\—

MSAYS FOR UDDER. IRRETATION fiASED UPON
DEOXYRWG-NUCLEIC AGED CONTEN? OF MILK
SOMATEC CELLS

That: for flu Degree 0% M 5‘
MEW§MN SfATE UNNERSETY
Max 1. Papa
W63

 

LIBRARY

Michigan State
University ‘

 

ABSTRACT

ASSAYS FOR UDDER IRRITATION
BASED UPON DEOXYRIBONUCLEIC ACID CONTENT

OF MILK SOMATIC CELLS
by.Max J. Paape

A Michigan Mastitis Test (MMT) for udder irrita-
tion was developed which produced results equivalent to
the California Mastitis Test (CMT) and the Milk Quality
Test (MOT). The MMT is an Open formula gelation test and
the ingredients can be purchased and the reagent prepared
with the aid of a pan balance.

In the course of a study of milk leucocytes, the
commonly used stains were found to be unsatisfactory be-
cause background staining masked the leucocytes, or be-
cause of a lack of nuclear-cytoplasmic resolution within
the leucocytes. A pyronin Y - methyl green staining
procedure was develOped which greatly improved the micro-
scopic resolution of milk somatic cells. Furthermore, the
pyronin Y - methyl green stain resulted in an approximate
25% reduction in smear variance and in an approximate 50%
reduction in count variance when compared with Wright's
stain.

In View of the evidence that deoxyribonucleic acid
(DNA) was the material reSponsible for the gelation in the

CMT, MOT, and EMT reactions, the DNA-Specific Feulgen

Max J. Paape

reaction was modified, and was shown to be less variable
than the MOT and more closely related to leucocyte

numbers than the MOT.

ASSAYS FOR UDDER IRRITATION
BASED UPON DEOXYRIBONUCLEIC ACID CONTENT

OF MILK SOMATIC CELLS

BY

Max J. Paape

A THESIS

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

MASTER OF SCIENCE

Department of Dairy

1963

BIOGRAPHICAL SKETCH

Max J. Paape was born at Port Chester, New York,
on December 26, 1936. He received his elementary educa-
tion in Our Lady of Mercy Parochial School, and was
graduated from Port Chester Senior High School in June,
1954.

In September, 1954, he enrolled at the Long
Island Agricultural and Technical Institute, Farmingdale,
New York, majoring in Animal Husbandry. He was graduated
in June, 1956, with the degree of Applied Arts and Science.

From September, 1956 to June 1959, he completed
his undergraduate requirements leading to a Bachelor of
Science in Agriculture Extension at Michigan State Univer-
sity.

In September, 1960, he enrolled in the graduate
school of Michigan State University, Department of Dairy.

He received the Master of Science degree in March, 1963.

ii

ACKN 03-91. ‘2‘ DG EMBIK'ST S

The author wishes to express his gratitude and
thanks to Prof. W. W. Snyder and Dr. H. D. Hafs for
guidance and help in conducting this study. Their gene-
rous assistance with the manuscript is greatly appreciated.
The interest and assistance of Dr. H. Allen Tucker is also
acknowledged.

Most sincere appreciation is extended to Dr. M. J.
Gordon for his very helpful direction and assistance in
the formative part of this study.

The author is grateful for the financial support
provided by his parents, and by Michigan State Univer-
sity through a Graduate Research Assistantship.

Gratitude is extended to Dr. C. A. Lassiter,
Chairman of the Department of Dairy, for making available
the facilities and animals.

The interest and encouragement shown by Sue Paape

is appreciated.

iii

TABLE OF CONTENTS
Page

BIOGRAPHICAL SKETCH . . . . . . . . . . . . . . . . ii
ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . iii
INTRODUCTION . . . . . . . . . . . . . . . . . . . 1
EXPERIMENT I. THE MICHIGAN MASTITIS TEST

Review of Literature . . . . . . . . . . 2

Materials and Methods . . . . . . . . . . 4

Results and Discussion . . . . . . . . . . 6
EXPERIMENT II. THE EFFECTS OF DEOXYRIBONUCLEASE

ON THE MILK GELATION TESTS

Materials and Methods . . . . . . . . . . 10

Results and Discussion . . . . . . . . . . ll
EXPERIMENT III. VARIATION OF ESTIMATED NUMBERS OF

MILK SOMATIC CELLS STAINED WITH WRIGHT'S

STAIN OR PYRONIN Y - METHYL GREEN STAIN

Review of Literature . . . . . . . . . . . 12

Materials and Methods . . . . . . . . . . 14

Results and Discussion . . . . . . . . . . l6
EXPERIMENT IV. FEULGEN-DNA IN MILK AS A JEASURE

OF UDDER IRRITATION

Review of Literature . . . . . . . . . . . 22

Materials and Methods . . . . . . . . . . 23

Results and Discussion . . . . . . . . . . 26
SUMMARY . . . . . . . . . . . . . . . . . . . . . . 33
LITERATURE CITED . . . . . . . . o . . . . . . . . 37

APPQIDIX . O O O O O O O O O O C O O O 0 O O O C O O 41

LIST OF TABLES
Page
TABLE
1. Average CMT, MOT, and MMT gelation test scores
and within sample standard deviations and co-
efficients of variation . . . . . . . . . . . . 7
2. Average CMT and LMT gelation test scores and
within sample standard deviations and coeffic-
ients of variation . . . . . . . . . . . . . . 8
3. Averages and coefficients of variation for
each milk cell classification with Wright's
stain or pyronin Y - methyl green . . . . . . . l7
4. Level of significance (PI) for each measurable
source of variation for each milk cell class—

ification analysis . . . . . . . . . . . . . . 18

2 2 )

ham) and count (s czbm

5. Estimated smear (s
variances for granulocytes and total cells in
each stain . . . . . . . . . . . . . . . . . . l9
6. Expected 95% confidence intervals for a mean
number of milk cells obtained from differing
numbers of smears per milk sample and counts
per smear . . . . . . . . . . . . . . . . . . . 21
7. Averages, coefficients of variation, and
estimated smear (Szbsm) and count (Szcxbm)
variances for each milk cell classification . . 27

8. Correlation coefficients between somatic

cell numbers and Feulgen-DNA or MQT scores . . 28

LIST OF FIGURES

FIGURE
1. Standard color chart for Feulgen-DNA test .
2. Relationship of concentration of total
somatic cells to Feulgen-DNA score and to
MOT score . . . . . . . . . . . . . . . . .
3. Regression of logarithm of average number of
somatic cells (Y) on the logarithm of the

coded Feulgen-DNA score (X) . . . . . . . .

vi

Page
. 25
. 30
. 31

LIST OF APPENDICES
Page
APPENDIX
Procedure I. Staining with Wright's stain . . . . 41
Procedure II. Method used in finding the

area of a microsc0pic field . . . . 41

vii

INTRODUCTION

The term mastitis is defined as an inflammation
of the tissues in the cow's udder or teats. Many
methods have been suggested for the detection of udder
inflammation based on Changes in the milk composition
(lactose, chlorides, solids),pH, cell count, while
other methods are based on the organisms involved. From
the standpoint of reliability leucocyte counts are con-
sidered the most accurate method of detecting mastitis
(2, 36).

Certain management practices such as the im-
proper use of the milking machine will increase the
leucocyte count of milk. If corrected in time mastitis
could be prevented. One objection to most tests for
mastitis is that they do not detect the condition in the
early stages. The leucocyte count detects the condition
mudh sooner than some of the other tests (2).

The purpose of the researdh reported in this
thesis was to develop several new methods for detecting
udder irritation based upon the leucocyte numbers of
milk. These methods are presented as four experiments
with each experiment designed to study a Specific possib—

ility of improving tests for mastitis.

EXPERIMENT I. THE MICHIGAN MASTITIS TEST

A review of the literature dealing with general
mastitis tests and the gelation tests will be presented
in this section of the thesis while the literature dealing
specifically with the other tests for mastitis will be

reviewed in the apprOpriate sections.
Review of Literature

Cole and Easterbrooks (13) observed a leukOpenia
in dairy cows with acute mastitis. Similarly, Schalm (40)
and Theilen gg_gl. (45) reported that a leucocytic res-
ponse to mastitis is characterized by a rapid reduction of
leucocytes in the blood as they move into the area of in-
jury in the udder. Pattison g£_21. (35) found a predom-
inance of leucocytes in the udders of goats infected with
Streptococcus agalactiae: Hughes (19) showed that a high
cell count was associated with inflammed quarters.
McFarlane g2_gl. (28) and Chu (12) concluded that high
cell counts were an indication of mastitis. Furthermore,
it has been shown by many workers, Barnum and Newbould (4),
Jensen (21), Leidl and Schalm (24), Leidl et a1. (25),
and Schalm (41), that the California Mastitis Test (GMT)
(42) and the Milk Quality Test (MOT) (34) reactions, used
so widely by dairymen and by veterinarians, are indirect
measures of the concentration of leucocytes in the milk.

Both the CMT and MQT have received widespread

acceptance for field diagnosis of mastitis. Albright (1)
describes the use of the GMT as an effective tool in a
herd management mastitis control program. In California
(38, 39) the GMT is currently being applied to about
80,000 cows each month. Reports by Jackson (20), McKay
(30), and Steere (43) indicate favorable results by using
the GMT in mastitis control programs.

Marshall and Edmondson (29) suggested that the
GMT can be of great value to the practitioner in his
evaluation of the herd mastitis situation. Barnum (3) sug-
gested that when the CMT is applied at monthly intervals
it is possible to determine the infection status of each
quarter. He also presented evidence to indicate that
quarters infected with Streptococcus agalactiae,
Staphylococcus aureus, or other organisms that produce a
chronic infection, consistently show a positive GMT re-
action. Furthermore, he suggested that the GMT be employ-
ed as a screening test when introducing new cows into a
herd, and that a negative GMT on all quarters of a milk
cow is a reasonable criterion for the animal's introduc-
tion into the milking herd.

The GMT has also been used as a quality test for
milk samples collected from bulk tanks. Temple and Haller
(44) have shown that GMT scores were lower on bulk milk
for those herds enrolled in the New York mastitis control

program. Leidl and Schalm (24) suggested using the GMT

4

on bulk tank samples to detect the majority of the herds
which produce abnormal milk. Gray and Schalm (15) report
that a reduction of bulk milk scores to negative values
could lead to an increase of 14 to 20% in milk production,
as well as a significant improvement in the quality of
milk. Holt (18) reported that one might eXpect an in-
crease in butterfat production through the prOper use of
the GMT.

This evidence indicates that the GMT and MQT have
an important place in a mastitis control program. However,
the cost of the GMT and the MQT reagents may be a deter-
rent to their more extensive use. A less expensive re-
agent may encourage a more effective mastitis control
program because more dairymen would participate. The
purpose of this experiment was to develOp an Open formula
gelation test for mastitis similar in reaction to the
closed formula GMT and MQT, and to present statistical
evidence to determine if there was any difference in the

results obtained when using these three reagents.
Materials and Methods

The Open formula gelation test, hereafter called
the Michigan Mastitis Test (MMT), was prepared.by diluting
19.0 g of sodium alkylarylsulfonatel, 13.5 g of sodium

hydroxide1 (analytical reagent), and 1.5 g of methylene

 

1 Available from Fisher Scientific Co., 1458 N.
Lamon Ave., Chicago 51, Illinois.

5

blue Chloride to one gallon (3,785 ml) in softened tap
water. These reagents were mixed to obtain complete dis-
solution.

Ninety—one milk samples, each consisting of about
30 m1 of foremilk, were obtained daily over a period of
two weeks from.Holstein cows in various stages of lact-
ation. The samples were immediately tested in duplicate
using the CMT,.MQT, and MMT within 30 minutes after the
milk was obtained. The GMT and NOT procedures were per-
formed by adding 3.50 ml of reagent to 3.00 ml of milk in
a paddle and scoring the degree of gelation according to
the manufacturer's recommendation, except that gelation
scores of 0, 1, 2, 3, and 4 were substituted for 0, T, l,
2, and 3, respectively, to facilitate statistical analysis.
The procedure employed for the MMT was identical except
that the MMT solution was used in place of the GMT or MST
test reagents.

In a second trial, which was designed to determine
whether or not more readily available reagents could be
substituted for those in the MMT solution, 52 foremilk
samples were tested in duplicate with GMT and with a Lye
Mastitis Test solution, hereafter called LMT. This LMT
solution was prepared by diluting 2.5 level teaspoons

(13.5 g) of lye2 and four level teaSpoons (19 g) of liquid

 

2 Red Seal, high test Lyons New Flake Lye. Active
ingredients: 96% NaOH, 2% Sodium Carbonate, and 2% inert
ingredients.

detergent1 in one gallon of soft tap water. After gentle
mixing to dissolve the lye, the solution was allowed to
stand for 24 hours to permit clearing. The clear portion
was then decanted into another container with care to
avoid remixing the settled out impurities, which were
discarded. One level teaspoon (6 g) of navy blue dye2

was added to the clear test solution to produce the de-

sired color.
Results and Discussion

The average scores for 91 milk samples in the
initial experiment are shown in Table l with the within
sample standard deviations and coefficients of variation.
The average scores for the GMT, MQT, and MMT do not differ
significantly (P7>.50). Although the use of the MMT
resulted in lower total and relative variables, the
differences among the standard deviations and among the
coefficients of variation for the three tests are small
relative to the average variabilities. The correlation
of the average scores of GMT with MQT was 0.975: of GMT
with MMT, 0.962: and of MQT with MMT 0.969. Each of
these correlation coefficients were highly significant

(P (.001).

 

1 Swan liquid detergent.

2 Rit tints and dyes, navy blue 30.

7

TABLE 1

Average GMT, MOT, and MMT gelation test scores and with-
in sample standard deviations and coefficients of variation

   

-k .___. we ..

Within Sam 1e

 

TeSt Average Standard Coeffic-
. gelation
deviation ient of
scores

variation
California Mastitis Test (GMT) 1.68 0.33 20%
Milk Quality Test (MOT) 1.76 0.38 21%
Michigan Mastitis Test MMT) 1.81 0.32 18%
Average 1.75 0.34 20%

 

This statistical evidence demonstrates that little
difference exists among the test scores obtained from the
three tests employed, and that a choice among them may be
made based upon other factors such as personal preference,
availability of reagents, or cost. The MMT solution may
be made with a substantial reduction in cost. On the
other hand, prepared MOT and GMT solutions may be more
practical for those with no facilities for accurate
weighing.

Since the Methylene Blue Chloride simply serves
to make the milk gel more visible, other coal tar dyes
may be used in its place at the discretion of the user.

In fact, when the test procedure was performed in black
bakelite jar lids, no color additive was necessary because
the milk itself provided adequate contrast with the dark

background.

8

In the second trial, the average test score for

1T was 1.72 which was not significantly different
(P > .50) from the 1.77 score for LMT. The within sample
standardxhviations were 0.22 and 0.36, and the coefficients
of variation were 13% and 20%, reSpectively, for the GMT
and the LMT. Although these measures showed greater vari-
ability between duplicate LMT determinations, the differ-
ences were relatively small for most practical purposes.
The correlation between GMT and LMT score was 0.951
(P < .001) .

The data from the second trial demonstrate that
one has considerable flexibility in the grade and source
of reagents used in preparing the milk gelation test solu-
tions. Undoubtedly, many other readily available brands
of lye and detergents could perform equally well, al-
though they were not included in this experiment. In any

event, the test solution for the LMT was prepared at a

TABLE 2

Average GMT and LMT gelation test scores and within
sample standard deviations and coefficients of variation

Within sample

 

 

Test AZIriign Standard Coeffic-
geoges deviation ients of
variation
California Mastitis Test (GMT) 1.72 0.22 13%

Lye Mastitis Test (LM'I‘) 1.77 0.36 20%

 

9

small fraction of the cost of the GMT and MOT test solu-
tions. Furthermore, the procedure for preparing the LMT
solution is simple, the ingredients are readily available,
and the equipment used in preparing the test solution is

present in most households.

EXPERIMENT II. THE EFFECTS OF DEOXYRIBONUCLEASE ON THE

MILK GELATION TESTS FOR MASTITIS

The gelation of the GMT, MOT, and MM were remin-
iscent of the behavior of deoxyribonucleic acid (DYA) in
solution. The present experiment was designed to test
the hypothesis that DNA of milk somatic cells was the

material reSponsible for the gels.
Materials and Methods

A deoxyribonuclease (DNAse)1 solution was prepared
by dissolving 2 mg of DNAse in 2 ml of 0.008 molar ace-
tate buffer and 0.025 molar MgCl2 (pH 4.6).

The gels to be tested were obtained by adding the
GMT, MOT, or MMT reagents to a known mastitic milk sample
and then centrifuging for 15 minutes at 26,000 gravities.
After centrifugation, the gels were removed and each divid-
ed into approximately two equal portions and placed in
individual glass petri dishes. The pH of the gels was
11.5 and was adjusted to 5.3 by the addition of 6 N HCl.
One-hundredth ml of the DNAse solution was then added to
one portion of each sample and 0.01 ml of distilled water
to the other portion (control). According to the

Worthington Biochemical Corp., from whom the enzyme was

 

1 Worthington DNAse II.

10

ll

purchased, the DNAse shows optimum activity at this lower
pH. The pH was then adjusted to the original pH of 11.5
by the addition of one drop of 6 N NaOH. The trial was

repeated for six different mastitic milk samples.
Results and Discussion

No gelation recurred in the DNAse treated sample,
but a heavy gel recurred in the control sample after
elevation of the H to 11.5. The same results were ob-
served for each of the six different mastitic milk samples.

The addition of trypsinl, pepsinz, proteinase3,
papain4, and DNAse5 had no effect on the viscosity of the
gels as measured by viscometers. These findings and the
enzymatic specificity of DNAse for DNA indicated that DNA
of milk somatic cells was reSponsible for the gel in the
gelation tests.

Since this work has been completed similar find-
ings have been reported by Caroll and Schalm (10) for the

CMT .

 

Worthington Biochemical Corp.
General Biochemicals Corp.
Nutritional Biochemicals Corp.
Fisher Scientific Co.

Worthington Biochemical Corp.

EXPERIMENT III. VARIATION OF ESTIMATED NUMBERS OF MILK
SOMATIC CELLS STAINED WITH WRIGHT'S STAIN OR PYRONIN Y -

METHYL GREEN STAIN
Review of Literature

Several researchers Guallini & Leali (16),
Guallini & Vallis (l7), and Varrier-Jones (46) have at-
tempted to make differential counts of the leucocytes in
milk as an indication of mastitis. Macadam (27) concluded
that the pr0portion of granulocytes usually exceeded 70%
of the total cells from acutely infected quarters, where-
as it was usually less than 40% during mammary involution.
Blackburn gg_gl. (5) indicated that milk samples with a
low total cell count generally contain less than 45%
granulocytes and claimed that the differential cell count
is a valuable criterion to confirm the conclusions drawn
from the results of a total cell count, eSpecially Where
doubts arise owing to the presence of bacteria in the milk.

Because of the opacity of milk, it is necessary
to stain the leucocytes before microscopic observation.

In the course of a study of milk leucocytes in the authorS'
laboratory, the commonly used stains such as Wright's (48),
Newman's (33), Newman Modification (26), Broadhurst-Paley
(9) and its modification (11) were used and found to be
unsatisfactory because background staining masked the

leucocytes, or because of a lack of nuclear-cytoplasmic

12

13

resolution within the leucocytes.

The results of the previous experiment showed
that the GMT, MOT, and MMT gelation reactions were due to
the DNA in the milk somatic cells. This fact indicated
that a staining procedure Specific for nucleic acid would
be preferable to the previously used stains because the
background would not be eXpected to take appreciable
quantities of stain. Brachet (6, 7, 8) has shown that
the stain methyl green is specific for DNA and that pyronin
Y is specific for ribonucleic acid (RNA). Kaufmann (23),
using a staining mixture composed of methyl green and
pyronin Y, found that Chromosomes stain blue-lavender,
whereas nucleoli and RNA-containing cytOplasmic particles
stain red.

Preliminary investigations in the author's labor-
atory indicated that the pyronin Y - methyl green stain
resulted in considerably improved microscopic resolution
of milk somatic cells. The backgrounds of smears stained
with pyronin Y - methyl green were clear with a light
lavender mottling. The cytOplasm of epithelial cells con-
tained large quantities of pyronin Y-positive material,
whereas agranulocytes contained small quantities and
granulocytes contained none.

The purpose of this experiment was to compare the
variation of estimated numbers of somatic cells in milk
stained with pyronin Y - methyl green with that of cells

stained with Wright's stain.

14

Materials and Methods

A milk sample was obtained from each of 27 Holstein
cows in various stages of lactation. Fourteen of these
cows were considered to have mastitis, based upon GMT
observations. Within 30 minutes after obtaining the
samples, four smears were prepared from each milk sample
by the direct-smear method of Prescott and Breed (37),
which consisted of transferring 0.01 ml of milk in a stan-
dard platinum loop to a microscope slide and distributing
the milk over one square centimeter. The smears were
stained with Wright's stain according to the procedure out-
lined by Schalm (41) (Appendix, Procedure I). The remain-
ing two smears were stained with pyronin Y - methyl green
stain according to the following procedures:

a) Fix in Carnoy's for ten minutes.

b) Hydrate for two minutes each in 50% ethyl

alcohol, 30% ethyl alcohol, and distilled water.

c) Stain for six minutes in a freshly prepared

solution of 0.50% pyronin Y (National Analine)
and 0.30% methyl green (National Analine) in
distilled water.

d) Immerse for three minutes in eaCh of two

changes of N-butyl alcohol.
e) Clear for three minutes in each of two changes

of xylol.

15

f) Mount in 50% piccolyte dissolved in xylol.

After preparation, the smears were microsc0pically
examined, using immersion oil, by a person who was un-
aware of the identity of the smear. The diameter of a
microsc0pe field was 0.178 mm, resulting in about 4,000
fields per smear (Appendix, Procedure II). The number of
somatic cells per ml of each milk sample was estimated
from the number in 25 fields randomly chosen from a smear
and multiplied by 16,000. This constituted a count. The
number of agranulocytes, granulocytes, leucocytes (the
sum of agranulocytes and granulocytes), epithelial cells,
and total somatic cells (the sum of leucocytes and epi-
thelial cells) was determined in duplicate for each smear.

Five identical analyses of variance, one for each
of the cell classifications, considered the following
sources of variation: milk samples, stains, smears,
counts, and an interaction between samples and stains.
Samples were considered to be random and factorial, and
stains to be fixed and factorial. Smears were considered
to be random and nested within samples and stains, and
counts to be random and nested within smears.

To compare the variation of determinations of cell
numbers within each of the two staining procedures, the
variance of duplicate smears and of duplicate counts were
estimated for each stain for the number of total cells
and the number of granulocytes. These variances were

used to estimate the standard errors of determinations of

16

mean cell numbers for future milk samples with varying
numbers of smears per sample and counts per smear, as

shown in the following formula:

 

2 2
S bsm + S cxbm

521‘s— @552)

where 32b:m is the smear variance, 32

 

czbm is the count
variance, nb is the number of smears per sample, and nC is
the number of counts per smear. The standard errors were
then used to predict the 95% confidence intervals for
determinations of mean cell numbers as follows:
95% confidence interval = (2) (t.05) (sx),

where t.05 has nbnc-l degrees of freedom. Similar calcu-
lations were not made for the other three cell classifica-

tions because they have not been extensively used in

mastitis research.
Results and Discussion

The average number of cells for each of the five
cell classifications as determined after staining with
Wright's stain or with pyronin Y - methyl green are list-
ed in Table 3 with their correSponding coefficients of
variation. The coefficients of variation were determined
by taking the square root of the sum of the estimated
smear and count variances and dividing by the mean. The
milk sample variance was excluded from these estimates

because the samples included in the present experiment

17

would not necessarily be representative of those chosen
for a future eXperiment where, for example, one may begin
with either infected cows or cows with no history of
mastitis. Consequently, the coefficients of variation in
Table 3 are minimal estimates. A summary of the prob-
abilities (pI's) obtained in the analysis of variance for
eaCh of the five cell classifications is presented in

Tafle4.

TABLE 3

Averages and coefficients of variation for each milk cell
classification with Wright's stain or pyronin Y - methyl green

 

 

Cell classification Wright's stain Pyronin Y - methyl green

 

 

E C.V. i C.V.

(x 104) (96) (x 104) (as)

Agranulocytes 8 62 15 100
Granulocytes 306 58 344 41
Leucocytes 314 57 359 41
Epithelial cells 64 33 63 30
Total cells 378 49 422 36

 

The number of cells determined in smears stained
with pyronin Y - methyl green was considerably larger than
that determined in smears stained with Wright's stain for
agranulocytes (P’5 .13), granulocytes (P’5 .07), leucocytes
(Pg .03), and total cells (P‘3 .03), but not for epi-

thelial cells (P'g'.50). Apparently, the epithelial cells'

18

are equally recognizable in the two stains, a conclusion
which is supported by the similarity of the coefficients
of variation of epithelial cell counts. In contrast,
these data suggest that a considerable portion of the
leucocytes was masked by the background of those smears
prepared with Wright's stain. Since the interaction be-
tween samples and stains did not approach significance

for granulocytes, leucocytes, or total cells, we may con-
clude that where these cells are of primary importance one
may expect to obtain larger cell numbers in most milk
samples when the cells are stained with pyronin Y - methyl

green.

TABLE 4

Level of significance (P ) for each measurable source
of variation for each mil cell classification analysis.

Cell classification

 

 

Source Degrees

of of Agranul- Granul- Leuco- Epithe- Total
variation freedom ocytes ocytes cytes lial cells

cells

Milk sample 26 .01 .01 .01 .01 .01
Stain 1 g .13 ”X .07 91.03 .50 ’=" .03
Interaction 26 .01 .50 .50 :1: .16 .50
Breed smears 54 .01 .01 .01 .01 .01

 

g'Approximately equal to

As expected, milk samples differed significantly
for each of the five cell classifications. However,

duplicate smears also differed significantly (P‘(.01) in

19

each analysis, a result that was not anticipated in view
of the Opinions of Blackburn §E_al. (5), McKenzie (31),
Meigs gg_gl. (32), and Waite & Blackburn (47) that the
"Breed“ smear method gave very consistent results.
Estimated smear and count variances for both
stains are shown in Table 5 for granulocytes and for
total cells. Although, asshown above, significantly more
cells were counted in the pyronin Y - methyl green stain,
this stain resulted in an approximately 25% reduction in
the variance between duplicate smears and in an approx-
imate 50% reduction in the variance between duplicate
counts when compared with Wright's stain. Also, the fact
that the smear variance is at least two to four times
larger than the count variance suggests that some major
errors are inherent in the smear technique used. For in-
stance, the use of a platinum wire lOOp to deliver 0.01 ml
of milk could be a major source of variation contributing

to the large smear variance.

Tmmss

Estimated smear ($2. ) and count (52 ) variances for

oxm csbm
granulocytes and total cells in milk stained with Wright's
stain or pyronin Y - Methyl Green

 
 

.__._.- _, __ _. -_____._ ». -_._._. ————._
_ .- - _

 

-___.__ -,__ _,. 1.“... ___{_3__ a... 2 _-1_, - _- _- __ ,
Variance Wri9ht s stain Pyronin Y Methyl Green

 

 

 

 

 

Component Granulocytes Total cells Granulocytes Total cells
(x 108)

szb‘m 22,702 25,191 16,152 19,580

32c:bm 9,241 9,883 4,003 3,975

 

20

Expected 9 % confidence intervals for mean numbers
of milk somatic cells which are obtained from varying
numbers of smears and counts are tabulated in Table 6.

If one count on the total cells is made on each of two
smears from a milk sample, the eXpected 95% confidence
interval is 3,354 x 104. In other words, in about 95 out
of 100 such observations,/4 will be within the range R :

1,677 x 104. This range is far too large, because most

researChers agree that milk with 50 x 104 to 100 x 104
cells per ml is mastitis positive (22, 36).

Although inspection of Table 6 reveals somewhat
smaller values for the pyronin Y - methyl green stain, the
differences between the stains are small relative to the
magnitude of the confidence intervals. Furthermore,
these data show that the number of smears has more influ-
ence on the magnitudes of the confidence intervals than
does the number of counts. If one were to make an infin—
ate number of counts of total cells on only one smear
with Wright's stain, the 95% confidence interval would be
about 623 x 104, indicating that elimination of the count
variance does not bring the interval close to the import-
ant value of 50 x 104 cells. Consequently, it may be
generalized that one should make upwards of 200 smears
from a milk sample in order to obtain a 95% confidence
interval of about 50 x 104.

In view of this evidence, it appears that the

magnitude of the error variance must be reduced before

21

TABLE 6

Expected 95% confidence intervals for a mean number of
milk cells obtained from differing numbers of smears per
milk sample and counts per smear

 
  

-— ..._._. _ ,-, -7 Wm— _ ‘._.~ -,__1..._“ —-——-.——-—’ ._.__.—. .—

Numbers of Number of wright's stain Pyronin Y ' Methyl

 

 

 

 

 

 

counts smears per Granul- Total Granul-Greegotal
per smear milk sample ocytes cells ocytes cell§__
(x 104)

l 2 3,202 3,354 2,542 2,770
5 444 466 356 384

20 168 176 134 142

200 52 52 40 44

10 2 456 478 380 418
5 278 290 230 254

20 134 142 114 126

200 44 44 36 40

 

estimations of milk cell numbers can be considered to be

a practical indicator of the inflammatory state of the
udder. One possible method of reducing the magnitude of
the error variance would be to deliver the 0.01 ml volume
of milk with a ten-lambda pipette, which would be eXpected

to result in a reduced smear variance.

EXPERIMENT IV. FEULGEN-DNA IN MILK AS A MEASURE OF UDDER

IRRITATION
Review of Literature

Babel (2) and Plastridge (36) state that leucocyte
counts are generally regarded as being superior to other
laboratory tests for diagnosing inflammation of the mammary
gland. However, the previous experiment demonstrated
that present methods for direct microsc0pic estimations
of leucocyte numbers have serious limitations. Also, in
View of the evidence presented in experiment II showing
that DNA was the material reSponsible for the GMT reaction,
it seemed likely that one of the standard DNA tests would
also be useful for the detection of udder irritation.

The Feulgen nuclear reaction is commonly used as
a histochemical test for DNA. The reaction was introduced
by Feulgen and Rossenbeck (14) as a Specific test for
DNA and is dependent on the release of aldehyde groups
from the deoxypentose sugar of DNA by acid hydrolysis
with l N HGl at 60°C. The exposed aldehyde groups then re-
act with the Schiff's reagent to produce a color reaction
at the site of any DNA present. Such a reaction could be
easily standardized and would therefore be more objective
than the GMT.

The purpose of this experiment was to adapt the
Feulgen-DNA test to milk and to determine its relation-

ship to the concentration of somatic cells in milk.

22

23

Materials and Methods

Seventy-five milk samples were obtained from
Holstein cows in various stages of lactation. Duplicate
Feulgen-DNA and duplicate MQT determinations were perform-
ed as an indication of the milk somatic cell concentra-
tion within 30 minutes after obtaining the milk sample.
The MOT was used as previously described. Milk samples
were selected at random until 15 were obtained for each
MOT score. Feulgen-DNA determinations were performed
according to the following procedure:

a) Each milk sample (2 ml) was hydrolyzed with an

equal amount of 1 N BCl at 600 for 24 minutes.

b) Schiff's reagent was then added in an amount

equal to twice the volume of milk. Schiff's
reagent was prepared by dissolving 1.00 g of
basic Fuchsinl in 200 ml boiling distilled
water, cooling to 50° 0. adding 20 ml of 1 N HCl,
colling to 240 C, and dissolving 2.00 g of
sodium metabisulfite. This solution was stor-
ed in the dark for 10-15 hrs and then 0.5 g
activated charcoal was added, the mixture stir-

red for one minute, and rapidly filtered

 

1 National Aniline 0.1. No. 42500.

24

through Whatman No. 1 paper. The filtrate
should be stored in a full, well-stOppered,
dark bottle at 50 C. The solution should be
colorless; deveIOpment of a pink color indi-
cates that it must be discarded.

c) Thirty minutes was allowed for color develop-
ment.

d) The degree of inflammation was scored by
comparing the color intensity of the treated
milk sample with a standard color chart similar

to the one shown in Fig. 1.1
Two smears were prepared from each of the 75 milk
samples by the direct smear method of Prescott and Breed

(37). The previous experiment on the direct microscopic

estimation of number of somatic cells in milk, using a

standard platinum loop in conjunction with the "Breed"

smear, showed that duplicate smears were highly variable.

In an attempt to reduce this error in the present experi-

ment, a lO-lambda pipette was used in place of the plati-

num 100p to transfer the milk to the microsc0pe slides.

Between samples the pipette was washed in detergent,

rinsed with water, acetone, then dried with forced air.

The smears were air-dried on a level surface for 24 hours

and then stained with pyronin Y - methyl green by the pro-

cedure described in experiment III. The concentration

 

1
request.

A standard color chart is available upon

COLOR STANDARD FOR FEULGEN-DNA IN MILK

Department of Dairy
MICHIGAN STATE UNIVERSITY

East Lansing, Michigan

'0
"1 -

i}:\'\/Il

a --.’° :6 :g’" I.‘ 3:",
Jul Jigfitfafi "

r‘V-T u}, — 7, Uri:
‘ , I‘ 4 sxrfi" ’ ' “’, . ’
.'“ ‘.9tt“’$q‘yfi 7‘ ' ‘0 L7 *4
'41 I *4.
.' r14 ' ’In " . .J:.);..:.. 4.._ an

n 1.45"“.

 

 

3
l
l
I

 

 

 

Directions for Use

Cut one half-inch hole in each color black.

Mount chart in clear plastic folder.

(I)

(2)

Move sample behind color blocks to match color.

To score

(3)

Standard color chart for Feulgen-DNA test

Figure 1.

26

of somatic cells per milliliter of each milk sample was
estimated from the number in 25 microsc0pic fields. The
number of agranulocytes, granulocytes, leucocytes (the
sum of agranulocytes and granulocytes), epithelial cells,
and total somatic cells (the sum of leucocytes and epi-
thelial cells) were determined on two smears from each
milk sample by each of two persons who were unaware of

the identity of the smears.
Results and Discussion

The correlation of Feulgen—DNA score with MOT
score was 0.907 (P<(.Ol). The standard deviation of
duplicate Feulgen-DNA scores was 0.114, considerably less
than the 0.230 for duplicate MQT scores. The coefficients
of variation were 7 and 12% for Feulgen-DNA and MQT
scores, reapectively.

The average numbers of cells for each of the five
cell classifications are listed in Table 7 with their
corresponding coefficients of variation. The coefficients
of variation were determined by taking the square root of
the sum of the smear and count variances and dividing by
the mean, for the reasons outlined in experiment III.
These values were generally lower than those in Table 3
from slides prepared by means of a platinum loop, but were
much'higher than the value of 0.114 for duplicate Feulgen-

DNA scores.

27

Each of the estimated variances for duplicate
smears (szb‘m) was negative, and since negative variances
are not possible these values are considered to be esti-
mates of parameters which are very close to or equal to
zero. These results were in direct contrast to the re-
sults of experiment III in which the smear variance
accounted for at least two-thirds of the total within
sample error. The reduced smear variation in the present
experiment was probably due to the use of a lO-lambda
pipette to transfer milk to microsc0pe slides. The esti-

mated variances for duplicate counts (52 ) in the

csbm
present experiment were two to three times larger than

those in experiment III, probably because one person made

TABLE 7

Averages, coefficients of variation, and estimated
smear (Szbzm) and count (s2 ) variances for each milk

   

 

cell clgngfication
Cell classification - CV s2 2
brm cxbm
(x 104) (as) (x 108)
Agranulocytes 24 83 -266 666
Granulocytes 254 34 -544 7,882
Leucocytes 278 33 -l,866 10,434
Epithelial cells 72 40 -79 942

Total cells 350 32 -l,667 13,829

 

28

the duplicate counts in the earlier eXperiment, whereas
in the present experiment, a count was made on each smear
by each of two persons. In other words, count variance
included person to person variance in the present experi-
ment.

The correlations of the estimated number of each
cell type with Feulgen-DNA score and with MOT score are
tabulated in Table 8. Each of these correlations was sig-
nificant (P<(.01) and those for the relationship of
Feulgen-DNA with cell numbers were consistently higher
than those for MOT and cell numbers. However, the high-
est correlation of 0.876 between Feulgen score and total
cells accounts for only 77% of the total variation, indi-
cating that these relationships are low for prediction

purposes.

Correlation coefficients between somatic cell numbers
and Feulgen—DNA or MOT scores

 
    
 

—.——_- "1.. _ -,,_ _ _ _ -______ __ __ __ —— —. . _, —. ‘__.‘. ,___'_ __. — - — —— -- ———

¢ell classification Feulgen-DNA score MOT score

 

Agranulocytes .805 .769
Granulocytes .814 .730
Leucocytes .834 .751
Epithelial cells .816 .790
Total cells .876 .799

 

29

Feulgen-DNA and MOT scores were plotted against
the estimated number of total somatic cells per milliliter
of milk, resulting in the curves shown in Figure 2. Since
these relationships were not linear, an attempt was made
to create a linear relationship by logarithmic transform-
ation of Feulgen-DNA scores and total somatic cell numbers.
It was necessary to code the Feulgen scores by adding the
number one to each score to eliminate the score of zero.
The curve resulting from a double logarithmic transforma-
tion of the codeddata is illustrated in Figure 3.

The relationship between MOT scores and cell
numbers was not transformed because of the subjectivity
of the MOT scores. Although MOT scores are quite repeat-
able within a single person at any given time, they are
much more variable between people because there is no
standard for eadh score. This lack of objectivity limits
the value of MOT in the prediction of cell numbers.

The correlation between logarithms of coded
Feulgen-DNA scores and logarithms of estimated numbers of
total somatic cells was 0.909 (P*<.Ol), a value which,
although higher than that for the non-transformed data,
accounts for only 83% of the total variation. Neverthe-
less, the regression equation shown in Figure 3 should be
useful for estimating total somatic cell numbers from
Feulgen-DNA score. In view of the high variation (C.V.=

32%) associated with microsc0pic estimations of total

30

38008 NBS-IDEA

whoom BOX 09 odd ouoow «zmusomfismm

op wHHoo ofipmsow Haven mo cowpmnpcoocoo no afinmsowpwamm .m onswwm

3.2 can fo. xv 330 2:28

 

 

00m. 000. com
a a i in a _ LII IA L Li a q a 4i 4
— Yln
N T
m” T.
zwomnwm
' v
has.
n I

 

 

 

38008 10w

_ . . . .-~ q
L‘u“°.‘.fl 1 .. -Q " ‘v

 

31

38008 N39303:!

Cs onoon addressee soaoo one .3 55,212

9.3 so C; mmaoo empress mo songs: owmpog mo .Efiwnmloa mo comanmpmox .n whims.

A: 1:2 mud foo. xv mujmo Orr/“Sow do .02 mo<mw>< 5.: “.0 00..
O.m ON 0..

 

 

9. \\..

 

had + XOOM new

 

 

 

Nd

Md

0.0

0.0

Nd

md

(X) 38003 N39103:.I 03000 3H130 90‘l

32

cell numbers, this regression equation may be at least as
accurate in estimating milk cell numbers as micrOSCOpic
estimations in the manner routinely performed in most
laboratories.

The colors in Figure 1 were selected from a large
number of possibilities which ranged from white to the
most dense color shown in Figure l. The selection of the
colors was arbitrary since it was based upon the ability
to distinguish small differences in the intensity of this
color. A preferable procedure, which would eliminate all
subjectivity, would.be to measure the intensity of the
color of the Feulgen-treated milk samples by reflectance
Spectrophotometry. Although such a procedure would be ex-
pected to be more reliable, it would be more costly than
the Feulgen scoring procedure used to obtain the Feulgen-

DNA data in this experiment.

SUMMARY

An Open formula gelation test for mastitis, the
Michigan Mastitis Test (MMT), was develOped which produced
results equivalent to the GMT and MOT. In addition, the
MMT ingredients can be purchased, and the reagent prep-
ared with the aid of a pan balance. The average test
scores for 91 samples of milk tested in duplicate with
GMT, MOT, and MMT were 1.68, 1.76, and 1.81, respectively.
The comparable within sample standard deviations were
0.33, 0.38, and 0.32, respectively. These data indicated
that the results of the three tests are essentially equiv-
alent and that one may choose among them on a basis of
cost, availability, or personal preference.

In an effort to make the EMT open formula test
solution more practical to prepare, a test solution was
prepared using ingredients that are readily available.
This test solution may be prepared with the equipment
available in most households. Although the results of
the use of this test were someWhat more variable than
those for GMT, they demonstrate that one has considerable
flexibility regarding the source and quality or purity of
the ingredients of the test solutions employed in the
mastitis gelation tests.

Since the gelation of the GMT, MOT, and EMT was

reminiscent of the behavior of DEA, DfiAse was added to

33

34

the gels with the result that the gels disappeared, indi-
cating DNA was reSponsible for the gels.'

In the course of a study of milk leucocytes, the
commonly used stains were found to be unsatisfactory be-
cause background staining masked the leucocytes, or because
of a lack of nuclear-cytOplasmic resolution within the
leucocytes. In view of the evidence that DEA was the
material responsible for the gelation in the GMT, MOT, and
MMT, and since leucocytes are the primary source of
nucleic acids in milk, he nucleic acid-specific stain
pyronin Y - methyl green resulted in considerably improved
microsc0pic resolution of milk somatic cells.

To compare the variation of estimated numbers of
somatic cells in milk, four "Breed" smears were made from
each of 27 milk samples, including 14 mastitic samples.
Two smears were stained with Wright's stain and two with
pyronin Y - methyl green stain. The average (x 104)
obtained from duplicate determinations of agranulocytes,
granulocytes, leucocytes, epithelial cells, and total
somatic cells were 8,306, 314, 64, and 378, respectively,
for Wright's stain: and 15, 344, 359, 63, and 422, res-
pectively, for pyronin Y - methyl green stain. Although
the epithelial cells were equally recognizable in the two
stains, a considerable portion of the leucocytes were
masked by the background of the smears prepared with

Ziiright' s stain.

35

The pyronin Y - methyl green stain resulted in
an approximate 25% reduction in smear variance and in an
approximate 50% reduction in count variance when compared
with Wright's stain. Expected 95% confidence intervals
for determinations of mean cell numbers were computed
from the estimated variance components with varying
numbers of smears and counts. These values generally in-
dicated that one should make upwards of 200 smears from a
milk sample to obtain a 95% confidence interval of approx-
imately 50 x 104 cells per ml of milk, a result leading
to the conclusion that the magnitude of the error variance
must be reduced before estimations of milk cell numbers
can be considered to be a practical indicator of the in-
flammatory state of the udder.

In view of the evidence that DNA was the material
responsible for the gelations in the GMT, MOT, and MMT
reactions the Feulgen-DNA test was modified as follows:

A milk sample was hydrolyzed with an equal amount of l N
HGl at 600 G for 24 minutes. Schiff's reagent was then
added in an amount equal to twice the volume of milk.
Maximum color deve10pment occurred within 30 minutes
after the addition of Schiff's reagent. The developed
color was compared with a color chart Which contained
seven colors, corresponding to seven possible scores for
udder inflammation.

To determine the reliability of the Feulgen-DEA

test in estimation of the concentration of milk somatic
cells, duplicate MOT's were made for each of 75 milk
samples. Milk somatic cell concentrations were determined
in duplicate from "Breed“ smears after transferring milk
to each of two micrOSCOpe slides with a lO-lambda pip-
ette and staining the smears with pyronin Y - methyl green
stain.

The correlation of the number of each cell type
with Feulgen-DNA score was consistently higher than with
MOT score, but the highest correlation of 0.876 between
Feulgen-DNA score and total somatic cell numbers account-
ed for only 77% of the total variation. Plotting the ob-
served values revealed a curvilinear relationship, Which
became linear when both variables were transformed to
logarithms. The correlation of the logarithm of the
coded Feulgen-DNA scores with numbers of total somatic
cells was 0.909, including that Feulgen-DNA has practical
value in the estimation of somatic cell numbers in milk.

The coefficient of variation for Feulgen-DNA
scores was 7%, mudh less than the 32% for the microscopic
estimations of somatic cell concentration, indicating
that the Feulgen-DNA socres may be at least as accurate

in estimating milk cell numbers as microscopic estimations.

(l)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

LITERATURE CITED

Albright, J. L. Herd Management in Los Angeles
County Commercial Dairies. J. Dairy Sci., 43:1696.
1960.

Babel, F. J. Problems of the Dairy Bacteriologist
Resulting from Mastitis. J. Am. Vet. Med. Assoc.
133:161. 1958.

Barnum, D. A. A Method for the Control of Staphlycoccal
and Streptococcal Mastitis in Ontario. Can. Vet. J.
3:161. 1962.

Barnum, D. A., and Newbould, F. H. s. The Use of
the California Mastitis Test for the Detection of
Bovine Mastitis. Can. Vet. J., 2:83. 1961.

Blackburn, P. 8., Laing, G. M., and Malcolm, D. F.
A Comparison of the Diagnostic Value of the Total
and Differential Cell Counts of Bovine Milk. J.
Dairy Researdh, 22:37. 1955.

Brachet, J. c. R. Soc. Biol., Paris, 133:88. 1940.
(Cited by Pearse, A. G. E., Histochemistry Theo-
retical and Applied, 2nd ed., p. 205, 1960. Little,
Brown & Co., Boston, 1961).

Brachet, J. La Localisation des Acides Pentosenucl-

eiques dans les Tissus Animaux et les Oeufs d'Amphib-
iens en Voie de Developpement. Arch. 8101., 53:207.

1942.

Brachet, J. Chemical Embryology, 2nd ed., Inter-
science Publishers, Inc., New York, 1950.

Broadhurst, J., and Paley C. A Single-Dip Stain for
the Direct Examination of Milk. J. Am. Vet. Med.
Assoc., 94:525. 1939.

Carroll, E. J., and Schalm, O. W. Effect of Deoxy-
ribonuclease on the California Test for Mastitis, J.
Dairy Sci., 45:1094. 1962.

Charlett, S. M. An Improved Staining Method for the
Direct Microsc0pical Counting of Bacteria in Milk.
Dairy Industry, 19:652. 1954.

Chu, S. J. Bovine Mastitis: A Comparison of the
Value of Diagnostic Methods. J. Comp. Pathol. Therap.,
59:81. 1949.

37

(l3)

(14)

(15)

(l6)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

Cole, E. J., and Easterbrooks, H. L. LeukOpenia
Associated with Acute Bovine Mastitis. J. Am. Vet.
Med. Assoc., 133:163. 1958.

Feulgen, R., and Rossenbeck. H. MikroskOpish-
Chemischer Nachweis einer Nucleinsaure vom Typus der
Thymonucleinsaure und die darauf beruhrende elektive
Farbung von Zellkernen in mikroskopischen Praparaten,
HOppe-Seyler's Ztschr. f. Physiol. Chemie., 135:203.
1924.

Gray, D. M., and Schalm, O. W. Interpretation of the
California Mastitis Test Results on Milk from Indi-
vidual Mammary Quarters, Bucket Milk and Bulk Herd
Milk. J. Am. Vet. Med. Assoc., 136:195. 1960.

Guallini, L., and Leali, L. Differential Cell Pic-
ture of Milk from Cows with Mastitis. Part II.
Arch. Vet. Italiano, 10:1. 1959.

Guallini, L., and Vallis, P. Differential Cell
Picture of Milk from Cows with Mastitis. Part I.
Arch. Vet. Italiano, 8:575. 1959.

Holt, D. Mastitis Can Be Licked Now. Hoards Dairy-
man. March 10, p. 231. 1959.

Hughes, D. L. Some Reflections on the Mastitis
Problem. Vet. Record, 66:235. 1954.

Jackson, R. A. A Practical Mastitis Control Program.
Vet. Med., 56:143. 1961.

Jensen, P. T. Investigations into the Whiteside-
Test and the C.M.T.-Test for the Detection of Patho-
logical Secretions in Herd Milk samples. Nord.

Vet. Med., 9:590. 1957.

Johns, C. K. Applications and Limitations of Quality
Tests for Milk and Milk Products. A Review. J.
Dairy Sci., 42:1625. 1959.

Kaufmann, B. P., McDonald, M., and Gay'H. Enzymatic
Degradation of Ribonucleoproteins of Chromosomes,
Nucleoli and CytOplasm. Nature, 162:814. 1948.

Leidl, W., und Schalm, O. W. Die Anwendung des
SchalmeMastitis-Testes (SMT) bei Sammelmilch zur
Bestimmung minderwertiger Milch, bedingt durch hohen
Zellgehalt. Tierarztl. Umschau., 16:219. 1961.

(25)

(26)

(27)

(28)

(29)

(30)

(31)

(32)

(33)

(34)

(35)

(36)

39

Leidl, W., Schalm, O. W., Krieger, A., und Lagemann,
H. Vergleich des Schalm-Mastitis-Testes mit dem
Zellgehalt, det Katalase-Thybromol-Probe und dem
Chloregehalt in der Milch. Tierarztl. Wochen-
schrift, 19:382. 1961.

Levowitz,
tion"

D., Weber, M. An Effective "Single Solu-
Stain. J. Milk and Food Tech., 19:121. 1956.

Macadam, I. The Pathology and Bacteriology of
Bovine Mastitis in Relation to Cell Counts. J.
Comp. Pathol. Therap. 68:106., 1958.

McFarlane, D., Blackburn, P. 8., Malcolm, J. F., and
Wilson, A. L. A Comparison of AnteeMortem and Post-
Mortem Findings in Bovine Mastitis. Vet. Record,
61:807. 1949.

Marshall, R. T., and Edmondson, J. E. Value of-
California Mastitis Test Records to the Practitioner.
J. Am. Vet. Med. Assoc.,,l40:45. 1962.

McKay, K. G. The California Mastitis Testing (CMT)
Program. J. Dairy Sci., 43:896. 1960.

McKenzie, D. A. A review of Cell Count and Solids-
Not-Fat Content of Milk. Dairy Industries, 26:184.
1961.

Meigs, E. B., Burkey, L. A., Sanders, G. P., Rogosa,
M., and Converse, H. T. The Relationship of Machine
Milking to the Incidence and Severity of Mastitis.
USDA, Tech. Bull. 992. 1949.

Newman, R. W. Calif. Dept. of Agr. Monthly Bu11.,
Nov. 16, 1927. (Cited by Little and Plastridge.
Bovine Mastitis, lst ed., p. 515. McGrawéHill Book
Co., New York. 1946.

Noorlander, D. 0.
Compton Press, Inc.

Milking Machines and Mastitis.
Chino, California. 1960.

Pattison, I. H., Taylor, J. 1., and Holman, H. H.
Studies on EXperimental Streptococcal Mastitis. I.
Innoculation of Large Numbers of Streptococcus
agarlatiae into the Teat Canal of Goats. J. Comp.
Pathol. Therap., 60:71. 1950.

Plastridge, W. N. Bovine Mastitis: A Review. J.

Dairy Sci., 41:1141. 1958.

(37)

(38)

(39)

(40)

(41)

(42)

(43)

(44)

(45)

(46)

(47)

(48)

4O

Prescott, S. C., and Breed, R. S. The Determination
of the Number of Body Cells in Milk by a Direct
Method. J. Infect. Diseases, 7:632. 1910.

Schalm, O. W. Udder Health Gain in Mastitis Control
Held 'Spectacular'. As quoted by Veterinary Dis-
patdh., 3:1. 1961-62.

Schalm, 0. W. Bovine Mastitis and a Program for Its
Control in California. Can. Vet. Jour., 3:90. 1962.

Schalm, 0. W. Leukocyte ReSponses to Diseases in
Various Domestic Animals. J. Am. Vet. Med. Assoc.,
140:557. 1962.

Schalm, O. W. A Syllabus on the Mammary Glands in
Health and Disease. University of Calif. Student
Book Store. Davis, California. 1959.

Schalm, 0. W., and Noorlander, D. 0. Experiments and
Observations Leading to Development of the California
Mastitis Test. J. Am. Vet. Med. Assoc., 130:199.
1957.

Steere, J. H. Mastitis Control in Denmark. Mod.
Vet. Practice, No. 5, 41:30. 1960.

Temple, H. C., and Haller, C. J. Quality Tests on
Bulk Milk to Determine Presence of Mastitis Secretion.
Milk Sanitarians Annual Report. 1960.

Theilen, G. H., SChalm, 0. W., Straub, O. C., and
Hughes, J. P. Bovine Hematology. I. Leukocyte
ReSponse to Acute Bovine Mastitis. J. Am. Vet. Med.
Assoc., 135:481. 1959.

Varrier-Jones, P. C. The Cellular Content of Milk:
Variations Met with under Physiological and Patho-
logical Conditions. Lancet, 207:537. 1924.

Waite, R., and Blackburn, P. S. The Chemical Comp-
osition and the Cell Count of Milk. J. Dairy
Research, 24:328. 1957.

Wright, J. H. A Rapid Method for Differential
Staining of Blood Films and Malarial Parasites. J.
Med. Research, 7:138. 1902.

APPENDIX

Procedure I: Staining with Wright's stain.

a) Defat and mix the milk smear in acetone-free
methyl alcohol for 2-3 minutes.

b) Place in Wright's stain for 3-5 minutes.

c) Slides are removed from the stain directly into
phOSphate buffer to which has been added 1 part
of Wright's stain solute to 10 parts of buffer.
After 3-4 minutes in the buffer the slides are
briefly rinsed in clear water and allowed to dry
without blotting, at room temperature. The
phOSphate buffer is prepared by diluting 3.8 g

NaZHPO P0

and 5.4 g of KH to 1,000 m1 of dis-

4' 4

tilled water.

2

Procedure II: Method Used in Finding the Area of a
Microsc0pic Field.

By using a stage micrometer the diameter of a field
was found to be 17.8 mm (0.0178 cm). Area equals 3.1416 x
(0.0089)2 = 0.00025 sq. cm or 1/4000 sq. cm.

The milk volume represented by each field equals
1/4000 x 1/100 = 1/400,000 m1. Each cell seen in a field
taken at random equals 400,000 cells/m1 of milk. In
examining 25 fields, each cell would then represent

16,000 cells.

41

Hllilllllllll

 

'
I
I
l
I
ll!
'-

1"
Ill!
0‘
l
l