MICHIGAN STATE UNIVERSlTY
TADD EVERETT DAWSON
' 1994 4

? P‘Romo'mc ACtD P’Rob'u'cmc 'BACTERiA As“ [5}
, atomocumms FOR THE PRESERVATION 0F,
' ENSILED HIGH" MOISTURE CORN .1{:;’::f:_:{52‘5:;:3:2;.fEff-331$:

 

 

!"

6

.3 - WWIIWHWHUUHI!HIM!MW”WWW”?!

1593 7984 '
(f 0 7

 

PLACE IN REI‘URN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

mﬁﬂ“

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1/98 chlRCJDdoDuapGS-p.“

 

PROPIONIC ACID-PRODUCING BACTERIA
AS BIOINOCULANTS FOR THE PRESERVATION
OF ENSILED HIGH-MOISTURE CORN

By

Tadd Everett Dawson

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Animal Science

1 994

ABSTRACT
PROPIONIC ACID-PRODUCING BACTERIA
AS BIOINOCULANTS FOR THE PRESERVATION
OF ENSILED HIGH-MOISTURE CORN
By

Tadd Everett Dawson

Aerobic instability is the major muse of spoilage, nutrient loss and feed refusal of ensiled,
high-moisture corn. In addition, some organisms responsible for aerobic deterioration of
ensiled, high-moisture corn can pose a health risk to livestock and agricultural workers
through production of harmful compounds such as mycotoxins. A series of studies was
conducted to address the hypothesis that addition of selected, propionic acid-producing
bacteria will increase propionic acid content of ensiled, high-moisture corn and
consequently decrease aerobic deterioration by controlling the growth of microorganisms
implicated in causing aerobic deterioration of ensiled feeds. In experiment one, a mixture
of two strains of cheese associated propionic acid-producing bacteria (Propionibacterium
acidpropionici and P. freudenreichii) was applied with or without a commercial lactic acid-
producing bacteria bioinoculant to reconstituted high-moisture corn. The mixture of P.
acidpropionici and P. freudenreichii was effective in increasing propionic acid
concentration to .26 g/ 100 g DM and decreasing numbers of yeasts and molds after 56
days of ensilement. Experiment two focused on enriching and isolating propionic acid-

producing bacteria from ensiled feeds. A method was developed which facilitated isolation

of a rapid growing, high acid-producing strain of P. acidpropionici (designated DH42)
from ensiled, high-moisture corn. Propionic acid production and growth rate of eight
strains of propionic acid-producing bacteria, including P. acidipropionici DH42, when
grown on substrates and competing with microorganisms found in ensiled, high-moisture
corn were evaluated in experiment three. Propionibacten’um acidipropionici DH42 had the
greatest propionic acid production and growth rate indicating it was more competitive in
ensiled, high-moisture corn. In the fourth experiment, P. acidipropionici DH42 was used
as a bioinoculant in high-moisture corn. Addition of P. acidipropionici DH42 increased
propionic acid content to .32 g/ 100 g DM after 42 days ensilement and to .61 .g/ 100 g
DM after ﬁve days aerobic exposure of the ensiled material compared to less than .05 g/
100 g DM in control corn. Aerobic deterioration was dramatically reduced in the treated,
high-moisture corn as evidenced by lower temperature, pH, numbers of yeasts and molds,
total aerobic bacteria and Acetobacter during ﬁve days aerobic exposure of the ensiled
material. A ﬁnal study was conducted in which a non-radioactive nucleic acid probe to
detect P. acidipropionici DH42 was developed. In conclusion, selected, propionic acid-
producing bacteria are capable of being effective bacterial inoculants for preservation of

ensiled, high-moisture corn.

This dissertation is dedicated to my Mother and the loving memories of my Father,

Grandfather and Great-Grandmother whom all instilled the desire to pursue and to learn.

ACKNOWLEDGMENTS

I would like to thank Dr. Melvin Yokoyama and Dr. Steven Rust for serving as co-
advisors during my doctoral program and providing laboratory facilities to complete the
research. In addition, gratitude is extended to my committee members Dr. Mike Allen,
Dr. Dennis Banks, Dr. John Breznak and Dr. Dave Odelson for their helpful input.
Special thanks are extended to Dr. Mike Woolford for serving as external examiner.

Special thanks are offered to Dr. Chad Risling of Chr. Hansen's Biosystems, Dr.
Bonita Glatz of Iowa State University and Dr. Mahendra Jain of Michigan Biotechnology
Institute for providing strains of Propionibacterium spp. used in this research. I am
indebted and thankful for the assistance of Mike Guettler in the analysis of bacterial
cellular fatty acids and use of photomicrographical equipment. I would also like to thank
Ron Southwick and Dr. Telmo Oleas for their assistance in nucleic acid methods.

The research progressed very well due to the help of Dr. Susan Hengemuehle's
extraordinary acquisition and morning beverage skills. Thanks are also extended to Tom
Pilbeam for discussion of the research and pertinent topics. Mike Schlegel, Shan Chung
and Abner Rodriguez were very helpful in the progress of the research and I am very
thankful. I am enormously indebted to Dr. Margaret Benson for being a fantastic teaching
mentor and inspirational in showing that lights do exist at the completion of the graduate

school tunnel.

I cannot offer enough thanks to the two very special ladies in my life for their
moral and ﬁnancial support during the research and preparation of this dissertation. The
long distance support of my Mother has been a very comforting support mechanism and
has helped me meet and successfully face many challenges. Words cannot offer worthy nor
sufﬁcient enough gratefulness to Veronica Groshko for all the ﬁnancial and emotional

support she has been during the preparation of this dissertation.

vi

TABLE OF CONTENTS

LIST OF TABLES .......................................... x

LIST OF FIGURES ........................................ xvi

INTRODUCTION .......................................... 1

CHAPTER

1 - REVIEW OF THE LITERATURE ............................. 6
Importance of ensiled feedstuffs ............................. 6
Phases of the ensilement process ............................. 7
Microbiology and biochemistry of the ensilement process ............. 7
Inﬂuence of oxygen on ensilement and aerobic deterioration .......... 10
Microbiology associated with aerobic deterioration ................ 14
Methods of preventing aerobic deterioration .................... 17
Effectiveness of propionic acid as a preservative .................. 20
Mechanisms of antimicrobial action of propionic acid .............. 21
Why use propionic acid-producing bacteria? .................... 23
Historical perspective for propionic acid-producing bacteria .......... 25
Previous research with propionic acid-producing bacteria ............ 26
Enrichment and isolation of propionic acid-producing bacteria ......... 30
Use of nucleic acid probes in microbial ecology .................. 31

2 INOCULATION OF RECONSTITUTED HIGH-MOISTURE CORN WITH
PROPIONIC ACID-PRODUCING BACTERIA WITH OR WITHOUT

ADDITION OF LACTIC ACID-PRODUCING BACTERIA ........... 34
Abstract ........................................... 34
Introduction ......................................... 35
Experimental Procedures ................................ 36
Results and Discussion .................................. 38
Implications ......................................... 49

vii

CHAPTER

3 ENRICHMENT AND ISOLATION OF PROPIONIC ACID-

PRODUCING BACTERIA FROM ENSILED FEEDSTUFFS ........... 50
Abstract ........................................... 50
Introduction ......................................... 51
Experimental Procedures ................................ 52
Results and Discussion .................................. 57
Implications ......................................... 58

4 CHARACTERIZATION OF PROPIONIBACIFRI UM ACmIPROPIONI CI DH42,
A PROPIONIC ACID-PRODUCING BACTERIA ISOLATED FROM

ENSILED, HIGH-MOISTURE CORN .......................... 59
Abstract ........................................... 5 9
Introduction ......................................... 60
Experimental Pro'cedures ................................ 60
Results and Discussion .................................. 63
Implications ......................................... 78

5 COMPETITION OF PROPIONIC ACID-PRODUCING BACTERIA
IN A MIXED MICROBIAL COMMUNITY FROM

ENSILED, HIGH-MOISTURE CORN .......................... 79
Abstract ........................................... 79
Introduction ......................................... 80
Experimental Procedures ................................ 80
Results and Discussion .................................. 83
Implications ......................................... 96

6 MANIPULATION OF THE FERMENTATION AND AEROBIC STABILITY
OF ENSILED, HIGH-MOISTURE CORN BY ADDITION OF

PROP] ONIBA CIERI UM ACmIPROPI 0N1 CI DH42 ................. 97
Abstract ........................................... 97
Introduction ......................................... 98
Experimental Procedures ................................ 99
Results and Discussion ................................. 103
Implications ........................................ 121

7 NON-ISOTOPIC NUCLEIC ACID PROBES FOR PROPIONIC ACID-

PRODUCING BACTERIA IN ENSILED, HIGH-MOISTURE CORN ..... 122
Abstract .......................................... 122
Introduction ........................................ 123
Experimental Procedures ............................... 125
Results and Discussion ................................. 131
Implications ........................................ 138

viii

CHAPTER

8 SUMMARY AND CONCLUSIONS .......................... 139

APPENDIX

A DIVERSITY OF THE MICROBIAL COMMUNITY ASSOCIATED WITH
AEROBICALLY-EXPOSED CORN SILAGE INOCULATED WITH

PROPIONBACYERI UM ACDIPROPIONI CI DH42 ................ 144
Abstract .......................................... 144
Introduction ........................................ 145
Experimental Procedures ............................... 146
Results and Discussion ................................. 147
Summary ......................................... 151
Implications ........................................ 151

B SUPPORTING SMALL AND LARGE SCALE STUDIES EVALUATING
PROPIONIBA CIFRI UM ACDIPROPIONI CI DH42 AS A BIOINOCULANT

FOR ENSILED, HIGH-MOISTURE CORN ..................... 152
C DATA TABLES ....................................... 154
Literature Cited .......................................... 165

ix

Table 1-1.

Table 1-2.

Table 1-3.

Table 1-4.

Table 1-5 .

Table 2-1 .

Table 2-2.

Table 2-3.

Table 2-4.

Table 2-5.

LIST OF TABLES

Source, magnitude, and causative factors of dry matter losses
during ensilement and aerobic exposure (Henderson et al. , 197 9;
Woolford, 1990) ................................ 11

Microorganisms previously isolated from aerobically deteriorating
corn silage and ensiled, high-moisture corn (Burmeister and Hartman
1966; Beck, 1978; and Seitz et al., l982a,b) ............... 16

Minimum inhibitory concentration of propionic and lactic acids
at pH 5 or selected microorganisms found in ensiled feeds
(Woolford, 1975) ................................ 20

Inﬂuence of pH on proportion of undissociated molecules of lactic,
acetic and propionic acids in aqueous solution (Ray and Sandine,
1992) ........................................ 22

Species of genus Propionibacterium categorized by ecosystem
(Cummins and Johnson, 1986) ........................ 26

Effect of propionic acid-producing bacteria (PAPB) with or
without addition of lactic acid-producing bacteria (LAPB) on the
pH of ensiled, reconstituted high-moisture corn ............. 43

Effect of propionic acid-producing bacteria (PAPB) with or

without addition of lactic acid-producing bacteria (LAPB) on the

number of lactic acid-producing bacteria of ensiled, reconstituted
high-moisture corn ................................ 44

Effect of propionic acid-producing bacteria (PAPB) with or

without addition of lactic acid-producing bacteria (LAPB) on the

number of lactic acid utilizing microorganisms of ensiled,

reconstituted high-moisture corn ....................... 45

Effect of propionic acid-producing bacteria (PAPB) with or

without addition of lactic acid-producing bacteria (LAPB) on the

number of yeasts and molds of ensiled, reconstituted

high—moisture corn ................................ 46

Effect of propionic acid-producing bacteria (PAPB) with or
without addition of lactic acid-producing bacteria (LAPB) on the
acetic acid content of ensiled, reconstituted high-moisture corn . . . . 47

Table 2-6.

Table 3-1

Table 4-1

Table 4-2

Table 4-3

Table 4-4

Table 5-1.

Table 5-2.

Table 5-3.

Table 54.

Effect of propionic acid-producing bacteria (PAPB) with or

without addition of lactic acid-producing bacteria (LAPB) on the
propionic acid content of ensiled, reconstituted

high-moisture corn ................................ 48

Source and type of sample used as enrichment sources for propionic
acid-producing moisture bacteria from ensiled feeds ........... 56

Differential characteristics of Propionibacterium acidipropionici DH42,
P. acidipropionici 22562, P. freudenreichii 1382, P. jensenii P25 ,
P. thoenii P15 and P. shermanii P92 .................... 67

Metabolic proﬁle of Propionibacterium acidipropionici DH42,
P. acidipropionici 22562, P. freudenreichii 1382, P. jensenii P25 ,
P. shermanii P92 and P. thoenii P15 as measured by BIOLOG-GP . 70

Cellular membrane fatty acid composition of
Propionibacterium acidipropionici DH42 ................. 75

Effect of pH on speciﬁc growth rate of P. acidipropionici DH42
grown on glucose or lactate under nitrogen ................ 77

Acetic acid production of selected, propionic acid-producing bacteria
grown in an aqueous extract of ensiled, high-moisture corn

(pH 4.1, .11 % lactic acid) with (mixed culture) or without

(pure culture) epiphytic microorganisms found in ensiled,

high-moisture corn ................................ 85

Acetic acid production of selected, propionic acid-producing bacteria

. grown in an aqueous extract of ensiled, high-moisture corn

(pH 4.6, .02% lactic acid) with (mixed culture) or without
(pure culture) epiphytic microorganisms found in ensiled,
high-moisture corn ............................... 86

Acetic acid production of selected, propionic acid-producing bacteria
grown in peptone-yeast extract-lactate media (pH 6.9,

1.1% lactic acid) with (mixed culture) or without (pure culture)
epiphytic microorganisms found in ensiled, high-moisture corn . . . 87

Propionic acid production of selected, propionic acid-producing

bacteria grown in an aqueous extract of ensiled, high-moisture corn

(pH 4.1, .11% lactic acid) with (mixed culture) or without

(pure culture) epiphytic microorganisms found in

ensiled, high-moisture corn .......................... 89

xi

Table 5-5.

Table 5-6.

Table 5-7.

Table 5-8.

Table 5-9.

Table 6-1.

Table 6-2.

Table 6-3.

Table 6-4.

Table 6-5.

Table 6-6.

Propionic acid production of selected, propionic acid-producing

bacteria grown in an aqueous extract of ensiled, high-moisture corn

(pH 4.6, .02% lactic acid) with (mixed culture) or without

(pure culture) epiphytic microorganisms found in ensiled,

high-moisture corn ................................ 90

Propionic acid production of selected, propionic acid-producing

bacteria grown in peptone-yeast extract-lactate media (pH 6.9,

1.1% lactic acid) with (mixed culture) or without (pure culture)
epiphytic microorganisms found in ensiled, high-moisture corn . . . 91

Speciﬁc growth rate of selected, propionic acid-producing bacteria
grown in an aqueous extract of ensiled, high-moisture corn (pH 4.1,

.11 % lactic acid) with (mixed culture) or without (pure culture)
epiphytic microorganisms found in ensiled, high-moisture corn . . . 93

Speciﬁc growth rate of selected, propionic acid-producing bacteria
grown in an aqueous extract of ensiled, high-moisture corn (pH 4.6,

.02 % lactic acid) with (mixed culture) or without (pure culture)
epiphytic microorganisms found in ensiled, high-moisture corn . . . 94

Speciﬁc growth rate of selected, propionic acid-producing bacteria
grown in peptone-yeast extract-lactate media (pH 6.9, 1.1%

lactic acid) with (mixed culture) or without (pure culture)

epiphytic microorganisms found in ensiled, high-moisture corn . . . 95

Effect of Propionibacterium acidipropionici DH42 on the pH during
the ensilement of high-moisture corn ................... , 106

Effect of Propionibacterium acidipropionici DH42 on the propionic
acid content during the ensilement of high-moisture corn ...... 106

Effect of Propionibacterium acidipropionici DH42 on the acetic
acid content during the ensilement of high-moisture corn ...... 107

Effect of Propionibacterium acidipropionici DH42 on the lactic
acid content during the ensilement of high-moisture corn ...... 107

Effect of Propionibacterium acidipropionici DH42 on the glucose
content during the ensilement of high-moisture corn .......... 108

Effect of Propionibacterium acidipropionici DH42 on the ethanol
content during the ensilement of high-moisture corn .......... 108

xii

Table 6-7.

Table 6-8.

Table 6—9.

Table 6-10.

Table 6-11.

Table 6-12.

Table 6-13.

Table 6—14.

Table 6-15.

Table 6-16.

Table 6—17.

Table 6-18.

Effect of Propionibacterium acidipropionici DH42 on the number of
lactic acid producing bacteria during the ensilement of
high-moisture corn ............................... 109

Effect of Propionibacterium acidipropionici DH42 on the number of
yeasts and molds during the ensilement of high-moisture corn . . . 109

Effect of Propionibacterium acidipropionici DH42 on the number
of coliforms during the ensilement of high-moisture corn ...... 110

Effect of Propionibacterium acidipropionici DH42 on the dry matter
and organic matter recovery of high-moisture corn
after 42 days ensilement ........................... 110

Effect of Propionibacterium acidipropionici DH42 on the propionic
acid content during the aerobic exposure of ensiled,
high-moisture corn ............................... 115

Effect of Propionibacterium acidipropionici DH42 on the acetic
acid content during the aerobic exposure of ensiled, ,
high-moisture corn ............................... 115

Effect of Propionibacterium acidipropionici DH42 on the ethanol
content during the aerobic exposure of ensiled, high-moisture com 116

Effect of Propionibacterium acidipropionici DH42 on the lactic
acid content during the aerobic exposure of ensiled,
high-moisture corn ............................... 116

Effect of Propionibacterium acidipropionici DH42 on the glucose
content during the aerobic exposure of ensiled, high-moisture com 117

Effect of Propionibacterium acidipropionici DH42 on the pH during
the aerobic exposure of ensiled, high-moisture corn .......... 117

Effect of Propionibacterium acidipropionici DH42 on the number of
yeasts and molds during the aerobic exposure of ensiled,
high-moisture corn ............................... 118

Effect of Propionibacterium acidipropionici DH42 on the number

of Acetobacter during the aerobic exposure of ensiled,
high-moisture corn ............................... 118

xiii

Table 6-19.

Table 6-20.

Table 6-21.

Table 6—22.

Table A-1
Table A-2.
Table A-3.
Table A-4.
Table A-5 .

Table B-1 .

Table C-l.

Table C-2.

Effect of Propionibacterium acidipropionici DH42 on the number
of total aerobic microorganisms during the aerobic exposure of
ensiled, high-moisture corn ......................... 119

Effect of Propionibacterium acidipropionici DH42 on temperature
changes during aerobic exposure of ensiled, high-moisture corn . . 119

Effect of Propionibacterium acidipropionici DH42 on dry matter
recovery during the aerobic exposure of ensiled,
high-moisture corn ............................... 120

Effect of Propionibacterium acidipropionici DH42 on organic
matter recovery during the aerobic exposure of ensiled,
high-moisture corn ............................... 120

Weight loss (% of DM) of corn silage inoculated with
P. acidipropionici DH42 after 24 h or 72 h growth or treated
with 1% propionic acid after 25 or 170 d ensilement ......... 148

Fermentation characteristics of corn silage inoculated with
P. acidipropionici DH42 after 24 h or 72 h growth or treated
with 1% propionic acid after 170 d ensilement ............. 149

Microbial characteristics of corn silage inoculated with
P. acidipropionici DH42 after 24 h or 72 h growth or treated
with 1% propionic acid after 25 d ensilement ............. 149

Microbial characteristics of corn silage inoculated with 24 h or 72 h
growth of P. acidipropionici DH42 or treated with 1% propionic acid
after 170 d ensilement ............................ 150

Percent of substrates in the BIOLOG-GN plate utilized by the
mixed microbial community of corn silage inoculated with a propionic

acid-producing bacteria or treated with propionic acid ........ 150

Fermentation characteristics of high-moisture corn ensiled
with P. acidipropionici DH42 (10‘S cfu-g"-DM“) for 180 d ..... 153

DatausedforanalysisinChapter2 ........... 154
Analysis of variance of the effects of propionic acid-producing

bacteria with or without lactic acid-producing bacteria and
period of ensilement on results reported in Chapter 2 ......... 156

xiv

Table C-3.

Table C-4.

Table C-5.

Table 06.

Data used in the analysis presented in Chapter 6 fermentation study 157

Analysis of variance of the effects of P. acidipropionici DH42 and
period of ensilement on results reported in Chapter 6 ......... 159

Data used in the analysis presented in Chapter 6 aerobic
stability study ................................. 161

Analysis of variance of the effects of P. acidipropiom’ci DH42 and
period of aerobic exposure reported in Chapter 6 ........... 163

Figure 1-1.

Figure 1-2.

Figure 4-1.

Figure 4-2.

Figure 7-1

Figure 7-2

LIST OF FIGURES

Aerobic and anaerobic metabolism of forage carbohydrates by
various microorganisms associated with the ensilement
fermentation (Pahlow, 1991 as referenced by Woolford, 1991). . . . . 8

Production of ATP, acetate and propionate by Propionibacterium
spp. utilizing reduced ﬂavoprotein (Cummins and Johnson, 1990). . 28

Phase contrast photomicrograph of P. acidipropionici DH42 grown in
.5X MRS under ambient gas phase in sealed Hungate tubes
(1500X magniﬁcation, each bacterium approximately .7 pm wide). . 66

Growth characteristics of P. acidipropionici DH42 grown in either
peptone-yeast extract-lactate or peptone-yeast extract-glucose at pH 7
under nitrogen. .................................. 76

Template for the hybridization of a non-isotopic nucleic acid

probe developed from Propionibacterium acidipropionici DH42

to total nucleic acids extracted from 1 ml of exponential growth of

P. acidipropionici (American Type Culture Collection; ATCC 225 62,
'ATCC 4965), P. pentosaceum (Chr. Hansen's Lab; CH P11), P.
thoenii (CH P15), P. jensenii (CH P25), P. shermanii (CH P92,

CH CDC 3094), P. ﬁeudenreichii (ATCC 1382, ATCC 6207),
Lactobacillus cortﬁtsus (ATCC 27646) , L. delbruckii (ATCC 9649),

L. sake (ATCC 15521), L. helveticus (ATCC 12046),

L. ﬁ'uctovorans (ATCC 8288) and L. bulgaricus (ATCC 8144). . . 134

Template for the hybridization of a nonisotopic probe developed from
P. acidipropionici DH42 to total microbial DNA extracted from

aqueous extracts of ensiled, high moisture corn inoculated with 10‘

cfu per g DM of P. acidipropionici DH42 or without (control) at

day 0 after 0, 2, 7, 14, 28 and 42 days of ensilement and 3 days of
aerobic exposure and slot blotted onto a nylon membrane ....... 137

INTRODUCTION

Storage of feedstuffs by ensilement is a major management practice used in ruminant
livestock production. In 1992, over eight million hectares of land were used for corn
production in Michigan producing greater than one hundred million bushels (approximately
twenty billion kg) of corn. On a national scale, greater than ten billion bushels of corn
were harvested with approximately half being fed to livestock. Three conventional methods
of preserving the corn crop are whole crop silage, dry grain and as ensiled, high-moisture
corn. Corn silage or ensiled, high-moisture corn is a major portion of the rations fed to
feedlot and dairy cattle (Wilkens, 1981).

Ensiling high-moisture corn is popular because it allows producers to economically
store and feed grain without added expenses incurred from drying. High-moisture corn is
easily handled and allows for a more rapid fall harvest at an earlier date (two to three
weeks), elimination of several rehandling operations, allows for greater mechanization,
rodent free, storage and decreased ﬁeld losses when compared to corn stored as dry grain.
However, some disadvantages associated with ensiling high-moisture corn are the need to
determine moisture content at harvest, spOilage during feeding, approximately 25 % more
weight to handle and grain will heat (aerobic instability) if removed from the silo to soon

before feeding (Jones, et al. 1974).

2

Aerobic deterioration, also called aerobic instability or poor bunk life, is due to the
biological activities of aerobic microorganisms that metabolize substrates produced during
the ensiling fermentation (e.g. , lactic acid) or residual carbohydrates (e. g. , glucose) of the
original crop. The main products associated with this aerobic metabolism are carbon
dioxide, water and heat. An inherent problem with the ensilement of high-moisture corn
is the loss of nutrients or dry matter as carbon dioxide and heat when exposed to
atmospheric oxygen. Loss of dry matter by aerobic deterioration is an inefﬁciency that
depletes digestible nutrients and prevents maximum use of the feedstuff by ruminant
animals. Previous research summarized by Henderson et al. (1979) and Honing and
Woolford (1990) showed that aerobic instability and deterioration of ensiled feeds can
account for losses in excess of 35 % of the original crop dry matter.

Generally, good silo management practices and maintaining anaerobic conditions
during ensilement will help prevent aerobic deterioration during storage (Woolford et al. ,
1982) . The natural fermentation that occurs in non-treated silage is a lactic acid
fermentation similar to that found in sauerkraut. Production of lactic acid is generally
sufﬁcient to decrease pH of the ensiled crop below 4.5. Low pH generally yields a
metabolically stable feed when stored anaerobically. However, lactic acid is a preferred
substrate for microorganisms implicated in the aerobic deterioration process.

Antibiotics (W oolford and Cook, 197 8) and direct acidiﬁcation (Jones et al. 1974;
Ohyama and McDonald, 1975) have been used to address problems associated with aerobic
deterioration. Ohyama and coworkers (1975 , 1979) directly acidiﬁed corn and grass silage

with mixtures of formic, propionic, butyric and other volatile fatty acids. Results showed

3

success in preventing aerobic deterioration with increasing chain length of the volatile fatty
acid used. Propionic acid is commonly used in the baking and dairy industries as a
mycostatic and bacteriostatic agent to help preserve products such as breads and cheese
(Glatz, 1992). Ohyama and McDonald (1975) showed that addition of propionic acid at
low levels (1 to 3 percent) to ensiled crops reduced aerobic deterioration and increased dry
matter intake when fed to dairy cattle. Bulk quantities of acetic, formic and propionic acid
are currently obtained from petroleum products. Costs associated with these processes
usually precludes their use as additives to ensiled crops for prevention of aerobic
deterioration. In addition, direct use of acids has inherent problems associated with
operator health and equipment damage.

Current research dealing with silage is focused on developing new methods to preserve
the harvested crop in order to achieve the greatest feed quality possible. To accomplish this
goal, new and innovative methods of controlling silage fermentation and aerobic instability
need to be developed and implemented. The overall objective of this dissertation research
was to develop and study a novel method of manipulating the fermentation and aerobic
deterioration of high-moisture corn by using selected, propionic acid-producing bacteria

as bioinOculants.

The speciﬁc aims of this dissertation research were:
1. Select species of propionic acid-producing bacteria capable of increasing propionic

acid content of ensiled, high-moisture corn.

4

2. Study and optimize conditions that affect the performance of the selected, propionic

acid-producing bacteria.

3. Develop nucleic acid probes for the selected, propionic acid-producing bacteria

that will allow more precise evaluation of their ecological interaction.

4. Evaluate the use of the selected, propionic acid-producing bacteria as bioinoculants

for ensiled, high-moisture corn.

Chapter 1 reviews pertinent literature related to losses associated with aerobic
deterioration of ensiled crops and previous attempts to reduce these losses. In addition,
Chapter 1 reviews interrelationships of various microbial ecosystems and niches as they
apply to the fermentation and aerobic instability of ensiled feeds. A study addressing the
ﬁrst and third objective using two classical strains of propionic acid-producing bacteria is
discussed in Chapter 2. Chapter 3 describes the methodology used for enrichment and
isolation of propionic acid-producing bacteria from ensiled feeds. A strain of
Pmpionibacterium acidipropionici, designated P. acidipropionici DH42, was isolated from
ensiled, high-moisture corn and is characterized in Chapter 4. The ability of propionic
acid-producing bacteria, with emphasis on P. acidipropionici DH42, to compete for
substrates against the epiphytic microbial population of high-moisture corn is evaluated in
Chapter 5. Chapter 6 discusses a study that uses results from previous chapters and
evaluates Propionibacterium acidipropionici DH42 as a bioinoculant for ensiled, high-
moisture corn. Chapter 7 deals with development and validation of a non-isotopic nucleic
acid probe for P. acidipropionici DH42. Research ﬁndings are summarized and

conclusions based on the objectives are discussed in Chapter 8. Pertinent data with

5

corresponding analysis of variance tables and supporting studies are presented in three
appendices.

The overall hypothesis for this research is that addition of selected, propionic acid-
producing bacteria as bioinoculants to ensiled, high-moisture corn will increase production
of propionic acid during the ensiling fermentation. Consequently, increased concentration
of propionic acid in ensiled, high-moisture corn will decrease extent of aerobic
deterioration and instability by controlling growth of microorganisms responsible for

aerobic deterioration.

CHAPTER 1
REVIEW OF THE LITERATURE

Principle objectives of any preservation method are adequate recovery of nutrients
and production of a product that is acceptable to livestock (Barnett, 1954; McCullough,
1978). Previous strategies for efﬁcient, silage fermentation encouraged high levels of lactic
acid production and lower pH (below 4.2) during the anaerobic fermentation phase. This
efﬁcient fermentation usually produces silage that is stable under anaerobic storage
conditions. However, once the ensiled mass is disturbed during feedout or other storage
perturbations, aerobic microorganisms will begin to metabolize the lactic acid and residual
carbohydrates. The resulting heat and mold formation has been deﬁned as poor bunk
stability or aerobic instability. Nutrient losses (as dry matter) during aerobic metabolism
(instability) have been reported to be as high as 35 % (Henderson et al. , 197 9). The Food
and Agriculture Organization of the United Nations estimates that grain losses during

storage are in excess of 5%. '

Importance of ensiled feedstuffs
In a summary of past silage research, McCullough (1993) reported that over 120

million tons of ensiled crops are produced on an annual basis in the United States costing

6

7

over 2.25 billion dollars (McCullough, 1978). Of this amount, a vast majority is used for
feeding ruminant animals (Wilkens, 1981). Importance of ensiled feedstuffs to animal

agriculture has been recently discussed by Pitt (1990) and Bolsen (1991).

Phases of the ensilement procses

Phases of the ensilement process that ultimately affect the resulting silage are
generally broken down into ﬁve phases (Barnett, 1954; Muck, 1988). These phases are
plant cellular respiration after harvest, initial acetic acid fermentation, initial lactic acid
fermentation, major lactic acid fermentation and the quiescent storage phase. The quiescent
phase can be interrupted by either a possible secondary fermentation initiated by either
clostridia or aerobic deterioration often due to air inﬁltration. Besides these ﬁve phases,
the quality of the silage preserved can be inﬂuenced by two additional factors or phases.
These two additional phases are the interactions of the epiphytic microbial population with
the crop in the ﬁeld and resulting microbial growth at feedout. Other physical factors
inﬂuence the seven phases of ensilement and therefore the quality of the silage fed by
inﬂuencing the pattern of the fermentation. Major factors include crop moisture, packing
density of the ensiled crop, ambient temperature at harvest, temperature during silo ﬁlling
and unloading and crop maturity.
Microbiology and biochemistry of the ensilement process

The microbial ecosystem associated with ensilement is very diverse and dynamic.
Pahlow (Figure 1-1 as referenced by Woolford, 1991) diagrammed the general interactions

associated with the major microbial groups associated with the silage fermentation.

ACETATE. LACTATE. Hp, 00,. ACETOIN. 2,3 BUTANE DDL. DIACETYL

 

 

 

 

 

o F
a 5
° 5
I:
'“ 2':
< ACETATE ‘
PROPIONATE
coa H10 2
9 E
in <
0 E
3 in
“i 2
g a:
‘ LACTATE. ACETATE. err-m 00,. MANMTOL. GLYoenoL aAcremocm E

Figure 1-1. Aerobic and anaerobic metabolism of forage carbohydrates by various
microorganisms associated with the ensilement fermentation (Pahlow, 1991
as referenced by Woolford, 1991).

9

Microorganisms of interest in silages include various aerobic bacteria, lactic acid-
producing bacteria, acetic acid-producing bacteria, propionic acid-producing bacteria,
endospore-forming bacteria, coliforms, listeria and yeasts and molds. Each of the above
described microbial groups affects the fate of carbohydrates and fermentation products
such as lactic acid and carbon dioxide.

Anaerobic fermentation of plant carbohydrates to lactic acid and the resulting
decrease in pH is the biological basis to the preservation of ensiled feeds. Lactic acid-
producing bacteria will produce a diverse array of end-products depending on species
involved, presence of oxygen and carbohydrate fermented. Some fermentation end-
products produced by lactic acid-producing bacteria include lactate, acetate, ethanol,
carbon dioxide, mannitol, glycerol and bacteriocins. Lactate, ethanol, acetate and carbon
dioxide are normally associated with heterofermentative lactic acid-producing bacteria,
whereas lactic acid is the predominant end-product of the homofermentative lactic acid-
producing bacteria (Scale, 1986).

Growth and metabolism of clostridia are generally not associated with high quality
silage. Clostridia are normally associated with poor ensiling conditions where a sufﬁciently
low pH was not attained. Fermentation products associated with clostridia are butyrate,
carbon dioxide and hydrogen. Coliforms, commonly referred to as enterobacteria, are also
associated with poor quality silage (Ostling and Lindgren, 1991). Acetic acid, ethanol,
2,3-butanediol, carbon dioxide, lactic acid and toxins have been found in silages where

coliforms dominated the ensiling fermentation. Yeasts have also been found to grow

10

anaerobically in silage and their main end-products of carbohydrate metabolism are ethanol
and carbon dioxide.

Nitrogenous compounds, such as amino acids and proteins, can be metabolized by
microorganisms associated with aerobic deterioration and poor ensilement to ammonia-
nitrogen and volatile amines. These compounds have been shown to have detrimental
effects on organoleptic properties and animal acceptance of resulting silages (McDonald,

1981; Woolford, 1984).

Inﬂuence of oxygen on ensilement and aerobic deterioration

Aerobic microbial metabolism in ensiled feedstuffs when exposed to oxygen has
been the focus of many studies (Woolford, 1990; Ruxton and Gibson, 1993). Signiﬁcant
losses are attributed to aerobic activity, with the majority due to aerobic deterioration and
surface wastage (Table 1-1). Aerobic instability, or poor bunk life, is characterized by
increases in silage temperature and pH, decreases in lactic and acetic acid and a
concomitant reduction in feeding value (Honing and Woolford, 1979).

Many physical changes occur to ensiled, high-moisture corn during aerobic
deterioration. The germ, embryo, entire seed or kernel can become discolored (Jones et
al. , 1974). Heating and mustiness is generally accompanied with losses of dry matter
during aerobic deterioration of ensiled, high-moisture corn (McDonald, 1981). In addition,
biochemical changes occur within the grain which lead to a decrease in the digestibility of
nitrogen containing nutrients, such as proteins and amino acids. Evidence of this

phenomena is the increase in losses of nitrogen in urine due to the increase in ammonia

11

Table 1-1. Source, magnitude, and causative factors of dry matter losses during
ensilement and aerobic exposure (Henderson et al. , 1979; Woolford, 1990)

 

Dry Matter Loss

 

Source (% ) Causative Factor

Initial plant respiration 1 - 2 Plant enzymes and aerobic
microorganisms

Primary fermentation o — 4 lactic acid bacteria

Secondary fermentation 0 - 5 Clostridia

Efﬂuent and/or wilting 2 - 7 Low dry matter or wilting technique

Surface waste 0 - 10 Aerobic microorganisms

Aerobic deterioration 0 - 35 Aerobic microorganisms

12

content from the deamination of amino acids (Pitt, 1986). In addition, the Maillard
reaction can cause increases in the amount of nitrogen containing compounds found in
feces, indicating decreased nitrogen retention (Pitt, 1986). Potential for production of
harmful metabolites by the microﬂora associated with the deterioration process, such as
mycotoxins, can be a problem (Lee et al., 1986; Woolford, 1984a; Widstrom, 1992.).

Majority of losses associated with aerobic deterioration occur on the surface during
storage in the silo and feed-out losses from the surface (exposed silage face) as feed is
removed (Woolford, 1984b). Several factors inﬂuence the extent of aerobic instability in
ensiled feeds. Main factors are moisture content of the grain, temperature during storage,
temperature of grain before storage, temperature at feeding, length of storage, oxygen
inﬁltration during storage, tighmess of packing, number of aerobic microorganisms and
substrates available for metabolism by aerobic microorganisms (Jones et al. , 1974,
Ohyama et al., 1981; Muck, 1988). Pitt (1986) summarized additional factors that have
an effect on poor bunk life. These factors are presence of oxygen, presence of carbon
dioxide, silage dry matter content, forage species ensiled and concentration of fermentation
acids.

Recent reports, reviews, and models relevant to aerobic deterioration of ensiled
crops include Courtin and Spoelstra (1990), Woolford (1990), Pitt et al. (1991a,b), Muck
and O'Keily (1992) and O'Keily and Muck (1992). The review by Woolford (1990)
explains the extent of the problems associated with aerobic instability of ensiled feeds and
is a good general reference for this area of research. Mathematical models have been

developed that combine research results that identify the causes of aerobic instability and

l3

simplify the prediction of possible aerobic deterioration in ensiled feeds. Courtin and
Spoelsua (1990) developed a mathematical model which predicts the stability of grass and
corn silages by simulating the growth of yeast and acetic acid bacteria (Acetobacter spp.),
oxidation of fermentation acids, consumption of oxygen and concomitant production of
carbon dioxide and rise in silage pH. Initiation of aerobic instability is attributed to the
growth of acetic acid bacteria with or without the parallel growth of yeasts in the model
of Courtin and Spoelstra (1990). In addition, this model and most other models associated
with the silage fermentation consider the effects of initial number of yeasts and molds, pH,
moisture content, lactic acid and acetic acid concentrations on stability. All of these factors
directly inﬂuence which microbial population is given an environmental advantage for
growth and therefore, which populations will predominate. Pitt and coworkers (1991b)
developed a mathematical model to predict aerobic instability as affected by the growth of
yeasts and molds. It was concluded that temperature, water-soluble carbohydrate
concentration, fungal populations and concentration of undissociated acidic fermentation
end-products had the largest effects on aerobic instability of corn silage.

Muck and O'Kiely (1992) studied the effect of fermentation end-products on the
initiation and extent of aerobic deterioration of alfalfa and corn silages. The effect of
yeasts on the aerobic deterioration of the same silages was reported in a companion paper,
O'Kiely and Muck (1992). In alfalfa silage, a ”stability factor" or deterioration inhibitor
was found to be present in certain silos after the fermentation was completed and was not
thought to be a major or principle product of the ensiling fermentation (O'Kiely and Muck,

1992).

l4

Wardynski (1991) evaluated aerobic instability of high-moisture corn stored at
different temperatures, moisture levels and inoculation rates. As moisture level increased
from 22 to 37 % , dry matter recovery from laboratory silos decreased. Additionally, as
corn moisture increased, aerobic instability increased. High-moisturecorn removed from
laboratory silos stored at 4°C was less stable than corn stored at 20°C. Inoculation with
lactic acid-producing bacteria increased dry matter recoVery from the silo but did not alter
aerobic instability.

Oxygen can dramatically inﬂuence the seven phases associated with ensilement
during the ﬁeld phase, initial aerobic phase in the silo, air inﬁltration during storage and
aerobic deterioration after opening the silo before feeding. This is due to enhancement of

the growth of microorganisms involved in processes associated with aerobic deterioration.

Microbiology associated with aerobic deterioration

Factors that determine the speciﬁc microorganisms that will proliferate in silage
when exposed to air are not fully understood (Woolford, 1990; Courtin and Spoelstra,
1991). Simultaneously with the growth of these organisms, silage pH and temperature will
increase (Courtin and Spoelstra, 1991). Increases in silage temperature are associated with
the growth of yeasts, molds and certain bacteria (W oolford, 1990; Rust and Yokoyama,
1992). The microbes responsible for aerobic instability metabolize many substrates, some
derived from the silage fermentation and others that are residues of the initial crop

(W oolford, 1990).

15

In a recent survey of aerobic instability of high-moisture corn from 29 farms in
Michigan (Rust and Yokoyama, 1992), aerobic instability was positively associated with
yeast and mold populations. Neither fermentation end-products nor pH was associated with
instability. In addition, propionic acid was not a major component of the fermentation end-
products. Similarly, in corn silage, lower pH and greater lactic acid did not improve
stability (Rust et al., 1989).

Aerobic respiration of microorganisms associated with instability of ensiled feeds
can result in formation of an inedible crusty or slimy layer and the possible formation of
toxins. Potentially pathogenic microorganisms have been found in aerobically deteriorating
silages (Table 1-2; Woolford, 1990). These include microorganisms that pose a potential
health hazard to livestock and agricultural workers (Ruxton and Gibson, 1993). Woolford
(1990) summarized research ﬁndings that speciﬁcally found the occurrence of the
organisms responsible for listeriosis, botulism and mycotoxin formation in aerobically
unstable silages.

Aerobic metabolism of carbohydrates by silage microorganisms (Figure l- 1) leads
to the production of diverse metabolic end-products. As a whole, the aerobic heterotrophic
microbial population associated With silages will metabolize forage carbohydrates to carbon
dioxide, water and heat. However, there are a few exceptions that produce other
compounds such as propionic acid-producing bacteria. In addition, some toxins associated
with enterobacteria and molds can be associated with the aerobic metabolism of forage
carbohydrates (W oolford, 1990). Yeasts aerobically metabolize carbohydrates to carbon

dioxide, water and heat as does the mixed heterotrophic aerobic microbial population.

16

Table 1-2. Microorganisms previously isolated from aerobically deteriorating corn
silage and ensiled, high-moisture corn (Burmeister and Hartman 1966;
Beck, 1978; and Seitz et al., 1982a,b)

 

Yeasts Molds

Candida intermedia Geotrichium spp.
Candida krusei Monascus spp.

Candida parapilosis Monilia spp.

Candida melini Mucor spp.

Hanselua anomala Penicillium spp.

Pichia fennentans

Pichia membranefaciens

Saccharomyces baili

Saccharomyces exiquis

T orulopsis candida

Proteolytic bacteria Lactic acid-producing bacteria
Bacillus cereus Lactobacillus buchneri
Bacillus ﬁrmus Lactobacillus bulgaricus
Bacillus lentus . Lactobacillus viridescens
Bacillus sphaericus Pediococcus cerevisiae

Aerobic deterioration inititators
Acetobacter spp.

17

Lactic acid-producing bacteria can aerobically metabolize soluble forage
carbohydrates to acetate, lactate, hydrogen peroxide, carbon dioxide, acetonin, 2,3-
butanediol and diacetyl. Acetic acid-producing bacteria have been implicated as
initiators of the aerobic deterioration process (Spoelstra, 1990). These microorganisms
will metabolize carbohydrates to acetate and carbon dioxide.

Many problems are associated with the diverse end-products produced by the
aerobic silage microbial population. This is mostly due to the concomitant production
of heat associated with aerobic instability. Heat production during aerobic instability
can easily cause the silo biomass to exceed 55 °C (Wardynski et al., 1993) which will
decrease feed acceptance by cattle and can reduce the availability of nitrogen containing

compounds.

Methods of preventing aerobic deterioration

Methods commonly used to reduce losses associated with aerobic deterioration deal
with either increasing the rate of pH decline or by attempting to inhibit growth of aerobic
microorganisms. Inorganic acids such as sulfuric, hydrochloric and phosphoric have been
used to preserve harvested crops by rapidly decreasing the pH below 4 which will
generally lead to a metabolically quiescent biomass during anaerobic storage (Woolford,
1978). Use of these additives is becoming less attractive because of hazards associated with
the use of strong acids. Besides health risks to personnel, other negative factors include
feed refusal by animals, destruction of harvesting and storage equipment and increased

environmental liability from the efﬂuent from silages treated with theses acids.

18

Organic acids have been effective in the inhibition of yeasts, molds, and bacteria
that are responsible for the deterioration of stored crops such as hay and silages (Ekstrom,
1973; Jones et al., 1974; Raeker et al., 1992). Formic acid has been widely used in
Europe with grass silages. Acetic acid is also effective in the preservation of ensiled feeds
(Crawshaw et al. , 1980). Many commercial hay and silage preservation amendments are
based on mixtures of formic and acetic or acetic and propionic acid. However, demand for
acetic acid as a preservative for human food generally precludes its use as a feed additive
on a price basis.

As previously mentioned, inhibition of the silage fermentation can be- a method
used to reduce losses associated with aerobic deterioration. Fermentation inhibitors such
as sulﬁtes, sulfur dioxide and sodium metabisulﬁte have been used with some success in
the preservation of feeds (Woolford, 1984). In addition, formaldehyde-containing additives
have been evaluated (W oolford, 1975; Barry et al. , 1980) for the prevention of aerobic
instability in grass and corn silages.

One of the most common additives for silages is lactic acid-producing bacterial
inoculants. Many commercial bacterial inoculants are based solely on use of lactic acid-
producing bacteria as starter cultures. A multitude of research has been done on the
efﬁcacy of bioinoculants based on the use of lactic acid-producing bacteria for preserving
ensiled feeds. It is generally accepted that the addition of lactic acid-producing bacteria
will lead to a more rapid decrease in the pH of the ensiled material (Scale, 1986). This is
accomplished by supplying a high amount, usually greater than 105 cfu-g'I-DM", of

homofermentative lactic acid-producing bacteria at harvest or when placing into storage.

19

It is not the scope nor the intent of this review to focus on the use of lactic acid-producing
bacteria due to many reviews on this subject. Scale (1986) conducted a comprehensive
review addressing the use of lactic acid-producing bacteria as silage amendments.

Other additives have been used to help reduce aerobic deterioration. Antibiotics
(W oolford and Cook, 1978) have been added to experimental silos to decrease
proliferation of yeasts and molds. Compounds that buffer or increase the pH of the ensiled
material, enhance the lactic acid production by extending the fermentation. Ammonia
(Phillip et al., 1985) and limestone (Woolford, 1984) are two common compounds used
in this manner. Ammonia serves two purposes in the ensilement process. Firstly,
increasing pH of the ensiled material above 8 (providing limited bacteriostatic and
mycostatic activity) and providing a source of non-protein nitrogen that can be used by
some silage microorganisms. However, the use of ammonia is usually for the improvement
of crude protein content of silages.

Increases in lactic acid concentration of ensiled fwds has not been consistent in
enhancing aerobic stability. Rust et al. (1989) found that inoculation of corn silage with
a homofermentative lactic acid-producing bacteria increased aerobic instability. Addition
of lactic acid-producing bacterial inoculants to high-moisture corn has also had mixed
results. Wardynski et al. (1993) showed an increase in aerobic instability of ensiled, high-
moisture corn inoculated with a lactic acid-producing bacterial inoculant. However, Phillip
and Fellner (1992) found that the addition of a lactic acid-producing bacterial inoculant

decreased aerobic instability.

20

Effectiveness of propionic acid as a preservative

Propionic acid has been shown to be one of the most effective and economical
mycostatic (mold inhibiting) and bacteriostatic (bacteria inhibiting) agents (Table 1-3;
Voclker et al., 1989; Pitt et al., 1991). The effectiveness of propionic acid is based on the
inhibitory properties of the undissociated acid and not the pH reducing properties
(W oolford, 1975) . Many commercial products based on propionic acid are available to
increase the shelf life of bakery and dairy products, and enhance the preservation of hay

and silages (Bolsen, 1991.; Wittenberg, 1991; Glatz, 1992).

Table 1-3. Minimum inhibitory concentration of propionic and lactic acids at pH 5 for
selected microorganisms found in ensiled feeds (Woolford, 1975)

 

 

Microorganisms Propionic acid lactic acid
(IBM) (mM)
Homofermentative LAB 250 94
Clostridia < 8 8
Yeasts 125 >250
Molds ' 63 >250

 

Propionic acid is commonly found in ensiled feeds at very low levels (less than
.05 % , compared to 2 to 4 % for lactic acid). Addition of propionic acid at low levels (less
than 1%) increased storage life of ensiled, high-moisture grains and resulted in similar or
improved animal performance when compared to non-treated control corn (Jones et al. ,
1974; Voclker et al. , 1989). Propionates are generally used for reducing losses within the

silo, decreasing soluble protein losses and decreasing aerobic deterioration of silages (Pitt, ,

21

1986). Woolford and Cook (1977) showed that propionic acid, at 2.5 % DM, inhibited
aerobic instability of corn silage for 19 days. Britt and coworkers (1975) showed that
propionic acid was effective in the inhibition of pH changes of aerobically-exposed corn
silage for 36 days. Two different studies using propionic acid in corn silage (Huber and
Socjono, 1976; Ham and Ohyama, 197 9) showed reduced heating and fungal counts after
7 days of aerobic exposure. Aerobic stability of unwilted grass silages was also improved
by addition of propionic acid as measured by more stable temperatures for 36 h (Mann and

McDonald, 197 6) and pH for 6 days (Crawshaw et al. , 1980) of aerobic exposure.

Mechanisms of antimicrobial action of propionic acid

The bacteriostatic and mycostatic properties of propionic acid are very beneﬁcial
in the preservation of high-moisture feedstuffs. The following description of the
antimicrobial action mechanisms of propionic acid is adapted from Ray and Sandine
(1992). Propionic acid affects different types of microorganisms in various ways. Extent
of control (slowing or inhibiting) of microbial growth, is inﬂuenced by factors such as
propionic acid concentration, pH of the environment, lipophilic properties, temperature,
resistance of the microbes present, amount of microbes present and anaerobic conditions.

At high concentrations of propionic acid, the pH outside the cells is decreased.
Propionic acid acts similarly to strong inorganic acids at these concentrations. Exposed cell
wall components, the periplasmic space and outer surface of the cytoplasmic membrane
are exposed to the low external pH. At low pH, enzymes associated with these components

can become denatured therefore leading to destabilization of the cell membrane. This in

22

turn can lead to increased permeability of the membrane to the acids in the external space.
Denatured cell membrane components and inﬂux of copious amounts of acidic compounds
can lead to cell death.

Bacteriostatic action of propionic acid is based on the lipophilic nature of the
undissociated acid. Propionic acid's pK, (4.87) allows a greater proportion of the
undissociated acid at a given pH when compared to acetic and lactic acids (Table 1-4). The
increased permeability of the lipid bilaycr associated with bacterial cell membranes to
propionic acid facilitates more undissociated acid to diffuse into the cytoplasm.
Undissociated propionic acid within the aqueous environment of the cytoplasm dissociates
into hydrogen ions and propionate ions. Once the concentration of the hydrogen ions
exceeds the buffering capacity of the cytoplasm, the internal pH will begin to decrease and
the excess hydrogen ions will nwd to be transported via membrane associated proton
pumps. Transport of these ions out of the cell requires energy. Depletion of energy to
remove excess hydrogen ions can lead to decreased internal energy stores that in turn limits

nutrient transport processes that require energy.

Table 1-4. Inﬂuence of pH on proportion of undissociated molecules of lactic, acetic
and propionic acids in aqueous solution (Ray and Sandine, 1992)

 

 

 

Undissociated acid
Acid pK, (% of total acid concentration)
pH 4 pH 5 pH 6 pH 7
lactic 3.85 39.2 6.1 0.6 0.1
Acetic 4.75 84.5 34.9 5.1 0.5

Propionic 4.87 87.6 41.7 6.7 0.7

 

23

Besides depleting microbial energy stores, once the proton pumps cannot keep up
with the excess hydrogen ions, the pH of the cytoplasm will further decrease. Low
cytoplasmic pH can lead to similar effects as high concentrations of propionic acid or low
external pH. Such effects as denaturation of structural and functional components of the
microbial cell will greatly affect cellular growth and viability.

The effect of propionic acid on microbial growth is enhanced by elevated
temperature, low microbial load and anaerobic conditions. As stated previously, propionic
acid affects groups of microorganisms differently. Yeast and molds are more resistant to
propionate followed in resistance by gram positive bacteria. The gram negative bacteria
are the least resistant to the bacteriocidal effects of propionic acid. In addition, acidophilic
and fermentative bacteria are more resistant to propionic acid than respiratory bacteria.
This may be due to greater tolerance to low cytoplasmic pH or changes in cell membrane

permeabilities to lipophilic compounds.

Why use propionic acid-producing bacteria?

The caustic nature of propionic acid is problematic for equipment and hazardous
for personnel when added before storage. Production of propionic acid by propionic acid-
producing bacteria will alleviate problems associated with the pre-storage application of
caustic materials. lactic acid-utilizing bacteria, which produce propionic acid, have been
isolated from grass silage (Woolford, 1975). However, these organisms did not account

for a large portion of the microbial population (Rosenberger, 1956; Woolford, 1975).

24

Beyond possibly affecting the microbial ecosystem associated with ensiled, high-
moisture corn and decreasing aerobic instability, the addition of propionic acid-producing
bacteria may serve as a vehicle to inﬂuence the rumen microbial ecosystem and possibly
help alleviate some men maladies such as acidosis. Diets that contain high concentrations
of fermentable sugars, such as 85 % grain diets that are commonly fed to feedlot cattle, can
lead to over-production of ruminal lactic acid that overwhelms the lactic acid-utilizing
population of the rumen (Slyter, 1976). Extent of acidosis can range from temporary lack
of appetite to animal death (Huber, 197 6). Megaspheara elsdenii is currently being
evaluated (Robinson et al. , 1991) as a possible direct feed microbial which is thought to
help reduce the amount of ruminal lactic acid accumulation. Additional in vitro studies
have been conducted looking at the possrble use of either M. elsdenii or Propionibacterium
shermanii to modulate changes in ruminal lactic acid production (Kung et al. , 1991).

Leng (1970) summarized research ﬁndings that found propionate to be the main
glucose precursor in ruminants. Amino acids, lactate, pyruvate and glycerol are all
possible gluconeogenic precursors. However, amino acids are generally required for
protein synthesis. Carbon from propionic acid is metabolized by liver tissue more
efﬁciently than lactic acid (Armento, 1992). Over 85 % of propionic acid absorbed by the
rumen is cleared by hepatic tissue (Armento, 1992). Theoretically, increasing the
proportion of propionic acid absorbed from the rumen may signiﬁcantly affect protein and
carbohydrate usage by ruminants to the extent of improving use of dietary and microbial

protein post-ruminaly.

25

Historical perspective for propionic acid-producing bacteria

Propionic acid-producing bacteria are commonly used in the production of dairy
products such as Swiss and Emmental cheeses (Drinan and Cogan, 1992) and can produce
up to 1% propionic acid in Swiss cheeses. Hettinga and Reinhold (1972a,b,c) provided a
series of classical reviews on the growth, metabolism and miscellaneous metabolic
activities of the propionic acid-producing bacteria which are very informative and the basis
for the majority of recent information. Bacteria of the genus Propionibacterium (Table 1-5)
are divided into two distinct categories based on the ecosystem. The classical (dairy)
species are normally associated with cheese production whereas, the cutaneous (skin
associated) are generally found on skin and can be responsible for facial acne (P. acnes)
and infections following surgery. Both types are gram positive and occur in a diverse array
of shapes. Most are branched or rod shaped (sometimes looking like Chinese alphabet
characters). In addition, Propionibacterium spp. are generally resistant to lysozyme,
sulfonamides and penicillin-G (Cummins and Johnson, 1990). Other bacteria which
produce propionic acid include species within the genera Bacteroides, Clostridium,
F usobacterium, Megasphaera, Prevotella, Selenomonas, T reponema, and Veillonella

(Johns l951a,b; Holdeman et al., 1977).

26

Table 1-5. Species of genus Propionibacterium categorized by ecosystem (Cummins
and Johnson, 1986) ' '

 

 

Classical Cutaneous
(dairy related) (skin associated)
P. acidipropionici P. acnes

P. freudenreichii subsp. freudenreichii P. avidium

P. freudenreichii subsp. shermanii P. granulosum
P. jensenii P. lymphopilium

P. thoenii

 

Previous research with propionic acid-producing bacteria

Early research in the 1930's was focused on using propionic acid-producing
bacteria for the commercial production of propionic acid (Hettinga and Reinbold, 1972a)
and vitamin B12 (Bullerman and Berry, 1965). Microbial fermentations became the popular
method to produce propionic acid. In the 1950's, there was a shift to the chemical
synthesis of propionic acid from petroleum products (Hettinga and Reinbold, 1972a) and
production of vitamin Bu by Pseudomonas (Glatz, 1992). These production systems were
less expensive and had higher yields of propionic acid and vitamin B12.

The elucidation of the fermentation pathways of propionic acid-producing bacteria
were by Johns (1951b,c; 1952), Wood et al. (1955) and Crow (1986; 1987a,b). There are
two pathways used by different microorganisms for the production of propionic acid.
Bacteria of the genus Clostridium and a other genus use the acrylate pathways which
incorporates acrylic acid as an intermediate in the fermentation of lactic acid to propionic

acid (Johns, 1952). In contrast, bacteria of the genus Propionibacterium use the succinate-

27
methylmalony-CoA pathway (Figure. 1-2). For a complete description of the enzymatic
pathways involved with production of propionic acid production by Propionibacterium
spp. the review by Hettinga and Reinhold (1972a) is very comprehensive and follows the
general diagram outlined in Figure 1-2. The general stoichiometry of the propionic acid
fermentation is (Wood, 1981):

3 glucose+12Pi+12ADPs-12ATP+4H,O+4C02+2acetate+4propionate

Hexoses are predominately converted to pyruvate by the Emden-Mcyeroff pathway.
lactic acid is oxidized in the presence of fumarate by lactic dchydrogenase. Three of the
six ATP are produced by the action of pyruvate kinase in which three mol of
phosphoenolpyruvate are converted to 3 mol pyruvate and 3 mol ATP. An additional ATP
is produced by the action of acetyl kinase in which acetyl-phosphate is converted to acetate
and ATP. The ﬁnal two ATP are produced by the action of fumarase in conjunction with
ﬂavoprotein and reduced NAD. Carbon dioxide arises in the fermentation from the activity
of pyruvate dehydrogenase that produces acetyl-CoA from pyruvatelntermediates of the
propionic acid fermentation not mentioned above include methyl-malonyl-CoA, methyl-
malonyl, propionyl-CoA, acetyl-CoA, oxaloacetate, malate, fumarate, succinate and
succinyl-CoA. Biotin and cyanocobalamin (vitamin B12) are two very important cofactors
along with NAD(H2) and FAD(I-12). A complex which include biotin (designated C1 in

Figure 1-2) facilitates the transfer of one carbon intermediates. '

28

Acetvl-P—-t- ACETATE + ATP

1“?”

Acetyl-COA + NADH + co2

1 1+ NAD. CoAl

Pyruvate T— 'C‘ ' .7 ﬂ

P-cnolpyruvate—COz—v- Oxalacetate Methylmalonyl-

 

 

 

 

 

 

 

P‘m PP, CoAA 4
1C NADH 1
Fumarate Succinyi-
ATP i
SF Succinate "CoA «4
PROPIONATE

Figure 1-2. Production of ATP, acetate and propionate by Propionibacterium spp.
utilizing reduwd ﬂavoprotein and biotin containing transfer complex (C1;
Cummins and Johnson, 1990).

29

Current research is focused on increasing propionic acid production from corn co-
products and the production of bacteriocins for food grade Operations. Bayer and Ryba
(1991) have studied the relatedness of Propionibacterium spp. by immunoblotting of cell-
associated proteins which can be presumptive bacteriocins and other cell associated
proteins. Research at Ohio State University (Lewis and Yang, 1992a,b,c) is currently
focused on developing new methods to use propionic acid-producing bacteria and dairy
byproducts for propionic acid formation. Propionic acid production by propionic acid-
producing bacteria is both associated with exponential growth and with the stationary phase
of growth. In addition, the optimal pH for propionic acid production is lower than the pH
optimum for growth. These two points favor the possibility of propionic acid-producing
bacteria's competitiveness in slightly acidic environments such as silages.

Parker and Moon (1982) showed a beneﬁcial (commensalistic) interaction between
lactic acid-producing bacteria and propionic acid-producing bacteria. It appears that
propionic acid-producing bacteria produce metabolites that beneﬁt the growth of the lactic
acid bacteria (Parker and Moon, 1982; Flores-Galarza et al. , 1985). Most propionic acid-
producing bacteria prefer lactic acid to other substrates (Hettinga and Reinbold, 1972a)
which partially alleviates the end-product inhibition of lactic acid on the growth of lactic
acid-producing bacteria (Parker and Moon, 1982). As mentioned earlier, Bullerman and
Berry (1965) and Hettinga and Reinbold (1972a) explained that propionic acid-producing
bacteria produce vitamins and other cofactors that may further beneﬁt the ensiling
fermentation. Another advantage of the classical propionic acid-producing bacteria is

inherent stability when grown in the laboratory. Commensalism of propionic acid-

30
producing bacteria with the epiphytic lactic acid-producing bacteria may serve as a major
factor for improving the silage fermentation.
Some possible disadvantages associated with the known classical propionic acid-
producing bacteria are slow growth, low acid tolerance, and low substrate afﬁnities (Russel
and Baldwin, 1979; Russel et al., 1979; Glatz, 1992). These disadvantages would

possibly exclude their use as bioinoculants.

Enrichment and isolation of propionic acid-producing bacteria

Previous efforts at deve10pment of selective media for propionic acid-producing
bacteria have been based on yeast extract, peptone and sodium lactate (V edamuthu and
Reinbold, 1967; Hettinga et al., 1968; Peberdy and Freyers, 1976). The selectivity of the
above media was based on the fermentation of lactate by propionic acid-producing bacteria
under anaerobic conditions. However, lactic acid-producing bacteria and other
microorganisms commonly found in silages can use the same media components. Reddy
and coworkers (1973) and Drinan and Cogan (1992) attempted to correct this problem by
addition of two antibiotics, cloxacillin and kanamycin. This alleviated the growth of most
lactic acid-producing bacteria except Leuconostocs. The concentration of antibiotics
necessary to inhibit the growth of Leuconostocs was also inhibitory to the growth of
propionic acid-producing bacteria of interest. Currently, enrichment and isolation of
propionic acid-producing bacteria are still based on the ability of propionic acid-producing

bacteria to use lactic acid in the production of propionic acid (Atlas and Parks, 1993).

31

Novel selection procedures (continuous or semi-continuous enrichment) and
possible genetic engineering may simplify selection of new strains that can overcome the
previously mentioned obstacles. Muck (1988) recommends that lactic acid-producing
bioinoculants be added at 10‘ cfu-g‘1-DM'1 for best results. This number, 10‘ to 10'I per g,
of propionic acid-producing bacteria at the time of storage may allow pr0pionic acid-

producing bacteria a competitive advantage when used as a silage bioinoculant.

Use of nucleic acid probes in microbial ecology

Reviews by Hazen and Jimenez (1988) and Sayler and layton (1990) discussed
probe technology, radioactive and non-radioactive labeling, colony and plaque
hybridization, environmental DNA recovery and hybridization as it pertains to the use of
nucleic acid probes in environmental microbiology. Currently, double-stranded DNA,
single-shanded DNA, RNA and oligonucleotides are used in the development of gene
probes for detecting speciﬁc sequences, and therefore speciﬁc species or strains of
microorganisms, in environmental applications.

Sensitivity and speciﬁcity of nucleic acid probes have been improved by the use of
nonradioactive labeling and detection (Martin et al. , 1990) with and without the use of the
polymerase chain reaction (Lion and Haas, 1990). Speciﬁc sequences of 16S rRNA have
been used for the ampliﬁcation of species and strain speciﬁc nucleic acid probes (Barry et
al. (1990). '

Nucleic acid probes are becoming increasingly popular for the evaluation of

interactions of different microorganisms in various ecosystems. Historically, procedures

32

for detection of nucleic acid sequences involved radioactive materials. Recently, emphasis
has been placed on the development and use of non-isotopic methods for detection of
nucleic acids. Digoxigenin is a steroid that can be cross-linked by a chemical spacer arm
to dexoxyuridine-triphosphate (DIG-dUTP) and is the basis for a commercial
nonradioactive labeling and detection kit (Genius-l , Boehringer Mannheim). The use of
DIG-dUTP in a random primed DNA labelling method can lead to its incorporation every
20 to 25 nucleotides in the newly synthesized DNA. The DIG-dUTP can then be detected
by' immunological methods and detected by colorimetric, ﬂuorescent or luminescent
techniques depending on the substrate chosen for the alkaline phosphatase conjugated to
the anti-DIG antibody fragment. Sensitivity of this method is comparable, if not better, to
radioactive techniques and can be done more safely and quickly.

Speciﬁc applications using nucleic acid probes are numerous. Most applications are
associated with pathogenic microorganisms such as the detection of Listeria monocytogenes
(Peterkine et al. , 1991; Wang, 1991). However, nucleic acid probes are commonly used
in microbial ecology studies. Examples in soil microbiology include the detection of
Bradyrhizobium japonicum (Holben et al. , 1988) and denitrifying bacteria (Smith and
Tiedje, 1992) in soil. Stahle et al. (1988) and Briesacher et al. (1992) have used nucleic
acid probes to monitor the rumen microbial ecosystem with emphasis on Fibrobacter
succinogenes S85 . Nucleic acid probes have been developed using regions of the 16S
rRNA DNA sequence that is speciﬁc to strains of Streptococcus bovis.

Very limited research has been done studying the use of nucleic acids for following

the persistence or interactions of silage microorganisms. Hill and Hill (1986) used plasmid

33

proﬁles for a bacterial inoculant based on a speciﬁc plasmid containing Lactobacillus
plantarum in corn silage. This approach was successful in differentiating lactic acid-
producing bacteria which proliferated during ensilement and identifying the organisms as
those which were added before ensilement as a lactic acid-producing bacterial inoculant.
Hendrich and coworkers (1991) developed a radioactive nucleic acid probe to a cryptic
plasmid of a strain of Bacillus pumilus that was being used as a bioinoculant for hay. This
was done to simplify monitoring the persistence of the inoculated strain in hay, soil and
plant material. In addition, the use of the probe was to relate efﬁcacy of the product to
organism numbers. Plasmid transfer between the inoculated strain of Bacillus pumilus and
epiphytic Bacillus pumilus was considered a disadvantage to this method.
Non-radioactive nucleic acid probes were developed to the total genomic DNA of
strains of lactic acid-producing bacteria involved in wine production by lonvaud-Funel
and coworkers (1991). This method showed that non-radioactive nucleic acid probes, based
on total genomic DNA, were species speciﬁc for most of the Lactobacillus spp. involved
in the fermentation of wine. In addition, Zarda and coworkers (1991) used non-
radioactive nucleic acid probes to detect single cells based on rRNA sequences. Generation
of nucleic acid probes from the genomic DNA of selected bacteria should allow for more
precise measurement of their status in diverse microbial ecosystems such as ensiled

feedstuffs.

CHAPTER 2

IN OCULATION OF RECONSTITUTED HIGH-MOISTURE CORN WITH
PROPIONIC ACID-PRODUCING BACTERIA WITH OR WITHOUT

ADDITION OF LACTIC ACID-PRODUCING BACTERIA

Abstract

Dry corn (90 % DM) was cracked in a roller mill and reconstituted with sterile
distilled deionized water to 70% DM. TYeatments consisted of control (no inoculum
added); PAPB (Propionibacterium acidipropionici ATCC 4965 and P. freudenreichii
ATCC 6207); LAPB (a commercial mixture of Lactobacillus plantarum, Streptococcus
faecium and Pediococcus acidilacticii); and PAPB + LAPB (a mixture of PAPB and
lAPB). All inocula was applied at 106 colony forming units per g DM. Approximately 1.5
kg of the treated reconstituted, high-moisture corn was placed into a laboratory silo.
Duplicate silos were prepared for each combination of treatment and storage interval. Silos
were opened on d 0, 2, 7, 14, 28, and 56. lactic acid-producing bacteria and non-mycotic
lactic acid- utilizing microorganisms were enumerated on LBS agar. or Na-lactate agar with
amphotericin-B. Yeasts and molds were enumerated on rose bengal agar supplemented
with chloramphenicol. The pH of the aqueous extract was lowest (P < .05) on d 5 6 for
LAPB and PAPB+LAPB (4.02 and 3.99) as compared to control and PAPB (4.31 and

34

35
4.24). Yeasts and molds were lowest (P < .05) in treatment PAPB on d 56. Propionic
acid content was highest (P < .01) in the PAPB treatment on all days except zero. In
addition, propionic acid concentration was highly correlated (r=-.81; P < .0001) with the

inhibition of the proliferation of yeasts and molds.

Introduction

Propionic acid is routinely used as a treatment for the prevention of mold
proliferation in high-moisture (80% DM or less) stored grains (Jones et al. , 1974). The
undissociated form of propionic acid is an effective bacteriostatic and mycostatic agent
(Crawshaw et al. , 1980). Propionic acid-producing bacteria are used in cheese production
and have been used as experimental amendments to silage fermentations. Lindgren and
coworkers (1983) used Propionibacterium shermanii to manipulate the fermentation of
grass silage; however, they were unsuccessful in changing the fermentation pattern.
F lores-Galarm and coworkers (1985) used a combination of selected propionic acid-
producing bacteria and lactic acid-producing bacteria to preserve high-moisture corn. A
beneﬁcial synergistic relationship between propionic acid-producing bacteria and lactic
acid-producing bacteria was observed in the above study (F lores-Galarza et al. , 1985) as
well as in an other study (Parker and Moon, 1982). The objective of this study was to
evaluate the effectiveness of a mixture of P. acidipropionici ATCC 4965 and P.
freudenreichii ATCC 6207 with or without added lactic acid-producing bacteria as possible

bioinoculants for the preservation of reconstituted, high-moisture corn.

36

Experimental Procedm’es

A 4 x 6 factorial arrangement of four treatments and six storage intervals (0, 2, 7,
14, 28, 56 d) was used to evaluate the effect of inoculation of reconstituted, high-moisture
corn using propionic acid-producing bacteria with or without lactic acid-producing
bacteria. Treatments were: 1) control (no inoculum added); 2) PAPB (a mixture of P.
acidipropionici ATCC 4965 and P. freudenreichii ATCC 6207); 3) LAPB (HMF- 10,
American Farm Products, Ypsilanti, MI, composed of Lactobacillus plantarum,
Pediococcus acidilacticii, and Streptococcus faecium); and 4) a mixture of treatments
PAPB and LAPB. Duplicate silos were used for each treatment by storage interval
combination. Dry com (90% DM) was cracked in a roller mill and reconstituted with
sterile deionized water to 70 % DM by hand agitiation for 15 min in a 20 1 plastic bucket.
Inocula were grown aerobically for 12 h to 109 cfu per ml. The PAPB inoculum was
grown in a medium modiﬁed from Malik et al. (1968) and Holdeman et al. (1977). This
media contained (g/l) Na-lactate, 16.0; peptone, 5.0; trypticase, 10.0; yeast extract, 10.0;
tween-80, .5; cysteine, .5; CaClz (anhydrous), 0.018; MgSO4-7H20, 0.018; KzHPO4,
0.04; KHZPO4, 0.04; NaHCO3, 0.4; NaCl, 0.08; MnSO4, 0.002; and amphotericin-B,
.001. The LAPB inoculum was reconstituted as directed by the manufacturer by diluting
the dry inoculant in distilled water to achieve a ﬁnal concentration of 10’ cfu/ml.
Approximately 1.5 kg of the reconstituted corn, along with 10‘ cfu/ g DM of the
appropriate organisms, was placed into a PVC laboratory silo equipped with a rubber
policeman for gas release release (Wardynski, 1991). Silos were emptied on d 0, 2, 7, 14,

28 and 56 and a homogenous 25 g sample was removed from each silo on the appropriate

37

day and extracted in 225 ml sterile saline (0.9% w/v) using a stomacher lab blender
(Tekmar 3500, Tekrnar, Cincinnati, OH) to dislodge the microﬂora from the com. A one
ml aliquot was serially diluted from 102 to 10“. lactic acid-producing bacteria were
enumerated on LBS agar medium (Difco, Detroit, MI). Non-mycotic lactic acid-utilizing
microorganisms were enumerated on the medium used for growth of the inoculum
supplemented with 1.5% agar. A similar medium was used by Flores-Glalarza et al.
(1985) for the presumptive enumeration of propionic acid-producing bacteria. However,
initial laboratory tests showed that lactic acid-producing bacteria were able to grow on this
media. Sterile agar media (12.5 ml molten agar at 45° C) was poured aseptically onto each
petri dish and the plates were gently swirled to disperse the microﬂora. Yeast and molds
were enumerated using rose bengal agar with chloramphenicol supplement (Difco, Detroit,
MI). Agar pour plates of the lactate producers and utilizers were incubated at 37°C in an
anaerobic glovebox (Coy laboratories, Troy, MI). Yeasts and molds plates were incubated
at 37° C aerobically. Microbial counts were performed using a digital colony counter
under a Quebec darkﬁeld counting chamber (American Optical, Buffalo, NY). Results are
expressed as log colony forming units per gram of dry matter (log cfu/ g DM). On (1 zero,
samples were processed for microbial counts within two hours of treatment application.
Acetic and propionic acid were determined on aliquots of the aqueous extracts by gas-
liquid chromatography as described by Baertsche et al. (1986). The pH of the aqueous
extract was measured within ﬁve minutes of extraction at 23 °C using a combination glass

electrode (AccupHast, Fisher Scientiﬁc, Chicago, IL) attached to a digital pH meter

38

(Fisher Scientiﬁc 825MP pH meter, Fisher Scientiﬁc, Chicago, IL) that was standardized

from pH 4 to 7 using commercial buffers (Curtis Mattheason, Woodale, IL).

The experiment was analyzed as a completely randomized design by analysis of

variance (Steele and Torrie, 1980) using the general linear models (GLM) procedure of

the Statistical Analysis System (SAS , 1987). Individual treatment means were compared

using the Bonferroni t-test (SAS , 1987). The AN OVA model for the fermentation study

was as follows:

Individual response variable measured (e. g. , pH, numbers
of lactic acid-producing bacteria, propionic acid).

Overall mean.

Effect of treatment (Control, PAPB, LAPB or
PAPB+LAPB).

Effect of period of ensiling (0, 2, 7, 14, 28 or 56 (1).
Interaction of treatment and period of ensiling.

Random residual effect (assumed normally distributed).

Results and Discussion

The pH of the aqueous extract of the ensiled, high-moisture corn (Table 2-1)

decreased over time in all treatments. An interaction of treatment and period of ensiling

(P < .0001) was observed. The interaction of treatment with period of ensiling is

39

generally observed in fermentation experiments due to the rapid growth and acid
production of the homofermentative lactic acid-producing bacteria as found in treatments
LAPB and PAPB+LAPB. Phillip and Fellner (1992), Schaefer et al. (1989) and
Wardynski et al. (1993) showed similar decreases in pH with the addition of
homofermentative lactic acid-producing bacteria. The pH decreased to a greater extent in
the LAPB and LAPB + PAPB treatments when compared to control or PAPB after 5 6 d
of ensilement. This indicates the PAPB did not increase acid production of the epiphytic
lactic acid-producing bacteria or the homofermentative lactic acid-producing bacteria in
treatment LAPB. The pH of PAPB treated corn was greater (P < .05)- than corn
inoculated with LAPB from d 2 to 56.

As seen with the change in silage pH there was a treatment by ensiling period (P
< .0001) interaction on numbers of lactic acid-producing bacteria (Table 2-2). All
treatments, including control, provided dramatic increases in numbers of lactic acid-
producing bacteria during the ﬁrst seven days of storage. Many studies have shown the
same effect when measuring the growth of added lactic acid-producing bacteria. lactic
acid-producing bacteria in the control silage numbers stabilized after d 7. Treatments
LAPB and LAPB + PAPB had decreases (P < .05) in numbers of lactic acid-producing
bacteria by d 56 to less than control or PAPB. Since the distribution or number of
homofermentative lactic acid-producing bacteria was not determined, care must be taken
in stating the decrease in numbers is only due to a switch in the type of lactic acid-

producing bacteria which were growing and predominating later in the fermentation.

4O

Lactic acid-utilizing bacteria (Table 2-3) followed the same general trend as
observed for lactic acid-producing bacteria. The lactic acid-utilizing bacteria were
enumerated using the medium that Flores-Galarza et al. (1986) which was presumptively
used for enumeration of propionic acid-producing bacteria in ensiled, high-moisture corn.
However, lactic acid-producing bacteria are capable of growing on this media as seen with
high numbers of lactic acid-utilizing bacteria in the LAPB treatment at d 0. Therefore, the
numbers of lactic acid-utilizing bacteria (as measured by this medium) followed the same
general trend as that followed by the lactic acid-producing bacteria. Control and PAPB
numbers of lactic acid-utilizing bacteria stabilized after d 7. By (I 28, LAPB and LAPB
+ PAPB had lactic acid-utilizing numbers less than (P < .01) control or PAPB.

The PAPB treatment was effective in decreasing the number of yeast and molds
(Table 2-4) when compared to all other treatments (P < .01) at d 14, 28, and 56. Yeasts
and molds were not enumerated on d 0, 2 or 7 due to problems in obtaining the rose
bengal agar. A treatment effect (P < .0001) was observed and attributable to the
decreased numbers of yeasts and molds observed in PAPB. The effect of adding propionic
acid-producing bacteria on decreasing the growth of yeasts and molds was also reported
by Flores-Galarza et al. (1985). However, Flores-Galarza et al. (1985) showed a much
greater decrease in yeasts to values less than 10‘ cfu-g"-DM". There was no estimate of
prop ionic acid production in that study and P. shermanii was used as an inoculant. In
addition, Flores-Galana et al. (1985) also reported decreases in yeasts and molds by using

lactic acid-producing inoculants. This was not found the case in this study or by Wardynski

41

(1991). Since initial numbers of yeasts and molds are not attainable for this experiment,
comparisons to the initial load of yeasts and molds cannot be made.

Acetic acid is a co-product in the fermentation pathways cf propionic acid-
producing bacteria (Wood, 1981). Therefore, the increase in the concentration of acetic
acid~ (Table 2-5) in the PAPB treated corn after 14 d of ensilement was expected. A
tendency (P < .13) was observed for an increase in acetic acid concentration during
ensilement for all treatments. Production of acetic acid within the silo can be advantageous
due to the antimycotic activities it possesses which is slightly less than that of propionic
acid. Amount of acetic acid produced during the ensiling fermentation was lower than that
reported by Wardynski et al. (1993). This may be attributed to slight differences in
moisture content of the com, the fact that the corn in this study was reconstituted and
possibly the length of ensilement. Decrease in pH and increases in the number of lactic
acid-producing bacteria tends to preclude the effect of reconstitution for it appears that
reconstituted corn follows a similar fermentation pattern as corn harvested at similar
moisture levels.

Propionic acid production (Table 2-6) was greatest (P < .0001) in corn treated
with the mixture of P. aadipropionici ATCC 4965 and P. freudenreichii ATCC 6207 after
2, 7, 14, 28 and 56 d of ensilement when compared to all other treatments. No propionic
acid was measured in either control or LAPB treated corn. Previous studies have found
that the propionic acid concentration of non-treated ensiled, high-moisture corn is
generally less than .1 g/ 100 g DM. A period of ensilement effect (P < .0001) indicates

production of propionic acid by treatment PAPB was time dependent and increased with

42

increasing length of storage as commonly seen in the lactic acid fermentations associated
with silage. Treatment PAPB+ LAPB had higher (P < .05) propionic acid concentration
from d 2 to d 56 than did control or LAPB. As expected, enhanced production of
prop ionic acid was not - observed in the PAPB + LAPB treatment as compared to PAPB.
The expectation was based on co-culture work of Parker and Moon (1982) and Lee et al.
(1976). Both of these studies showed increased propionic acid production when co-culturcs
of lactic acid-producing bacteria and propionic acid-producing bacteria were grown in
liquid media. Differences in substrates available for growth, water activity, pH of
environment and other factors associated with competition in the silage microbial
ecosystem probably did not allow for the same proliferation of the propionic acid-
producing bacteria when added in combination with LAPB.

Production of propionic acid within the silo is possible based on the results
presented. Recommendations for preserving corn containing 70% DM (moisture level used
in this study) state that 1.1% propionic acid should be added to preserve the corn for six
months (Jorgensen et al. , 1980) . Even though PAPB treated corn did not reach the 1.1%
propionic acid, there was a noticeable effect on the numbers of yeasts and molds. Further
studies are needed to select propionic acid-producing bacteria capable of competing for
substrates found in ensiled, high-moisture corn. In addition, future studies should address
what concentration is necessary to be produced within the silo to yield a stable product.
Many variables will need to be addressed such as temperature of storage, moisture level

of the ensiled crop, amount of oxygen ingression during storage, feedout rates and many

43

 

 

 

Table 2-1. Effect of propionic acid-producing bacteria (PAPB) with or without
addition of lactic acid-producing bacteria (LAPB) on the pH of ensiled
reconstituted, high-moisture corn

Time Treatment
(‘1) Control PAPB LAPB Pg; SEM
pH of aqueous extract

0 5.33' 5.52" 5.69" 5.52" .063
4.83" 4.96” 3.90‘ 4.01'

7 4.84c 4.35" 3.83‘l 3.95‘I

14 4.57c 4.31” 3.90' 3.99‘

28 4.46" 4.28" 3.93‘ 3.92'

56 4.31" 4.24b 4.02‘I 3.99‘

 

“’° Means within a row lacking a common superscript differ (P < .05)

 

 

 

Table 2-2. Effect of propionic acid-producing bacteria (PAPB) with or without
addition of lactic acid-producing bacteria (LAPB) on the number of lactic
acid-producing bacteria of ensiled reconstituted, high-moisture corn

Time Treatment
(‘1) Control PAPB LAPB P311231” SEM
log cfu-g'1-DM'l

0 5.10" 4.64‘I 8.11‘l 7.81c .091
7.94‘ 8.54" 9.34c 948°

7 9.00' 9.02“ 9.38° 9.27"c

14 9.13b 8.57‘ 880' 856|

28 8.96° 8.63" 8.18‘l 8.33‘I

56 8.79" 8.61" 7.19‘ 7.39‘

 

“’°" Means within a row lacking a common superscript differ (P < .05)

45

 

 

 

Table 2-3. Effect of propionic acid-producing bacteria (PAPB) with or without
addition of lactic acid-producing bacteria (LAPB) on the number of lactic
acid- utilizing microorganisms of ensiled reconstituted, high-moisture corn

Time Treatment
(‘1) Control PAPB LAPB P33; SEM
log cfu-g‘I-DM’l

5.74' 877’ 8.36" 8.49" .092
8.33‘ 8.61" 9.14c 9.39c

7 8.94‘ 8.97' 9.30b 9.20”

14 9.14" 8.99“” 8.84‘I 8.76‘

28 886“ 8.58" 8.17‘I 8.24'

56 874° 873‘ 7.09‘ 7.73"

 

“’° Means within a row lacking a common superscript differ (P < .05)

46

 

 

 

Table 2—4. Effect of propionic acid-producing bacteria (PAPB) with or without
addition of lactic acid-producing bacteria (LAPB) on the number of yeasts
and molds of ensiled reconstituted, high-moisture corn

T' Treatment
1me
SEM

(d) PAPB +

Control PAPB LAPB LAPB

log cfu-g'I-DM'l

14 6.55" 4.81' 6.15" 6.45” .207
28 6.37" 5.49‘ 6.35" 6.16"
56 6.01" 5.37' 6.18" 6.08"

 

‘5 Means within a row lacking a common superscript differ (P < .05)

47

Table 2-5. Effect of propionic acid-producing bacteria (PAPB) with or without
addition of lactic acid-producing bacteria (LAPB) on the acetic acid content
of ensiled reconstituted, high-moisture corn

 

 

 

Time Treatment
(‘1) Control PAPB LAPB Pg; SEM
g/ 100 g DM

.10 .13 .14 .06 .037
.18 .20 .10 .16
.14 .20 .11 .17

14 .13' .25b .10 .16“

28 .11 .16 .11 .15

56 .15 .22 .15 .20

 

“’ Means within a row lacking a common superscript differ (P < .05)

48

Table 2-6. Effect of propionic acid-producing bacteria (PAPB) with or without
addition of lactic acid-producing bacteria (LAPB) on the propionic acid
content of ensiled reconstituted, high-moisture corn

 

 

 

Time Treatment
(‘1) Control PAPB LAPB P33; SEM
g/ 100 g DM

0 o .00 .00 .00 .015
.00- .13° .00- .05”

7 .00- .23° .00- .05"

14 .00- .21c .00 .05b

28 .00- .14c .00- .05b

56 .00- .26c .00- .071

 

d” Means within a row lacking a common superscript differ (P < .05)

49

other factors addressed which impact the ensiling fermentation and aerobic stability of

ensiled feeds.

Implications

Addition of a mixture Propionibacterium acidipropionici ATCC 4965 and P.
freudenreichii ATCC 6207 as a bioinoculant to reconstituted, high-moisture corn was
effective in decreasing the numbers of yeast and molds when compared to control, lactic
acid-producing bacteria and the lactic acid-producing bacteria + propionic acid-producing
bacteria mixtm'e. Addition of P. acrdipropionici ATCC 4965 and P. freudenreichii ATCC
6207 was not effective in decreasing silage pH when compared to control, and reached a
higher terminal pH when compared to the lactic acid-producing bacterial inoculant.
Addition of propionic acid-producing bacteria to ensiled, high-moisture corn shows
promise as a bacterial inoculant based on decreased proliferation of yeasts and molds.
More studies need to be conducted to isolate propionic acid-producing bacteria that are

more suitable for use in ensiled crops.

CHAPTER 3

ENRICHMENT AND ISOLATION OF PROPIONIC ACID-PRODUCIN G

BACTERIA FROM ENSILED FEEDSTUFFS

Abstract

Propionic acid is an effective mold and bacterial-inhibiting agent commonly used in
human food products and animal feeds. Propionic acid-producing bacteria are commonly
used for cheese production. The objective of this study was to develop methods to isolate
indigenous propionic acid-producing bacteria from ensiled feedstuffs. Thirty-eight samples
of various ensiled feeds (corn silage, ensiled, high-moisture corn, alfalfa haylage and
oatlage) were obtained from Michigan farms. Samples were homogenized and enrichments
made in a medium containing: sodium lactate, peptone, yeast extract, minerals, buffers and
amphotericin-B. Primary enrichments were incubated at 30°C until turbid growth was
attained. Enrichments were analyzed for propionic acid concentration by high performance
liquid chromatography and primary enrichments containing greater than 10 mM propionic
acid were transferred to a secondary enrichment medium containing sodium propionate.
Progressive enrichments were allowed to grow until turbid and analyzed for propionate
production. Enrichments with greater than 10 mM propionate production were serially

diluted and plated on a homologous agar. Single colony isolates were transferred to

50

51

homologous agar with sterile toothpicks and streaked for purity and characterized by
fermentation proﬁles, catalase production, and gram stain. A bacterium was isolated from
ensiled, high-moisture corn that produced propionic acid and acetic acid from either lactate

or glucose and is further characterized in Chapter 4.

Introduction

Propionic acid is routinely used as a treatment for the prevention of mold
proliferation in high-moisture (i.e., less than 80% DM) stored grains (Jones et al., 1974).
The undissociated form of propionic acid is an effective bacteriostatic and mycostatic agent
(Crawshaw et al., 1980). Propionic acid-producing bacteria are used in cheese production
and limited commercial production of propionic acid. Lindgren and coworkers (1983) used
Propionibacterium shermanii to manipulate the fermentation of grass silage. However, they 4
were unsuccessful in changing the fermentation pattern. Flores-Galarza and coworkers
(1985) used a combination of selected propionic acid-producing bacteria and lactic acid
bacteria to preserve high-moisture com. A synergistic relationship between propionic acid-
producing bacteria and lactic acid-producing bacteria was observed in the above study
(Flores-Galena et al., 1985). Similarly, Parker and Moon (1982), reported enhance growth
of lactic acid-producing bacteria and increased production of propionic acid by propionic
acid-producing bacteria when grown in co-culture with lactic acid-producing bacteria. lactic
acid-utilizing bacteria, which produce propionic acid, have been isolated from grass silage
(Woolford, 19750). However, these organisms did not account for a large portion of the

microbial population (Rosenberger, 1951; Woolford, 1975c). Currently, enrichment and

52
isolation of propionic acid-producing bacteria are still based on their ability to use lactic
acid (Atlas and Parks, 1993). Selectivity of the above media is based on the fermentation
of lactate by propionic acid-producing bacteria under anaerobic conditions. However,
lactic acid-producing bacteria and other microorganisms commonly found in silages can
use the same media components. Reddy and coworkers (197 3) and Drinan and Cogan
(1992) attempted to correct this problem by addition of two antibiotics, cloxacillin and
kanamycin. This alleviated the growth of most lactic acid-producing bacteria except
Leuconostocs. The concentration of antibiotics necessary to inhibit the grth of
Leuconostocs was also inhibitory to the growth of propionic acid-producing bacteria of
interest. The objective of this study was to develop methods to simplify isolation of
indigenous propionic acid-producing bacteria from ensiled feedstuffs that might be useful

as bioinoculants for ensiled, high-moisture corn.

Experimental Procedures

Sample description. Thirty-eight samples (approximately 500 g) of various ensiled
feedstuffs were obtained from different Michigan farms. The geographical source and type
of ensiled material obtained for inoculum source for enrichment cultures are shown in A
Table 3-1.

Media composition. Enrichment and isolation media were similar to that of Malik
et al. (1968) and Holdeman and coworkers (1977). Modiﬁcations included the addition
of amphotericin-B (Sigma A-9528, St. louis, M0) to decrease proliferation of yeasts and

molds in enrichment cultures and use of Hungate tubes with nitrogen as gas phase. Primary

53

enrichment media contained (g/ L in distilled deionized water): Na-lactate syrup (60 %
syrup), 16.0; peptone, 1.0; yeast extract, 1.0; Tween-80, .5; cysteine, .5; CaCl,, .018;
MgSO,-7H20, .018; KZHPO“ .04; KH2P04, .04; NaHC03, .4; NaCl, .08; MnSO4, .002
and Amphotericin-B, .001 . Secondary enrichment media contained (g/ L in distilled
deionized water): Na-lactate syrup (60% syrup), 16.0; peptone, 1.0; Na-propionate, 10.0;
yeastextract, 1.0; Tween-80, .5; cysteine, .5; CaCl,_, .018; MgSO,-7H20, .018; K2HP04, .04;
KH2P04, .04; NaHC03, .4; NaCl, .08; MnSOg .002 and Amphotericin-B, .001. Primary and
scoondary enrichment media were prepared using nitrogen as the gas phase and 10 ml of
primary or secondary enrichment media was pipetted into 16 x 125 mm Hungate tubes
(Bellco 2047-00125 , Vineland, NJ) which were scaled by a manual crimper (Bellco,
Vineland, NJ). Both the primary and secondary enrichment media used lactic acid as the
main carbon source for fermentation. Sodium propionate was added to the secondary
enrichment medium to reduce competition of non-propionic acid-producing bacteria from
the primary enrichment meduim. Isolation media contained (g/ L in distilled deionized
water): Na-lactate syrup (60% syrup), 16.0; peptone, 5.0; Na-propionate, 10.0; trypticase,
10.0; yeast extract, 1.0; Tween-80, .5; cysteine, .5; CaClz, .018; MgSO4-7H20, .018;
K2HP04, .04; KHZPO,“ .04; NaHCO3, .4; NaCl, .08; MnSOi, .002; Amphotericin-B, .001 and
agar, 15.0.

Enrichment and isolation procedures. Twenty-ﬁve g of ensiled material described
in Table 3-1 was homogenized in 225 ml of autoclaved .9% NaCl (pH 7.0) in a
Stomacher (Model 3500, Tekmar, Cincinnati, OH) for 5 min. A 100 pl aliquot was

removed by a sterile one ml tuberculin syringe ﬁtted with a 5/8 inch 25 gauge needle

54

(Becton Dickinson 9626, Rutherford, NJ). This aliquot was transferred to a Hungate tube
containing 10 ml of primary enrichment media. Primary enrichments were incubated at 30
°C (optimum growth temperature for classical Propionibacterium spp. , Glatz, 1992) until
turbidity was observed. Primary enrichment media contained sodium lactate as an selective
agent for the growth of propionic acid-producing bacteria (Atlas and Parks, 1993). Since
yeasts and molds also metabolize lactic acid, growth of lactic acid-utilizing yeasts and
molds was reduced by the inclusion of the mycostatic agent, amphotericin-B. One ml
aliquots of enrichments were removed using one ml syringes (Becton Deckinson 9626,
Rutherford, NJ) and aseptic technique. Samples were analyzed for propionic acid
production by HPLC procedures as described later. Primary enrichment cultures
containing greater than 10 mM pr0pionic acid were transferred to Hungate tubes
containing 10 ml secondary enrichment media to help dilute non-propionic acid-producing
bacteria from the enrichment cultures. Secondary enrichment media contained sodium
propionate as an agent for the inhibition of non-desired microorganisms and to enhance the
growth of propionic acid-producing bacteria. Enrichments were incubated as before and
HPLC analysis was done on cultures that had turbid growth.

Secondary enrichments which had propionic acid content greater than 10 mM above
the added sodium propionate were serially diluted from 101 to 10‘ in autoclaved .9 %
NaCl. A 500 pl aliquot of each dilution was spread on the surface of 10 ml of isolation
media that was poured aseptically into sterile 100 x 15 mm Petri dishes (Baxter D1096,
Deerﬁeld, IL). The plates were incubated in an anaerobic glovebox (Coy laboratories,

Troy, MI) and visually checked for colony formation. Single colonies were aseptically

55

removed and streaked onto fresh isolation media plates to evaluate culture purity. Single
colonies were transferred to peptone-yeast extract-lactate media (Holdeman et al. , 1977)
and incubated as before. Culture extracts were analyzed by HPLC for propionic acid
production as were enrichment cultures. Pure cultures which produced prop ionic acid were
stored in peptone-yeast extract-lactate media mixed with 50 % glycerol at -20° C. Gram
stain and catalase reaction were performed as described by Holdeman and coworkers
(1977).

HPLC analysis of enrichment cultm'es. Analysis of progressive enrichment media
for propionic acid was done by ion exchange-exclusion HPLC (BIORAD aminex HPX-
87H, Richmond, CA). Mobile phase consisted of .005 N H2804 at a ﬂow rate of .9
ml/ min. Column temperature was maintained at 65° C by an external column heater
(Waters Millipore, Milford, MA). Three ml of enrichment cultures were ﬁltered through
.2 pm ion chromatography syringe ﬁlters (Gelman Acrodisks, 25 mm, Ann Arbor, MI)
into 3 ml HPLC sample vials (National Scientiﬁc, Atlanta, GA). Filtered samples were
stored at -20 °C until analysis. Twenty pl of each ﬁltered sample was injected by an
autoinjector (Waters WISP 712, MIlford, MA) and analytes were detected by refractive
index (Waters 410 refractive index detector, Milford, MA). Peak heights were quantiﬁed
by a commercial HPLC software package (Turbochrom 3, PE Nelson, Cupertino, CA) and

compared to standards of lactic, acetic and propionic acid (Supelco, Bellefonte, PA).

56‘

Table 3-1 Source and type of sample used as enrichment sources for propionic acid-
producing moisture bacteria from ensiled feeds

 

Farm or location
Michigan State University Beef Cattle Research Unit

Michigan State University Dairy Research Unit
Galesburg, MI
Parma, MI

Big Rapids, MI

Falmouth, MI

Reed City, MI
Chatam, MI
Muir, MI

Litchﬁeld, MI

Mason, MI

Sample Type
corn silage
ensiled, high-moisture corn
alfalfa haylage
corn silage
ensiled, high-moisture corn
alfalfa haylage
corn silage
ensiled, high-moisture com
corn silage
alfalfa haylage
oatlage
alfalfa haylage
alfalfa haylage + corn
silage
corn silage
corn silage
alfalfa haylage
alfalfa haylage
ammoniated corn silage
alfalfa haylage
corn silage
alfalfa haylage, corn silage

ensiled, high-moisture corn

 

57

Results and Discussion

Few samples of ensiled feeds had primary enrichments that produced propionic
acid. A total of 219 primary enrichments was prepared and analde using 38 different
inoculum sources. Only 34 % of the primary enrichments had propionic acid production
above the threshold limit. Eight of the primary enrichments yielded secondary enrichments
with propionic acid production. Growth of epiphytic microorganisms was observed in all
enrichments suggesting that the enrichment media used was not selective only for propionic
acid-producing bacteria. Many enrichments that did not produce propionic acid were
characterized by copious gas production and very putrid odor production. Analysis of these
primary enrichments by HPLC indicated high amount of n-butyric and i-butyric acids with
higher (unidentiﬁed) chain length volatile fatty acids. Production of butyric acid in
enrichments would suggest possible enrichment for lactate-utilizing clostridia.

Previous attempts by Woolford (1975c) were not successful in isolating propionic
acid-producing bacteria from fresh forages. Rosenberger (1956) found that numbers of
Propionibacterium spp. were very low in silage. Woolford (1975c) was able to isolate a
limited number of propionic acid-producing bacteria resembling Propionibacterium spp.
from (two out of six) ensiled samples. A recent review of the literature did not ﬁnd any
published studies regarding the isolation of propionic acid-producing bacteria from ensiled
feeds. Very few colonies of bacteria grew on the isolation media after transfer from
secondary enrichments that had propionic acid production. Although an antimycotic agent
was used, amphotericin-B, overgrowth of molds was observed on many plates of isolation

media.

58

Use of the techniques described facilitated the isolation of a bacterium from
ensiled, high-moisture corn (from the MSU Dairy Research and Teaching Facility) that
produced propionic acid and acetic acid from either lactate or glucose. The bacterium is
a gram-positive, facultative anaerobe, non-sporeforming, nonmotile, pleiomorphic rod
which forms clumps of cells when grown in peptone-yeast extract-lactate medium. Surface
colonies, when grown aerobically on agar plates are pin-point (approximately 1 mm) in
size, yellow, circular and possess a strong catalase reaction. End-products from the
fermentation of lactate and glucose are acetate and propionate (1:3 and 1:5 ratio,

respectively). This bacterium is further characterized in Chapter 4.

Implications
The use of progressive enrichments in a semi-selective media allows for the
selection and screening of many different types of microorganisms. Using the methods
discussed, a rapid growing propionic acid-producing bacteria was isolated from ensiled,
high-moisture corn. Adaptation of this selection process, or use of improved selection
media, will simplify enrichment and isolation of novel, propionic acid-producing bacteria
from ensiled feeds that should enhance selection of new strains of propionic acid-

producing bacterial inoculants.

CHAPTER 4

CHARACTERIZATION OF PROPIONIBACTERIUM ACIDIPROPIONICI DH42,
A PROPIONIC ACID-PRODUCING BACTERIA

ISOLATED FROM ENSILED, HIGH-MOISTURE CORN

Abstract

A bacterium was isolated from ensiled, high-moisture corn that produced propionic acid and
acetic acid from either lactate or glucose. The bacterium is a gram-positive, non-
sporeforming, non-motile, pleiomorphic rod, facultative anaerobe which forms clumps of
cells when grown in peptone-yeast extract-lactate medium. Surface colonies are pin-point
size after ﬁve days aerobic incubations at 30 ° C. In addition, surface colonies are yellow,
circular, and possess a strong catalase reaction. The end-products from the fermentation
of lactate and glucose are acetate and propionate (1:3 and 1:5 ratio, respectively). The
bacterium grows rapidly using glucose under anaerobic and aerobic conditions in liquid
media. However, the organism grows very slowly on the surface of aerobically incubated
solid media. The majority of the phenotypic characteristics of the isolated bacterium match
those of the ATCC type strain of Propionibacterium acidipropionici, therefore the name
Propionibacterium acidipropionici DH42 is proposed for the identiﬁcation of the isolated
propionic acid-producing bacterium.

59

60
Introduction

Classical propionic acid-producing bacteria, those associated with dairy products,
are generally characterized as being gram-positive, catalase-positive, non-sporeforming,
rod-shaped facultative anaerobes with pH optimum for growth between 6 and 7 (Hettinga
and Reinbold, 1972a; Hsu and Yang, 1991). Propionibacterium shermanii (Lindgren et
al., 1983; Flores-Galarza et al., 1985) and P. jensenii (Tomes, 1991) have been used with
mixed results as inoculants in corn silage, grass silage and ensiled, high-moisture corn. In
addition, P. acidipropionici and P. freudenreichii were used in Chapter 2 as inoculants for
reconstituted ensiled, high-moisture corn. However, most of the above mentioned
Propionibacterium spp. were previously isolated from dairy products. The objective of this
study was to phenotypically characterize and identify a fast growing, propionic acid-
producing bacteria, designated DH42, isolated from ensiled, high-moisture corn in Chapter

3.

Experimental Procedures
Bacterial strains. A propionic acid-producing bacteria previously isolated from
ensiled, high-moisture corn was maintained by growing in .5X (half-strength) MRS
(Difco, Detroit, MI) and storing at -20° in glycerol. Reference strains of propionic acid-
producing bacteria were obtained from the following sources to facilitate phenotypic
comparisons of the isolated bacterium to known strains. Propionilracteriurn jensenii P25 ,
P. shermanii P92 and P. thoenii were kind gifts from Chr. Hansen's Biosystems

(Milwaukee, WI). Propionibacterium acidipropionici 22562 and P. freudenreichii 1382

61

were obtained from Michigan Biotechnology Institute (Lansing, MI). All cultures were
grown in 16 x 125 mm Hungate tubes (Bellco 2047-00125, Vineland, NJ) sealed with a
manual crimper (Bellco, Vineland, NJ) under ambient gas phase at 30 'C in .5X MRS
(Difco, Detroit, MI) and maintained at -20 'C stored in glycerol.

Media composition. General media used for this study included .5X MRS (Difco,
Detroit) for fermentation proﬁles and maintenance of cultures. Cellobiose, erythritol,
lactose, maltose, mannitol, sorbitol, soluble starch, sucrose and trehalose were added to
a ﬁnal concentration of 1% as sterile solutions to purple broth base (BBL 11558,
Cockeysville, MD) to assess fermentation characteristics. Peptone-yeast extract-lactate and
peptone-yeast extract-glucose media were used for pH optima determination (Holdeman
et al. , 1977).

HPLC analys’m of fermentation end-products. Aliquots of fermentation media
were analyzed for acidic end-products by ion exchange-exclusion HPLC (BIORAD aminex
HPX-87H, Richmond, CA). Mobile phase consisted of .005 N HZSO4 at a ﬂow rate of .9
ml/ min. Column temperature was maintained at 65°C by an external column heater
(Waters Millipore, Milford, MA). Three ml of enrichment cultures were ﬁltered through
.2 pm ion chromatography syringe ﬁlters (Gelman Acrodisks, 25 mm, Ann Arbor, MI)
into 3 ml HPLC sample vials (National Scientiﬁc, Atlanta, GA). Filtered samples were
stored at -20 °C until analysis. Twenty p1 of the ﬁltered samples was injected by an
autoinjector (Waters WISP 712, Milford, MA) and analytes were detected by refractive

index (Waters 410 refractive index detector, Milford, MA). Peak heights were quantiﬁed

62

by a commercial HPLC software package (Turbochrom 3, PE Nelson, Cupertino, CA) and
compared to standards of lactic, acetic and propionic acid (Supelco, Bellefonte, PA).

Metabolic proﬁle determination. BIOLOG-GP (BIOLOG systems, gram positive
microplate, Hayward, CA) was used to evaluate the metabolic proﬁle of the propionic
acid-producing bacteria isolated from high-moisture corn and reference strains of
Propionibacterium spp. listed previously. The following procedure was modiﬁed for liquid
cultures as compared to cultures on solid agar described in the BIOLOG user's manual
(Hayward, CA). Cultures were grown in .5X MRS (Difco, Detroit) in Hungate tubes at
30 °C to mid-exponential phase (ODm = .8) and transferred to fresh .5X MRS. These
cultures were grown for 12 h and 1 ml was aseptically removed and placed into a sterile
microcentrifuge tube. Cultures were harvested by centrifugation (2 min @ 16,000 x g) and
resuspended in 18 ml sterile .9% NaCl. An eight-channel multipipetter was used to
dispense 150 pl of the resuspended culture per well of the BIOLOG-GP microplate.
BIOLOG-GP microplates were incubated at 30 °C for 24 h. Optical density (ODsgo) of
each well was measured by a BIOLOG microstation plate reader and results transformed
to compare to known strains of gram positive bacteria using the BIOLOG microstation
gram positive database and the reference strains of Propionibacterium spp.

Cellular membrane fatty acid composition determination. Fatty acid content of
the cellular membranes was analyzed as fatty acid-methyl esters using the MIDI Microbial
Identiﬁcation System of bacteria by gas chromatography of Cellular fatty acids (MIDI,

Newark, DE) at Michigan Biotechnology Institute (Lansing, MI). The proﬁle of fatty acid-

63

methyl esters was compared to known proﬁles of reference strains of different bacteria to
help in identiﬁcation.

Growth characteristics. Optimal pH for growth on either glucose or lactate was
determined using a 1% inoculum of the isolated bacterium taken from an exponentially
growing culture and inoculated into peptone-yeast extract-lactate and peptone-yeast extract-
glucose (Holdeman et al., 1977) at pH values of 4.9, 5.3, 5.7, 5.8, 6.2, 6.5, 6.7, 6.9, 7.4
and 7.9. The pH of the media was measured after autoclaving. Initial pH values before
autoclavmg were 4.5, 5.0, 5.5, 6.0, 6.5, 6.8, 6.9, 7.0, 7.5 and 8.0. Duplicate tubes of
each media and pH were incubated at 30 °C and OD“ was measured in a Specironic 70
(Bausch and Lomb, Newark, NJ) at 60 min intervals during growth. Speciﬁc growth rate
was determined for each pH and carbon source (glucose or lactate) by determining the
slope of the natural log of OD“ against time by regression procedures (PROC REG) of
SAS (SAS, 1987) as described by Russel et al. (1979). Therefore, the speciﬁc growth rate
of the isolated organism is reported with the units or h". These units correspond to the
change in absorbance at 600 nm (a unitless measurement) per change in time (h) which is
the slope of the line which describes exponential growth. Nitrate reduction, esculin
hydrolysis, catalase production, gelatin digestion and gram stain were performed using the

procedures described by Holdeman and coworkers (1977).

Results and Discussion
Propionic acid was the major fermentation product from glucose and lactic acid as

previously discussed in Chapter 3 of the fast growing isolate (DH42) from ensiled, high-

64

moisture corn. The differential characteristics of classical species of the genus
Propionibacterium are listed in Table 4-1. The propionic acid-producing bacteria (DH42)
isolated from ensiled, high-moisture corn hydrolyzed esculin, reduced nitrate and produced
acidic end-products from cellobiose, erythritol, lactose, maltose, mannitol, sorbitol,
soluble starch, sucrose and trehalose. In addition to these fermentation characteristics, the
isolated organism did not digest gelatin. This allowed for the classiﬁcation of isolate DH42
as a classical propionic acid-producing bacteria and not a cutaneous strain. These
characteristics of the isolated bacteria, DH42, matched the differential characteristics of
P. acidipropionici 225 62. Based on these characteristics, DH42 was presumptively
identiﬁed as P. acidipropionici. Future discussion of results will refer to the organism as
Propionibacterium acidipropionici DH42.

Phase contrast microscopy (Figure 4-1) revealed that P. acidipropionici DH42 was
a pleiomorphic rod that followed the description of Propionibacterium spp. (Holdeman et
al. , 1977). Propionibacterium acidipropionici DH42 was approximately .5 to .9 pm wide
and up to 10 pm in length. Cells of P. acidipropionici DH42 generally appeared in pairs
or were aggregated into small clumps containing up to 100 cells when grown in .5X MRS.
The general morphology was similar to that of the type strain of P. acidipropionici (ATCC
22562).

Propionibacterium acidipropionici DH42 grew slowly on the surface of aerobic
tryptic soy agar plates supplemented with 1% glucose. Colonies were pinpoint in size after
ﬁve days growth at 30 °C. In addition, surface colonies were a translucent yellow, circular

with entire margins and were catalase positive. Propionibacterium acidipropionici DH42

65

grew well in liquid media aerobically or anaerobically. Maximum growth was generally
attained after 12 to 24 h in .5X MRS incubated at 30 °C. No noticeable gas was produced,
as measured by inverted Durham tubes, after ﬁve d aerobic incubation in .5X MRS.
Terminal pH (48 h growth) of cultures of P. acidipropionici DH42 was 3.9 and 4.1 for
cultures grown in purple base-glucose and .5X MRS, respectively. The sediment of cells
of P. acidipropionici DH42 after 72 h growth was ropy with some granularity. This
deviates from the type strain of P. acidipropionici ATCC 22562 that had a ropy sediment
with no noticeable granularity. No noticeable pigment was produced by P. acidipropionici
DH42 when grown in .5X MRS. All cultures were a creamy white color in contrast to the

dark reddish-brown color of P. theonii P15.

66

 

Figure 4-1. Phase contrast photomicrograph of P. acidipropionici DH42 grown in .5X
MRS under ambient gas phase in sealed Hungate tubes (1500X
magniﬁcation, each bacterium approximately .7 pm wide).

67

Table 4-1 Differential characteristics of Propionibacterium acidipropionici DH42, P.
' acidipropionici 22562, P. freudenreichii 1382, P. jensenii P25, P. thoenii
P15 and P. shermanii P92

 

Characteristic DH42 acidiprop- “wail" jensenii thoenii shermanii
lonlcl reichu

esculin

hydrolysis "' + + + + ‘

gelatin
digestion
nitrate reduced + + + .. - -
Acidic end-products from:
cellobiose +

erythritol

lactose

++++
+++

maltose

+++++

mannitol
sorbitol
soluble starch

SUCI'OSC

++++++++
+++++++++

+
+++
++++

trehalose

 

68

The BIOLOG-GP library (gram positive database) does not contain metabolic
information for Propionibacterium spp. However, use of the microplate reader and the
pre-ﬁlled BIOLOG-GP allows for a rapid determination of metabolic characteristics of
isolates. Since P. acidipropionici DH42 grew slowly on the surface of agar plates, cells
were harvested from liquid culture and washed in saline. Procedures followed from that
step are identical to those used for the BIOLOG system as described in the experimental
procedures section. Reference strains of Propionibacterium spp. and P. acidipropionici
DH42 were capable of metabolizing a vast array of compounds found in the BIOLOG-GP
microplate (Table 4-2). The BIOLOG software could discriminate the different
Propionibacterium spp. evaluated, but was unable to differentiate a species identiﬁcation
from its database. Propionibacterium acidipropionici DH42 matched the metabolic proﬁle
of P. acidipropionici 22562 with few discrepancies. Propionibacterium acidipropionici
DH42 metabolized D-arabitol, a-D-lactose, mono-methyl-succinate, alaniniamide,
putrescine, adenosine-S '-monophosphate and uridine-5'-monophosphate, whereas P.
acidipropionici 225 62 was unable to fully metabolize these compounds. In contrast, P.
acidipropionici 22562 was able to metabolize acetic acid and a-hydroxybutyric acid better
than P. acidipropionici DH42.

Fatty acid compositional analysis of the cellular fatty acids (Table 4-3.) conﬁrmed
the preliminary identiﬁcation of DH42 as P. acidipropionici. Most of the cellular fatty
acids consisted of the isomers a 15-carbon fatty acid. Greater. than 80 % of the total
cellular fatty acids were composed of saturated fatty acids of 15 carbons in length. This

matches the description of the cellular lipids seen in classical propionic acid-producing

69
bacteria (Hettinga and Reinbold, 1972a). Comparison of the fatty acid composition to the
database of the MIDI system found that P. acidipropionici DH42 had a similarity index
of .58 to the type strain of P. acidipropionici. An index greater than .6 is considered an
excellent match using the MIDI system, whereas indices greater than .5 are considered to
be good matches. Slight differences in the cellular fatty acids are to be expected between
different isolates, especially when the two strains are from drastically different ecosystems.
Strains of Propionibacterium spp. in the MIDI database are based on the
Priopionibacteriurn spp. from Virginia Polytechnical Institute Anaerobe Laboratory that are
predominately associated with dairy products.

Growth characteristics of P. acidipropionci DH42 using glucose or lactate are
shown in Figure 4-2. Previous research by Hsu and Yang (1992) using P. acidipropionici
ATCC 4875 had similar results for growth characteristics when grown on glucose or
lactate at pH 6.6. In that study, P. acidipropionici ATCC 4875 grew slower on lactic acid
than glucose, however it was capable of producing similar amounts of propionic acid
which indicated a higher speciﬁc substrate consumption rate. The use of glucose and lactic
acid over a range of media pH is described in Table 4—4. Propionibacterium
acidipropionici DH42 grew more rapidly (P < .01) on glucose than lactic acid at all pH
values evaluated. The effect of pH on the speciﬁc growth rate of P. acidipropionici DH42
was greater when the cells were grown on lactic acid. The optimal pH for growth on lactic
acid was from 6.2 to 7.8, whereas maximal growth was obtained at pH 5.3 when grown

on glucose with acceptable growth rates between pH 4.9 and 7.8.

70

Table 4-2 Metabolic proﬁle of Propionibacterium acidipropionici DH42, P.
acidipropionici 22562, P. freudenreichii 1382, 'P. jensenii P25, P.
shermanii P92 and P. thoenii P15 as measured by BIOLOG-GP microplates

 

acidiprop- freuden-

Substrate DH42 jensenii shermanii thoenii

 

ionici reichii

e-cyclodextrin - - .. - - -
p-cyclodenrin - - / - .. -
demin + + + + + +
glycogen + / / / / /
mannan - - .. - - -
tween40 + + + + + +
tween 80 + + + + + +
N-acetyl-D- + + + + + +
glucosaminc

N-acetyl-P- + + + + + +
mannosamme

amygdalin - - - - - -
L-arabinose + + + + + +
D-arabitol + - + + + +
arbutin + + + + + +
cellobiose + + + + + +
D-fructose + + + + + +
bﬁicose + + + + + +
D-galactose + + + + + +
D-galacmronic acid + + + + + +
gentiobiose + + + + + +

 

71

Table 4-2 (cont'd).

 

Substrate DH42 “dim" ("New jensenii shermanii thoenii

 

ionici reichii

D-gluconic acid + + + + + +
a-D-glucose + + + + + +
m-inositol + + + + + +
u-D-lactose + / - / + /
lactulose - - - - + _
maltose + + + + + +
maltotriose + + + + + +
D-mannitol + + + + + +
D-mannose + + + + + +
D-melezitose - — + .. .. -
D-melibiose + + - + + +
III-methyl D- + + + + + +
galactoside

ﬂ-methyl D- + + + + + /
galactoside

3-methyl glucose / / / / / -
a-methyl D- + + / - / _
glucoside

ﬂ-methyl D- + + + + + +
glucoside

a-methyl D- - — - - - -
mannoside

palatinose + + + + + +
D-psicose + + / + + /

D-raffmose + + + + + +

 

72

Table 4-2 (cont'd).
—

 

Substrate DH42 “ﬁg?“ {333” jensenii shermanii thoenii
Irrhamnose + + + + + +
sedoheptulosan - - .. - - -
D-sorbitol + + + + + +
stachyose + + + + + +
sucrose + + + + + +
D-tagatose - - - - - -
D-trehalose + + + + + +
turanose + + + + - -
xylitol - - - - - -
D-xylose + + + + + +
acetic acid / + / + + /
a-hydroxybutyric - / / - + -
acid
B-hydroxybuwric + + + + + +
acid
v-hydroxybutyric - - .. - - -
acid
p-hydroxrphenyl + + + + + +
acetic acid
a-keto glutaric - - - - - -
acid
a-keto valeric - - - - - -
acid
lactamide .. - .. - / ..
D-lactic methyl + + + + + +

ester

73
Table 4-2 (cont'd).
Substrate DH42 acidiprop- freuden- jensenii shermanii thoenii

 

ionici reichii
L-lactic acid + + + + + +
methyl pyruvate + + + + + +
mono-methyl + / + + + +
succinate
propionic acid .. - - / + -
pyruvic acid + + + + + +
succiniamic acid .. .. .. .. / ..
succinic acid + + + + + +
N -acetyl L- + + + + + +
glutamic acid
alaniniamide + / - / - -
D-alanine + + + + + +
lralacinc + + + + + +
lralanyl-glycinc + + + + + +
Irasparagine + + + + + +
L-glutamic acid + + + + + +
glycyl-lrglutamic / / + / / /
acid
L—pyroglutamic .. - .. - - ..
acid
L-serine + + + + - + +
putrescine + - - - - -
2,3-butanediol - - - .. - ..

glycerol + + + ' + + +

 

74

Table 4-2 (cont'd).

 

 

Substrate DH42 “1:52p. fggfﬁn' jensenii shermanii thoenii
adenosine + + + + + +
thymidinc + + + + + +
uridinc + + + + - +
adenosine-5’- + / - / / /
monophosphate
thymidine-5'- - - - - - -
monophosphate
uridine-5'- / - - - - /
monophosphate
fructose-6 + + + + + +
phosphate
glucose-l- + + + + + +
phosphate
glucose-6- + + + + + +
phosphate
Dim-glycerol + + + + + +
phosphate

 

+ = organism able to metabolize compound
/ = organism metabolizes compound very weakly
- = organism does not metabolize compound

Table 4-3 Cellular membrane fatty acid composition of Propionibacterium

 

 

acidipropionici DH42
Fatty Acid ceI‘Irrlgrtci'thy 1353:1533:
acrds
13:0 ISO FAME 0.35 0.01
13:0 ANTEISO FAME 0.28 0.00
13:0 FAME 0.20 0.01
14:0 FAME 0.44 0.03
15:0 ISO FAME 38.28 0.07
15:0 ANTEISO FAME 22.68 0.02
15:0 FAME 12.15 0.25
15:0 ISO DMA 5.86 0.08
15:0 ANTEISO DMA 1.81 0.02
16:0 FAME 0.99 0.01
16:0 DMA 0.70 0.01
17:0 ISO FAME 2.68 0.14
17:0 ANTEISO FAME 0.55 0.01
17:0 FAME 1.65 0.07
un 17.103 17:0i DMA 0.69 0.01
17:0 ANTEISO FAME 0.32 0.01
17:0 DMA 0.81 0.01
18:1 C18 9 FAME ' 0.35 0.05

 

76

Absorbance (600 nm)

 

 

Them)

Figure 4—2. Growth characteristics of P. aczdlpropionici DH42 grown in either peptone-
yeast extract-lactate or peptone-yeast extract-glucose at pH 7 under
nitrogen.

 

77

 

 

Table 44. Effect of pH on speciﬁc growth rate of P. acidipropionici DH42 grown on
glucose or lactate under nitrogen
- Media pH Carbon source SEM P <

Glucose lactate
Speciﬁc growth rate (h")

4.9 .124""‘1 .025‘ .0056 .0001

5 .3 . 141° .087" .0001

5.7 .134dc .068be .0001

5.8 .114‘” .055" .0001

6.2 .112” .076“1 .0002

6.5 .132°“° .080“ .0001

6.7 .101‘ .080“ 0135

6.9 . 1 17“” .080"d .0002

7.4 .116‘be .085d .0008

7.8 .107‘ .085“ .0125

 

M Means within a column lacking a common superscript differ (P < .05)

78
Implications

A propionic acid-producing bacterium previously isolated from ensiled, high-
moisture corn had similar phenotypic characteristics as Propionibacterium acidipropionici
and was designated P. acidipropionici DH42. Propionibacterium acidipropionici DH42
had slight dissimilanties from the type strain of P. acidipropionici in the content of cellular
lipids and the ability to metabolize putrescine. Propionibacterium acidipropionici DH42
can grow aerobically and anaerobically on lactic acid or glucose. Further studies need to
be done to evaluate its competitiveness in the microbial ecosystem from which it was
isolated to help predict efﬁcacy as a bacterial inoculant for the preservation of high-

moisture corn.

CHAPTER 5

COIWPETITION OF PROPIONIC ACID-PRODUCING BACTERIA
IN A MIXED MICROBIAL COMMUNITY FROM

ENSILED, HIGH-MOISTURE CORN

Abstract

Substrate competition is a major factor inﬂuencing the diversity or pr0portional ‘
distribution of microbial communities. A study was conducted to evaluate competitiveness
of propionic acid-producing bacteria when introduced into an environment with a mixed
microbial community and substrates commonly found in ensiled, high-moisture corn. Eight
strains of propionic acid-producing bacteria (Propionibacterium acidipropionici, P.
freudenreiChii (2 strains), P. jensenii, P. pentosaceum, P. shermanii, P. thoenii and P.
acidipropionici DH42, an isolate from ensiled, high-moisture corn) were inoculated into
three media (peptone-yeast extract-lactate and two aqueous extracts from ensiled, high-
moisture corn samples) with and without the addition of the mixed epiphytic microbial
community. All strains were capable of metabolizing at least 80% of the lactic acid present
to propionic acid and acetic acid in all three media within 7 d. Propionibacterium
acidipropionici DH42 was the most competitive for lactic acid and substrates found ' in
extracts of ensiled, high-moisture corn when compared to the other strains. These results

79

80

suggest propionic acid-producing bacteria possess features that may enhance production
of propionic acid and decrease aerobic deterioration of ensiled crops. Therefore, these

microorganisms may prove beneﬁcial as bacterial inoculants to ensiled crops.

Introduction

Prop ionic acid-producing bacteria have been isolated from ensiled crops by
Woolford (1975c) and in Chapter 3. Selected strains of propionic acid-producing bacteria
previously isolated from cheese have been used in experimental silos (Flores-Galarza et
al. , 1992) with limited success. A problem addressed for the classical propionic acid-
producing bacteria, which might limit their use as silage bioinoculants, is their slow
growth rate (Glatz, 1992) and possible poor substrate afﬁnity (Russel et al. , 1979). The
objective of this study was to evaluate propionic acid production and growth rate of
propionic acid-producing bacteria, with emphasis on P. acidipropionici DH42, when
introduced into an environment similar to a mixed microbial community and substrates

commonly found in ensiled, high-moisture corn.

Experimental Procedm'es
Media and bacterial strains. Three media were used to evaluate growth rate,
propionic acid production and competitiveness of eight propionic acid-producing bacteria.
Peptone-ycast extract-lactate broth (pH 6.9, 1.1% lactic acid; Holdeman, 1977) and
ﬁltered (.2 pm) aqueous extracts of two ensiled, high-moisture corn samples (Extract 1:

pH 4.1, 0.11% lactic acid; Extract 2: pH 4.6, 0.02% lactic acid) were prepared in

81

Hungate tubes under N2 and sealed with serum stoppers. Propionibacterium
acidipropionici DH42 was isolated from ensiled, high-moisture corn in Chapter 3 and
characterized in Chapter 4. Propionibacterium acidipropionici 4965 and P. freudenreichii
6207 were purchased from American Type Culture Collection (ATCC, Rockville, MD).
Pmpionibacteriumpentosaceum P11, P. thoenii P15, P. jensenii P25, P. shermanii CDC
3094 and P. freudenreichii P72 were kind gifts from Chr. Hansen's Biosystems
(Milwaukee, WI). Cultures were inoculated in duplicate (.1 ml per 10 m1 media,
approximately 107 organisms) with and without the addition of .1 ml of a mixed microbial
community from ensiled, high-moisture corn. An aqueous homogenate of two different
samples of ensiled, high-moisture corn was prepared by homogenizing 25 g of the sample
in 225 ml sterile .9% NaCl for 5 min. The mixed microbial community was harvested by
cennifugation (15 min @ 400 x g) of 50 ml of the aqueous extracts of the ensiled, high-
moisture corn samples.

Acetic and propionic acid were quantiﬁed by ion exchange-exclusion HPLC
(BIORAD aminex HPX-87H, Richmond, CA). Mobile phase consisted of .005 N HZSO.
at a ﬂow rate of .9 ml/min. Column temperature was maintained at 65°C by an external
column heater (Waters Millipore, Milford, MA). Three ml of culture extracts after 150
h of growth was ﬁltered through .2 pm ion chromatography syringe ﬁlters (Gelman
Acrodisks, 25 mm, Ann Arbor, MI) into 3 ml HPLC sample vials (National Scientiﬁc,
Atlanta, GA). Filtered samples were stored at -20 °C until analysis. Twenty pl of the
ﬁltered samples was injected by an autoinjector (Waters WISP 712, Milford, MA) and

analytes were detected by refractive index (Waters 410 refractive index detector, Milford,

82

MA). Peak heights were quantiﬁed by a commercial HPLC software package (Turbochrom
3, PE Nelson, Cupertino, CA) and compared to standards of lactic, acetic and propionic
acid (Supelco, Bellefonte, PA).

Specific growth rate of the mixed microbial community was determined by
measuring 0D,,00 in a Spectronic 70 spectrophotometer (Bausch and lomb, Newark, NJ)
during incubation at 30° C for 150 h. The slope of the linear portion of the growth curve
was determined using regression procedures (PROC REG) of SAS (SAS, 1987) and
compared using the general linear models method of SAS (SAS , 1987). Speciﬁc growth
rate of the mixed microbial community is presented as change in OD“ (a unit-less
measurement) per h during the linear portion of growth.

The experiment was analyzed as a completely randomized design by analysis of
variance (Steele and Torrie, 1980) using the general linear models (GLM) procedure of
the Statistical Analysis System (SAS, 1987). Preplanned comparisons of individual
treatment means within strain of Propionibacterium spp. with or without the addition of

the epiphytic microbial population were done using the PDIFF procedure (SAS, 1985).

83

The AN OVA model for the fermentation study was as follows:

Individual response variable measured (e.g. , speciﬁc growth
rate, acetic acid and propionic acid).

Overall mean.

Effect of strain of Propionibacterium spp. (P.
acidipropionici DH42, P. acidipropionici 4965 , P.
freudenreichii 6207, P. freudenreichii P72, P. shermanii
CDC 3094, P. pentosaceurn P11, P. thoenii P15 and P.
jensenii P25).

Effect of epiphytic microbial population (added epiphytic
population or pure).

Interaction of strain and presence of epiphytic population.

Random residual effect (assumed normally distributed).

Results and Discussion

Acetic acid production by the propionic acid-producing bacteria in the ﬁrst extract

(Table 5-1, pH 4.1, .11% lactic acid) was similar for all strains tested. All of the lactic

acid was metabolized by the strains evaluated with 150 h. The amount of acetic acid

produced was greater than what could be expected based on the initial amount of lactic

acid indicating metabolism of other substrates not detected by HPLC analysis. The extract

84

of ensiled, high-moisture corn with pH 4.6 (Table 5-2) had similar values for acetic acid
production with or without addition of the epiphytic microbial population which was also
seen in Table 5-1. Propionibacterium shermanii CDC 3094 tended to have the highest
concentration of acetic acid production in the extract with the higher pH (Table 5 -2).
Propionibacterium acidipropionici DH42 had the highest concentration of acetic acid after
150 h when grown with the epiphytic microbial population in peptone-yeast extract-lactate
media (Table 5-3). Propionibacterium jensenii P25 did not grow in peptone-yeast extract-
lactate media. This was unanticipated since transfers were made from exponential growth
in the same media. With the exception of P. freudenreichii 6207 , all of the other propionic
acid-producing bacteria had greater production of acetic acid when grown with the
epiphytic microbial population extracted from ensiled, high-moisture corn. Interaction of
propionic acid-producing bacteria with lactic acid-producing bacteria has been shown to
increase growth of both species (Parker and Moon, 1982). This may explain the enhanced
production of acetic acid which could be attributed to either the increased growth of the
epiphytic bacteria and(or) the propionic acid-producing bacteria themselves. Amounts of
acetic acid production in peptone-yeast extract-lactate media corresponds with
concentrations of acetic acid obtained for P. acidipropionici DH42 when grown on this

media in previous experiments.

85

Table 5-1. Acetic acid production of selected propionic acid-producing bacteria grown
in an aqueous extract of ensiled, high-moisture corn (pH 4.1, . l 1 % lactic
acid) with (mixed culture) or without (pure culture) epiphytic
microorganisms found in ensiled, high-moisture corn

 

 

 

Culture
Strain
Mixed Pure SEM P <
Acetic acid
(% w/v)
P. acidipropionici DH42 .094 .087 .008 .5265
P. acidipropionici 4965 .096 .103 .5761
P. freudenreichii 6207 . 114 .107 .5789
P. shermanii CDC 3094 .099 .089 .4398
P. pentosaceum P11 .101 .080 .0933
P. thoenii P15 .084 .072 .3180
P. jensenii P25 .071 .114 .0024

P. freudenreichii P72 .104 .115 .3584

 

86

Table 5 -2. Acetic acid production of selected propionic acid-producing bacteria
grown in an aqueous extract of ensiled, high-moisture corn (pH 4.6,
.02 % lactic acid) with (mixed culture) or without (pure culture)
epiphytic microorganisms found in ensiled, high-moisture corn

 

 

 

Culture
Strain
Mixed Pure SEM P <
Acetic acid

(% w/v)
P. acidipropionici DH42 .022 .016 .002 .1146
P. acidipropionici 4965 .023 .018 .1656
P. freudenreichii 6207 .026 .017 .0161
P. shermanii CDC 3094 .031 .032 .8099
P. pentosaceum P11 .020 .027 .0718
P. thoenii P15 .024 .028 .3495
P. jensenii P25 .026 .025 .6019

P. freudenreichii P72 .021 .016 .1299

 

87

Table 5 -3. Acetic acid production of selected propionic acid-producing bacteria
grown in peptone-yeast extract-lactate media (pH 6.9, 1.1% lactic acid)
with (mixed culture) or without (pure culture) epiphytic microorganisms

found in ensiled, high-moisture corn

 

 

 

Culture
Strain Mixed Pure SEM P <
Acetic acid
(% w/v)
P. acidipropionici DH42 .502 .326 .023 .0001
P. acidipropionici 4965 .385 .328 .1004
P. freudenreichii 6207 .338 .292 .1764
P. shermanii CDC 3094 .373 .268 .0062
P. pentosaceum P11 .426 .289 .0009
P. thoenii P15 .363 .258 .0059
P. jensenii P25 ND ND
P. freudenreichii P72 .346 .278 .0574

 

88

Propionibacterium acidipropionici DH42 had the greatest production of propionic
acid in extract with .11% lactic acid and pH 4.1 (Table 5-4) with the epiphytic microbial
population. The low pH of the extract was well below the optimal pH (between 6.5 and
6. 9) determined for P. acidipropionici DH42 in Chapter 3. Propionibacterium jensenii
P25 was able to grow in this extract as indicated by the higher (P < .01) production of
propionic acid without the addition of the epiphytic microbial population (.075 vs
.048 %). All other cultures evaluated had similar values for production of propionic acid
with or without the addition of the epiphytic microbial population. Propionic acid
production was similar for all strains evaluated in the extract with .02 % lactic acid and
pH 4.6 (Table 5-5). As seen in Table 5-4, P. jensenii P25 was the only pr0pionic acid-
producing bacteria evaluated which had differences in the amount of pr0pionic acid
produced. However, the addition of the epiphytic bacterial population enhanced
production of propionic acid in this extract.

Propionibacterium acidipropionici DH42 had the greatest amount of propionic
acid produced when grown with the epiphytic microbial population in peptone-yeast
extract-lactate media (Table 5-6). As stated previously, P. jensenii P25 did not grow in
this peptone-yeast extract-lactate media. P. freudenreichii P72 had similar amount of
prop ionic acid produced with or without the addition of the epiphytic microbial
population. All of the other strains evaluated showed enhanced (P < . 10) production of

propionic acid when grown with the epiphytic microbial population.

89

Table 5-4. Propionic acid production of selected propionic acid-producing bacteria
grown in an aqueous extract of ensiled, high-moisture corn (pH 4.1, .11%
lactic acid) with (mixed culture) or without (pure culture) epiphytic
microorganisms found in ensiled, high-moisture corn

 

 

 

Culture
Strain -
Mixed Pure SEM P <
Propionic acid

(% w/v)
P. acidipropionici DH42 .073 .066 .007 .4384
P. acidipropionici 4965 .050 .041 .3396
P. freudenreichii 6207 .050 .045 .6384
P. shermanii CDC 3094 .062 .055 .4781
P. pentosaceum P11 .063 .052 .2898
P. thoenii P15 .055 .054 .9175
P. jensenii P25 .048 .075 .0103

P. freudenreichii P72 .052 .036 .1126

 

90

Table 5-5. Prop ionic acid production of selected propionic acid-producing bacteria
grown in an aqueous extract of ensiled, high-moisture corn (pH 4.6, .02%
lactic acid) with (mixed culture) or without (pure culture) epiphytic
microorganisms found in ensiled, high-moisture corn

 

 

 

Culture
Strain
Mixed Pure SEM P <
Propionic acid

(% w/v)
P. acidipropionici DH42 .018 .017 .002 .7939
P. acidipropionici 4965 .017 .012 .1461
P. freudenreichii 6207 .018 .019 .6599
P. shermanii CDC 3094 .022 .023 .7441
P. pentosaceum P11 .017 .017 _ .8446
P. thoenii P15 .016 .013 .3037
P. jensenii P25 .017 .009 .0161

P. freudenreichii P72 .018 .022 .1821

 

91

Table 5-6. Prop ionic acid production of selected propionic acid-producing bacteria
grown in peptone-yeast extract-lactate media (pH 6.9, 1.1% lactic acid)
with (mixed culture) or without (pure culture) epiphytic microorganisms
found in ensiled, high-moisture corn

 

 

 

Culture
Strain
Mixed Pure SEM P <
Propionic acid
(% w/v)
P. acidipropionici DH42 .688 .573 .019 .0008
P. acidipropionici 4965 .601 .519 .0093
P. freudenreichii 6207 .597 .497 .0022
P. shermanii CDC 3094 .589 .541 .0940
P. pentosaceam P11 .574 .526 .0946
P. thoenii P15 .546 .504 .1401
P. jensenii P25 ND ND

P. freudenreichii P72 .472 .487 .5652

 

92

Addition of the epiphytic microbial population with the propionic acid-producing
bacteria tended to increase speciﬁc growth rate of the total mixed microbial community
in extract 1 (Table 5-7). Propionibacterium pentosaceum P11 and P. jensenii P72 showed
the greatest increases in speciﬁc growth rate of the mixed microbial community when
grown in combination with the epiphytic microbial population. Speciﬁc growth rate of
the propionic acid-producing bacteria and the total mixed community in extract 2 (Table
5 -8) was lower than that seen for extract 1 (Table 5-7) with the exception of P. thoenii
P15 which showed a marked increase in speciﬁc growth rate when grown with the
epiphytic microbial population.

Addition of the epiphytic microbial population decreased (P < .01) speciﬁc
growth rate of the propionic acid-producing bacteria when grown in peptone-yeast extract-
lactate media (Table 5-9). The speciﬁc growth rate for P. acidipropionici DH42 was the
greatest for the strains evaluated when grown in pure culture. The speciﬁc growth rate
in peptone-yeast extract-lactate media in Chapter 3 was slightly higher than what was
obtained by P. acidipropionici DH42 in this study. When grown in combination with the
epiphytic microbial population, all strains had similar growth rates. As shown in Table

5-3 and 5-6, P. jensenii P25 did not grow in peptone-yeast extract-lactate media.

93

Table 5 -7 . Speciﬁc growth rate of selected propionic acid-producing bacteria grown
in an aqueous extract of ensiled, high-moisture corn (pH 4.1, .11 % lactic
acid) with (mixed culture) or without (pure culture) epiphytic
microorganisms found in ensiled, high-moisture corn

 

 

 

Culture
Strain
Mixed Pure SEM P <
Speciﬁc growth rate
(11")

P. acidipropionici DH42 .0167 .0144 .0011 .1548
P. acidipropionici 4965 .0162 .0134 .0959
P. freudenreichii 6207 .0164 .0148 .3180
P. shermanii CDC 3094 .0167 .0158 .5406
P. pentosaceum P11 .0173 .0156 .2845
P. thoenii P15 .0167 .0139 .0900
P. jensenii P25 .0175 .0147 .0922

P. freudenreichii P72 .0167 .0160 .6595

 

’94

 

 

 

Table 5-8. Speciﬁc growth rate of selected propionic acid-producing bacteria grown
in an aqueous extract of ensiled, high-moisture corn (pH 4.6, .02% lactic
acid) with (mixed culture) or without (pure culture) epiphytic
microorganisms found in ensiled, high-moisture corn

Culture
Strain
Mixed Pure SEM P <
Speciﬁc growth rate
(11")

P. acidipropionici DH42 .0107 .0122 .0008 .1898
P. acidipropionici 4965 .0126 .0123 .7952
P. freudenreichii 6207 .0108 .0100 .4494
P. shermanii CDC 3094 .0121 .0119 .8735
P. pentosaceum P11 .0123 .0113 .3359
P. thoenii P15 .0183 .0082 .0316
P. jensenii P25 .0118 .0115 .8376
P. freudenreichii P72 .0112 .0103 .4184

 

Table 5-9.

95

in ensiled, high-moisture corn

 

Speciﬁc growth rate of selected propionic acid-producing bacteria grown
in peptone-yeast extract-lactate media (pH 6.9, 1.1% lactic acid) with
(mixed culture) or without (pure culture) epiphytic microorganisms found

 

 

Culture
Strain
Mixed Pure SEM P <
Speciﬁc growth rate
(11“)

P. acidipropionici DH42 .0098 .0301 .0013 .0001
P. acidipropionici 4965 .0105 .0159 0122
P. freudenreichii 6207 .0099 .0234 .0001
P. shermanii CDC 3094 .0111 .0273 .0001
P. pentosaceum P11 .0111 .0215 .0001
P. thoenii P15 .0111 .0252 .0001
P. jensenii P25 .0111 .0000 .0001
P. freudenreichii P72 .0120 .0238 .0001

 

96
Implications

Selected strains of propionic acid-producing bacteria can metabolize substrates
commonly found in ensiled, high-moisture corn. Propionibacterium acidipropionici DH42
(a propionic acid-producing bacteria isolated from ensiled, high-moisture corn) and P.
shermanii CDC 3094 were more capable of competing and producing elevated amounts
of propionic acid in peptone-yeast extract-lactate and aqueous extracts of ensiled, high-
moisture corn than the other snains of Propionibacterium spp. evaluated. With these
factors in consideration, it is reasonable to expect bioinoculants for ensiled crops to start
using selected strains of propionic acid-producing bacteria to improve the' quality of

ensiled feedstuffs.

CHAPTER 6

MANIPULATION OF THE FERMENTATION AND AEROBIC STABILITY OF
EN SILED, HIGH-MOISTURE CORN BY ADDITION OF

PROPIONBACIFRIW ACDH’ROPIONICI DH42

Abstract

Aerobic instability is the major cause of spoilage, nutrient loss and feed refusal of
ensiled, high-moisture com. A 2 x 6 factorial arrangement with two inoculant treatments:
Propionibacterium acidipropionici DH42 (previously isolated from ensiled, high-moisture
corn) added at 10‘ cfu-g'I-DM'1 and control (sterile water) over six periods of ensiling (0, 2,
7, 14, 28 and 42 d) was used to evaluate the effect of a propionic acid-producing bacterial
bioinoculant on fermentation characteristics and aerobic stability of ensiled, high-moisture
corn. Addition of P. acidipropionici DH42, increased propionic acid content (.35 vs .03
g/ 100 g DM, P < .01) after 42 d ensilement when compared to non-treated ensiled, high-
moisture corn. In addition, corn treated with P. acidipropionici DH42 had a higher acetic
acid concentration (.63 vs .29 g/100 g DM, P < .01), lower pH (4.17 vs 4.32, P < .01),
lower numbers of yeasts and molds (10"2 vs 10"° cfu-g“- M", P < .05) and higher dry
matter recovery (98.7 vs 96.8%, P < .05) after 42 d ensilement when compared to non-
treated ensiled, high-moisture corn. Temperature increased (P < .01) to 30 °C in the control

97

98

corn after three days of aerobic exposure, whereas the temperature of P. acidipropionici
DH42 treated corn remained unchanged from ambient temperature (21° C). After ﬁve days
of aerobic exposure, increases (P < .01) in pH (6.6 vs 4.4), numbers of yeasts and molds
(107'8 vs 10"2 cﬁr-g“-DM"), numbers of Acetobacter (107'9 vs 10"2 cfll-g'10DM“) and
numbers of total aerobic bacteria (10“ vs 10“:0 cfu-g"-DM") were observed in control
indicating poorer aerobic stability compared to treated, high-moisture corn. Propionic acid
concentration was higher (.61 vs 0 g/ 100 g DM, P < .01) after ﬁve days of aerobic exposure
in high-moisture corn ensiled with P. acidipropionici DH42. In addition, organic matter
recovery of inoculated corn was greater (94.4 vs 93.5%, P < .01) after ﬁve days aerobic
exposure. Presence and persistence of P. acidipropionici DH42 in the treated, high-moisture
corn were veriﬁed using a non-radioactive nucleic acid probe discussed in Chapter 7.
Inoculation of high-moisture com with P. acidipropionici DH42 improves the ensiling

fermentation and reduces aerobic instability of ensiled, high-moisture corn.

Introduction
Aerobic deterioration of ensiled feeds is detrimental to the production of livestock
feeds for many reasons. Aerobic microorganisms such as yeasts, molds and certain bacteria
metabolize the highly-digestible fraction of the dry matter and can account for losses of dry
matter in excess of 30% (W oolford, 1990). In addition, the microbial activity associated with
aerobic deterioration can increase the temperature of the feed and generate feed that is less
palatable and reduced digestibility (Ruxton and Gibson, 1993). Propionic acid has been used

quite successfully in reducing losses associated with aerobic instability (Lindgren et al.,

99

1985). Flores-Galarza and coworkers (1985) added strains of propionic acid-producing
bacteria commonly found in cheese and decreased proliferation of yeasts and molds
associated with high-moisture corn. However, Lindgren and coworkers (1983) were
unsuccessful in manipulating the fermentation of grass silage by the addition of P.
shermanii. A strain of P. acidipropionici (DH42) was isolated from high-moisture corn and
characterized in Chapter 3. The objective of this study was to evaluate the effectiveness of
P. acicb‘propionici DH42 on the fermentation and aerobic stability of ensiled, high-moisture

corn.

Experimental Procedures

High-moisture com (74% DM) was obtained from the Mason elevator (Mason, MI)
and ensiled in laboratory silos (45.7 x 10.2 cm dia) constructed of PVC pipe and ﬁtted with
rubber caps as described by Wardynski (1991). The time delay from harvest until ensiling
was not determined since the corn came from a commercial elevator. The corn appeared
fresh and did not have any characteristics associated with pre-fermented com. A rubber
policeman (Fisher Scientiﬁc, Chicago, IL) was slit and inserted with copper tubing into the
top cap to allow for gas release during ensiling. Thirty silos were used in a 2 x 5 factorial
arrangement of treatments consisting of two treatments (control and P. acidipropionici
DH42) and six periods of ensilement (0, 2, 7, 14, 28 and 56 d) with three silos for each
treatment and period of ensilement combination.

Propionibacterium acidipropionici DH42, a propionic acid-producing bacteria

isolated using methods described in Chapter 3 and characterized in Chapter 4, was grown

100

aerobically in a shaker water bath (60 rpm, 30 °C) to late log phase in Actinomyces broth
(BBL, Cockeysville, MD). A 1.5 ml aliquot was applied to the corn using a 3 ml syringe
with a 22 gauge needle while hand mixing, to assure homogenous application. The culture
was applied to a ﬁnal concentration of 10‘ cfu-g"-DM". Inoculation rate was based on the
recommended amounts of lactic acid-producing bacteria to be added as silage inoculants
(Muck, 1988). Culture density of the inoculum was 109 Cfll / ml as determined by decimal
dilutions in purple base broth (BBL 11558, Cockeysville, MD) with 1% erythritol (Sigma,
St. Louis, MO). Bacteria in the genus Propionibacterium are unique in their ability to
ferment erythritol to propionic and acetic acids (Holdeman et al., 1977). Therefore, use of
erythritol as a fermentation substrate and bromocresol purple as an acidic pH indicator dye
allows for estimating numbers of pr0pionic acid-producing bacteria when used as a most
probable number dilution scheme.

Aqueous extracts of the ensiled, high-moisture corn were obtained by homogenizing
50 g of sample in 450 ml sterile .9% NaCl (pH 7.0) using a stomacher (T ekmar 3500,
Cincinnati, OH) for 5 min. The pH of the aqueous extract was measured within ﬁve
minutes of extraction at 23 °C using a combination glass electrode (AccupHast, Fisher
Scientiﬁc, Chicago, IL) attached to a digital pH meter (Fisher Scientiﬁc 825MP pH meter,
Fisher Scientiﬁc, Chicago, IL) that was standardized from pH 4 to 7 using commercial
buffers (Curtis Mattheason, Woodale, IL). Serial dilutions of the aqueous extract of corn
after addition of the inoculum in the same media supplemented with amphotericin-B
(Sigma), a mycostatic agent used in mammalian cell culture, allowed estimation of the

number of added P. acidipropionici DH42. Serial dilutions were incubated aerobically at 30

101

° C and the highest dilution that had growth and yellow color (indication of erythritol
fermentation) was used as an estimate of the number of propionic acid-producing bacteria.
An appropriate control consisted of 1.5 ml of sterile .9% NaCl. Lactic acid-producing
bacteria were enumerated on pour plates of LBS agar medium (Difco, Detroit). Yeasts and
Molds were enumerated on pour plates of Rose Bengal agar medium base supplemented with
chloramphenicol (Difco, Detroit, MI). The media described by Spoelstra (1988) was used
as pour plates to enumerate Acetobacrer spp. Total aerobic bacteria were enumerated on
pour plates of tryptic soy agar (Difco, Detroit, MI).

Dry matter and ash were determined according to AOAC (1984) by drying at 60°C
for dry matter and ignition at 550 °C for ash. Glucose, lactic acid, acetic acid, propionic acid
and ethanol were determined by ion exchange-exclusion HPLC (BIORAD aminex HPX-
87 H, Richmond, CA) following the general procedures of Canale et al. (1984). Mobile
phase consisted of .005 N H2S04 at a ﬂow rate of .9 ml/min. Column temperature was
maintaimd at 65°C by an external column heater (Waters Millipore, Milford, MA). 'lhree
ml of aqueous extracts of each sample of ensiled, high-moisture corn were ﬁltered through
.2 pm ion chromatography syringe ﬁlters (Gelman Acrodisks, 25 mm, Ann Arbor, MI)
into 3 ml HPLC sample vials (National Scientiﬁc, Atlanta, GA). Filtered samples were
stored at -20 °C until analysis. Twenty pl of the ﬁltered samples was injected by an
autoinjector (Waters WISP 712, Milford, MA) and analytes were detected by refractive
index (Waters 410 refractive index detector, Milford, MA). Peak heights were quantiﬁed
by a commercial HPLC software package (Turbochrom 3, PE Nelson, Cupertino, CA) and

compared to standards of lactic, acetic and propionic acid (Supelco, Bellefonte, PA).

102

Statistical analyses. The fermentation experiment was analyzed as a completely
randomized design by analysis of variance (Steele and Torrie, 1980) using the general linear
models (GLM) procedure of the Statistical Analysis System (SAS, 1985). Individual
treatment means were compared using the Bonferroni t-test (SAS, 1985). The ANOVA
model for the fermentation study was as follows:

Yin = [.1 ‘1' Ti + Dj +(T’D)ﬁ + eijk

Y5], = Individual response variable measured (e.g., pH, numbers of

lactic acid-producing bacteria, propionic acid).

p = Overall mean.

T, = Effect of treatment (Control or P. acidipropionici DH42)
Dj = Effect of period of ensilement (0, 2, 7 , 14, 28 or 42 d).
(T-D),,- = Interaction of treatment and period of ensilement .

em, = Random residual eﬁ'ect (assumed normally distributed).

Aerobic stability was analyzed as above but using the repeated measure option of the
general linear models procedure (SAS, 1987) since repeated measurements were made on
each silo opened after 42 d ensilement and not on individual silos used in the fermentation

experiment.

103

Results and Discussion

Fermentation characteristics. Both the control corn and corn treated with P.
acidipropionici DH42 showed the normal fermentation pattern observed in ensiled, high-
moisture corn except increased propionic acid in corn inoculated with P. acidipropionici
DH42. The pH (Table 6-1) showed the characteristic decrease (P < .0001) seen during
typical ensilement of high-moisture corn. However, addition of P. acidipropionici DH42
resulted in high-moisture corn with a lower pH (P < .05) after 14 d ensiling. The pH values
obtained in this study followed the same pattern observed in the study reported in Chapter
2 and the general values reported for the fermentation of similar corn by Wardynski (1991).

The hypothesis to be tested was that the addition of P. acidipropionici DH42 would
increase production of propionic acid during ensilement . A treatment by ensiling period
interaction (P < .0001) was observed for propionic acid concentration (Table 6-2). No
propionic acid was detected in aqueous extracts of control or treated high-moisture corn
during the ﬁrst 2 d of ensiling. Propionic acid concentration increased in P. acidipropionici
DH42 to values greater (P < .0001) than control corn after 14 d ensiling. Control corn had
measurable amounts (.03 g/ 100 g DM) of propionic acid after 42 d ensilement which has
also been observed in control and corn treated with a lactic acid-producing bacterial
inoculant (Wardynski et al., 1993). Propionic acid production was greatest after 14 d in
reconstituted, high-moisture corn treated with P. acidipropionici ATCC 4965 and P.
freudenreichii ATCC 6207 in the study reported in Chapter 2. This was not the case with the

addition of P. acidipropionici DH42 in which the maximum amount of propionic acid was

104

reached after 42 (1. Future studies may need to be conducted to evaluate the effect of longer
periods of ensilement on the production of propionic acid in ensiled, high-moisture corn.
Acetic acid (Table 6-3) increased (P < .01) during ensilement and was greatest (P
< .01) in corn treated with P. acidipropionici DH42 aﬁer 42 d ensilement . The higher
amount of acetic acid is to be expected in the treated corn because acetic acid is a co-product
of the fermentation of carbohydrates and lactic acid by P. acidipmpionici DH42. Another
fermentation acid which is highly correlated to decreases in silage pH is lactic acid. The
production of lactic acid (T able 6-4) increased (P < .01) through 14 d in both the control and
treated corn following the same general fermentation pattern observed in non-treated ensiled,
high-moisture corn. Glucose (Table 6-5) was rapidly fermented (period of ensiling P.<
.0001) within the ﬁrst 2 (1 post ensiling. Very little residual glucose remained in either the
control or treated corn after 42 d ensiling. Ethanol (Table 6-6) increased throughout the ﬁrst
14 (1 (period of ensiling P < .0001) of ensiling in both control and treated corn. No
differences (P > .1) were observed between treatments, possibly due to the large variation
in ethanol content within treatments and periods of ensiling. .
Numbers of lactic acid-producing bacteria (Table 6-7) increased (period of ensiling
P < .0001) at an exponential rate during the ﬁrst 7 d ensilement . The increase in numbers
of lactic acid-producing bacteria corresponds to the decreases in pH and increases in lactic
acid concentration. Addition of P. acidipropionici DH42 prior to ensilement resulted in
lower (P < .01) number of lactic acid-producing bacteria after 42 d ensilement . It can only

be surmised about what type of lactic acid-producing bacteria were present. Therefore,

105

conclusions of the effect of P. acidipropionici DH42 on homofermentative lactic acid-
producing bacteria cannot be made.

Yeast and mold numbers (Table 6-8) increased throughout the ﬁrst 14 d ensilement
(period of ensiling P < .0001) to maximum values of 105'2 for both control and treated corn.
Numbers of yeasts and molds are important because they have been implicated as having a
major role in the aerobic deterioration of ensiled, high-moisture corn (Rust and Yokoyama,
1992). The increase in numbers of yeasts and molds could help explain the observed increase
in ethanol, a major end-product of their metabolism. However, numbers of yeasts and molds
decreased after 14 d and tended (P < .1) to be lower after 28 d in the corn treated with P.
acidipropionici DH42. Addition of P. acidipropionici DH42 was effective in decreasing
number of yeasts and molds after 42 d ensiling.

Presence of coliform bacteria is generally an indicator of poor quality silage.
Numbers of coliforms (Table 6-9) were slightly higher at d 0 for control corn. However,
after 2 d ensilement both the treated and control corn had similar numbers of coliforms
indicating the d 0 values did not cause a noticeable effect on their numbers after 2 d ensiling.
Addition of P. acidipropionici DH42 decreased proliferation of coliforms as shown by lower
(P < .05) numbers after 14, 28 and 56 d ensilement.

Dry matter and organic matter recovery were measured after 42 d ensilement . Dry
matter recovery (Table 6-10) was improved (P < .05) from 96.8% for control corn to 98.7%
by addition of P. acidipropionici DH42. Organic matter recovery (Table 6-10) was

numerically higher (P < .25) in treated corn (99.6%) than in control corn (99.3%).

106

Table 6-1. Effect of Propionibacterium acidipropionici DH42 on the pH during the
ensilement of high-moisture corn

 

 

 

. Treatment
T1311; Control P. acidipropionici SEM P <
DH42
pH of aqueous extract

0 5.39 5.45 0.033 0.2714
2 5.24 5.28 0.4068
7 4.84 4.78 0.2433
14 4.57 4.46 0.0338
28 4.35 4.20 0.0042
42 4.32 4.17 0.0042

 

Table 6-2. Effect of Propionibacterium acidipropionici DH42 on the propionic acid
content during the ensilement of high-moisture corn

 

Treatment

 

 

T153; Control P. acidipmpionici SEM P <
DH42
g/ 100 g DM
0 0.00 0.00 0.013 1.0000
2 0.00 0.00 1.0000
7 0.00 0.02 ' 0.3059
14 0.00 0.12 0.0001
28 0.00 0.11 0.0001

42 0.03 0.35 0.0001

107

Table 6-3. Eﬂ‘ect of Propionibacterium acicb'propionici DH42 on the acetic acid content
during the ensilement of high-moisture corn '

 

 

 

. Treatment
T152; Control P. acidipropionici SEM P <
DH42
g/ 100 g DM

0 0.00 0.00 0.079 10000
2 I 0.18 0.26 0.4910
7 0.18 0.32 0.2450
14 0.29 0.34 0.6249
28 0'14 0-15 0.9290
42 0.29 0.63 0.0051

 

Table 6-4. Eﬁ‘ect of Propionibacterium acidipropionici DH42 on the lactic acid content
during the ensilement of high-moisture corn

 

 

 

. Treatment
T1311; Control P. acidipropionici SEM P <
DH42
g/ 100 g DM
0 0.00 0.00 .146 1.0000
2 0.46 0.52 0.7593
7 0.68 0.74 ' 0.8008
14 1.26 0.89 0.0871
28 0.64 0.43 0.3163

42 0.72 0.75 0.8832

108

Table 6-5. Effect of Propionibacterium acidipropionici DH42 on the glucose content
during the ensilement of high-moisture corn '

 

 

 

_ Treatment
“(1:11; Control P. acidipropionici SEM P <
DH42
g/ 100 g DM

0 0.34 0.29 0.027 0.1914
2 0.01 0.02 0.9106
7 0.01 0.00 0.7043
14 0.01 0.01 0.9724
28 0.02 0.03 0.8090
42 0.00 0.04 0.2898

 

Table 6-6. Effect of Propionibacterium acidipropionici DH42 on the ethanol content
during the ensilement of high-moisture corn

 

 

 

. Treatment
T18; Control P. acidipropionici SEM P <
DH42

. g/lOOgDM
0 0.00 0.00 0.289 1.0000
2 0.85 0.84 0.9871
7 1.44 1.63 0.6581
14 2.21 1.60 . 0.1517
28 0.96 0.72 0.5478

42 1.67 1.13 0.1969

 

109

Table 6-7. Eﬁ‘ect of Propionibacterium acidipropionici DH42 on the number of lactic
acid-producing bacteria during the ensilement of high-moisture corn

 

 

 

. Treatment
T13; Control P. acidipropionici SEM P <
DH42
. log cfu per g DM

0 3.39 ' 3.49 0.068 0.4034
2 6.39 6.35 0.6311
7 8.41 8.34 0.4939
14 8'57 3-47 0.3081
23 8-45 8.47 0.8366
42 8-46 8.19 0.0117

 

Table 6-8. Eﬁ‘ect of Propionibacterium acidipropionici DH42 on the number of yeasts
and molds during the ensilement of high-moisture corn

 

 

 

. Treatment
T18; Control P. acidipropionici SEM P <
DH42
log cfu per g DM
0 3.57 3.9 0.199 0.3503
2 5.28 5.42 0.6331
7 4.43 4.68 . . 0.3855
14 6.22 6.25 0.9163
28 5.34 4.80 0.0705

42 4-93 4.16 0.0126

 

110

Table 6-9. Effect of Propionibacterium acidipropionici DH42 on the number of
coliforms during the ensilement of high-moisture corn

 

 

 

. Treatment
T153; Control P. acidipropionici SEM P <
DH42
log cﬁi per g DM

0 6.9 6.63 .051 0.0048
2 7.06 7.02 0.5516
7 6.45 6.45 0.8903
14 6.38 5.98 0.0001
28 6.27 6.12 0.0529
42 8.05 7.51 0.0001

 

Table 6-10. Effect of Propionibacterium acidipropionici DH42 on the dry matter and
organic matter recovery high-moisture corn after 42 days ensilement

 

 

 

Treatment
Item Control P. acidipropionici SEM P <
DH42
% recovered
Dry matter 96.8 98.7 0.45 0.0403

Organic matter 99.3 99.6 0.17 0.2470

1 1 1

Aerobic stability. Since propionic acid is commonly used as a chemical aid to
prevent aerobic deterioration, then the addition of P. acidipropionici DH42 prior to
ensilement would be beneﬁcial through an increase propionic acid content of ensiled, high-
moisture com. This was the case during 42 d of fermentation (Table 6-2) and during the
aerobic exposure period of 5 (1 (Table 6-11). Propionic acid disappeared from the control
corn after 5 d aerobic exposure. To the contrary, corn treated with P. acidipropionici DH42
showed increased propionic acid concentration with the greatest amount after 5 d (.61 g/ 100
g DM). The amount of propionic acid that is recommended for the addition to corn with
70% DM is 1.1% for 6 months storage. Future research will be needed to decide what
concentration is necessary to be reached when propionic acid is being produced in the silo.

No trend was observed due to the variability of the amounts of acetic acid (Table 6-
12) across treatments and time of aerobic exposure. The concentration of acetic acid was
lower than that observed in'the study by Wardynski et al. (1993) where concentration as high
as 3 g/ 100 g DM were determined in high-moisture corn inoculated with lactic acid-
producing bacteria. Ethanol (Table 6-13) is usually associated with coliform or yeast
fermentations. There was a large variation in the amount of ethanol present in the ensiled,
high-moisture corn and the only effect was due to time of aerobic exposure (P < .01).
Ethanol concentration decreased during aerobic exposure.

An interaction (P < .01) of treatment and time of aerobic exposure was present for
lactic acid concentration (Table 6-14) in the aerobically exposed high-moisture corn. Greater
(P < .05) concentration of lactic acid was found in the control corn after 3 d exposure and

less (P < .01) after 5 d as compared to the corn treated with P. acidipropionici DH42. Loss

112

of lactic acid is generally associated with aerobic instability and the differences observed in
the P. acidzpropionici DH42 treated corn at 5 d were anticipated. However, the lower
amount of lactic acid at d 3 is puzzling. Glucose concentration (Table 6-15) was slightly
increased (P < .10) by addition of P. acidipropionici DH42 after 3 d of aerobic exposure.
Future studies should focus on changes in total carbohydrates present prior to and after
ensilement or aerobic exposure.

A classical response seen in high-moisture corn undergoing aerobic deterioration is
an increase in pH over time of aerobic exposure. The pH of control corn (Table 6-16)
increased (P < .0001) from 4.32 to 6.62 after 5 d of aerobic exposure. Corn treated with P.
acic¢ropionici DH42 had a stable pH (4.17) throughout the ﬁrst 3 d of aerobic exposure and
a slight increase to 4.38 alter 5 d of aerobic exposure. Addition of P. acidipropionici DH42
prior to ensilement was effective in reducing the change in pH normally observed when
ensiled feeds are exposed to oxygen.

Yeasts and molds have bear implicated in the aerobic deterioration of ensiled feeds,
especially ensiled, high-moisture corn (Rust and Yokoyama, 1992). Propionibacterium
acidpropionici DH42 addition to ensiled, high-moisture corn was effective in maintaining
lower (P < .0001) numbers of yeasts and molds (Table 6-17) in aerobically exposed ensiled,
high-moisture corn. Control corn had initial yeast and mold numbers of 10"93 cfu-g'I-DM'l
and this increased (P < .0001) dramatically to 107'83 cfu-g'I-DM'l after 5 d aerobic exposure.
Numbers of yeasts and molds were 100 times lower (P < .01) in corn treated with P.
acidipropionici DH42 after 3 and 5 days of aerobic exposure that was shown with a

treatment by period of aerobic exposure interaction (P < .0001). Besides yeasts and molds,

113

acetic acid bacteria (Acetobacter) have been implicated in the initiation of aerobic
deterioration (Spoelstra, 1990). Numbers of Acetobacter followed the same trends observed
with theproliferation of yeasts and molds in the control corn.

Numbers of total aerobic bacteria over 5 d of aerobic exposure (Table 6-19) were
decreased (P < .01) by addition of P. acidipropionici DH42. Whereas, the number of total
aerobic bacteria increased in the control corn. Total aerobic bacteria are also indicators of
the inherent stability of the ensiled feedstuffs. Therefore, the greater number of aerobic
bacteria present, the higher the chance the ensiled material will be unstable.

Temperature of aerobically exposed silage will increase rapidly in unstable silages.
The temperature of high-moisture corn ensiled with a lactic acid-producing bacterial
inoculant increased to 50 ’C within 5 d aerobic (exposure in a study by Wardynski et al.
(1993). The temperature of the control corn in this studyh(T able 6-20) increased to 30 °C
within 3 d of aerobic exposure. Even though this indicated poor stability, it was not as
unstable as the corn treated with a lactic acid-producing bacteria in the study by Wardynski
et al. (1993).

Overall stoichiometry cannot be directly evaluated using the techniques presented
in this study. All substrates and end-products must be accounted for in order to precisely
estimate the overall balance of carbon. Dry matter (Table 6-21) and organic matter
recoveries (Table 6-22) were reduced (P < .05) when ensiled, high-moisture corn was
exposed to oxygen. In other words, dry matter and organic matter are being metabolized to
carbon dioxide, water and heat which is decreasing the amount recovered. Addition of P.

acidipropionici DH42 did not change the amount of dry matter loss after 5 d of aerobic

114

exposure (Table 6-21). However, organic matter recovery was increased by addition of P.
acidipropionici DH42 after 5 d aerobic exposure indicating a slower rate of nutrient loss.

The aerobic stability of the control corn would have been considered adequate since
temperature of the aerobically exposed ensiled, high-moisture corn was stable over 48 h.
Temperature reached a peak (30.6 °C) at 3 d and then declined near ambient. Additionally,
the maximum temperature recorded for the treated corn was 5 °C less (25.4 °C vs 30.6 °C) I
than control corn. The control corn in the study by Wardynski et al. (1993) did not reach a
maximal temperature until 9 d of aerobic exposure. Addition of P. acidipropionici DH42
prior to ensilement delayed increases in temperature, pH, numbers of yeasts, molds, aerobic
bacteria and Acetobacter indicating a dramatic improvement in aerobic stability over control

corn that would have been considered normal ensiled, high-moisture corn.

115

Table 6-11. Effect of PrOpionibacterium acidipmpionici DH42 on the propionic acid
content during the aerobic exposure of ensiled, high-moisture corn

 

 

 

Aerobic Treatment

EXPOS‘HB P. acidipropionici SEM P <
(d) Control DH 42

g/100 g DM

0 0.03 0.35 0.034 0.0001
1 0.04 0.44 0.0001
3 0.07 0.28 ' 0.0008
5 0.00 0.61 0.0001

 

Table 6-12. Eﬁ‘ect of Propionibacteﬁwn acidipropionici DH42 on the acetic acid content
during the aerobic exposure of ensiled, high-moisture corn

 

 

 

Aerobic Treatment
EXPOSUR’ P. acidipropionici SEM P <
(d) Control DH 42
g/ 100 g DM
0 0.29 0.63 0.153 0.1332
1 0.49 0.65 0.4654
3 0.12 0.32 _ 0.3540

5 0.17 0.48 0.1743

 

116

Table 6-13. Effect of Propionibacterium acidipropionici DH42 on the ethanol content
during the aerobic exposure of ensiled, high-moisture corn

 

 

 

Aerobic Treatment
Exposure P. acidipropionici SEM P <
(d) Control DH 42
g/100 g DM
0 1.67 1.13 0.325 0.2624
1 1.44 1.42 0.9725
3 0.02 0.49 0.3298
5 0.05 0.30 0.5980

 

Table 6-14. Eﬂ‘ect of Propionibacteﬁwn acidipropionici DH42 on the lactic acid content
during the aerobic exposure of ensiled, high-moisture corn

 

 

 

Aerobic Treatment
Exposure P. acidipropionici SEM P <
(d) Control DH 42
g/100 g DM
0 0.72 0.75 0.215 0.9208
1 1.31 1.20 0.7317
3 1.38 0.58 ' 0.0179
5 0.50 1.65 0.0016

 

117

Table 6-15. Effect of Propionibacterium acidipropionici DH42 on the glucose content
during the aerobic exposure of ensiled, high-moisture corn

 

 

 

Aerobic Treatment

Exposure P. acidipropionici SEM P <
((1) Control DH 42

g/ 1 00 g DM

0 0.00 0.04 0.021 0.1924
1 0.05 0.08 0.3929
3 0.09 0.04 0.0671
5 0.04 0.10 0.1104

 

Table 6-1 6. Effect of Propionibacterium acidipropionici DH42 on the pH during the
aerobic exposure of ensiled, high-moisture corn

 

 

 

Aerobic Treatment
Exposure P. acidipropionici SEM P <
(d) Control DH 42
pH of aqueous extract
0 4.32 4.17 0.040 0.0184
1 4.37 4.22 0.0184
3 4.63 4.22 ' 0.0001

5 6.62 4.38 0.0001

 

118

Table 6-17 . Eﬁ‘ect of Propionibacterium acidipropionici DH42 on the number of yeasts
and molds during the aerobic exposure of ensiled, high-moisture corn

 

 

 

Aerobic Treatment
Exposure P. acidipropionici SEM P <
(d) Control DH 42
log cfu per g DM
0 4.93 4.17 .223 0.0272
1 5.23 4.00 0.0012
3 6.93 4.67 0.0001
5 7.83 5.17 0.0001

 

Table 6-18. Effect of Propionibacterium acidipropionici DH42 on the number of
Acetobacter during the aerobic exposure of ensiled, high-moisture corn

 

 

 

Aerobic Treatment
Exposure P. acidipropionici SEM P <
((1) Control DH 42
log cfu per g DM
0 5.00 4.00 .182 .0013
1 5.00 3.93 .0008
3 6.93 4.37 ' .0001

5 7.93 5.20 .0001

 

119

Table 6-19. Effect of Propionibacterium acidipropionici DH42 on the number of total
aerobic microorganisms during the aerobic exposure of ensiled, high-
moisture corn

 

 

 

Aerobic Treatment

Exposure P. acidipropionici SEM P <
(d) C°mr°1 DH42

log cfu per g DM

0 8.33 8.13 0.061 0.0346
1 8.00 8.00 1.0000
3 8.17 7.63 0.0001
5 8.57 8.03 0.0001

 

Table 6-20. Effect of Propionibacterium acidipropionici DH42 on temperature changes
during aerobic exposure of ensiled, high-moisture corn

 

 

 

Aerobic Treatment
Exposure P. acidipropionici SEM P <
((1) Control DH 42
°C
0 23.9 23.7 .88 .8958
1 21.7 ' 21.5 .8953
2 21.7 21.8 .9580
3 30.6 21.4 .0001
4 27.6 21.7 , .0001
5 23.7 25.4 .1862
6 21.9 21.1 .5289

 

Ambient temperature averaged 21.4

120

Table 6-21. Effect of Propionibacterium acidipropionici DH42 on dry matter recovery
during the aerobic exposure of ensiled, high-moisture corn

 

 

 

Aerobic Treatment
Exposure P. acidipropionici SEM P <
((1) Control DH 42
% recovered
1 100.5 99.1 0.40 0.0285
3 100.7 99.8 0.1449
5 99.7 99.5 0.7186

 

Table 6-22. Effect of Propionibacterium acidipropionici DH42 on organic matter
recovery during the aerobic exposure of ensiled, high-moisture corn

 

 

 

Aerobic Treatment
Exposure P. acidipropionici SEM P <
(d) Control DH 42
% recovered
1 98.7 99.1 0.21 0.2497
3 98.5 98.8 0.3765

5 93.5 94.4 0.0108

 

121

Care should be exercised in extrapolating the effect of P. acidipropionici DH42 in
laboratory silos on the fermentation of high-moisture corn. Future research is needed to
evaluate the optimal inoculation rate, moisture levels, growth media, environmental (i.e.,
temperature, relative humidity) and method of application. Other additives such as lactic
acid-producing bacteria inoculants and ammonia addition may affect the efﬁcacy of
propionic acid-producing bacteria when grown in high-moisture corn. In addition, the effect
of P. acidipropionici DH42 on the fermentation and aerobic stability of different ensiled
crops (i.e., alfalfa haylage, corn silage) cannot be directly estimated from this research since

it was performed using only high-moisture corn.

Implications

Addition of Propionibacterium acidipropionici DH42 as a bacterial inoculant to
high-moisture corn was successful in manipulating the ensiling fermentation and improving
aerobic stability. Propionic acid content was increased to .3 5g/ 100 g DM during 42 d
ensilement and to .61 g/ 100 g DM during ﬁve d of aerobic exposure after ensilement.
Aerobic stability was increased dramatically as evidenced by a more stabilized temperature
and pH, reduced numbers of yeasts, molds and Acetobacter in high-moisture corn
inoculated with P. acicgvropionici DH42 prior to ensilement. F arm-scale silo experiments
should be conducted to determine if P. acidzpropionici DH42 can have the same effect in

a less controlled environment.

CHAPTER 7

NON-RADIOACTIVE NUCLEIC ACID PROBES FOR PROPIONIC ACID-

PRODUCING BACTERIA IN ENSILED, HIGH-MOISTURE CORN

Abstract

Use of nucleic acid probes for assessment of ecological interactions in microbial
ecosystems is becoming popular with the advent of non-radioactive detection methods for
nucleic acids. The objective of this study was to develop a protocol for a non-radioactive
nucleic acid probe to monitor the persistence of a propionic acid-producing bacterial
inoculant. Total DNA extracted from a propionic acid-producing bacteria previously
isolated from ensiled, high-moisture corn, designated Propionibacterium acidipropionici
DH42, was randomly labeled by the incorporation of digoxigenin-labeled deoxyuridine-
triphosphate and used for detection of the organism previously added as a bioinoculant to
ensiled, high-moisture corn. Speciﬁcity of the probe was evaluated by slot-blotting
extracted DNA from exponential growth of nine Propionibacterium spp. and six
Lactobacillus spp. No cross-hybridization was observed with. DNA extracted from
Propionibacteriumjensenii, P. thoenii, Laaobacillw bulgaricus, L. cory‘tuus, L. delbrukii,
L. fmctovorans, L. helveticus, or L. sake. However, there was slight cross-hybridization
with DNA extracted from P. freudenreichii and P. shermanii. The probe was used

122

123

successfully in detecting the presence and persistence of P. acidipropionici DH42 when
added to high-moisture corn after 2, 7, 14, 28, and 42 d of ensiling. In addition, DNA
extracted from ensiled, high-moisture corn Heated with P. acidipropionici DH42 after
three d of aerobic exposure, hybridized to the probe. Nucleic acids extracted from the
mixed, microbial community associated with non-treated, high-moisture corn did not
hybridize to the probe on any of the samples taken from the ensiling fermentation or
aerobic exposure phases. These results infer that non-radioactive DNA prObes can be used
to monitor persistence of a propionic acid-producing bacterial inoculant in ensiled, high-
moisture corn and support the fermentation data presented in chapter 6, in which addition
of the propi0nic acid-producing bacteria increased propionic acid content after 14 days of
ensiling. Further optimization needs to be performed to increase the sensitivity of detection

for this probe.

Introduction

Nucleic acid probes are becoming popular for the evaluation of interactions of
different microorganisms in various ecosystems. Historically, procedures for detection of
nucleic acid sequences involved radioactive materials. Recently, emphasis has been placed
on the development and use of non-radioactive methods for detection of nucleic acids.
Digoxigenin is a steroid which can be cross-linked by a chemical spacer arm to
deoxyuridine-triphosphate (DIG-dUTP) and is the basis for a commercial nonradioactive
labelling and detection kit (Genius-1, Boehringer Mannheim, Indianapolis, IN). The use

of DIG-dUTP in a random primed DNA labeling method can lead to incorporation of

124

DIG-dUTP every 20 to 25 nucleotides in the newly synthesized DNA. The DIG-dUTP can
then be detected by immunologieal methods and visualized by colorimetric, ﬂuorescent or
luminescent techniques. The method of visualization is dependent on the substrate chosen
for the alkaline phosphatase which is conjugated to the anti-DIG antibody fragment.
Sensitivity of this method is comparable, if not better, to radioactive techniques and can
be done in a safer and quicker manner.

Previous efforts at development of selective media for propionic acid-producing
bacteria have been based on yeast extract, peptone and sodium lactate (V edamuthu and
Reinbold, 1967; Hettinga et al. , 1968; Peberdy and Freyers, 197 6). The selectivity of the
above media was based on the fermentation of lactate by propionic acid-producing bacteria
under anaerobic conditions. However, lactic acid-producing bacteria and other
microorganisms commonly found in silages can metabolize lactic acid. Reddy and
coworkers (1973) and Drinan and Cogan (1992) attempted to correct this problem by
addition of two antibioties, cloxacillin and kanamycin. This alleviated the growth of most
lactic acid-producing bacteria except Leuconostocs. The concentration of antibiotics
necessary to inhibit the growth of Leuconostocs was also inhibitory to the growth of
propionic acid-producing bacteria.

Hendrich and coworkers (1991) developed a radioactive nucleic acid probe to a
cryptic plasmid of a main of Bacillus pumilus that was being used as a bioinoculant for
hay. This was done to simplify monitoring the persistence of the inoculated strain in hay,

soil and plant material. Furthermore, the probe was used to relate efﬁcacy of the product

125

to organism numbers. Plasmid transfer between the inoculated strain of Bacillus pumilus
and epiphytic Bacillus pumilus was considered a disadvantage to this method.

Non-radioactive nucleic acid probes were developed to the total genomic DNA of
strains of lactic acid-producing bacteria involved in wine production by Lonvaud-Funel
and coworkers (1991). This method showed that non-radioactive, nucleic acid probes,
based on total genomic DNA, were species speciﬁc for most of the Lactobacillus spp.
involved in fermentation of wine. Generation of nucleic acid probes from the genomic
DNA of selected bacteria should allow for more precise measurement of their status in
diverse microbial ecosystems such as ensiled feedstuffs.

A propionic acid-producing bacterium capable of competing for substrates against
epiphytic microorganisms found in ensiled, high-moistm'e corn was isolated using methods
described in chapter 3 and characterized in chapter 4. This organism, designated
Propionibacterium acidipropionici DH42, was successfully used as a bioinoculant to
improve the fermentation and aerobic stability of ensiled, high-moisture corn in the
experiment described in chapter 6. The objective of this study was to develop a protocol
for use of a non-radioactive nucleic acid probe to detect the presence of P. acidipropionici

DH42 when used as a bioinoculant for ensiled, high-moisture corn.

Experimental Procedures
Bacterial strains. Propionibacterium acidipropionici DH42 was previously
isolated from ensiled, high-moisture corn and stored in 8 ml of 50% glycerol and .5X

(half strength) MRS (Difco, Detroit, MI) in 10 ml serum bottles (Bellco, Vineland, NJ)

126

at -20 ° C. Propionibacterium acidipropionici 4965 and P. freudenreichii 6207 were
purchased from American Type Culture Collection (ATCC, Rockville, MD).
Pmpionibacteﬁwnjemenii P25, P. "pentosaceum " P11 (strain of P. acidipropionici), P.
shermanii CDC 3094, P. shemtanii P92 and P. thoenii were kind gifts from Chr. Hansen's
Biosystems (Milwaukee, WI). Propionibacterium acidipropionici 22562, P. freudenreichii
1382, Lactabacillus bulgaricus 8144, L. confusus 27646, L. delbrukii 9649, L.
ﬁ'uctovorans 8288, L. helviticus 12046 and L. sake 15521 were obtained from Michigan
Biotechnology Institute (Lansing, MI). All cultures were grown in crimper sealed 16 x 125
mm Hungate tubes (Bellco 2047-00125) under ambient gas phase at 30 C in .5X MRS
(Difco, Detroit, MI) and stored as previously described.

Nucleic acid extraction. The freeze-thaw DNA extraction method described by
Tsai and Olson (1991) was modiﬁed by using a commercial DNA puriﬁcation system
(Wizard Minipreps DNA Puriﬁcation System, Promega, Madison, WI). One ml of mid-
exponential growth of P. acidipropionici DH42 was harvested by oentrifugation (16,000
x g for 2 min) in a sterile 1.5 ml microcentrifuge tube. Supernatant was carefully decanted
and the pellet was resuspended in 200 pl cell resuspension solution (50 mM Tris-HCL pH
7.5, 10 mM EDTA, 100 ug/ml RNase A; Wizard Minipreps DNA Puriﬁcation System,
Promega, Madison, WI). Three cycles of freezing in a -70 °C dry ice-ethanol bath and
thawing in a 65 °C water bath were used to disrupt the cell membrane and release DNA.
Two hundred pl of cell lysis solution (.2 M NaOH, 1% SDS; Wizard Miniprep DNA
Puriﬁcation system, Promega, Madison, WI) was added with complete mixing of the lysis

solution achieved by repeated inversion of the microcentrifuge tube. Cell membrane

127

disruption was facilitated by heating the lysis mixture in a 95 °C water bath for 10 min.
Potassium acetate (1.32 M, pH 4.8; Wizard Miniprep, Promega, Madison, WI) was added
to the lysis solution to neutralize the cellular lysis material. Cellular debris was removed
by centrifugation (16,000 x g for 5 min). A 500 pl aliquot of the supernatant was carefully
removed and transferred to a sterile 1.5 ml microcentrifuge tube. One ml of DNA
puriﬁcation resin (Wizard Minipreps DNA Puriﬁcation Resin, Promega, Madison, WI)
was added and mixed by inversion of the microcenn'ifuge tube. A DNA puriﬁcation
minicolumn (Wizard Miniprep, Promega, Madison, WI) was afﬁxed to a 3 ml disposable
syringe and the resin mixture was slowly passed through the column with pressure applied
by the syringe plunger. Two ml of column wash solution (200 mM NaCl, 20 mM Tris-
HCl pH 7.5, 5 mM EDTA in 50% ethanol; Wizard Miniprep, Promega, Madison, WI)
were passed through the column. The DNA puriﬁcation minicolumn was transferred to a
microcentrifuge tube and centrifuged (16,000 x g for 20 s) to remove residual column
wash solution and then transferred to a sterile microcentrifuge tube. Fifty [1.1 of autoclaved
water (heated to 80 ’C) was placed into the DNA puriﬁcation column and allowed to sit
for 60 sec. Genomic DNA was then harvested by centrifugation of the DNA puriﬁcation
column (16,000 x g for 20 s) to a ﬁnal volume of 50 al.

Synthesis of non-radioactive probes for genomic DNA of P. acidipropionici
DH42. Total DNA extracted from P. acidipropionici DH42 was randomly labeled with
digoxigenin (Genius-1 Digoxigenin-dUTP labeling and detection kit, non-radioactive Cat.
No. 1093 057, Boehringer Mannheim, Indianapolis, IN) using the general procedures

outlined by Lonvaud-Funel and coworkers (1991) with modiﬁcations to the puriﬁcation

128

of the labeled probe using a commercial DNA puriﬁcation kit (Wizard Miniprep,
Promega). A scaled up labeling reaction was performed as described in the Genius
(Boehringer Mannheim # 101 023, Indianapolis, IN) manual. Thirty-ﬁve pl of the
extracted DNAwas denatured by heating ina95 °C water bath for 10 minand then placed
in a -70 °C dry ice-ethanol bath for 30 s. Denatured DNA template was added to a 1.5 ml
microcentrifuge tube on ice. Ten pl of hexanucleotide mix (all possible combinations of
hexanucleotides to serve as primers for DNA strand synthesis, Boehringer Mannheim #
127 081, Indianapolis, IN), 10 pl dNTP labeling mix (dATP, dCTP, dGTP, dTTP and
DIG-dUTP, Boehringer Mannheim # 1277 065, Indianapolis, IN), 40 pl sterile water and
5 pl Klenow enzyme (DNA polymerase 1, labeling grade, Boehringer Mannheim # 1008
404, Indianapolis, IN) were added to the denatured DNA template. The reaction mixture
was incubated in a 37 °C water bath for 20 h. Synthesis of the nucleic acid probe was
terminated by addition of 10 pl EDTA (.2 M, pH 8.0). A 5 pl aliquot of a glycogen
solution (20 mg/ml, Boehringer Mannheim #901 393, Indianapolis, IN) was added to
facilitate DNA recovery. Nucleic acid puriﬁeation was done as previously described in the
extraction procedure using a commercial DNA puriﬁcation kit (Wizard Miniprep,
Promega, Madison, WI).

Yield of DIG-dUTP labeled genomic DNA was estimated by spotting 1 pl of DIG-
dUTP standard DNA (DNA labeling and detection kit, nonradioactive Genius-l ,
Boehringer Mannheim 1093 657, Indianapolis, IN) at concentrations of 1 ng/pl, 100
pg/pl, 10 pg/pl, 1 pg/pl and .1 pg/pl on a 5 cm square piece of nylon membrane (Zeta-

Probe GT membrane, Bio-Rad # 162-0190, Richmond, CA). One pl of the dilutions (1:5,

129

1:50, 1:500, 125,000 and 1:50,000) of DIG-dUTP labeled DNA were spotted on the same
membrane. The membrane was dried in a vacuum oven at 80 °C for 30 min to afﬁx the
DNA to the membrane. Blocking areas of the membrane where DNA was not attached was
accomplished by incubating the membrane in a heat sealed bag with 20 ml blocking
solution (2% Boehringer Mannheim # 1096 176 blocking reagent , Indianapolis, IN; in
100 mM Tris-HCl, 150 mM NaCl, pH 7.5) on a shaker (Gyrotory water bath shaker G-76,
New Brunswick Scientiﬁc, New Brunswick, NJ) at 150 rpm for 30 min. Blocking solution
was removed and 30 ml of a 1:5000 dilution (in blocking solution) of anti-digoxigenin Fab
fragment (anti-DIG antibody-alkaline phosphatase conjugate, 150 U/200 pl, Boehringer
Mannheim #1093 274, Indianapolis, IN) was added then incubated for 5 min at 23 °C
(ambient temperature). The membrane was washed twice in 25 ml buffer (Genius buffer
1, 100 mM Tris-HCl, 150 mM NaCl, pH 7.5) to remove unbound antibody. Alkaline
phosphatase was then activated by incubating the membrane in alkaline buffer (Genius
buffer 3, 100 mM Tris-HCl, 100 mM NaCl, 50 mM MgC12, pH 9.5) for 2 min. Forty-ﬁve
pl NBT solution (75 mg/ml nitroblue tetrazolium salt in 70% v/v dimethyl formamide,
Boehringer Mannheim #1383 213, Indianapolis, IN) and 35 pl X-phosphate (50 mg/ml 5-
bromo—4—chloro—3-indoyl phosphate, toluidinium salt in 100 % dimethyl formamide,
Boehringer Mannheim # 1383 221 , Indianapolis, IN) were added to alkaline buffer
(Genius buffer 3, 100 mM Tris-HCl, 100 mM NaCl, 50 mM MgC12, pH 9.5) and color
development allowed to proceed for 30 min in the dark. A blue-black precipitate,
formazine, started to form immediately and the color development was normally sufﬁcient

within 30 to 60 min. Amount of DIG-dUTP labeled probe was estimated by comparing the

130

intensity of the color to the standard DIG-dUTP DNA. Approximately 30 ng of genomic
DNA was labeled with DIG-dUTP using these conditions.

Speciﬁcity of probe. Total nucleic acids were extracted from 1 ml of exponential
growth (ODmo > = . 8) of cultures previously described, grown in .5X MRS , denatured
by boiling in 500 111 .4 M NaOH and slot blotted (Bio-Dot SF Microﬁltration Apparatus
170—6542, Bio-Rad laboratories, Richmond, CA) onto nylon membranes (Zeta Probe GT
membrane 162-0190, Bio—Rad laboratories, Richmond, CA). The membrane was washed
with 12.5 ml of 2X SSC (20X SSC = 3M NaCl, .3 M Na-citrate, pH 7.0) and allowed
to air dry. The dried membrane was placed into a heat scalable bag with 30 ml of the
standard hybridization solution (5X SSC, 1% blocking reagent, .1 % Na-lauroylsarcosine,
.02% SDS) and incubated at 65 °C for 2 h. An aliquot of the labeled probe (25 ng) was
denatured by heating at 95 °C for 10 min and placing in a -70 °C dry ice-ethanol bath for
30 s. Denatured probe was added to fresh hybridization solution and transferred into the
sealed bag with the membrane. Hybridization of the probe with homologous sequences was
allowed to proceed for 12 h at 65 C. Unbound probe was removed by washing the
membrane twice in 2X wash solution (2X SSC, .1% SDS) for 5 min and then twice in .5X
wash solution (.5X SSC, .1 % SDS) for 15 min at 65 C. Non-speciﬁc binding of the anti-
DIG antibody was reduced by incubating the membrane in 30 m1 Genius buffer 2 on a
shaker (150 rpm) for 60 min. Anti-DIG antibody was diluted 1:5000 in Genius buffer 2
and used to replace the blocking buffer by incubating for 30 min. The membrane was
transferred to a new heat sealable bag. Unbound anti-DIG antibody was removed by two

washes with 100 ml of Genius buffer 1 for 15 min per wash. Alkaline phosphatase

131

conjugated to the bound antibody was activated by incubation of the membrane in 10 ml
Genius buffer 3 (pH 9.5) for 10 min. Forty-ﬁve pl of NET solution (75 mg/ml nitroblue
tetrazolium salt in 70% v/v dimethyl formamide, Boehringer Mannheim #1383 213,
Indianapolis, IN) and 35 pl X-phosphate (50 mg/ml 5-bromo-4-chloro—3-indoyl phosphate,
toluidinium salt in 100% dimethyl formamide, Boehringer Mannheim # 1383 221,
Indiarnpolis, IN) were added to alkaline buffer (Genius buffer 3, 100 mM Tris-HCl, 100
mM NaCl, 50 mM MgC12, pH 9.5) and color development allowed to proceed for 30 min
in the dark. A blue-black precipitate, formazine, started to form immediately and the
color development was sufﬁcient for visualization after 30 to 60 min.

Detection of P. acidipropionici DH42 in ensiled, high-moisture corn. Aqueous
extracts obtained from ensiled, high-moisture corn samples from the study reported in
chapter 6 were stored at -20 °C until extraction of total microbial DNA. The DNA from
the microbial community associated with the ensiled, high-moisture corn samples was
extracted and slot blotted using methods previously described for the speciﬁcity

experiment. The color reaction was allowed to proceed for 60 min.

Results and Discussion
Speciﬁcity of probe. Using the scaled up procedure outlined in the Genius kit
user's manual (Boehringer Mannheim #101 023, , Indianapolis, IN), a total of 25 ng DIG-
dUTP labeled probe was obtained. Lonvaud-Funel and coworkers (1991) were able to
obtain 300 ng DIG-dUTP labeled probe from Lactobacillus spp. used in wine production.

Propionibacterium spp. are very recalcitrant to lysis by lysozyme as used in the procedure

132

of Lonvaude-Funel and coworkers (1991). Less than 25 pg DIG-dUTP labeled probe for
P. acidipropionici DH42 was obtained using their procedure. The freeze thaw method of
cellular lysis, as outlined by Tsai and Olson (1991), was effective in increasing yield of
DNA extracted from P. acidipropionici DH42 to 30 ng when used in combination with
heating the lysate for 10 min at 95 “C.

The slot blot of samples for the speciﬁcity experiment is shown in Figure 7-1. The
probe developed from P. acidipropionici DH42 hybridized with DNA extracted from a
subculture of itself as expected. Propionibacterium acidipropionici 22562 showed the
strongest hybridization of the cultures evaluated. Cross-hybridization was not surprising
since P. acidipropionici 225 62 is the type strain for P. acidipropionici at ATCC.
Propionibacterium "pentosaceum " P11 is a strain of P. acidipropionici and showed
hybridization with the probe but not to the same extent of P. acidipropionici 225 62. The
cultures of P. acidipropionici 4965 and P. freudenreichii 6207 did not grow well (ODwo
= .1) in this experiment. The pellets of cells harvested by centrifugation were hardly
visible. Therefore, the lack of hybridization of the probe to the DNA extracted from P.
acidipropionici 4965 and P. freudenreichii 6207 was expected. All other cultures grew
well (ODGOO > = .8 after 20 h).

Nucleic acids extracted from P. shermanii P92, P. shermanii CDC 3094 and P.
freudenreichii 1382 showed slight color development indicating hybridization less than the
hybridization of the probe to P. acidipropionici 225 62. Since the probe was based on total
genomic DNA, cross-hybridization of the two species to the probe deve10ped from P.

acidipropionici DH42 may not be such an anomaly. These three species, P.

133
acidipropionici, P. freudenreichii and P. shermanii are common isolates from dairy
products and share many physicochemical characteristics based on fermentation proﬁles
(Holdeman et al., 1977). Propionibacterium jensenii P25 and P. thoenii P15 did not
hybridize to the probe from P. acidipropionici DH42.

None of the Lactobacillus spp. used in this study showed any hybridization under
the conditions evaluated. A concern that may arise is the absence of L. plantarum,
Pediococcus acidilactici and Streptococcus faecium from the list of microorganisms
evaluated. Future studies should look at the possible cross-hybridization with these species
since they are commonly used as bioinoculants in ensiled feedstuffs. Lack of hybridization
to P. jensenii P25, P. thoenii P15 and the diverse array of the Lactobacillus spp. evaluated
may refute the possrbility of cross hybridization with the three species commonly used as
silage inoculants based on the diversity of the strains evaluated.

Future experiments should address improving the speciﬁcity of the probe toward
P. acidipropionici DH42. Manipulation of the hybridization and washing conditions
(higher stringency washes) may alleviate or control some of the species cross-
hybridization. Other methods, such as development of probes based on variable regions
of the 16S rRNA genes have been quite successful in being species and strain speciﬁc to
organism such as Streptococcus bavis (Odelson et al. , 1993). Future work is warranted in
development of strain speciﬁc probes for silage bioinoculants to address efﬁcacy and

persistenCe of the introduced strains.

134

22562 22562 ”DH42 .6207. 6207 .
13le P11, ' DH42 527648 27648

' 'P15’-.-".P15',. our}; 964979649”
925* 11525 ‘DH4221155211. 15521

P92 P92 DH42': 12046 , '12046 i .‘
CDC3094 DH42... 8288 8288
y 1382 .1382 DH42 8144. 8144

u..-‘ —-
, o

'4965 . 4965 DH42 snar- Blank

. 1*.

Figure 7-1 Template for the hybridization of a non-radioactive nucleic acid probe
developed from Propionibacterium acidipropionici DH42 to total nucleic
acids extracted from 1 ml of exponential growth of P. acidipropionici
(American Type Culture Collection; ATCC 22562, ATCC 4965), P.
pentosaceum (Chr. Hansen's lab; CH P11), P. thoenii (CH P15), P.
jensenii (CH P25), P. shermanii (CH P92, CH CDC 3094), P.
freudenreichii (ATCC 1382, ATCC 6207), Lactobacillus confusus (ATCC
27646), L. delbruckii (ATCC 9649), L. sake (ATCC 15521), L. helveticus
(ATCC 12046), L. fructovorans (ATCC 8288) and L. bulgaricus (ATCC
8144).

135

Detection of P. acidipropionici DH42 in ensiled, high-moisture corn. The slot
blot corresponding to DNA extracted from the ensiled, high-moisture corn treated with P.
acidipropionici DH42 or without (control) is shown in Figure 7 -2. Propionibacterium
acidipropionici DH42 was added to achieve a concentration of 106 cfu per g DM at the
time of ensiling. A low intensity band of color was observed after 60 min color reaction
in samples from 0 and 2 d post-ensilement in the P. acidipropionici DH42 treated corn
indieating hybridization. The slight hybridization was probably due to insufﬁcient amount
of probe in the hybridization solution or inadequate time for color development. The color
reaction was stopped after 60 min due to the concomitant development of background color
development which could have been attributed to insufﬁcient blocking. A culture of P.
acidipropionici DH42 was diluted and extracted to give equivalents of 103, 107, 10‘, 10’
and 10‘ cfu per g DM. The hybridization conditions and color development allowed for
detection of approximately 106 cfu P. acidipropionici DH42 per g DM of ensiled, high-
moisture corn.

Nucleic acids extracted from the microbial community associated with the control
corn did not show any hybridization at any of the ensiling times. This suggested that the
presence of P. acidipropionici DH42 or similar organisms was below detection levels
using the conditions presented. Samples of corn treated with P. acidipropionici DH42
revealed easily noticed color indicating hybridization of the probe after 7, 14, 28 and 42
days of ensiling. The results presented from the fermentation experiment in chapter 6
showed that propionic acid content did not increase until 7 days of ensiling. Hybridization

of the probe after 7 days of ensiling supports the observation that the production of

136

propionic acid in the high-moisture corn inoculated with P. acidipropionici DH42 was due
to the addition of the organism and not some other factor. If all the propionic acid was
present at the time of addition of the organism (d 0), one could possibly conclude that the
fermentation was inﬂuenced by the media added with the bioinoculant and not the
microorganism being added. Intensity of the bands in the slot blot increased from d 7 to
d 42 suggesting an increase in the number of P. acidipropionici DH42 participating in the
fermentation. The increase in growth of P. acidipropionici DH42 after 2 days of ensiling
suggests that the organism is not a major participant in the early phase of the ensiling
fermentation as would be the case with most Lactobacillus spp.

Propionic acid content increased after aerobic exposure in the corn treated with P.
acidipropionici DH42 as reported in chapter 6. Aqueous extracts of the ensiled, high-
moisture corn were obtained from samples exposed to air for 3 days after 42 days of
anaerobic storage. As seen during the fermentation, the probe developed from P.
acidipropionici DH42 did not hybridize to any of the control samples (Figure 7-2).
However, DNA extracted from corn treated with P. acidipropionici DH42 did hybridize
with the probe, indicating the presence and possible participation of P. acidipropionici
during the production of propionic acid seen during the aerobic exposure of the ensiled,

high-moisture corn.

Figure 7-2

137

Control day P. acidipropionici DH42

.14..
28
4'2
_ +3. .
DH42 8 7 6 ,5 4 blank
logcfu '— —" . '

all

. .‘1
Template for the hybridization of a nonradioactive probe developed from
P. acidipropionici DH42 to total microbial DNA extracted from aqueous
extracts of ensiled, high-moisture Corn inoculated with 10‘ cfu per g DM
of P. acidipropionici DH42 or without (control) at day 0 after 0, 2, 7, 14,

28 and 42 days of ensiling and 3 days of aerobic exposure and slot blotted
onto a nylon membrane.

138

Extracts of ensiled, high-moisture corn contain substances which clogged the nylon
membranes used for slot blotting. The commercial DNA preparation system (Wizard
Miniprep, Promega, Madison, WI) allowed for a rapid extraction and cleaning procedure
that yielded DNA free of other contaminants that interfered with blotting. Some of
thepropionic acid-producing bacteria evaluated had very noticeable colors.
Propionibacterium thoenii P15 has a very distinguishing reddish-yellow pigment which
interfered with prior steps when trying to visualize the color deve10pment associated with
the procedure. All DNA preps were clear and amendable to be blotting onto the nylon

membranes after the extraction and puriﬁcation methods described.

Implications

A non-radioactive DNA probe was developed for a strain of Propionibacterium
aa'dipropionici (DH42) previously isolated from ensiled, high-moisture corn. The probe
detected the presence and proliferation of Propionibacterium acidipropionici DH42 when
added as a bioinoculant to high-moisture corn. The probe did not hybridize to six strains
of Laaobacillus spp., Propionibacten'wn thoenii or Propionibacterium jensenii. However,
there was limited cross-hybridization with DNA from two related propionic acid-producing
bacteria, Propionibacterium freudenreichii and Propionibacterium shermanii. The use of
non-radioactive DNA probes will allow more precise evaluation of the efﬁcacy of silage

bioinoculants.

CHAPTER 8
SUMMARY AND CONCLUSIONS

Studies conducted to evaluate selected, propionic acid-producing bacteria as
bioinoculants for ensiled, high-moisture corn were successful in manipulating the ensiling
fermentation and reducing the extent of aerobic deterioration. Speciﬁc aims of this
research were to 1) select species of propionic acid-producing bacteria capable of
increasing propionic acid content of ensiled, high-moisture corn; 2) study and optimize
conditions that affect performance of the selected, propionic acid-producing bacteria; 3)
develop nucleic acid probes for the selected, propionic acid-producing bacteria to allow
for more precise evaluation of the ecological interactions; and 4) evaluate the use of the
selected, propionic acid-producing bacteria as bioinoculants for ensiled, high-moisture
corn.

The study reported in Chapter 2 showed that a mixture of Propionibacterium
acidipropionici ATCC 4965 and P. freudenreichii ATCC 6207 was able to increase
propionic acid content of ensiled, reconstituted, high-moisture corn to .26 g/ 100 g DM
after 56 (1. Increased propionic acid concentration was highly correlated to decreased
numbers of yeasts and molds. Because P. acidipropionici ATCC 4965 and P.

freudenreichii ATCC 6207 are isolates from dairy products, progressive enrichment

139

140

techniques were developed and used in Chapter 3 to isolate propionic acid-producing
bacteria from ensiled feeds. A rapid growing propionic acid-producing bacteria was
isolated from ensiled, high-moisture corn and characterized in Chapter 4. The isolated
propionic acid-producing bacterium had similar phenotypic characteristics to P.
acidipropionici and was designated P. acidipropionici DH42. Propionibacterium
acidipropionici DH42 was then evaluated for its competitiveness and growth on substrates
commonly found in ensiled, high-moisture corn in Chapter 5 . Eight strains of propionic
acid-producing bacteria were screened in the competition study. Propionibacterium
acidipropionici DH42 was the most competitive as seen by higher growth rates 'and greater
propionic acid production in two different aqueous extracts of samples of ensiled, high-
moisture corn and in peptone-yeast extract-lactate medium.

Manipulation of the fermentation and aerobic stability of high-moisture corn by P.
acidipropionici DH42 was evaluated in Chapter 6. Addition of P. acidipropionici DH42,
as a bacterial inoculant, increased propionic acid concentration to .35 g/ 100 g DM after
42 days of ensiling and .61 g/ 100 g DM after ﬁve days of aerobic exposure after ensiling.
Increased propionic acid was most likely responsible for decreased yeasts and molds during
the ensiling fermentation and aerobic exposure. Consequently, dry matter recovery and
aerobic stability were improved as evidenced by stable pH and temperature during aerobic
exposure. Total DNA extracted from P. acidipropionici DH42 was used to generate a non-
radioactive nucleic acid probe to monitor persistence of P. acidipropionici DH42 when
added as a silage bioinoculant by a protocol described in Chapter 7. The non-radioactive,

nucleic acid probe detected the presence of P. acidipropionici DH42 in exn'acts of ensiled,

141
high-moisture corn inoculated with P. acidipropionici DH42 prior to ensiling (Chapter 6) .
Nucleic acids extracted from the microbial community associated with the non-treated corn
did not hybridize with the probe. This conﬁrmed those changes in the fermentation and
aerobic stability of the treated corn were due to the addition of P. acidipropionici DH42.

Limitations of experimental results are inherent to all biological experimentation.
The isolation of P. acidipropionici DH42 was not meant to be the end to isolation of
propionic acid-producing bacteria from ensiled feedstuffs. Fmther replication of each study
would further validate results. It is quite possible that more competitive strains of
propionic acid-producing bacteria are present in diverse types of silages. Future
experiments need to address efﬁcacy of propionic acid-producing bacteria in farm scale
silos. In addition to manipulation of the ensiling fermentation, animal performance and
feed acceptability should be evaluated.

Previous experimental results obtained using laboratory scale silos have generally
correlated with responses observed in large scale silos. Exuapolation of results obtained
in the two small scale silos experiments after d 42 and d 56 may not occur after 100 or
more days of ensiling. This point should be veriﬁed by future experiments. Care should
also be exercised is trying to directly apply results from these studies to fermentation of
corn silage, alfalfa haylage or other silages treated with additives.

Positive experimental results with P. acidipropionici DH42 and use of non-
radioactive nucleic acid probes to monitor population levels will allow further study of
ecological interactions of microbial species in silage and rumen microbial ecosystems. The

largest daily source of microbial matter to ruminant animals being fed ensiled feedstuffs

142

is the microbial population associated with the ensiled feeds. Many efforts have been made
trying to develop direct-fed microbial feed-additives that may influence the rumen
microbial population. An interesting future study could involve following persistence or
establishment of P. acidipropionici DH42, or possible more competitive propionic acid-
producing bacteria, in the rumen of eattle being fed a diet consisting of a large proportion
of ensiled, high-moisture corn that was treated with P. acidipropionici DH42. Besides
possible establishment of P. acidipropionici DH42, future studies should look at its
influence on fwdlot problems such as ruminal lactic acidosis.

Enrichment and isolation conditions detailed in Chapter 3 were successful in the
isolation of propionic acid-producing bacteria from ensiled crops. However, in view of
recent ﬁndings on the use of erythritol for selective growth of propionic acid-producing
bacteria, the author recommends evaluation of different minimal media based on erythritol
and selected antibiotics for the isolation of new pr0pionic acid-producing bacteria from
ensiled feeds. Sequential enrichment schemes discussed in Chapter 3 will still prove
helpful if used in a controlled manner. Limiting the source of ensiled feeds to those that
contain high levels of propionic acid may be an alternative to decreasing the number of
enrichments to be made. However, this may also limit the likelihood of isolating a diverse
array of propionic acid-producing bacteria that may not be dominating the crop due to
unknown factors.

Use of the non-radioactive nucleic acid probe used in Chapter 7 in future silage
fermentation studies should prove quite useful. Depending on what aspects of the

fermentation need to be addressed, the nucleic acid probe can be used with few

143

modiﬁcations. Cross-hybridization with different strains of Propionibacterium spp. should
be addressed in future studies. Use of higher stringency washes for removal of unbound
nucleic acids or use of selected sequences of DNA may help address the cross-
hybridization of related species. In addition, a larger array of common silage
microorganisms should be evaluated for similar cross-hybridization.

The overall hypothesis for this research was that the addition of selected, propionic
acid-producing bacteria as bioinoculants to ensiled, high-moisture corn would increase
production of propionic acid druing the ensiling fermentation and, consequently, decrease
the extent of aerobic deterioration and instability by controlling the growth of the
microorganisms responsible for the aerobic deterioration process. Addition of P.
acidipropionici DH42 was capable of increasing propionic acid in ensiled, high-moisture
corn sufﬁciently to increase aerobic stability. In addition, the persistence of P.

acidipropionici DH42 was conﬁrmed with a non-radioactive nucleic acid probe.

APPENDICES

APPENDIX A

DIVERSITY OF THE MICROBIAL COMMUNITY ASSOCIATED WITH
AEROBICALLY-EXPOSED CORN SILAGE INOCULATED WITH

PROPIONIBACIERIUIW ACIDIPROPIONICI DH42

Abstract

The concentration of lactic acid in ensiled feedstuffs is generally sufﬁcient to
exclude the growth of detrimental microorganisms during anaerobic storage. Once the
ensiled biomass is exposed to air, lactic acid is quickly metabolized resulting in aerobic
deterioration. Propionic acid has been shown to be effective in inhibiting the growth of
microorganisms responsible for the aerobic deterioration of silages. An experiment was
conducted to evaluate the effects of a propionic acid-producing bacteria on the diversity
of the microbial community associated with aerobically exposed corn silage. Treatments
included sterile water (control), PAPB-24 (10‘S cfu-g‘I-DM‘1 after 24 h growth of
Propionibacterium acidipropionici DH42), PAPB-72 (106 cfu-g'1-DM‘1 after 72 h growth
of P. acidipropionici DH42) and C3 (1% propionic acid added) were applied to 50 g of
fresh chopped whole plant corn (28.6% DM) and stored in 75 ml test tubes with gas
release valves (10 tubes per treatment). Five tubes were opened on d-25 and d-170 for

each treatment. Aerobic bacteria, yeast and molds, lactic acid-utilizing bacteria, and lactic

144

145

acid-producing bacteria were enumerated at opening and on 7 and 14 d after opening.
During ensiling, weight loss was minimized (P < .01) during the initial 25 d by C3. All
treatments had similar (P > .10) weight loss and temperature changes when exposed to
air after 25 (1 storage and similar pH values (3.88 :1: .07). Yeasts and molds were
decreased (P < .01) in PAPB-72 (105'1 cfu/g) and C3 (103'8 cfu/g) when compared with
control (10” cfu/ g) after 25 d of storage. The diversity of the microbial community as
measured by, the proportion of substrates metabolized by the mixed population of the
BIOLOG-GN plate decreased (P < .05) in C3 and PAPB-72 (55 %) compared with control
(74 %) and the epiphytic population (79 %) after 25 d of ensiling. Addition of P.
acidipropionici DH42 and propionic acid decreased the diversity of substrates metabolized

by the microbial community after ensiling.

Introduction

Previous strategies for efﬁcient silage fermentation encouraged high levels of lactic
acid production and lower pH development (below 4.2) during the fermentation phase.
This type of fermentation generally produces silage which is stable under anaerobic storage
conditions. However, once the ensiled mass is exposed to oxygen, large amounts of
nutrients ean be metabolized or destroyed. Resulting heat and mold formation is generally
deﬁned as aerobic instability or poor bunk life. Propionic acid is a very effective
mycostatic (mold inhibiting) and bacteriostatic (bacteria inhibiting) agent which has been
shown to be beneﬁcial in decreasing the losses associated with aerobic instability of ensiled

feeds (V oelker et al., 1989). Propionic acid is more inhibitory than lactic acid to the

146
organisms generally considered to be initiators of silage spoilage. Objectives of this study
were to evaluate the effects of P. acidipropionici DH42 on the microbial diversity of corn
silage and the use of the BIOLOG microplate system for assessing changes in microbial

communities associated with ensiled feeds.

Experimental Procedures

Four treatments, control (sterile water), PAPB-24 (24 h growth of P.
acidipropionici DH42 grown in peptone-yeast extract-lactate) , PAPB-72 (72 h growth of
P. aa'tﬁpmpionici DH42) and C3 (1% prOpionic acid added) were applied to 50 g chopped
whole plant corn (28.6% DM). The material was ensiled in 75 ml test tubes with a gas
release valve. Five tubes per treatment were stored for 25 and 170 d. Microbial
enumeration was performed utilizing pour plates of LBS (Difco, Detroit, MI) for lactic
acid-producing bacteria, Rose Bengal agar supplemented with chloramphenicol (Difco,
Detroit, MI) for yeasts and molds, tryptic soy agar (Difco, Detroit, MI) for total aerobic
bacteria and peptone-yeast extract-lactate agar for lactic acid-utilizing microbes (Holdeman
et al. , 1977). Volatile fatty acids and lactic acid were quantiﬁed by ion exchange HPLC
(Bio-Rad aminex HPX-87H, Richmond, CA) using refractive index detection as described
in previous chapters (Canale et al. , 1984).

Community diversity was assessed by a BIOLOG-GN (96 well microplate, Biolog
Systems, Hayward, CA) using a procedure modiﬁed from Drasar and Roberts (1991) and
Garland and Mills (1991) with modiﬁcations. Ten g of a homogenous sample of silage

was removed from each experimental silo and homogenized in 90 ml of sterile .9% NaCl

147

in a Stomacher (Tekmar 3500, Cincinnati, OH) for 5 min. The homogenate was
centrifuged (5,000 x g for 5 min) to remove plant debris. Supernatant was removed and
the mixed microbial community was harvested by centrifugation (25,000 x g for 10 min).
The mixed microbial community was resuspended in 30 ml sterile . 9 % NaCl to an OD“,
of approximately .5. The suspension was vortexed and 150 pl was pipetted into each well
of a BIOLOG-GN microplate. Microplates were incubated for 24 h at 30 °C and color
reactions in each well were visually observed and recorded. BIOLOG-GN microplates
contain ninety-ﬁve different carbon substrates with an indicator dye and a blank control
in each well of a standard ninety-six well microplate. A metabolic 'fmgerprint" is
generated by the different substrates which are utilized by either a pure culture of bacteria

or a mixed microbial community

Results and Discussion

Weight loss of the corn silage (Table A-1) was slightly reduced by C3 and PAPB-
72 after 25 d ensiling. The variation in the small scale silos was such that no treatment
differences were observed after 170 d ensiling. In addition, the silos used in this
experiment did not provide good ensiling conditions after 60 days of ensiling as evidenced
by mold proliferation. Fermentation characteristics after 170 d ensiling (Table A—2)
indicated similar pH and concentrations of lactic acid and ethanol. Propionic acid was
greatest (P < .05) in C3 and slightly higher in PAPB-24 and PAPB-72. Acetic acid was
lowest in C3. Microbial characteristics after 25 d ensiling (Table A-3) indicated lower

numbers of yeasts and molds in C3 and PAPB-72 (P < .05). In addition there were

148
higher numbers of lactic acid-producing bacteria and lactic acid-utilizing bacteria in control
silages than the three treated silages. All microbial characteristics were similar after 170
d ensiling (Table A—4). The microbial communities were different as evaluated by the
BIOLOG-GN microplates after 25 d ensiling (Table A-5). As seen with silage weight loss,
variation in the small silos after 170 d ensiling was sufﬁciently great to negate possible

treatment effects.

Table A-1 Weight loss (% of DM) of corn silage inoculated with P.
acidipropionici DH42 after 24 h or 72 h growth or treated with 1%
propionic acid after 25 or 170 d ensilement

 

 

 

Period of Ensiling (d)

Treatment
25 170

Condo] 6.36“ 19.37
P. acidipropionici DH42 (24 h growth) 7.26‘ 21.29
P. acidipropionici DH42 (72 h growth) 6.33“ 19.27
Propionic acid (1 %) 4.55” 23.36
SEM 0.56 3.53
ab

 

Means within a column lacking a common superscript differ (P < .05).

149

Table A-2. Fermentation characteristics of corn silage inoculated with P.
acidipropionici DH42 after 24 or 72 h growth or treated with 1%

propionic acid after 170 d ensilement

 

 

 

Fermentation Treatment

Characterimc Control PAPB-24 PAPB-72 (:3 SEM
Lactic acid (% DM) 5.85 3.7 4.27 5.67 1.44
Acetic acid (% DM) .66‘ .53" .56" .33" 0.06
Propionic acid (% DM) .07‘ .23be .14‘” .95° 0.04
Ethanol (% DM) 1.29 1.25 1.66 0.27 0.47
pH 3.59 4.09 3.8 3.44 0.28

 

“’° Means within a row lacking a common superscript differ (P < .05).

Table A-3. Microbial characteristics of corn silage inoculated with P.
acidipropionici DH42 after 24 h or 72 h growth or treated with 1%

 

 

propionic acid after 25 d ensiling
name... 13°33: Yﬁgdgnd T036338 173$
bacterra bacteria
log cfu-g'I-DM'1

Control 6.87‘I 4.90‘ 6.03 6.56‘
PAPB-24 6.50" 4.58‘” 5.68 6.43b
PAPB-72 6.72‘ 4.31b 5.67 6.49"D
C3 618‘ 3.00c 5.55 . 6.34°
SEM 0.05 0.17 0.14 0.03

 

M Means within a column lacking a common superscript differ (P < .05).

150

Table A-4. Microbial characteristics of corn silage inoculated with P.
acidipropionici DH42 after 24 h or 72 h growth or treated with 1%
propionic acid after 170 d ensiling

 

 

Treatment 1306;121:111:- YES: (1:11 d 1103:0253? ic 121331;]??-
bacterla bacteria
log cfu-g'1-DM'1
Control 4.58 4.35 6.02 5.91
PAPB-24 4.82 3.35 5.25 5.41
PAPB-72 5.77 4.04 6.02 6.04
C3 5.87 3.87 5.51 5.86
SEM .51 .62 .51 .53

 

Table A—5. Percent of substrates in the BIOLOG-6N plate utilized by the mixed
microbial community of corn silage inoculated with a P.
acidipropionici DH42 or treated with propionic acid

 

 

Ensiling Period ((1)
Treatment 25 170

substrates used (%)
Control 77‘ 68
P. acidipropionici DH42 (24 h) 71‘ 52
P. acidipropionici DH42 (72 h) 55" 76
Propionic acid (1%) 55" 17
SEM ,5 24

 

“’ Means within a column lacking a common superscript differ (P < .05).

151
Summary .
The treatments PAPB-72 (72 h growth of P. acidipropionici DH42) and C3 (1%
propionic acid) decreased yeast and mold number in corn silage after 25 d of ensiling. In
addition, PAPB-72 and C3 decreased the diversity of the microbial community as

measured by the diversity of substrates metabolized by the total microbial community.

Implications
Pmpionibacteriwn acidipropionici DH42 is capable of decreasing the diversity of
the microbial community in corn silage ensiled in laboratory scale silos. In addition,

BIOLOG microplates allow rapid assessment of changes in silage microbial communities.

APPENDIX B

SUPPORTING SMALL AND LARGE SCALE STUDIES EVALUATING
PROPIONIBACTERIUM ACEIPROPIONI CI DH42 AS A BIOINOCULANT FOR

ENSILED HIGH-MOISTURE CORN

A pilot study was conducted using high-moisture corn from the Michigan State
University Dairy Research and Teaching Center. Two PVC lab-scale silos were ﬁlled for
the control and P. acidipropionici DH42 treatments with 1.5 kg corn (73 %DM) obtained
from the bunker silos on March 3, 1993. This pilot study was conducted to evaluate the
effect of P. acidipropionici DH42 on the fermentation of corn harvested and placed into
bunker silos in the winter (January). The corn was harvested late in the season because
of immaturity, low test weight and high-moisture content the previous fall. Analysis of
aqueous extracts was conducted as in previous chapters. High-moisture corn ensiled with
P. acidipropionici DH42 had greater residual glucose, less lactic acid, greater acetic acid,
ethanol and propionic acid that did the control non-heated corn (Table B—l).

A large scale study was performed using high-moisture corn (75 % DM) inoculated
with P. acidipropionici DH42. Two n'uck loads of high-moisture corn were purchased
from the Mason Elevator Company (Mason, MI) and delivered to the Michigan State
University Beef Cattle Research and Teaching Center. Propionibacterium acidipropionici

152

153

DH42 was applied to 35,900 kg corn and stored in silo 3, whereas water was added as a
control to 38,200 kg and stored in silo 4. This high-moisture corn will be used in a feeding
study to evaluate feed acceptability and animal performance. Initial results indicated less
surface losses during ensiling. However, variation existed in aerobic stability of the
ensiled, high-moisture corn. Cattle were consuming both control and treated corn at
similar amounts indicating acceptable feed quality. Problems in obtaining the corn,
freesing temperatures at during silo ﬁlling and partial ﬁlling of the silos possibly lead to
the variation within the silos. Future experiments should be performed using uniform corn

in numerous types of silos.

Table B-1. Fermentation characteristics of high-moisture corn ensiled with P.
acidipropionici DH42 (10‘ cfu-g"-DM‘1) for 180 d

 

 

 

Treatment

Fermentation

characteristic Control P. acidipropionici DH42
Glucose (% DM) .14 ' .19

Acetic acid (% DM) .26 .46

Ethanol (% DM) .10 .18

Lactic acid (% DM) .68 .43

Propionic acid (% DM) .02 .32

 

Values are averages of two replicate per treatment

APPENDIX C

DATA TABLES

154

Table C-l. Data used for analysis in Chapter 2

 

ACE

re
tn
"U .
8

DAY TRT

 

*
0 CON A 5.11 5.88 ND 5.50 .000 .037
o CON B 5.08 5.59 ND 5.15 .000 .170
0 PAPB A 4.76 8 84 ND 5.71 .000 .112
0 PAPB B 8.08 8.40 ND 5.69 .000 .146
0 LAPB A 8.08 8.40 ND 5.69 .000 .134
0 LAPB B 8.13 8.32 ND 5.69 .000 .155
0 PA+LA A 7.74 8.36 ND 5.51 .000 .037
0 PA+LA B' 7.87 8.62 ND 5.52 .000 .077
2 CON A 7.67 8.10 ND 4.94 .000 .238
2 CON B 8.21 8.55 ND 4.72 .000 .121
2 PAPB A 8 63 8.73 ND 4.92 .171 .277
2 PAPB B 8.44 8.48 ND 5.00 .080 .121
2 LAPB A 9.34 9.12 ND 3.86 .000 .102
2 LAPB B 9.34 9.16 ND 3.94 .000 .096
2 PA-t-LA A 9.60 9.48 ND 4.04 .046 .164
2 PA+LA B 9.36 9.29 ND 3.97 .056 .148
7 CON A 8.93 8.94 ND 4.85 .000 .145
7 CON B 9.07 8.93 ND 4.82 .000 .130
7 PAPB A 9.02 8.91 ND 4.34 .227 .154
7 PAPB B 9.02 9.02 ND 4.37 .225 .237
7 LAPB A 9.43 9.34 ND 3.86 .000 .112
7 LAPB B 9.33 9.25 ND 3.80 .000 .102
7 PA+LA A 9.21 9.17 ND 3.94 .053 .179
7 PA+LA B 9.33 9.22 ND 3.96 .047 .165

 

Table C-l.

(cont)

155

 

DAY TRT REP LAB LAU YM PH PRO ACB
l4 CON A 9.34 9.24 6.58 4.49 .000 .142
14 CON B 8.92 9.04 6.52 4.65 .000 .118
14 PAPB A 8.57 8.96 4.46 4.32 .238 .327
14 PAPB B 8.56 9.02 5.15 4.30 .191 .182
14 LAPB A 8.90 8.84 6.02 3.90 .000 .103
14 LAPB B 8.69 8.83 6.27 3.89 .000 .099
14 PA+LA A 8.60 8.89 6.65 3.99 .037 .152
14 PA+LA B 8.51 8.63 6.25 3.99 .053 .163
28 CON A 9.01 8.87 6.35 4.47 .000 .108
28 CON B 8.90 8.84 6.39 4.44 .000 .110
28 PAPB A 8.62 8.64 5.64 4.27 .184 .197
28 PAPB B 8.63 8.52 5.34 4.29 .104 .115
28 LAPB A 8.22 8.29 6.20 3.93 .000 .115
28 LAPB B 8.13 8.05 6.49 3.93 .000 .102
28 PA+LA A 8.30 8.15 6.05 3.92 .060 .187
28 PA+LA B 8.36 8.33 6.26 3.91 .037 .112
56 CON A 8.80 8.76 6.51 4.33 .000 .145
56 CON B 8.77 8.72 5.50 4.28 .000 .155
56 PAPB A 8.59 8.81 5.30 4.25 .294 .299
56 PAPB B 8.62 8.64 5.43 4.23 .229 .150
56 LAPB A 7.14 6.97 6.29 4.02 .000 .186
56 LAPB B 7.23 7.20 6.07 4.02 .000 .109
56 PA+LA A 7.32 7.68 5.96 3.97 .068 .207
56 PA+LA B 7.46 7.78 6.19 4.01 .065 .189

 

AbbreviationszDay: Ensiling period (d); TRT: Treatmm; CON = Control, LAPB = lactic acid-producing
bacteria, PAPB = Propionic acid-producing bacteria, PA+LA = PAPB+LAPB REP: replicate; LAB: lactic
acid-producing bacteria (log cfu-g‘oDM"), LAU: lactic acid-utilizing bacteria (log cfu-g‘aDM"); YM: yeasts
and molds (10g cfrr-g‘oDM"); PH = pH of atpeous extract; PRO = propionic acid (g/ 100 g DM); ACE =

aceticacid(gl100gDM)

156

TableC-2. AmlysisofvaﬁmceofheﬁecBofpropkmtacﬁ-producmgbaaeﬁawhhorwummhcﬁc
acid—producingbacteﬁaandpeﬁodofensilingonresuhsreportedmChapterz

 

 

Characteristic Source df &nn of Squares P

pH of aqueous extract Trt 3 2.46 .0001
Time 5 11.41 .0001
Trtﬂ‘ime 15 1.86 .0001
Error 24 .189

lactic acid-producing bacteria Trt 3 2.16 .0001
Time 5 39.60 .0001
Trtﬂ‘ime 15 25.72 .0001
Error 24 .39

Lactic acid-utilizing bacteria Trt 3 1.55 .0001
Time 5 10.31 .0001
Trt‘Time 15 16.71 .0001
Error 24 .41

Yeasts and molds Trt 3 4.83 .0001
Time 2 . 14 .4760
Trt*Time 6 .91 .1904
Error 12 1.03

Acetic acid Trt 3 .037 .0110
Time 5 .025 .1346
TrtVI‘ime 15 .031 .7021
Error 24 . .065

Propionic acid Trt 3 .211 .0001
Time 5 .033 .0001
Trt*Time 15 .061 .0001

Error 24 .01 1

 

157

Table C-3. Data used in the analysis presented in Chmter 6 fermentation study

 

 

T D R DM OM
R A E PH GLC LAC ACE PRO ETH LAB YM COL

r Y P REC REC
C 0 A 5.5 0.38 0.00 0.00 0.00 0.00 3.5 3.2 6.9 ND ND
C 0 B 5.3 0.44 0.00 0.00 0.00 0.00 3.3 3.9 6.9 ND ND
C 0 C 5.4 0.20 0.00 0.00 0.00 0.00 ND ND ND ND ND
D 0 A 5.3 0.19 0.00 0.00 0.00 0.00 3.5 4 6.8 ND ND
D 0 B 5.5 0.33 0.00 - 0.00 0.00 0.00 3.5 3.8 6.5 ND ND
D 0 C 5.5 0.35 0.00 0.00 0.00 0.00 ND ND ND ND ND
C 2 A 5.3 0.01 0.28 0.11 0.00 0.59 6.2 5.2 7.1 ND ND
C 2 B 5.2 0.02 0.55 0.24 0.00 1.13 6.3 5.2 7 ND ND
C 2 C 5.3 0.00 0.54 0.20 0.00 0.82 6.7 5.4 7 ND ND
D 2 A 5.2 0.05 0.49 0.18 0.00 0.92 6.2 5 6.9 ND ND
D 2 B 5.3 0.00 0.53 0.38 0.00 0.84 6.3 5.4 7.1 ND ND
D 2 C 5.4 0.00 0.54 0.23 0.00 0.77 6.6 5.8 7 ND ND
C 7 A 4.8 0.00 0.42 0.15 0.00 0.96 8.4 4.9 6.4 ND ND
C 7 B 4.8 0.00 0.72 0.20 0.00 1.63 8.5 4.6 6.4 ND ND
C 7 C 4.8 0.04 0.91 0.19 0.00 1.74 8.4 3.8 6.5 ND ND
D 7 A 4.8 0.00 0.70 0.26 0.03 1.49 8.4 4.7 6.4 ND ND
D 7 B 4.8 0.00 0.74 0.33 0.00 1.68 8.3 4.7 6.4 ND ND
D 7 C 4.8 0.00 0.77 0.35 0.03 1.70 8.3 4.7 6.6 ND ND
c 14 A 4.6 0.00 0.88 0.18 0.00 1.43 8.6 6.1 6.4 ND ND
C 14 B 4.5 0.00. 1.80 0.40 0.00 3.12 8.5 6.3 6.3 ND ND
C 14 C 4.5 0.02 1.09 0.28 0.00 2.07 8.6 6.3 6.4 ND ND
D 14 A 4.5 0.00 0.99 0.48 0.13 1.64 8.4 6.6 5.9 ND ND
D 14 B 4.5 0.01 1.12 0.29 0.14 2.08 8.5 5.8' 6 ND ND
D 14 c 4.5 0.01 0.56 0.25 0.08 1.09 8.5 6.4 6 ND ND

 

Table C-3.

(cont)

158

 

 

T D R DM OM
R A E PH GLC LAC ACE PRO ETH LAB YM COL

T Y B REC REC
C 28 A 4.3 0.02 0.68 0.14 0.00 1.17 8.4 5.2 6.3 ND ND

C 28 B 4.3 0.01 0.59 0.08 0.00 0.50 8.5 5.7 6.3 ND ND

C 28 C 4.3 0.02 0.66 0.20 0.00 1.23 8.4 5.1 6.3 ND ND

D 28 A 4.3 0.02 0.31 0.12 0.08 0.64 8.5 4.4 6.2 ND ND

D 28 B 4.2 0.04 0.59 0.19 0.14 0.87 8.5 4.6 6.1 ND ND

D1 28 C 4.2 0.02 0.40 0.14 0.10 0.63 8.4 5.4 6 ND ND

C 42 A 4.3 0.00 0.57 0.57 0.04 3.05 8.4 4.9 8.2 95.8 99.3
C 42 B 4.3 0.00 1.20 0.22 0.06 1.29 8.5 4.9 8.1 98.0 99.4
C 42 C 4.3 0.00 0.41 0.09 0.00 0.67 8.5 5 7.9 96.7 99.2
D 42 A 4.2 0.05 1.00 0.42 0.29 1.21 8.2 4.6 7.5 98.7 99.3
D 42 B 4.2 0.08 0.98 0.47 0.36 1.31 8.2 3.9 7.6 98.7 99.5
D 42 C 4.2 0.00 0.28 1.01 0.39 0.87 8.3 4 7.5 98.8 100.1
Abbreviations:

TRT : C = control ; D = P. aa'dipropionici DH42 DAY: Ensiling period ((1)
REP: Replicatepertreatmentbyensilingperiodcombination

PH: pH of aqueous extract
LAC: Lactic acid (g/100 g DM)
ETH: Ethanol (g/100 g DM)
LAB: lactic acid-producing bacteria (log cfu-g‘obM")
YM: Yeast and molds (log cfu-g"-DM")
DMREC: Dry matter recovery (g/100 g DM)

OMREC: Organic matter recovery (g/100 g DM)

ND: Not determined

GLC: Glucosecontent(g/100gDM)

ACE: Acetic acid (g/100 g DM)

COL: Colifomrs (log cfu-g"-DM")

 

159

 

Table C-4. Analysis of variance of the eﬁects of P. acidipropionici DH42 and period of ensiling on
results reported in Chapter 6
Characteristic Source df Sum of Squares P

pH of aqueous extract Trt 1 .034 .0040
Time 5 7.517 .0001
Trt*Time 5 .062 .0131
Error 24 .081

Glucose Trt 1 .00002 .9185
Time 5 .457 .0001
Trt*Time 5 .007 .6728
Error 24 .052

Lactic acid Trt 1 .047 .4011
Time 5 3.780 .0001
Trtﬂ‘ime 5 .235 .6035
Error 24 1.535

Acetic acid Trt 1 .095 .0326
Time 5 .730 .0002
Trt‘I‘ime 5 .120 .2967
Error 24 .444

Propionic acid Trt 1 .077 .0001
Time 5 .157 .0001
Trt*'1‘ime 5 .109 .0001
Error 24 .012

Ethanol Trt 1 .372 .2347
Time 5 13.613 .0001
Trt‘Time 5 .761 .6944
Error 24 6.003

 

160

 

Table C-4. (cont)
—
Characteristic Source df Sum of Squares P
Lactic acid-producing bacteria Trt 1 .040 .1014
Time 5 94.160 .0001
Trt*Time 5 .099 2490
Error 22 .303
Yeasts and molds Trt 1 .118 .3320
Time 5 18.789 .0001
Trt‘Time 5 1.428 .0711
Error 22 2.631
Coliforms Trt 1 .448 .0001
Time 5 1 1.21 1 .0001
Trt'Time 5 .336 .0001
Error 22
Dry matter recovery Trt 1 5.549 .0403
Error 4 2.481
Organic matter recovery Trt 1 .150 .2470
Error 4 .328

 

161

Table 05. Data used in Chapter 6 aerobic stability study

 

PH TMP YM TSA EYA GLC LAC ACE PRO ETH DMR OMR

 

4.3 23.3 4.9 8.5 4.9 0.00 0.57 0.57 0.04 3.05
4.3 24.4 4.9 8.2 5.0 0.00 1.20 0.22 0.06 1.29
4.3 23.9 5.0 8.3 5.1 0.00 0.41 0.09 0.00 0.67

ND ND
ND ND
ND ND
4.2 22.8 4.6 8.2 4.0 0.05 1.00 0.42 0.29 1.21 ND ND
4.2 24.4 3.9 8.1 4.0 0.08 0.98 0.47 0.36 1.31 ND ND

ND ND

ooccco<>u

UUUOOOUUUOOOUUUOOOUUUOOO—IWH
nw>nw>ow>nw>nw>nw>nw>nw>wmw
is
N

4.2 21.7 4.0 8.1 4.0 0.00 0.28 1.01 0.39 0.87

4.4 21.7 5.5. 7.9 5.1 0.03 1.04 0.48 0.06 0.75 101.4 98.5
4.3 22.2 5.3 8.0 5.0 0.08 1.54 0.50 0.07 1.87 100.6 98.3
4.3 21.1 4.9 ' 8.1 4.9 0.05 1.35 0.48 0.00 1.69 99.6 99.3
4.3 21.1 4.7 8.1 3.9 0.09 1.12 0.60 0.38 1.22 99.8 99.1
20.6 3.5 8.0 4.0 0.05 0.94 0.53 0.37 1.12 99.3 98.9
4.2 22.8 3.8 7.9 3.9 0.10 1.53 0.83 0.56

s—s

21.7 ND
22.8

21.1
20.0 ND

§§§§§§§

ND
ND ND
ND 20.6 ND
ND ND
ND
ND

22.2 ND
4.8 31.1 7.2 8.3 7.1 0.16 0.85 0.10 0.06 0.00 101.2 98.3
4.7 35.0 6.4 8.1 7.1 0.05 1.37 0.00 0.08 0.00 98.7 98.4
4.5 25.6 7.2 8.1 6.6 0.08 1.92 0.26 0.07 0.06 102.7 98.4
4.2 21.1 4.4 7.6 4.0 0.06 1.03 0.56 0.37 0.85 100.0 98.4
4.3 18.9 5.0 7.7 4.8 0.04 0.69 0.41 0.28 0.61 100.4 98.9

wwmwwWNNNNNN

4.2 21.7 4.6 7.6 4.3 0.00 0.00 0.00 0.17 0.00 101.2 98.3

 

162

Table C-3. (cont)

PH TMP YM TSA EYA GLC LAC ACE PRO ETH DMR OMR

 

UUUOOOUUUOOOUUUOOOHWH
aaaaaauuuuuueeheee<>u

nw>nw>nw>nw>nw>nw>~emw
a
O\

ND 26.7 ND
ND 28.3 ND
ND 27.8 ND
ND 21.7 ND
ND 18.9 ND
ND 21.7 ND
6.7 23.9 7.7 8.7 7.9 0.00
24.4 7.9 8.6 8.0 0.07
6.5 22.8 7.9 8.4 7.9 0.06
4.5 26.7 5.5 8.1 5.8 0.04
4.3 20.0 4.6 8.0 4.4 0.13
4.3 26.7 5.4 8.0 5.4 0.11
ND 2.2 ND
ND 21.7 ND
ND 21.7 ND
ND 21.1 ND
ND 18.3 ND
ND ND

21.1

666666

 

Abbreviations

TRT :
REP:
TMP:
PH:
LAC:
E'TH:
')
DMR:
ND

C = control ; D = P. acidipropioniciDHtiz

DAY: Aerobic exposure period ((1)

Replicate per treatment by aerobic exposure period combination

Temperature of biomass

pH of aquoeus extract GLC:
lactic acid (g/100 g DM) ACE:
Ethanol (g/ 100 g DM) TSA:
YM: Yeast and molds (10g cfu-g‘oDM“)
dry matter recovery (g/ 100 g DM) OMR:

Not determined

Glucose content (g /100 g DM)

Acetic acid (g/ 100 g DM)

Total aerobic bacteria (log cfu-gloDM'
COL: Acetobacter (log cfu-g‘oDM")
organic matter recovery (g/100 g DM)

163

Table C-6. Analysis of variance of the eﬁects of P. acidipropionici DH42 and period of aerobic
exposure results reported in Chapter 6

 

Characteristic Source df Sum of Smares P

pH of aqueous extract Trt 1 3.263 .0001
Time 3 6.379 .0001
Trt*Time 3 4.546 .0001
Error 16 .078

Glrrcose Trt 1 .001 .3406
Time 3 .009 .1179
Trt't‘I‘ime 3 .011 .0741
Error 16 .022

Lactic acid Trt l .027 .6651
Time 3 .834 .1529
Trt‘Time 1 2.943 .0030
Error 16 2.210

Acetic acid - Trt 1 .388 .0317
Time 3 .423 .1531
Trt'Time 3 .032 .9265
Error 16 1.121

Propionic acid Trt ' 1 .862 .0001
Time 3 .063 .0064
Trt*Time 3 .132 .0002
Error 16 .056

Ethanol Trt l .010 .8647
Time 3 8.686 .0011
Trt*Time 3 .853 .4753

Error 16 5.210

 

164

 

Table C-6. (cont.)
m

Yeasts and molds Trt 1 18.027 .0001
Time 3 16.210 .0001
Trt*Time ' 3 3.510 .0019
Error 16 2.387

Total aerobic bacteria Trt 1 .602 .0001
Time 3 .645 .0001
Trt*Time 3 .312 .0009
Error 16 .180

Acetobacter wp. Trt 1 ' 13.368 0.0001
Time 3 20.350 0.0001
Trt*'I‘ime 3 3.945 0.0001
Error 16 1.587

Temperature Trt 1 44.847 .0002
Time 6 116.61 .0001
Trt*Time 6 139.71 .0001
Error 28 66.107

Dry matter recovery Trt 1 3.059 .0255
Time 2 1.268 .2967
Trt*Time 2 1.062 .3557
Error 12 i 5.646

Organic matter recovery Trt 1 1.171 .0117
Time 2 93.981 .0001
Trt*Time 2 .342 .3115
Error 4 1.594

 

LITERATURE CITED

Literature Cited

AOAC. 1984. Ofﬁcial Methods of Analysis (14th Ed.) Association of Ofﬁcial Analytical
Chemists. Washington, DC.

Armento, LE. 1992. Ruminant hepatic metabolism of volatile fatty acids, lactate and
pyruvate. J. Nutr. 1222838.

Ashbell, G., 2.6. Weinberg, H.Y. Azrielea, and B. Horeu. 1991. A simple system to
study aerobic determination of silages. Can. Agric. Engineer. 33:391.

Atlas, RM., and LC. Parks. 1993. Handbook of Microbiological Media. CRC Press.
Boca Raton. FL. '

Atwall, AS, and LC. Heslop. 1987. Effectiveness of propionic acid for preserving
alfalfa hay in large round bales. Can. J. Anim. Sci. 67:75.

Ayers, J.W., W.E. Sandine, and GR Weber. 1988. Preserving foods using metabolites
of Propionibacteria other than propionic acid. US. Patent 5,096,718.

Baer, A., and I. Ryba. 1991. Identiﬁcation of propionibacteria and streptococci by
immunoblotting. Milchwissenschaﬁ. 46:292.

Baertsche, SR, M.T. Yokoyama, and J.W. Hanover. 1986. Short rotation, hardwood tree

biomass as potential ruminant feed-chemical composition, nylon bag ruminal degradation
and ensilement of selected species. J. Anim. Sci. 63:2028.

Barnett, A]. 1954. Silage Fermentation. Academic Press, New York.

Barnett, A.J., and RE. Duncan. 1954. Volatile fatty acids in laboratory and ﬁeld silages.
J. Sci. Food Agric. 5:120. .

Baron, T., R Powell, and F. Gannon. 1990. A general method to generate DNA probes
from microorganisms. Bio/Techno]. 8:233. '

Baron, V.S., KR Stevenson, and JG. Buchanan-Smith. 1986. Proteolysis and

fermentation of grain-com ensiled at several moisture levels and under simulated storage
conditions. Can. J. Anim. Sci. 66:451.

165

166

Barry, T.N., M.E. Di Menna, PR Webb, and J.N. Parle. 1980. Some observations on
aerobic deterioration in untreated silages and in silages made with formaldehyde-
containing additives. J. Sci. Food Agric. 31:133.

Barry, T., R Powell, and F. Gannon. 1990. A general method to generate DNA probes
for microorganisms. Bio/Technology. 8:233.

Beaham, LR, and BA. Glatz. 1989. Protoplast production and regeneration in
Propionibacterium. J. Dairy Sci. 7222877.

Beck, T. 1978. The microbiology of silage fermentation. IN: ME. McCullough (Ed)
Fermentation of Silage - A Review. pp 61-115. National Feed Ingredients Association.
West Des Moines, IA

Blane, P., and G. Goma. 1987. Propionic acid fermentation: improvement of
performance by coupling continuous fermentation and ultraﬁltration. IN: Bioprocessing.
Engineer. Springer Verlage. Vol (2). 137.

Bodie, E.A., T.M. Andersson, N. Goodman, and RD. Schwartz. 1987a. Propionic acid
fermentation of ultra-high-temperature sterilized whey using mono- and mixed cultures.
Appl. Microbiol. Biotechnol. 25:434.

Bodie, E.A., N. Goodman, and RD. Schwartz. 1987b. Production of propionic acid by
mixed cultures of Propionibacterium shermanii and Lactobacillus casei in autoclaved-
sterilized whey. J. Industr. Microbial. 1:349.

Bolsen, K.K. 1991. Field Guide for Hay and Silage Management in North America. West
Des Moines, IA: National Feed Ingredients Association.

Bolsen, K..K, c. Lin, B.E. Brent, AM. Feyerherm, J.E. Urbana, and W.R Aimutis.
1992. Effect of silage additives on the microbial succession and fermentation process of
alfalfa and corn silages. J. Dairy Sci. 75:3066.

Bothast, RJ., G.H. Adams, E.E. Hatﬁeld, and EB. Lancaster. 1975. Preservation of
high-moisture com. A microbiological evaluation. J. Dairy Sci. 58:386.

Boyaval, P., and C. Corre. 1987. Continuous fermentation of sweet whey permeate for
propionic acid production in a CSTR with UF recycle. Biotechnol. Lett. 9:801.

Briesacher, S.L., T. May, K.N. Grigsby, M.S. Kerley, RV. Anthony, and J .A Patterson.
1992. Use of DNA probes to monitor nutritional effects on ruminal prokaryotes and
Fibrobacter succinogenes 885. J. Anim. Sci. 70:289.

167

Britt, D.G. 1973. Effect of organic acids and nan-pratein-nitragen an fungal growth,
nutritive value, fermentation, and refermentation of corn silage and high-moisture corn.
Ph. D. Dissertation, Michigan State University, East Lansing.

Britt, D.G., J .T. Huber, and AL. Rogers. 1975. Fungal growth and acid production
during fermentation and refermentation of organic acid treated corn silage. J. Dairy. Sci.
58:532.

Britz, T.J., and RC. Steyn. 1980. Comparative studies an propionic acid bacteria.
Phytophylactica. 12:89.

Britz, T.J., and K.H.J. Riedal. 1991. A numerical taxonomic study of Propionibacterium
strains from dairy sources. J. Appl. Bacterial. 71:407.

Brook, 1. 1991. Pathogenicity of Prepionibacterium acnes in mixed infections with
facultative bacteria. J. Med. Microbial. 34:249.

Brook, 1., and EH. Frazier. 1991. Infections caused by Propionibacterium species. Rev.
Infectious Dis. 13:819.

Buazzi, M.M., and EH. Marth. 1991. Mechanisms in the inhibition of Listeria
monocytogenes by potassium sarbate. Food. Microbial. 8:249.

Buazzi, M.M., and EH. Marth. 1992. Sites of action by propionate an Listeria
monocytogenes. Int. J. Food Microbial. 15:109.

Buckmaster, DR, C.A. Ratz, and RE. Muck. 1989. A comprehensive model of forage
changes in the silo. Trans. ASAE 32: l 143.

Bullerman, LB, and EC. Berry. 1965. Use of cheese whey for vitamin B12 production.
Appl. Microbial. 14:353.

Burmeister, HR, and RA. Hartman. 1966. Yeasts in ensiled high-moisture corn. Appl.
Micro. 14:35.

Burmeister, HR, P.A. Hartman, and RA. Saul. 1966. Microbiology of ensiled high-
maisture corn. Appl. Micro. 14:31.

Canale, A, M.E. Valente, and A. Ciatti. 1984. Determination of volatile carboxylic acids
(Cl-CSi) and lactic acid in aqueous acid extracts of silage by high performance liquid
chromatography. J. Sci. Food Agric. 35: 1 178.

Carranda, M.J., J.P. Crespa, and M.J. Maura. 1988. Production of propionic acid using a
xylase utilizing prapianibacterium. Humana Press. 295.

168

Cema, J ., Z. Pech, and H. Hrabava. 1975. Enrichment of fermented milk by vitamin B12
by means of precultivatian with Propionibacterium shermanii. Falia Microbialagica
20:7 1 .

Chatterj ee, D. 1990. Inhibition of fungal growth and infection in maize grains by spice.
ails. Letters Appl. Microbial. 11:148.

Chatterj ee, D., B.K. Chattopadhyay, and SK. Mukherje. 1990. Storage deterioration of
maize having pre-harvest infection with Aspergillusﬂavus. Letters Appl. Microbial.
1 121 1.

Champagne, C.P., C. Cate-Baillaergean, and J. Gaulet. 1989. Whey fermentation of
immobilized cells of Propionibacterium shermanii. J. Appl. Bacterial. 66:175.

Charfreitag, O., and E. Stackebrandt. 1989. Inter- and intragenic relationship of the
genus Propionibacterium as determined by 168 rRNA sequences. J. Gen. Microbial.
13522065.

Christensen, CM. 1949. Deterioration of stored grains by fungi. Botanical Rev. 23:108.

Christensen, C.M., and DR Gordon. 1948. The mold ﬂora of wheat and corn and its
relation to heating of moist grain. Cereal Chem. 25:40.

Chung, K., and CA. Murdock. 1991. Natural systems for preventing contamination and
growth of microorganisms in foods. Faad Struct. 10:361.

Clancy, M., P.J. Wagness, and ER Baumgardt. 1977. Effects of silage extract on
voluntary intake, rumen ﬂuid constituents and rumen motility. J. Dairy Sci. 60:580.

Courtin, M. G., and SF. Spaelstra. 1990. A simulation model of the microbiological and
chemical changes accompanying the initial stage of aerobic deterioration of silage. Grass
Forage Sci. 45:153.

Crawshaw, R, M. Tharne, and R Llewewlyn. 1980. The eﬂ‘ects of formic acid and
propionic acid an the aerobic deterioration of grass silage in laboratory units. J. Sci.
Food Agric. 312685.

Crow, V.L. 1986. Utilization of lactate isomers by Propionibacterium ﬁeudenreichii
subsp. shermanii: regulatory role of intracellular pyruvate. Appl. Environ. Microbial.
522352.

Crow, V.L. 1987. Citrate cycle intermediates in the metabolism of aspartate and lactate
by Propionoibacterium ﬁeudenreichii subsp. shermwrii. Appl. Environ. Microbial.
53 :2600.

169

Crow, V.L. 1987. Properties of alanine dehydrogenase and aspartase from
Propionibacteriumfreudenreichii subsp. shermanii. Appl. Environ. Microbial. 53: 1885.

Cullisan, AE. 1960. A simple, effective type of minature silo for conducting silage
investigations. J. Anim. Sci. 19:198.

Cummins, C.S., and J .L. Johnson. 1986. Propionibacterium. In: P.H. Sneath, N.S. Main,
and ME. Sharpe (Ed) Bergey's Manual of Systematic Bacteriology. (lst Ed.) Vol. 2.
Williams and Wilkins, Baltimore. 121346.

Cummins, C.S., and E.W. Mass. 1990. Fatty acid composition of Propionibacterium
propionicum (Arachnia propionica). Int. J. Syst. Bacterial. 40:307.

Dalmacia, I.F., D.Y. Fung, and J .H. Macneil. 1982. Microbiology af high-moisture corn
treated with a combination of "cold ﬂow" ammonia and propionic acid. Phillip.
Phytopathal. 182102.

Daescel, M.A, RE. Andersson, and HP. Fleming. 1987. Microbial ecology of
fermenting plant materials. FEMS Micro. Rev. 4621987.

Dawson, T.E., L.A. Miranda, M.T. Yokoyama, and SR Rust. 1991. Inoculation of
reconstituted high-moisture corn with propionic acid producing bacteria with or without
the addition of lactic acid bacteria. let Biennial Conference on Rumen Function.
Chicago, IL. Abst. 62.

Dawson, T.E., M.T. Yokoyama, and SR Rust. 1992. Enrichment, isolation and
characterization of propionic acid producing bacteria from ensiled feedstuffs J. Anim.
Sci. 70286 (Suppl. 1).

Dawson, T.E., SR Rust, and M.T. Yokoyama. 1993. Manipulation of silage
fermentation and aerobic instability by propionic acid producing bacteria. In: Silage
Production From Seed to Animal. Pracwdings from the National Silage Production
Conference. Syracuse, New York. pp .96-105.

Delwiche, EA. 1948. Mechanism of propionic acid formation by Propionibacterium
pentosaceum. J. Bacterial. 56:81 1.

Drasar, BS, and AK. Roberts. 1991. Methods of the study of anaerobic micraﬂara. In
P.N. Leveth (Ed), Anaerobic Microbiology A Practical Approach. Oxford University
Press, New York.

170

Driessen, AJ.M., and W.N. Kanings. 1990. Energetic problems of bacterial
fermentations: Extrusion of metabolic and products. The Bacteria, Vol. XII. Academic
Press. p 449.

Drinan, ED, and T.M. Cogan. 1992. Detection of propionic acid bacteria in cheese. J.
Dairy Res. 59265.

Eklund, T. 1985. Inhibition of microbial growth at different pH levels by benzoic and
propionic acids and esters of p-hydroxy benzoic acid. Int» J. Food Microbial. 2:159.

Ekstram, N. 1973. Preservation of moist feed grain by treatment with organic acids. Ann.
Technol. Agric. 222621.

Ellis, W.O., J .P. Smith, B.K. Simpson, and J .H. Oldham. 1991. Aﬂatoxins in food:
Occurence, biasythesis, eﬁ‘ects an organisms, detection and methods of control. Critical
Rev. Food Sci. Nutr. 30:403.

El-Shenawy, M.A, and EH. Marth. 1992. Behavior of Listeria monocytogenes in the
presence of sodium propionate together with food acids. J. Food Prat. 55:241.

Ely. L.O., N.J. Moon, and EM. Sudweeks. 1982. Chemical evaluation of Lactobacillus
addition to alfalfa, corn, sorghum and wheat forage at ensiling. J. Dairy Sci. 65: 1041.

Edwards, RA, and P. McDonald. 197 8. The chemistry of silage fermentation. In: M.E.
McCullough (Ed) Fermentation of Silage a Review. pp 27-60. National Feed Ingredients
Assoc. West Des Moines, IA

Faber, D.A, J .G. Linn, and DE. Otterby. 1989. Effect of bacterial inoculant on the
preservation of high-moisture shelled and car com. J. Dairy Sci. 7221234.

Fenton, MP. 1987 . An investigation into the sources of lactic acid bacteria in grass
silage. J. Appl. Bacterial. 62:181.

Fernandez, A, R Stroshine, and J. Tuite. 1985. Mold growth and carbon dioxide
production during storage of high-moisture corn. Cereal Chem. 62:137.

Fernandez-Garayzabal, J.F., M. Blanca, J .A. Vazquez-Boland, V. Briones, J .A Garcia,
C. Delgado, M. Domingo, J. Marco, and L. Dominguez. 1992. A direct plating method
for monitoring the contamination of Listeria monocytogenes in silage. J. Vet. Med.

39 2 S 13.

Fleming, HP. 1991. Mixed cultures in vegetable fermentations. In: Zeikus, J.G. and EA.
Johnson (Eds) Mixed Cultures in Biotechnology. McGraw-Hill Environmental
Biotechnology Series. 69.

171

Flores-Galarza, R, B. Glatz, C. Bern, and L. Van Fossen. 1985. Preservation of high-
maisture corn by microbial fermentation. J. Food Prat. 48:467. ’

Forsyth, J .G., J.B. Stone, and DM. Mowat. 1970. Propionic acid preserved corn in cattle
rations. J. Anim. Sci. 3121038.

Forsyth, J.G., D.M. Mowat, and J.B. Stone. 1972. Feeding value for beef and dairy cattle
of high-moisture corn preserved with propionic acid. Can. J. Anim. Sci. 52:73.

Foster, B.C., H.L. Trenholm, D.W. Friend, B.K. Thompson, and KB. Harten. 1987. The
effect of a propionate feed preservative in deoxynivalenal (vomitoxin) containing corn
diets fed to swine. Can. J. Anim. Sci. 6721159.

Fax, D.G. 197 6. Systems for storing and handling high-moisture corn. High Moisture
Grains Symp. p. 15. Oklahoma State Univ. Stillwater.

Freese, E., C. Shaw, and E. Galliers. 1973. Function of lipophilic acids as antimicrobial
food additives. Nature, London. 242321.

Furtner, K. 1974. Preservation of moist cereals with propionic acid and its effect on the
nutritional value of the feed. Nutr. Abstr. Rev. 44:281. (Abstr).

Garland, J .L., and AL. Mills. 1991. Classification and characterization of heterotrophic
microbial communities on the basis of patterns of community-level sale-carban-saurce
utilization. Appl. Envir. Microbiol. 5722351.

Gatze, G., and G. Gottschalk. 1991. Limitation of growth and lactic acid production in
batch and continuous cultures of Lactobacillus helviticus. Appl. Microbiol. Biotechnol.
34:446.

Gaudy, AF, and ET. Gaudy. 1980. Microbiology for Environmental Scientists and
Engineers. pp 540-543. McGraw-Hill, New York.

Gautier, M., A. Rouault, S. Lartal, Y. Lelour, and D. Patureau. 1992a. Characterization
of a phage infecting Propionibacteriumﬁeudenreichii. Lait. 72:431.

Gautier, M., N. Mouchel, A. Rouault, and P. Sanseau. 1992b. Determination of the
genome size of four Propionibacterium species by pulsed-ﬁeld gel electrophoresis. Lait.
72:421.

Gautier, M., S. Lartal, P.Boyaval, F. Girard, R Lemee, AF. de Carvalho, and C. Dupuis.
1993. Les bacteries propioniques laities. Lait. 732257 .

172

Ghash, J., and P. Hoggblom. 1985. Effects of sublethal concentrations of propionic and
butyric acid on growth of aﬂatoxin production by Aspergillus ﬂaws. Int. J. Food
Microbiol. 22323.

Gibson, T., AC. Stirling, RM. Keddie, and RF. Rosenberger. 1958. Bacteriological
changes in silage made at controlled temperatures. J. Gen. Microbiol. 19:112.

Giraud, E., B. Lelong, and M. Raimbault. 1991. Influence of pH and initial lactate
concentration on the growth of Lactobacillus plantarum. Appl. Microbiol. Biotechnol.
36:96.

Glatz, BA, and K1. Anderson. 1988. Isolation and characterization of mutants of
Propionibacterium strains. J. Dairy. Sci. 7121769.

Glatz, BA. 1992. The classical prapianibacteria: Their past, present, and future as
industrial organisms. ASM News. 58: 197.

Goering, H.K., and CH. Gordon. 1973. Chemical acids to preservation of high-moisture
feeds. J. Dairy Sci. 5621347.

Games, J ., and AS. Menawat. 1992. Estimation of fermentation parameters using partial
data. Biotechnol. Prog. 8:118.

Gordon, R]. 1991. Improving the feeding value of silage through biological control. In:
P.T. Lyons (Ed) Biotechnology in the Feed Industry. Alltech, Nicholasville, Kentucky. p
87.

Gould, G.W. 1991. Antimicrobial compounds. In: Goldberg, I. and R Williams (Eds)
Biotechnology and Food Ingredients. Van Nostrand Reinhold, New York. p 461.

Grinstead, DA, and SF. Barefoot. 1992. Jensenii G, a heat-stable bacteriocin produced
by Propionibacteriumjensenii P126. Appl. Environ. Microbiol. 582215.

Halarnkar, RR, and G.J. Blomquist. 1989. Comparative aspects of propioniate
metabolism. Comp. Biochem. Physiol. 92B2227.

Hansen, T.A., H.E. Nienhuis, and AI. Stams. 1990. A rod-shaped, gram-negative,
propionigenic bacterium with a wide substrate range and the ability to ﬁx molecular
nitrogen. Arch. Microbiol. 155242.

Hara, 8., and Y. Ohyama, 1979. Microorganisms concerning aerobic deterioration of
silages after opening silo, with reference to propionic acid treatment. Jap. J. Zootech. Sci.
50:288.

173

Hara, 8., M. Itoh, and Y. Ohyama. 1979. Aerobic deterioration of silages and changes in
temperature, gas metabolism, heat production and microflora. Jap. J. Zootech. Sci.
50:549.

Hazen, TC, and L. Jimenez. 1988. Enumeration and identification of bacteria from
environmental samples using nucleic acid probes. Micro. Sci. 5:340.

Henderson, AR, P. McDonald, and MK. Woolford. 1972. Chemical changes and losses
during the ensiling of wilted grass treated with formic acid. J. Sci. Food Agric. 2321079.

Henderson, AR, J .M. Ewart, and GM. Robertson. 1979. Studies on the aerobic
stability of commercial silage additives. J. Sci. Food Agric. 26:219.

Henderson, AR 1987. Silage making: biotechnology on the farm. Outlook on Agric.
16289.

Hendrich, C.A., B.K. Smiley, T.H. Shelley, and NJ. Tomes. 1991. Use of a plasmid
DNA probe to monitor populations of Bacillus pumilis inoculant strains in hay. Appl.
Environ. Microbiol. 57 2686.

Herting, DC, and EB. Drury. 1974. Antiﬁmgal activity of volatile fatty acids in grains.
Cereal Chem. 51274.

Herting, D., E. Drury, and S. Ames. 1974. Antifungal activity of volatile fatty acids on
grains. II. Effect of aqueous dilutions. Cereal Chem. 5123 82.

Hettinga, D.H., E.R Vedamuthu, and G.W. Reinbold. 1968. Pouch method for isolating
and enumerating propionibacteria. J. Dairy Sci. 5121707.

Hettinga, DH, and G.W. Reinbold. 1972a. The propionic acid bacteria - a review 1.
Growth. J. Milk Food Technol. 352295.

Hettinga, DH, and G.W. Reinbold. 1972b. The propionic acid bacteria - a review H.
Metabolism. J. Milk Food Technol. 35:358.

Hettinga, DH, and G.W. Reinbold. 1972c. The propionic acid bacteria - a review 111.
Miscellaneous metabolic activities. J. Milk Food Technol. 35:436.

Hill, HA, and J .E. Hill. 1986. The value of plasmid profiling in monitoring
Lactobacillus plantarum in silage fermentations. Curr. Microbial. 13291.

Hofherr, L.A, B.A Glatz, and EA. Hammond. 1983. Mutagenesis of strains of
propionibacteria to produce cold-sensitive mutants. J. Dairy Sci. 6622482.

174

Holben, W.E., J.K. Jannson, B.K. Chelm, and J .M. Tiedje. 1988. DNA probe method for
the detection of specific microorganisms in the soil bacteria community. Appl. Envir.
Microbiol. 54:703.

Holdeman, L., E. Cato, and W. Moore. 1977. Anaerobe Laboratory Manual. 4th Ed.
Virginia Polytechnic Institute, Blacksburg.

Holten, D., and H]. From. 1961. Puriﬁcation and properties of erythritol kinase from
Propionibacterium pentosaceum. J. Biol. Chem. 23622581.

Honing, H., and MK. Woolford. 1979. Changes in silage on exposure to air. In: C.
Thomas (Ed) Forage Conservation in the 80's, Brit. Grassl. Ocassional Symp. No. 11.,
Brighton, UK. p. 76.

Houwen, F.P., C. Dijkewa, AJ. Stams, and AJ. Zehnder. 1991. Prapionate metabolism
in anaerobic bacteria: determination of carboxylation reactions with l"‘C-NMR
spectroscapy. Biach. Biophys. Acta. 10562126.

Hsu, S.T., and ST. Yang. 1991. Propionic acid fermentation of lactose by
Propionibacterium acidipropionici 2 effects of pH. Biotech. Bioengineer. 26:217.

Huber, J .T., and M. Saejono. 1976. Organic acid treatment of high dry matter corn silage
fed to lactating dairy cows. J. Dairy Sci. 5922063.

Huber, TL. 1976. Physiological effects of acidosis on feedlot cattle. J. Anim. Sci.
43:902.

Ibragimova, S.I., N.M. Neronova, and LL. Rabotnova. 1969. Kinetics of growth
inhibition in Propionibacterium shermanii by hydrogen and hydroxyl ions. Mikrobiol.
382933.

Jackson, D.A., C.L. Johnson, and J .M. Forbes. 1991. The effect of compound

composition and silage characteristics on silage intake, feeding behaviour, production of
milk and live-weight change in lactating dairy cows. Anim. Prod. 52211.

Jain, D.K., RD. Tyagi, 1D. Kluepfrl, and TI. Agbebavi. 1991. Production of propionic
acid from whey ultrafiltrate by immobilized cells of Propionibacterium shermmtii in
batch process. Process Biochem. 262217 .

Jaster, E.H., and K]. Moore. 1992. Hay dessication and preservation with potassium
sorbate, potassium carbonate, sorbic acid and propionic acid. Anim. Feed Sci. Technol.
38: 175.

 

175

Johns, AT. 1951a. Isolation of a bacterium, producing propionic acid, from the rumen of
sheep. J. Gen. Microbiol. 52317.

Johns, AT. 1951b. The mechanism of propionic acid formation by Veillonela gazogenes.
J. Gen. Microbiol. 52326.

Johns, AT. 1951c. The mechanism of propionic acid formation by propionibacteria. J.
Gen. Microbial. 5:337.

Johns, AT. 1952. The mechanism of propionic acid formation by Clostridium
propionicum. J. Gen. Microbiol. 6:123.

Jones, GM. 1973. Performance of dairy cows fed propionic acid-treated high-moisture
shelled corn ration for complete lactation. J. Dairy Sci. 562207 .

Jones, G.M., E. Danefer, and J. 1. Elliot. 1970. Feeding value for dairy cattle and pigs of
high-moisture corn preserved with propionic acid. Can. J. Anim. Sci. 502483.

Jones, G.M., D.M. Mowatt, J .1. Elliot, and ET. Moran. 1974. Organic acid preservation
of high-moisture corn and other grains and the nutritional value: a review. Can. J.
Anim. Sci. 54:499.

Jones, G.M., RE. Larsen, and NM. Lanning. 1980. Prediction of silage digestibility and
intake by chemical analyses or in vitro fermentation techniques. J. Dairy Sci. 63 2579.

Jorgensen, N.A, H.J. Larsen, and GP. Barrington. 1980. Storing and utilizing high-
maisture corn preserved with organic acids. Wisconsin Res. Bull. R3057.

Jauany, J .P., G. Fontz, B. Lassalas, J. Dore, P. Gouet, and G. Berten. 1991. Effect of live
yeast cultures on feed degradation in the rumen as assessed by in vitro measurements.
21st Biennial Conference on Rumen Function. Chicago, IL. Abst. 6.

Kambar, C.S., G.W. Reinbold, and RV. Hussoy. 1952. A plating method for the
isolation and enumeration of propionibacteria. J. Dairy Sci. 35:915.

Katagiri, H., and Y. Ichikawa. 1953. Propionibacterium. Chem. Abstr. 4723923.

Keenan, T.W., and DD. Bills. 1968. Volatile compounds produced by
Propionibacterium shermanii. J. Dairy Sci. 51:797.

Klausner, RB., and CW. Donnelly. 1991. Environmental sources of Listeria and
Yersenia in Vermont dairy plants. J. Food Prat. 54:607.

176

Krebs, HA, and L.V. Eggleston. 1941. Biological synthesis of oxaloacetic acid from
pyruvic acid and carbon dioxide. H. The mechanism of carbon dioxide ﬁxation in
propionic acid bacteria. Biochem. J. 352676.

Kung, L. Jr., A Hessian, RS. Tung, and K. Maciarowski. 1991. Effects of
Propionibacterium shermanii on ruminal fermentations. let Biennial Conference on
Rumen Function. Chicago, IL. Abst. 56.

Langsrud, T., G.W. Reinbold, and E.G. Hammond. 1978. Free proline production by
strains of propionibacteria. J. Dairy Sci. 61:303.

Leaver, F.W., H.G. Wood, and R Stjeinholm. 1955. The fermentation of three carbon
substrates by Clostridium propionicum and Propionibacterium. J. Bacteriol. 7 02521.

Lee, I.H., AG. Fredrickson, and HM. Tsuchiya. 1974. Diauxic growth of
Propionibacterium shermanii. Appl. Microbiol. 28:831.

Lee, J.H., AG. Fredrickson, and HM. Tsuchiya. 1976. Dynamics of mixed cultures of
Lactobacillus plantarum and Propionibacterium shermanii. Biotechnol. Bioengineer.
18 2 5 13.

Lee, 8., MA Hanna, and LB. Bullerman. 1986. Carbon dioxide and alfatoxin
production in high-moisture corn treated with potassium sorbate. Cereal Chem. 63 :82.

Leibensperger, RY., and RE. Pitt. 1987. A model of clostridia] dominance in silage.
Grass Forage Sci. 422297 .

Leng, RA. 1970. Glucose synthesis in ruminants. Adv. Vet. Sci. Comp. Med. 14:209.

Lewis, V.P., and ST. Yang. 1992a. Propionic acid fermentation by Propionibacterium
acitivropionici: Effect of growth substrate. Appl. Microbiol. Biotechnol. 37:437.

Lewis, V.P., and ST. Yang. 1992b. Continuous propionic acid fermentation by
immobilized Propionibacterium acidipropionici in a novel packed-bed bioreactor.
Biotech. Bioeng. 40:465.

Lewis, V.P., and ST. Yang. 1992c. A novel extractive fermentation process for
propionic acid production from whey lactose. Biotechnol. Prog. 82104.

Lin, C., K.K. Bolsen, B.E. Brent, RA. Hart, J.T. Dickerson, AM. Feyerherm, and W.R
Aimutis. 1992. Epiphytic microﬂora an alfalfa and whole-plant corn. J. Dairy Sci.
7522484.

177

Lindgren, S., P. Lingvall, A Kaspersson, A de Kartzaw, and E. Rydberg. 1983. Effects
of inoculants, grain, and formic acid on silage fermentation. Swed. J. of Agric. Res.
13291.

Lindgren, S., K. Pettersson, A. Kaspersson, A Johnson, and P. Lingvall. 1985. Microbial
dynamics during aerobic deterioration of silages. J. Sci. Food Agric. 36:765.

Lindgren, S., A. Bromander, and K. Petterssan. 1988. Evaluation of silage additives
using scale-model silos. Swed. J. Agric. Res. 18241.

Lion, T., and 0A. Haas. 1990. Nonradioactive labeling of probe with digoxigenin by
polymerase chain reaction. Anal. Biochem. 1882335.

Little, W., R Manston, J.I.D. Wilkinson, and ME. Tarrant. 1991. Some factors related
to the voluntary intake of silage by individual dairy cows housed as a group during two
winter-feeding periods. Anim. Prod. 53219.

Liu, J .A, and NJ. Moon. 1982. Commensalistic interaction between Lactobacillus
acidophilus and Pr0pionibacterium shermanii. Appl. Environ. Microbiol. 442715.

London, J. 1976. The ecology and taxonomic‘status of the lactobacilli. Ann. Rev.
Microbiol. 302279.

Lonvaud-Funel, A, A. Jayeux, and O. Ledaux. 1991. Speciﬁc enumeration of lactic acid
bacteria in fermenting grape must and wine by colony hybridization with nan-isotopic
DNA probes. J. Appl. Bacterial. 712501.

Luchansky, J .B., P.M. Muriana, and TR Klaenhammer. 1988. Application of
electraporation for transfer of plasmidic DNA to Lactobacillus, Lactococcus,
Leuconostoc, Listeria, Pediococcus, Bacillus, Staphylococcus, Enterococcus and
Propionibacterium. Mol. Microbiol. 22637.

Lyon, W.A, and BA Glatz. 1991. Partial puriﬁcation and characterization of a
bacteriocin produced by Propionibacterium thoenii. Appl. Environ. Microbiol. 57 2701.

Lyon, W.J., and BA Glatz. 1993. Isolation and puriﬁcation of propionicin PLG-I, a
bacteriocin produced by a strain of Propionibacterium thoenii. Appl. Environ. Microbiol.
59283. -

Mader, T.L., RA. Britton, V.E. Krause, and DE. Panlaskie. 1985. Effect of additive on
alfalfa silage fermentation characteristics and feedlot performance of steers. J. Dairy Sci.
6821744.

178

Malik, A, G. Reinbold, and E. Vedamuthi. 1968. An evaluation of the taxonomy of
Propionibacterium. Can. J. Micro. 1421185.

Mann, E.M., and P. McDonald. 1976. The eﬁ‘ect of formalin and lower fatty acids on
silage fermentation. J. Sci. Food Agric. 27 2612.

Mantere-Alhonen, S. 1983. On the survival of Propionibacteriumﬁeudenreichii culture
during in vitro gastric digestion. Meijeretiet. Aikak. 41 :19.

Mantere-Alhoune, S. 1987 . A new type of sour milk with propionibacteria. Meij eretiet.
Aikak. 45:49.

Mareoux, V., Y. Beaulieu, C.P. Champagne, and J. Gaulet. 1992. Production of
Propionibacterium ﬁeudenreichii subsp. shermanii in whey-based media. J. F erm.
Bioengineer. 74:95.

Martin, R, C. Hoover, S. Grimme, C. Crogan, J. Holtke, and C. Kessler. 1990. A highly
sensitive, nonradioactive DNA labeling and detection system. BioTechniques. 9:762.

Martin, SA, and DJ. Nisbet. 1992. Effect of direct-fed microbials an rumen microbial
fermentation. J. Dairy Sci. 7521736.

Martin, S.L., J. Tuite, and MA Diekman. 1989. Inhibition radioimmunoassay for
Aspergillus repens compared with other indices of fungal growth in stored corn. Cereal
Chem. 66:139.

Mattey, M. 1992. The production of organic acids. Crit. Rev. Biotechnol. 12:87.

Mayne, CS. and R]. Gordon. 1986. The effect of harvesting system on the nutrient
losses during silage making. 1. Field losses. Grass Forage. Sci. 41 : 17.

McCullough, ME. 1978. Silage - some general considerations. In: M.E. McChrllough
(Ed) Fermentation of Silage - a Review. pp 3-28. National Feed Ingredients Assoc. West
Des Moines, IA.

McCullough, ME. 1984. Silage Management. National Feed Ingredients Assoc. West
Des Moines, IA

McCullough, ME. 1994. What silage is to a cow. In: Silage Production From Seed to
Animal. Proceedings from the national silage production conference. Syracuse, New
York. pp.1-15.

McDonald, P. 1981. The Silage Fermentation. John Wiley and Sons, New York.

179

McDonald, P., and AR. Henderson. 1974. The use of fatty acids as grass silage
additives. J. Sci. Food Agric. 25:791.

McDonald, P., AR. Henderson, and SJ. Heron. 1991. The Biochemistry of Silage.
2nd ed. Chalcombe Publications, Marlow, England.

Middlehoven, W.J., I.M. de Jong, and ME. de Winter. 1990. Yeasts and fungi
occuring in ensiled whole-crop maize and other ensiled vegatable crops. Antoine Van
Leeuwenhoek. 57:153.

Moon, NJ. 1983. Inhibition of the growth of acid tolerant yeasts by acetate, lactate
and propionate and their synergistic mixtures. J. Appl. Bacterial. 55:453.

Moon, N .J ., O.E. Lane, and EM. Sudweeks. 1980. Aerobic deterioration of wheat,
luceme and maize silages prepared with Lactobacilius acidophilic and Candida spp. J.
Appl. Bacterial. 49:75.

Muck, RE. 1988. Factors inﬂuencing silage quality and their implications for
management. J. Dairy Sci. 7122992.

Muck, RE. 1991. Predicting lactic acid bacterial numbers on alfalfa at six farms.
Trans. ASAE. 3421647.

Muck, RE, RE. Pitt, and R.Y. Leibensperger. 1991., A model of aerobic fungal
growth in silage. 1. Microbial characteristics. Grass Forage. Sci. 462283.

Muck, R.E. , and P. O'Keily. 1992. Aerobic deterioration of lucerne (Medicago sativa)
and maize (Zea mais) silages-effects of fermentation products. J. Sci. Food Agric.
59:145.

Nakagami, S. , H. Matsunaga, K. Miyoski, and A. Yamane. Nonradioactive detection
of nucleic acid by the universial probe system. Analyt. Biochem. 192211.

Nanba, A, R. Nakuda, and S. Nagai. 1983. Inhibition by acetic and propionic acids of
the growth of Propionibacterium shermanii. J. Ferment. Tech. 61:551.

Neronova, N.M., S.I. Ibragimova, and ND. Ierusalemski. 1967. Effect of propionate

concentration on the speciﬁc growth rate of Propionibacterium shermanii. Microbial.
36:330.

Nieuvenhof, F .F . , J. Stadhouders, and G. Hup. 1969. Stimulating effect of lactobacilli
on the growth of propionibacteria in cheese. Neth. Milk Dairy J. 23:287.

180

Nout, M.J., and F .M. Rombouts. 1992. Fermentative preservation of plant foods. J.
Appl. Bacterial. Supp. 73:136S.

Odelson, D.A, L. Nekms, S.F. Kotarski, and RB. Hespell. 1993. Development of a
DNA probe for the detection of ruminal Streptococcus bovis. 22nd Biennial Rumen
Function Conference. Chicago, IL. Abst. l4.

Oelberg, T.J., AK. Clark, R.K. McGuffy, and DJ. Schingoethe. 1983. Evaluation of
covering, dry matter and preservative at ensiling of alfalfa in bunker silos. J. Dairy
Sci. 66: 1057.

Ohyama, Y. 1984. Measuring silage management through research. p. 18-41. In: M.E.
McCullough and K.K. Bolsen (Ed) Silage Management. National Feed Ingredients
Assoc., West Des Moines, IA.

Ohyama, Y. , and S. Hara. 1975. Growth of yeasts isolated from silages on various
media and the relationship to aerobic deterioration of silage. Jap. J. Zootech. Sci.
46:713.

Ohyama, Y. , and P. McDonald. 1975. The effect of some additives on the aerobic
deterioration of silages. J. Sci. Food Agric. 26:941.

Ohyama, Y., S. Masaki, and S. Hara. 1975. Factors inﬂuencing aerobic deterioration

of silages and changes in chemical composition after opening silo. J. Sci. Food Agric.
26:1 137.

Ohyama, Y., S. Hara, and S, Masaki. 1979. The inﬂuence of hydrochloric acid on the
effect of caproic acid in the prevention of aerobic deterioration of silages. J. Sci. Food
Agric. 30: 101.

Ohyama, Y., T. Morichi, and S. Masaki. 1975. The effect of inoculation with
Lactobacillus plantarum and addition of glucose at ensiling on the quality of aerated
silages. J. Sci. Food Agric. 2621001.

O'Keily, P., and R.K. Wilson. 1991. Comparison of three silo types used to study in-
silo processes. Irish. J. Agric. Res. 30:53.

O'Keily, P. , and RE. Muck. 1992. Aerobic deterioration of lucerne (Medicago sativa)
and maize (Zea mais) silages-effects of yeasts. J. Sci. Food Agric. 59:139.

Ostling, C.E,. and S.E. Lindgren. 1991. Bacteria in manure and on manured and
NPK-fertilised silage crops. J. Sci. Food Agric. 55:579.

181

Pannon, G. 1990. Puriﬁcation and characterization of a proline iminopeptidase from
Propionibacterium shermanii 13673. Lait. 70:435.

Parker, R. B. 1981. Methodology for Determining Quality of Silage. National Feed
Ingredients Assoc. West Des Moines, IA

Parker, J.A., and NJ. Moon. 1982. Interactions of Lactobacillus and
Propionibacterium in mixed culture. J. Food Prat. 45:326.

Peberdy, M.F . , and T.F. Fryers. 1976. Improved selective media for the enumeration
of propionibacteria from cheese. N.Z. J. Dairy Sci. Technol. 11:10.

Perez, C.A, H.A Pesce de Ruiz, and G. Oliver. 1988a. Effect of PH and temperature
on. the proteolytic activity of Propionibacteria. Microbial. Aliments. Nutr. 6:91 .

Perez, C.A, F. Sesma, H.A Pesce de Ruiz, and G. Oliver. 1988b. Screening of
plasmids in strains of Propionibacterium and mesophillic lactobacilli from swiss type
cheeses. Microbial. Aliments. Nutr. 62171.

Peterkin, P.I., E.S. Idziak, and AN. Sharpe. 1991. Detection of Listeria
monocytogenes by direct colony hybridization on hydrophobic grid-membrane ﬁlters by
using a chromogen-labeled DNA probe. Appl. Environ. Microbiol. 5725 86.

Pettersson, K.L. , and S. Lindgren. 1990. The inﬂuence of the carbohydrate fraction
and additives on silage quality. Grass Forage Sci. 45:223.

Phillip, L.E., H.J. Garino, I. Alli, and BE. Baker. 1985. Effects of anhydrous
ammonia an amino acid preservation and feeding value of high-moisture car com for
growing steers. Can. J. Anim. Sci. 65:411.

Phillip, L.E. , and V. Fellner. 1992. Effects of bacterial inoculation of high-moisture
ear corn on its aerobic stability, digestion, and utilization for growth by beef steers. J.
Anim. Sci. 70:3178.

Pitt, RE. 1986. Dry matter losses due to oxygen inﬁltration in silos. J. Agric. Eng.
Res. 35: 193.

Pitt, RE. 1990. Silage and Hay Preservation. Ithaca, NY: Northeast Regional
Agricultural Engineering Service.NRAES-S.

Pitt, RE, RE. Muck, and R.Y. Leibensperger. 1985. A quantitative model of the
ensilage process in lactate silages. Grass Forage Sci. 402279.

182

Pitt, R.E., Y. Liu, and RE. Muck. 1991. Simulation of the effect of additives an
aerobic of alfalfa and corn silages. Trans. ASAE. 34: 1633.

Pitt, RE, RE. Muck, and N .B. Pickering. 1991. A model of aerobic fungal growth
in silage. 2. Aerobic stability. Grass Forage Sci. 462301.

Playne, M.J. 1985. Propionic and butyric acids. In: Mao-Young, M. (Ed)
Comprehensive Biotechnology. Oxford. Pergamon Press. pp. 731-759.

Prins, RA 1971. Isolation, culture, and fermentation characterisatics of Selenomonas
ruminantium var. bryanti var. n. from the rumen of sheep. J. Bacteriol. 105:820.

Racine, M., J. Dumount, C.P. Champagne, and A. Marin. 1991. Production and
characterization of the polysaccharide from Propionibacterium acidipropionici on
whey-based media. J. Appl. Bacterial. 71:233.

Ray, B., and W.E. Sandine. 1992. Acetic, propionic, and lactic acids of starter culture
bacteria as biopreservatives. In: B. Ray and M. Daeschel (Eds) Food Biopreservatives
of Microbial Origin. pp. 103-136. CRC Press, Boca Raton, FL.

Raeker, M.O., C.J. Bern, L.A Johnson, and RA Glatz. 1992. Preservation of high-
moisture maize by various propionate treatments. Cereal Chem. 69:66.

Reddy, M.S., G.W. Reinbold, and F .D. Williams. 1973. Inhibition of propionibacteria
by antibiotic and antimicrobial agents. J. Milk Food Technol. 362564.

Rehberger, T.G. 1993. Genome analysis of Propionibacterium freudenreichii by
pulsed-ﬁeld gel electrophoresis. Curr. Microbial. 27:21 .

Rehberger, T.G. , and B.A Glatz. 1990. Characterization of Propionibacterium
plasmids. Appl. Environ. Microbiol. 56:864.

Renterghem, B.V., F. Huysman, R. Rygole, and W. Verstraete. 1991. Detection and
prevalence of Listeria monocytogenes in the agricultural ecosystem. 71:211.

Ricke, S.C., D.M. Schaefer, M.E. Cook, and K.H. Kang. 1988. Differentiation of
ruminal bacterial species by enzyme-linked immunosorbent assay using egg yolk
antibodies from immunized chicken hens. Appl. Environ. Microbiol. 54:596.

Robinson, J.A., R.C. Greening, W.J. Smolenski, K. Barsuhn, M.M. Johnson, and
RC. Bell. 1991. Effects of ruminal inoculation of Megasphaera elsdenii strain 407A on
ruminal pH and organic acids in beef cattle. let Biennial Conference on Rumen
Function. Chicago, IL. Abst. 5.

183

Rooke, J .A 1990. The numbers of epiphytic bacteria on grass at ensilage on
commercial farms. J. Sci. Food Agric. 51:525.

Rosenberger, R.F . 1951. The development of methods for the study of obligate
anaerobes in silage. J. Appl. Bacterial. 14:161.

Rosner, B., and B. Schink. 1990. Propionate acts as a carboxylic group acceptor in
aspartate fermentation by Propionibacterium freudenreichii. Arch. Microbiol. 155246.

Russel, J.B., W.M. Sharp, and R.L. Baldwin. 1979. The effect of pH on maximum
growth rate and its possible role as a determinant of bacterial competition in the rumen.
J. Anim. Sci. 482251.

Russel, J. B. ,and R. L. Baldwin. 1979. Comparison of substrate afﬁnities among
several rumen bacteria. a possible determinant of rumen microbial competition. Appl.
Environ. Microbiol. 37. 531.

Rust, S.R., H.S. Kim, and G.L. Enders. 1989. Effects of a microbial inoculant on
fermentation characteristics and nutritional value of corn silage. J. Prod. Agric. 2:235 .

Rust, S.R. , and M.T. Yokoyama. 1992. Fermentation characteristics associated with
aerobic instability of high-moisture corn. J. Prod. Agric. 5:454.

Ruxton, G.D., and G.J. Gibson. 1993. A critical introduction to mathematical
modelling of the aerobic deterioration of silage. J. Sci. Food Agric. 62:209.

Sahlstrom, S., E. Espinjoza, T. Langmd, and T. Sorhaug. 1989. Cell wall,
membranes, and intracellular peptidase activities of Propionibacterium shermanii. J.
Dairy Sci. 72:342.

Salisbury, R.I., R.E. Mather, and CB. Bender. 1949. Various carbohydrates as energy
sources from some mixed cultures of silage organisms. J. Dairy Sci. 322901.

SAS. 1987. SAS/STAT' Guide for Personal Computers (6th Ed). SAS Inst. , Inc.
Gary, NC.

Sauer, D. B. ,.R Burroughs, and J. A Simon. 1972. Control of grain storage fungi with
propionic and acetic acids. Phytopath. 62: 787.

Sauer, D.B., and R. Burroughs. 1974. Efﬁcacy of various chemicals as grain mold
inhibitors. Trans. ASAE. 17:557.

Sayler, G.S. , and AC. Layton. 1990. Environmental application of nucleic acid
hybridization. Annu. Rev. Microbiol. 44:625.

184

Schaefer, D.M., P.G. Bratz, S.C. Arp, and D.K. Cook. 1989. Inoculation of corn
silage and high-moisture corn with lactic acid bacteria and its effeCts on the subsequent
fermentations and on feedlot performance of beef steers. Anim. Feed Sci. Technol.
25:23.

Schuppert, B. , B. Schink, and W. TrOsch. 1992. Batch and continuous production of
propionic acid from whey permeate by Propionibacterium acidi-propionici in a three-
electrode amperometric culture system. Appl. Microbiol. Biotechnol. 37 2549.

Schutz, M. , M. Sievers, and K. Klerm. 1987. Metabolically controlled continuous
fermentation of lactose to propionate, acetate and carbon dioxide. Milchwissenschaft.
42:71.

Scale, DR. 1986. Bacterial inoculants as silage additives. J. Appl. Bacterial. Symp.
Supp. 9S.

Seitz, L.M., D.B. Sauer, and H.E. Mohr. 1982a. Storage of high-moisture corn:
fungal growth and dry matter loss. Cereal Chem. 59: 100.

Seitz, L.M., D.B. Sauer, H.E. Mohr, and D.F. Aldis. 1982b. Fungal growth and dry
matter loss during bin storage of high-moisture corn. Cereal Chem. 59:9.

Sharp, R., H.J. Gilbert, and AG. O'Donnell. 1991. An improved method for the
extraction and recovery of micro-organisms from grass silages. Letters Appl.
Microbiol. 13:227.

Sharp, R., AG. O'Donnell, H.G. Gilbert, and GP. Hazlewood. 1992. Growth and
survival of genetically manipulated Lactobacillus plantarum in silage. Appl. Environ.
Microbiol. 58:2517.

Shaver, R.D., R.A Erdman, AM. O'Connor, and J.H. Vandersall. 1985. Effects of
silage pH on voluntary intake of corn silage and alfalfa haylage. J. Dairy Sci. 68:338.

Siegfried, B.R., H. Ruckemann, and G. Stumpf. 1984. Method of determination of
organic acids in silage by high performance liquid chromatography. Landwersch
Foschung 37:298.

Slyter, LL. 1976. Inﬂuence of acidiosis on rumen function. J. Anim. Sci. 43:910.
Smith, G.B., and J .M. Tiedje. 1992. Isolation and characterization of a nitrate

reductase gene and its use as a probe for denitrifying bacteria. Appl. Environ.
Microbiol. 58:376.

185

Spaelstra, S.F., M.G. Courtin, and J .A. VanBeers. 1988. Acetic acid bacteria can
initiate aerobic deterioration of whale-crop maize silage. J. Agric. Sci. 111:127.

Spaelstra, SF. 1990. Comparison of the content of clostridial spores in wilted grass
silage ensiled in either laboratory, pialt scale or farm silos. Neth. J. Agric. Sci.
38:423.

Stahl, D.A., B. Flesher, H.R. Mansﬁeld, and L. Montgomery. 1988. Use of
phylogenetically based hybridization probes for studies of ruminal microbial ecology.
Appl. Environ. Microbiol. 5421079.

Stallings, C.C., R.K. McGuffy, T.R. Middleton, and J.W. Thomas. 1979. Responses
of sheep and dairy cows to propionic acid treatment of alfalfa haylage fed with or
without corn silage. J. Dairy Sci. 6221264.

Steel, R.G., and J .H. Torrie. 1980. Principles and Procedures of Statistics (2nd Ed).
A Biometrical Approach. McGraw-Hill, New York.

Steele, D.B., and MD. Stawers. 1991. Techniques for selection of industrially
important microoganisms. Ann. Rev. Microbiol. 45289.

Sterling, AC., and R. Whitenbury. 1963. Sources of lactic acid bacteria occurring in
silage. J. Appl. Bacterial. 26286.

Stevenson, KR. 1976. Stability and bunk life of high-moisture corn. High Moisture
Grains Symp. p. 105. Oklahoma State Univ., Stillwater.

Stinson, E.E., and K.A Naftulin. 1991. Effect of pH on organic acid production by
Clostridium propionicum in test tube and fermentar cultures. J. Ind. Microbial. 8:59.

Suzuki, 1. , M. Nomura, and T. Morichi. 1991. Isolation of lactic acid bacteria which
suppress mold growth and show antifungal action. Milchwissenschaft. 462635 .

Suzzi, G., L. Grazia, and G. Ferri. 1990. Studies on isobutyric acid-producing bacteria
in silage. Letters Appl. Microbiol. 10:69.

Tanaka, K., K. Nakamura, and E. Mikami. 1990. Fermentation of maleate by a gram-
negative strictly anaerobic non-spore-former, Propionivibrio dicarboxylicus gen. nov. ,
sp. nov. Arch. Microbiol. 154:323.

Tomes, NJ. 1991. Bacterial treatment to preserve silage. U.S. Patent 4,981,705.

Tsai, Y.L., and B.H. Olson. 1991. Rapid method for direct extraction of DNA from
soil and sediments. Appl. Environ. Microbiol. 5721070.

186

Turner, K.W., H.A Morris, and F.G. Martley. 1983. Swiss-type cheese 11. The role
of thermophilic lactobacilli in sugar fermentation. New Zeal. J. Dairy Sci. Technol.
18: 1 17.

Tyree, R.W., E.C. Clausen, and J .L. Gaddy. 1991. The production of propionic acid
from sugars by fermentation through lactic acid as an intermediate. J. Chem. Tech.
Biotechnol. 50:157.

Vedamuthu, B.R., and G.W. Reinbold. 1967. The use of candle oats jar incubation for
the enumeration, characterization and taxonomic study of propionibacteria.
Milchwissenschaft. 22:428.

Voclker, H.H., D.P. Casper, F.C. Ludens, and DJ. Schingoethe. 1989. High
moisture corn preserved with esters of propionic acid for lactating cows. J. Dairy Sci.
72:89.

Voordouw, G., J .K. Voordouw, T.R. Jack, A. Foght, P.M. Fedorak, and D.W.
Westiche. 1992. Identiﬁcation of distinct communities of sulfate-reducing bacteria in
oil ﬁelds by reverse sample genome probing. Appl. Environ. Microbial. 5823542.

Varob'eva, L. I. 1958. Carbon balance in fermentation of lactic and pyruvic acids by
propionic acid bacteria. Mikrobiologiya. 27:286.

Vorob 'eva, L. I. 1959. Effect of aeration on propionic acid fermentation.
Mikrobiologiya. 28:209.

Walker, H. 1977. Spoilage of foods by yeasts. Food Tech. 31:57.

Wang, C. , and LA Johnson. 1992a. Wet-milling characteristics of prapionate-treated
high-moisture maize. I. Yields and composition of products. Cereal Chem. 69:43.

Wang, C. , and LA Johnson. 1992b.Wet-milling characteristics of propionate-treated
high-moisture maize. II. Qualities of starch and gluten. Cereal Chem. 69:47.

Wang, R.F., W.W. Cao, and M.G. Johnson. 1991. Development of a 16S rRNA-
based oligomer probe speciﬁc for Listeria monocytogenes. Appl. Environ. Microbiol.
5723666.

Wardynski, F .A. 1991. Effects of Microbial Cultures and Various Other Additives on
the Feeding Quality, Fermentation Pattern, Dry Matter Recovery and Aerobic Stability
of High Moisture Corn. M.S. Thesis. Michigan State University, East Lansing, MI.

187

Wardynski, F .A., S.R. Rust, and M.T. Yokoyama. 1993. Effect of microbial
inoculation of high-moisture corn on fermentation characteristics, aerobic stability, and
cattle performance. J. Anim. Sci. 71:2246.

Wawszkiewicz, EL, and H.A Barker. 1968. Erythritol metabolism of
Propionibacterium pentosaceum. The overall reaction sequence. J. Biol. Chem.
24321948.

Whittenbury, R., P. McDonald, and D.G. Bryan-Jones. 1967. A short review of some
biochemical and microbiological aspects of ensilage. J. Sci. Food Agric. 18:441.

Wilkens, RJ. 1981. The nutritive value of silages. In: W. Haresign and DJ. Cole
(Eds) Recent Developments in Ruminant Nutrition. Butterworths, London.

Wilson, RR, and RJ. Wilkins. 1972. An evaluation of laboratory ensiling techniques.
J. Sci. Food Agric. 23:377.

Wittenberg, K.M. 1991. Preservation of high-moisture hay in storage through the use
of forage additives. Can. J. Anim. Sci. 71:429.

Wood, H.G., R. Stjernholm, and F.W. Leaver. 1955. The metabolism of labeled
glucose by the propionic acid bacteria. J. Bacteriol. 70:510.

Wood, H.G. 1981 . Metabolic cycles in the fermentation by propionic acid bacteria.
Curr. Top. Cell. Regul. 182255.

Woolford, M.K. 1975a. Microbiological screening of the straight chain fatty acids
(Cl—Cu) as potential silage additives. J. Sci. Food. Agric. 26:219.

Woolford, M.K. 1975b. Microbiological screening of food preservatives, cold
sterilants and speciﬁc antimicrobial agents as potential silage additives. J. Sci. Food
Agric. 26:229.

Woolford, M.K. 1975c. The signiﬁcance of Propionibacterium species and
Micrococcus lactityticus to the ensiling process. J. Appl. Bacterial. 39:301.

Woolford, M.K. , and J .E. Cook. 1977 . Investigations into the prevention of the
aerobic deterioration of maize silage. Proc. 13th Int. Grassl. Congr. Leipzig, 232.

Woolford, M.K. 1978. Antimicrobial effects of mineral acids, organic acids, salts and
sterilizing agents in relation to their potential as silage additives. J. Br. Grassl. Soc.
33:131.

 

188

Woolford, M.K., and J.B. Cook. 1978. A note on the effects on the aerobic
deterioration of maize silage of the manipulation of the micraﬂora by means of
antibiotics. Anim. Feed Sci. Tech. 3289.

Woolford, M.K., K.K. Bolsen, and LA Peart. 1982. Studies on the aerobic
deterioration of whole crop cereal silages. J. Agric Sci. Camb. 98:529.

Woolford, M.K. 1983. Gamma radiation in the preparation of maize silage for studies
on the aerobic deterioration process. J. Agric. Sci. Camb. 1022105.

Woolford, M.K. 1984. The Silage Fermentation. Microbiology Series. Vol. 14.
Marcel Decker, New York.

Woolford, M.K. 1984. Managing aerobic deterioration in silage. In: M.E. McCullough
(Ed) Silage Management. Natl. Feed Ingred. Assoc. West Des Moines, IA.

Woolford, M.K., and AC. Wilkie. 1984. Investigations into the role of speciﬁc micro—
organisms in the aerobic deterioration of maize silage. J. Agric. Sci. Camb. 102:97.

Woolford, M.K. 1990. A review. The detrimental effects of air on silage. J. Appl.
Bacterial. 68:101.

Woolford, M.K. 1991. Feed preservation methods. In: P.T. Lyons (Ed)
Biotechnology in the Fwd Industry. Alltech, Nicholasville, KY. 227.

Woolford, M.K. 1992. Silage: rethinking silage fermentation. What happens? Is it
Necessary? Can we change it to control efﬂuent and make it more suitable for rumen
fermentation? In: P.T. Lyons (Ed) Biotechnology in the Feed Industry. Alltech,
Nicholasville, KY. p. 227 .

Woskow, S.A , and BA Glatz. 1991. Propionic acid production by a propionic acid-
talerant strain of Propionibacterium acidipropionici in batch and semicontinuous
fermentation. Appl. Environ. Microbiol. 57:2821.

Youngsmith, B., K. Sonomoto, S. Tanaka, and S. Fukui. 1982. Production of vitamin
B12 by immobilized cells of a propionic acid bacterium. Appl. Microbiol. Biotechnol.
16:70.

Zarda, B., R. Amann, G. Wallner, and K.H. Schleifer. 1991. Identiﬁcation of single
bacterial cells using digoxigenin-labelled, rRNA-targeted oligonucleotides. J. Gen.
Microbiol. 13722823.