
ultimate goal is to evaluate a forage routinely for an individual producer, and determine
the likely protein degradation when fed to a specific animal on a specific ration.

Considering this goal, it is much more advantageous to determine likely rates of

degradation of protein fractions than directly determining the extent. These rates can be

 

 

 

215

used to calculate the ruminal availability of protein by including the rates of passage
into the model. Future work will require estimation of passage rates of fractions, and
factors that affect them. Reduction of the problem of measuring ruminal. protein
degradation to the two sub-problems, rate of degradation and rate of passage, allows for
potentially greater understanding than could be obtained by trial and error evaluation of
an infinite number of rations.

Each of these sub-problems can be divided further into additional studies.
Factors that affect protein degradation include the susceptibility of the feed to
degradation, enzyme profiles in the rumen, and enzyme activity levels. The
susceptibility of the feed to degradation was the part of the puzzle that was addressed
directly in this dissertation. The hypothesis was that certain individual protein fractions
separated based on solubility would degrade completely at uniform rates. If these
fractions were uniform across forages, values for rates of degradation of the fractions
could be compiled, and forages could be analyzed routinely for protein degradation
pr0perties based on the composition of fractions. Further research could address factors
that affect rates of degradation of fractions, and the susceptibility of fractions to be
affected.

The research presented in this dissertation identifies several additional problems
that must be solved to analyze forages as described. The methods used to fractionate
proteins require further deve10pment. The methods used to solubilize CP in buffers
described in the Appendix and in Chapter 2 appeared to solubilize more true protein
than the methods used in Chapter 7. This comparison is made across studies with
different forages, but the differences might warrant further investigation. Differences in
the methods used included use of fresh, undried material that was macerated in a tissue
homogenizer for the first studies vs. use of freeze-dried material that was ground in a

Wiley mill. In the studies on solubility described in the Appendix, even hay samples

 

 

 

 

 

 

216

which were ground in the Wiley mill were also macerated in the tissue homogenizer to
be consistent with treatment of fresh forages. Another difference is that simpler buffers
were used in earlier studies, but bicarbonate-based in vitro media was used with later
studies. This difference was introduced so that solubility could be determined after in
vitro degradation in the same buffer as used for degradation. A preliminary study did
compare solubilization of alfalfa hay in phosphate buffer to in vitro media and showed
no differences. Solubility of CP in buffers and TCA was also shown to increase with
longer duration in the solvent, especially at higher incubation temperature (Appendix).
Serine-protease inhibitors failed to prevent this solubilization in TCA, but other
protease inhibitors may have been more successful and may be needed for these
analyses.

The means to preserve fresh forage samples before chemical analysis requires
further attention. It was shown to be important not to allow samples to be frozen and
thawed before analysis, and freeze drying was preferred (Chapter 2). However, further
work will require comparison of fresh material to freeze dried samples to evaluate
whether even this treatment is adequate. Addition of chemical reducing agents was
needed to prevent precipitation of protein in buffers (Chapter 2). Methods to preserve
samples on the farm for laboratory analysis will need to be developed.

The laboratory fermentation method to measure protein degradation was labor
intensive and had high variation within treatments. Some improvements in this method
were made, and should be further validated. The importance of the method relates to
validation of other procedures with a more physiological approach. The in vitro
fermentation offers advantages over in situ fermentation in that total microbial growth
can be evaluated, and end-products can be measured. The degradation of protein using
crude extract from rumen contents offered several interesting observations. It appeared

that proteolytic activity was resistant to self—degradation, and appeared to be auto-

 

 

 

 

 

217

activated. None-the—less, protein degradation rate was slow and not uniform within
feeds. Rate of buffer solubilization was greater than TCA solubilization when degraded
with enzyme, suggesting that proteins become filterable before complete degradation.
Protein degradation of alfalfa hay was not dependent only on proteolytic activity, but
responded to addition of cellulase. These observations suggest that part of the reason
for the slow degradation of protein with enzyme extract compared to degradation by
fermentation may have been due to absence of non-proteolytic enzymes that are
required for degradation. If certain proteins depend on fiber digestion for their own
degradation, then ruminal availability of these must take into account rate of
degradation of fiber. This is another nutrient that degrades at differing rates across
forages and is difficult to predict.

These observations are important to consider when evaluating the degradation
of protein fractions. While soluble protein was immediately degraded with enzyme
extract, other fractions were degraded to differing extents within early points of the
incubation, but degradation of some of these fractions slowed thereafter. The slowed
degradation rate must have been due to lack of uniformity of the fractions with respect
to degradation by rumen extract. Would these fractions have been uniform with respect
to degradation by ruminal fermentation? The broader degradation by fermentation may
have made parts of proteins available for degradation and prevented the slowing in
degradation rate. In any case, the fractions selected were not uniform for degradation
with enzyme, and so none of the fractions appeared to isolate only proteins that were
dependent on non-proteolytic enzymes for their digestion.

The problem of a lack of uniformity of protein fractions isolated can also be
explained by the possibility that proteins cannot be separated by solubility into uniform
fractions. It is possible that even a purified protein may degrade more slowly over time

when in the presence of excess enzyme. The proteolytic enzymes cleave proteins

 

 

 

 

 

 

 

218

between certain residues as they become available. Might the number of available sites
decline over time, and in fact might certain resistant peptides accumulate after
degradation? If this is the case, isolating uniform fractions by solubility is not possible.
The extent that this hypothesis is true if at all requires further investigation because it
may be a severe limitation to development of solubility methods.

The problem of predicting protein degradation of forages routinely in the
laboratory is far from settled. This dissertation describes research that is more basic
and systematic than many previous efforts that merely correlated simple laboratory
methods with in situ data. Due to this approach, some gains have been made into
understanding how proteins are degraded. Further efforts will have to address some of
the new questions. What means of macerating and solubilizing proteins from forages
are appropriate to isolate uniform fractions? How should forages be preserved before
analysis? What are the structural limitations from non-protein forage components to
degradation, and how are these measured? What is the nature of proteolytic enzymes in
the rumen? Do individual proteins degrade completely at the same rate? The methods
developed in this dissertation with further refinement might be used to answer some of

these questions.

 

 

 

219

APPENDIX : PRELIMWARY EXPERIMENTS AND OBSERVATIONS

Observations on Protein Solubility

Measurement of Forage CP Solubility. An experiment using bovine albumen and
bromegrass hay demonstrated that TCA was inefficient at precipitation of CP when
samples were filtered through Whatman #54 filter paper instead of precipitated by
centrifugation. The TCA soluble CP was higher for both samples, but more so for
albumen. While the TCA addition to albumen solution turned the solution from clear to
white, this protein was entirely filterable. Centrifugation speed (20,000 g vs. 1500 g)
was compared for TCA precipitation of CP from bromegrass hay and albumen, with no
difference between the two. Greater than 96% of the albumen was precipitated at either
speed. There was no difference between centrifugation for 15 min or 30 min on CP
precipitation of these same samples, and there was no improvement in precipitation by
addition of deoxycholate (2 mM final concentration) to the solution. Lowry's method to
determine CP in orchard grass was also attempted, but soluble CP was 45% higher
when measured this way compared to Kjeldahl. Precipitation of protein and
measurement of precipitate by Lowry's method resulted in 67% as much protein
predicted compared to Kjeldahl. Interferences with Lowry's method appear to be
abundant in forages.

Bromegrass hay was solubilized in water (60 ml/ g DM) by adding water to
sample at 25 °C and processing immediately. Samples (3 replicates of each) were
strained through cheese cloth or filtered through Whatman #54 filter paper. The filtrate

was sampled for N and the remaining sample was centrifuged at 1500 g for 30 min.

 

 

 

;— h

_._.___

 

 

 

 

 

 

 

 

220

True protein was precipitated with TCA (final conc. 6%, va01) and samples were
centrifuged.

Results are shown in Appendix Table 1. Filtering removed protein from
solution compared to straining through cheese cloth, and centrifuging after filtering
removed additional CP from the water. There was no effect of filtering before
centrifuging, compared to straining before centrifuging. However, TCA precipitated
less CP for filtered samples. Filtering was a slow process, and it was speculated that
filtering may have provided time for plant proteolytic enzymes to degrade some water
soluble protein to become TCA soluble. A subsequent experiment demonstrated an
increase in TCA soluble CP in bromegrass by keeping it for 4 h in water at 25 °C vs. 15
min, but these samples were not different in buffer soluble CP.

APPENDD( TABLE 1. Effect of straining through cheese cloth, filtering through

Whatman # 54, or centrifuging (1500 g) on water soluble CP, or trichloroacetic acid
(T CA) soluble CP in orchard grass hay.

 

 

Treatment Soluble CP (% of DM)
Filtered 5.7b
Strained ~ 6.7a
Filtered, centrifuged 4_9c
Strained, centrifuged 4.5C
Filtered, TCA precipitated 3.6d
Strained, TCA precipitated 3.06
Standard Error of Mean (n = 3) .14

 

adeeMeans with different superscripts are significantly different, P < .05.

Effect of pH and Eh on CP Solubility. An experiment was to evaluate effect of pH on
buffer- and TCA-soluble protein of orchard grass hay. Soluble— and TCA—soluble

crude protein in orchard grass bay (.5 g DM) were determined by storage with in vitro

 

 

 

 

221

media (30 ml) for 4 h at 25 °C under C02 pressure. The in vitro media was prepared
with or without cysteine as a reducing agent (.6 m g/ml) and at 3 pH levels: 3, 7 and 10.
Buffer solubility was determined by measurement of N in the supernatant after
centrifuging (1500 g, 15 min) with subtraction of N from cysteine. In retrospect,
sulfide would have been a better reducing agent. Solubility in TCA was determined by
adding 10 m1 of 24% TCA (wt/vol) to sample that had been in 30 m1 of media for 4 h,
and measuring N in supernatant after centrifuging. The pH and Eh of each tube were
measured after centrifugation.

Buffer pH and Eh (-340 to -250 mV) were maintained after soaking the sample.
Buffer solubility increased at higher pH in unreduced samples, and solubility in both
buffer and TCA seemed to increase with reduction, but since cysteine was used as
reducing agent, this result may have occurred because of a mis—estimate of cysteine.
The addition of TCA lowered pH to about 1.0 regardless of initial pH, but after addition
of TCA, Eh ranged from +290 mV for unreduced samples and +180 mV for reduced
samples. Buffering from media appeared to have no effect on TCA solubility. After
addition of acid and efflux of C02, volume of media decreased, thus final concentration
of TCA differed from expected by adding the sum of the parts. Higher buffer pH
resulted in greater solubilization of true protein from orchard grass hay.
Effect of Temperature on Proteolytic Activity of Alfalfa, Effect of soaking mid-
maturity freeze-dried alfalfa (CP = 20.5%) in reduced in vitro media on TCA soluble
CP was determined. Alfalfa samples (8 mg N) were soaked in media (30 ml) at 4°C or
38 °C in the absence or in presence of serine-protease inhibitor (3 mM
phenylmethylsulfonyl fluoride, PMSF). These sammes were incubated for 3 h, but
additional samples were incubated without protease inhibitor at 4 °C for 0 or 16 h.

The results are shown in Appendix Table 2. There was no effect of adding

PMSF to samples incubated at either 4 °C or 38 °C. However, samples stored at 4 °C

 

 

222

had lower TCA soluble CP than samples stored at 38 °C, and this value was not
different from samples that were not stored in buffer (Treatment 5). Storage at 4 °C
resulted in a slight increase in. TCA soluble CP at 16 h compared to 0 h. These results
demonstrate that alfalfa samples may be soaked for up to 3 h at 4 °C without affecting
the CP solubilized in TCA. However, soaking at 38 °C resulted in solubilization of CP
in TCA which could be associated with plant proteases. However, these potential
proteases were not inhibited by PMSF.

APPENDD( TABLE 2. Effect of soaking one tenth bloom freeze-dried alfalfa on
protein solubilized in TCA.

 

 

 

Treatment Time Temperature Inhibitor % TCA sol.
1 3 h 4 °C PMSF 24,20
2 3 h 4 °C None 23_5C
3 3 h 38 °C PMSF 28.9a
4 3 h 38 °C None 30.021
5 0 h 4 °C None 24.2c
6 16 h 4 °c None 26.7b

 

Means of each treatment are shown. Standard error of mean (n = 3) was .57.

abCMeans with different superscripts differ using student T test (P < .05).

Observations on In Vitro Fermentation
Titration of in vitro media. Media (25 ml) used for in vitro rumen batch culture was
titrated .to determine the amount of HCl required to lower pH to 4.0 in order to st0p

activity after digestion and to remove H28 gas from the media. Addition of 3 m1 of 1 N

HCl lowered pH to 2.5 with the pH dropping rapidly after reaching 6.0. However, over

223
time the pH increased again and 1 ml 6 N HCl is recommended to lower pH if this step
is necessary.
Variation in pH and reducing potential for in vitro batch culture. This experiment
determined the stability in pH and reducing potential of the in vitro environment. Eight
treatments (2 reps each) compared pH and Eh over time for different energy and protein
contents. The treatments (expressed as g/tube) were low energy (.25 g cellulose, .25 g
starch) and high energy (.5 g starch, .5 g cellulose) arranged factorially with no protein
(.0156 g cysteine as only N), low protein (.0156 g cysteine + .0312 g casein) and high
protein (.156 g cysteine + .0625 g casein). An additional treatment included 1 g of each
starch and cellulose and .0625 g protein.

APPENDIX TABLE 3. Effect of starch and cellulose level (E) and casein level (P) on
pH and Eh of in vitro batch culture.

 

 

Imment 0.11 2h 21h 48h 12h l.QO_h
pH Eh Eh pH Eh pH Eh pH Eh pH Eh
low E no P ~289 6.8 ~318 6.7 ~281 6.3 ~210 6.1 ~141

~279 6.7 ~329 6.7 ~288 6.3 ~l99 6.2 ~178
lowElowP 6.9 ~278 ~248 6.6 ~382 6.2 283 6.1 ~353 5.9 ~188
6.9 ~286 ~290 6.2 ~363 6.1 ~242 5.9 ~237 5.9 ~175

lowEhiP 6.8 -240 ~l90 6.4 ~312 6.0 ~l94 5.7 ~102 ~~ ~-
6.8 ~272 ~280 6.4 ~341 6.1 ~24] 5.8 ~167 5.9 ~127

hiEnoP ~280 6.6 ~325 6.5 ~275 6.1 ~l91 5.8 ~l67
~285 6.5 ~341 6.5 ~324 6.0 ~224 5.8 ~174
hiElowP ~280 6.3 343 5.7 197 5.5 147 5.5 ~135
289 6.3 ~311 5.8 ~182 5.5 ~139 5.4 ~l30
hiEhiP ~265 6.3 -326 5.7 ~183 5.4 ~102 5.4 ~77
~25] 6.3 ~33? 5.8 ~l93 5.5 ~135 5.4 ~85
veryhiEhiP ~250 5.6 ~433 5.3 ~307 5.0 ~l81 5.0 ~119

~290 5.9 ~333 5.5 ~247 5.3 ~157 5.2 ~129

 

 

224

Both energy and protein level affected pH and Eh despite the fact that energy
was supplied in great excess of that of protein (Appendix Table 3). This suggests that
adding a lot of extra energy will change the fermentation. Simply oversupplying
energy is not appropriate. Instead, an amount of starch or fiber should be added so that
tubes testing both forages and concentrates have the same amount of each energy
source. The amount of cysteine added may need to be reduced and the total amount of

protein and energy should be at least at the low values shown here.

Effect of Reducing Agents on Microbial Fermentation, An experiment was
conducted to determine the effect of ascorbate on microbial growth in vitro. Media (30
ml) containing soybean meal (100 mg), starch (200 mg) and cellulose (160 mg) was
incubated for 3, 8 or 24 h with rumen inocula (8 ml). Media (30 ml) was previously
reduced with sodium salt of ascorbate (0.4 g/L) and cysteine (0.04 g/L), and was
labeled at reduction with 353-cysteine (5.0 x 105 dpm).

APPENDIX TABLE 4. Effect of sodium ascorbate addition on precipitation of 358
(dpm) with soybean meal fermented with rumen microorganisms.

 

 

Ascorbate Time Rep 1 Rep} Rep 3
No 3 h 31,447 32,596 23,617
8 h 41,404 49,872 50,426
24 h 49,872 42,511 35,021
Yes 3 h 14,766 16,468 17,617
8 h 30,851 31,149 32,553
24 h 34,638 36,638 41,191

 

 

Precipitable 35S (dpm) is given in Appendix Table 4. The level of N in the
pellet did not change consistently due to the treatment but was greater for 8 and 24 h

than for 3 h (data not shown). Protein degradation was not calculated because 35S:N

 

 

 

 

225

ratios of microbial protein could not be determined. The TCA precipitated 358
increased over time and was reduced by addition of sodium ascorbate. These results
imply that addition of ascorbate inhibited microbial growth and protein degradation
(since precipitated N was constant).

Effect of carbohydrate source on microbial growth and 3SS:N ratio.
Several carbohydrates were compared for their ability to support microbial growth on
trypticase (36 mg). Treatments were corn starch (200 mg), corn starch (400 mg),
gelatinized corn starch (200 mg DM), cellulose (160 mg) wheat straw neutral detergent
fiber (W SNDF) (200 mg), starch and cellulose, and no carbohydrate. Gelatinized starch
was prepared by heating 10% starch solution for 50 min until it became viscous.
Samples were incubated for 22 h in media containing rumen inocula, 35S-cysteine (8.9
x 105 dpm) and cysteine (lg/1). Samples were rinsed three times in solution containing
1% HBr and 1 g/L of each cysteine.HCI, Na2804 and NaSO3. Radioactivity in the
pellet was determined after digestion in H2804 with H202. The pH was reduced from
6.85 to 6.65 for gelatinized starch compared to non-gelatinized at 4 h of incubation. It
was reduced from 6.54 to 6.40 at 22 h of digestion for gelatinized and not gelatinized
starch receptively.

The results in Appendix Table 5 indicate that the presence of starch in the in
vitro media is important to support microbial growth. The various starch treatments
were not different from one another in precipitable N. There was no difference in
microbial N precipitation between cellulose, no energy and wheat straw NDF. The
dpm:N ratio was much higher for WS NDF suggesting that the precipitation of non-
microbial 358 would be a bigger problem with some feeds than others. The level of
unlabeled sulfur would have to be decreased to allow prediction of precipitable

microbial N from precipitable S.

 

226

APPENDD( TABLE 5. Precipitation of 35S after incubation of 358-cysteine with
rumen microorganisms, trypticase and the carbohydrate sources listed as treatments.

 

 

Treatment dpm N (mg) dpm:N

No addition 22,3974,b 1.5a 15,904a
200 mg starch 32,7801),C 3.1b 10,629a
400 mg starch 40,3120 35*) . 11,750a
Gelatinized starch 31,901b.c 3.4b 9,404a
Cellulose 15,7 30a 1.2a 13,047a
Starch + Cellulose 39,248c 3.1b 10,868a
ws NDF 55,078d 1.8a 30,860b
SEM (n = 3) 3284 0.2 2,089

 

abcMeans with different superscripts differ (P<.05).

Observations on Crude Rumen Fluid Extract

Reducing agents on protein degradation. Sodium sulphide (1.2 g/l),
cysteine.HCl (1.2 g/l), a combination thereof (.6 gll each) and no reduction were
compared for soybean meal degradation at 16 h. All samples were incubated with
perfusion of C02 over samples in the presence of crude extract from rumen fluid.
Samples were accidentally incubated at 30 °C instead of 38 °C. Appendix Table 6
shows that, though reducing agents are necessary, either sulfide, cysteine.HCl or both
were equally adequate. These results contrast those from azocascin degradation, which
showed cysteine to result in higher proteolytic activity than sulfide, because the method
used for feed degradation may not allow sulfide to escape as easily due to the larger
volume of media. However, if C02 is diffused over the samples for the entire time it

might result in a loss of reducing potential from evolution of H28.

 

 

227

Alfalfa hay was degraded using several methods of reducing the sample and
using other variations. In vitro media (10 ml) was reduced with .6 gll of each NaS and
cysteine.HCl except where indicated otherwise. Feeds were incubated with 0, 5 or 10
ml of enzyme for 16 h and compared to 0 or 5 ml of enzyme that was immediately
precipitated with feed. One treatment differed in that it was presoaked with reduced
media for 2 h before addition of enzyme and 16 h incubation. Variation in reduction of
media included no reduction, saturation with C02 by dispersion of gas in media for 30
min after reduction (allows removal of some H28 gas), use of 25 ml media instead of
10 ml, and use of 3 gll of each reducing agent instead of .6 gll. Each treatment was
replicated twice. '

APPENDIX TABLE 6. Effect of reductants on soybean meal (SBM) degradation with
enzyme extracted from the rumen.

 

 

Reduction SBM degradation (%)
No reduction 6.4b

Na2S (1.2 gll) 20.8a

Cysteine.HCl (1.2 g/l) 23.1a

Na28 + Cysteine.HCl (.6 gll each) 243a

 

aDifferent superscripts indicate differences among means, 11 = 3, P < .05).

Results of the experiment on reduction and other treatments of luceme are
difficult to interpret owing to the small sample size. These results are shown in
Appendix Table 7. Since there were only two replicates per treatment raw data are
shown, but specific contrasts were made with certain combined treatments.
Comparison of treatments 1~2 with 4-5 shows that alfalfa does in fact appear to degrade

over time by the plant's own proteases (P > .0001). One may interpret this increase in

228

APPENDIX TABLE 7. Effect of the presence of enzyme and reducing method on
protein degradation of alfalfa.

 

 

 

% Alfalfa degraded
Treat— Enzyme Time (h) Media Presoak Reduce Rep 1 Rep 2
ment (ml) (ml)
1 0 16 10 no yes 51.1 45.1
2 5 16 10 no yes 44 44.3
3 10 16 10 no yes 44.3 44.2
4 0 0 10 no yes 30.7 32.3
5 5 0 10 no yes 30.6 30
6 5 16 10 yes yes 48 46.7
7 0 16 10 no no 45.4 41.7
8 5 16 10 no no 42.4 42.6
9 0 16 10 no before 41.0 41.7
C02
10 5 16 10 no before 42.8 45.8
C02
1 1 0 16 25 no yes 47.0 46.8
12 5 16 25 no yes 44.5 46.5
13 0 16 10 no 3 gll 40.2 42.2
14 5 16 10 no 3 gll 46.5 46.2

 

 

229

degradation as the leaching of NPN out of the luceme forage. However, in another
experiment where alfalfa was soaked in the refrigerator, the increase in TCA soluble
protein was not observed. Contrast of treatments 1, 4, 7, 9, 11 and 13 with treatments
2, 5, 8, 10, 12 and 14 shows the effect of adding the ruminal enzyme, and it was not
significant P > .1). This degradation in the absence of enzymes was not found for
SBM. It leaves questions about the value of extracting ruminal enzyme for forage
degradation. Soybean meal protein did not degrade to any appreciable extent without
addition of enzyme, indicating that the enzyme does in fact degrade feed proteins.
Contrast of slightly reduced treatments, 7-10 with highly reduced treatments 11-14 was
significant (P < .02) indicating that careful reduction can affect luceme protein
degradation.

Centrifugation of rumen extract before use. Effect of centrifuging enzyme
extract after thawing and before incubation with enzyme on CP degradation of soybean
meal and alfalfa hay. Thawed enzyme extract (5 ml) was centrifuged at 26,000 g before
adding to soybean meal samples, or it was not centrifuged. The experiment was
repeated twice with enzyme extract from different days with two replicates per day.
The first time, 2.65 mg N precipitated after 16 h in samples incubated without feeds
(blanks) for uncentrifuged enzyme extract compared to non-detectable levels for
centrifuged enzyme. Soybean meal was 45.6% TCA soluble after 16 h degradation for
samples incubated with uncentrifuged enzyme compared to 54.3 % for centrifuged
enzyme. The second time, blank N was .75 mg for uncentrifuged vs. non-detectable
levels for centrifuged extract, and soybean meal was 54.2 vs. 58.8 for uncentrifuged and
centrifuged enzyme extract respectively. A different study measured the proteolytic
activity on azocascin on both centrifuged and uncentrifuged samples of enzyme extract.
There was no loss of proteolytic activity from centrifugation an additional time, but the

change in the amount of blank N indicates that the interference is greatly reduced.

 

 

 

 

 

230

Higher predictions of protein degradation are observed, especially when blank N is
high, for centrifuged enzyme. These observations led to centrifuging enzyme before
use for all experiments. I
Presoaking of Alfalfa. An experiment with two replicates of each treatment
per run for two runs, showed that alfalfa hay CP was more soluble in TCA for samples
that were soaked in cold buffer for 2 h before addition of enzyme extract for 2 h than
for samples that were added directly to enzyme extract. In this experiment, there was
no change in TCA soluble CP from 0 to 2 h for the samples that were not presoaked,
but 11.5 % of initially insoluble CP was solubilized for presoaked samples. There was

no significant effect of the same treatments on soybean meal. Since there was a

 

reduced lag when samples were presoaked in cold media before addition of enzyme
extract, and since there was negligible degradation of CP by plant enzymes at 4 °C,
presoaking at 4 °C was adapted at a treatment for all feeds subjected to degradation
with crude rumen extract.

Tungstic acid vs. TCA precipitation. An experiment was conducted to
compare tungstic acid vs. TCA precipitation of corn and soybean meal protein and
enzyme extract. Treatments included soybean meal or corn grain (8 mg N) or no added
feed incubated for 0 h or 16 h with 10 ml ruminal extract prepared as described in
Chapter 5, except it was not centrifuged before use. Each feed was also precipitated
immediately without addition of enzyme, and each treatment was replicated twice.

Tungstic acid precipitated 70% of the enzyme preparation before incubation
compared to 42% for TCA (Appendix Table 8). Corn protein appeared poorly
degraded when TCA was used, and actually appeared to increase when tungstic acid
precipitation was used. Feeds precipitated without extract and without incubation

resulted in greater precipitation of feed than that calculated from subtracting blanks

 

231

from feeds precipitated with enzyme (Appendix Table 9). Enzyme extract or feed CP
appeared to precipitate to a lessor extent when together than when separate.

232

APPENDIX TABLE 8. Precipitation of CP before and after 16 h incubation with crude
rumen extract using trichloroacetic acid (TCA) or tungstic acid (W).

 

 

 

 

0 h
Feed TCA w TCA w
mg N
Enzyme only 5.6b 9.5b 2.0c 5.4b
SBM w/ enzyme 11.19 15.08 6.0b 10.1a
Com w/ enzyme 11831 14.2a 7.7a 11.5a
SEM (n = 3) .35 .39 .12 .40
% of Total N Precipitated
Enzyme only 41.83 70.1%! 14.79 39.851
SBM w/ enzyme 51.7a 69.93 28.0b . 47.0a
Com w/ enzyme 54.7a 658%! 35.7a 53.3a
SEM (n = 3) 4.2 4.1 1.1 4.4
% of 0 h Precipitated
Enzyme only 35.2b 56.8c
SBM w/ enzyme --- --- 54.2a 67.3b
Com w/ enzyme 65.33 81.03
SEM (n = 2) -- 3.2 2.1

 

ab‘3Means with different superscripts differ P < .05.

 

233

APPENDD( TABLE 9. Interference of crude rumen extract with measurement of
precrpltable protein (mg N of total 8 mg) in corn or soybean meal (SBM) before
degradation.

 

 

Feed TCA Tungstic acid
Corn with rumen extract 6.1 4.7
Corn alone 6.8 6.3
SBM with rumen extract 5.5 5.6
SBM alone 6.6 6.3
SEM (n = 2) .35 .39

 

1Feed protein was precipitated from buffer that either contained rumen enzyme extract
or did not contain extract. Feed protein precipitated was calculated as total CP
precipitated minus precipitated CP in media with only extract for samples combined
with extract. Main effect of presence of extract was significant P < .05.

 

 

 

 

 

II

 

HUI/I21!

LIBRQRIES

W)

I

”thrust.

». . «in

HICHIGRN STATE UNIV

. . Iii?
stun...“ .42.... . .
.2 Kill...

. art. . a}.
rMHFEt

... ...“...
I» ...

4.111.?
...i. 1531
I 4.34..