The Pathogenic Significance of Various Intramammary Infections

BT. vet.

J.

(1970), 126, 260

THE PATHOGENIC SIGNIFICANCE OF VARIOUS INTRAMAMMARY INFECTIONS By D.

FORBES

Central Veterinary Laboratory, Weybridge, Surrey

SUMMARY

The presence of intramammary infections in all quarters of eighleen cows was determined by collecting milk samples by syringe from the teat sinus through the teat wall. The samples were taken within I week before calving, 8 weeks after calving and thereafter at monthly intervals throughout lactation. Cell counts of fore-milk taken each week were made with an electronic counter. All intramammary infections are reported and the type related to the mean cell count levels. C. bovis infected all but one quarter, often concurrently with other infections. C. bovis and Staph. epidermidis caused significantly higher cell counts than those in uninfected quarters but did not cause clinical mastitis which was only associated with Staph. aur-eus and Str. !tberis. The role of high cell counts in quarters infected with relatively non-pathogenic bacteria IS disc Llssed in relation to the susceptibili ty to other pathogenic infections. INTRODUCTION

It has been shown by Schalm and his co-workers (summarized below) that a pre-existing leucocytosis in milk within the mammary gland protects the gland from experimentally induced infections with pathogenic bacteria. In their experiments a leucocytosis was induced by intramammary inoculations with saline or other solutions injurious to the milk secreting tissue; but Schalm, Carrol & Lasmanis (1964) reported that a mild inflammatory reaction due to infection with Streptococcus uberis inhibited multiplication of 500,000 Aerobacter aerogenes introduced into the gland through the teat canal. In a later paper Schalm et at. (1967) found that when Streptococcus agalactiae was inoculated into mammary glands multiplication of the bacteria was sometimes delayed and this was not continuously associated with the infiltration of large numbers of neutrophilleucocytes. They presented a hypothesis that crossreacting streptococcal antibodies may exist and suppress Str. agalactiae in udders already infected with other streptococcal species. Jain, Lasmanis & Schalm (1967) showed that an induced leucopenia allowed A. aerogenes and a coagulasenegative staphylococcus to multiply rapidly. The latter bacterium then caused a n acute mastitis within 24 hours, which was controlled by a subsequent massive infiltration of neutrophils. Edwards & Jones (1966) reported that the number of white cells in the fore-milk of cows which contained coagulasenegative staphylococci did not exceed I X 10 6 cells/m!. Nevertheless these quarters were not easily infected with Staphylococcus aureus even though the

INTRAMAMMARY INFECTIONS IN COWS

bacterium was occasionally isolated from their milk. The authors thought that the coagulase-negative bacteria populated the teat duct, an environment considered more favourable to these types than to Staphylococcus aureus, and that the leucocyte level was not a factor in preventing the latter from establishing intramammary infections. However, Schalm et al. (1966) showed that 200,000 to 500,000 cells/ml. protected quarters from experimental streptococcal inoculations, a level of cell count found by Forbes & H ebert (1968) to be associated with intramammary infections of coagulase-negative staphylococci. In investigating the mechanism by which bacteria invade the udder parenchyma, it is therefore important to know the effect non-pathogenic bacteria have on the cell count of milk. This paper reports the cell count levels of foremilk associated with naturally occurring intramammary infections during one lactation of an experimental dairy herd. MATERIALS AND METHODS

Nine pairs of monozygous twin cows were used in this work. All but two cows calved within 6 weeks of each other. The others, one each of a different pair of twins calved 2 months later. Their management is described elsewhere (Forbes, 1970 ). Milk samples. The teats were thoroughly cleaned with cotton wool soaked in methylated spirits and then 10 ml. of fore-milk were collected. These samples were collected weekly and bacteriologically examined. The number of cells per ml. was determined using the electronic counting method of Cullen (1967). Intramammary injections. Milk samples were collected by syringe through the teat wall using the method of Murphy & Stuart (1954). These samples were collected when possible during the week before calving, 8 weeks after calving and thereafter at monthly intervals. Some colostral samples were too viscous for collection by syringe and later samples were occasionally missed because of difficulty in collection. Bacteriological examination. o· I ml. of milk collected by syringe was spread on each of five plates of 5 per cent sheep blood agar containing aesculin ( 1 / I 0,000), Fungazone* ( 10 (Jog.fm!.) and Actidionet ( 1 / 10,000). All plates were incubated at 30 °c for 48 hours and refrigerated before being examined. Micrococcaceae present on the plates were recognized by colonial characters and sub-cultured onto Tryptone yeast extract agar to obtain pure cultures. Coagulase-positive strains (Staph. aureus) were differentiated from coagulasenegative strains by the tube test using a one in five dilution of rabbit plasma. Corynebacterium bovis was recognized by colonial and morphological characters. Anarysis of cell counts. Milk cell counts from quarters with the same intramammary infections and at the same period oflactation were grouped together. Values were transformed to logarithms, and the geometric means determined. Because of varying lengths of lactations in different cows and because of variations in cell counts during lactation (Blackburn, 1966; Forbes & Hebert, 1968), • E. R. Squibb, New York. Upjohn, Kalamazoo.

t

BRITISH VETERINARY JOURNAL, 126,5

the cell counts from the first three and last three months of lactations were grouped together for comparison leaving an intermediate period of varying duration. The week numbers of the lactation are numbered progressively forward to the intermediate period and progressively backward from the end of lactation after the intermediate period. The latter week numbers are denoted by a minus sign before the number. RESULTS

The incidence of intra mammary irifections Table I shows the type of bacteria isolated in milk taken by syringe through the teat wall and consequently of intramammary origin. Only one quarter remained completely uninfected throughout the lactation. C. bovis was isolated from teat sinus milk in every other quarter in the herd at some time during the lactation, often concurrently with another bacterium. The incidence of Staph. aureus and coagulase-negative coccal infections was low. Streptococcal infections are not included in the table. Three Staph. aureus and one Str. uberis infections, present at the time of calving, caused clinical mastitis soon after calving and these were treated by intramammary antibiotics. Two further clinical mastitis cases associated with Staph. aureus were treated in the third month of lactation and one in the fifth month. One cows' quarter contracted an intramammary infection of Str. t[ysgalactiae near the end oflactation and this, although not causing clinical mastitis was eliminated by treatment. There were no other intramammary streptococcal infections in the herd during the lactation and no other types of infection required treatment. It can be seen from Table I that the predominant bacterium was C. bovis which infected most quarters in the first half of the lactation, whether or not the quarter had an existing intramammary infection. Two quarters which were infected with Staph. epidermidis originally, became infected with C. bovis and the staphylococcus then disappeared. A similar event occurred in one quarter originally infected with Staph. aureus. In other quarters the corynebacterium established concurrently with existing staphylococcal infections, in some both infections persisted and in others the staphylococci later disappeared. In two quarters originally infected with C. bovis and later with Staph. aureus the corynebacteria subsequently disappeared. In only these and another two quarters in which C. bovis disappeared leaving the gland uninfected, did C. bovis once present as an intramammary infection subsequently disappear. In all other quarters it persisted solely or in association with staphylococcal infections. Thus it appears that not only is C. bovis an extremely invasive organism but that once established within the udder parenchyma it is able to persist and is possibly associated with the elimination of existing infections. The pathogenicity of the intramammary infections as determined by cell counts Table II records the geometric mean cell counts of uninfected quarters up to the time at which they first became infected and also quarters which were

.....

II ~

TABLE I THE NUMBER OF Q.UARTERS AFFECTED, AND THE TYPE

OF INFECTION

PRESENT IN DIFFERENT PERIODS DURING

LACTATION

~

Period of lactation (week numbers )

Type of bacteria present as intramammary infections

2-8

9- 12

13- 16

17-20

12 6 3 2

10

10

0

2 3 2 3 51 0 72 0 0 72

2 2 4 2 8 49

None S. aureus S. aureus and C. bovis S. epidermidis S. epidermidis and C. bovis

34 6 0 5

C. bovis S. aureus and S. epidermidis

10

30

40

3 59 13 0 72

0

0

58 14 0 72

58 14 0 72

Total No. of quarters sampled No. of quarters not sampled No. of dry quarters Total No. of quarters

6

5

0

67 4 72

21-24

25-28

29-3 2

II ;J>~

::0

><

0

3 2 3 41 0 51 17 4 72

2

2 22 0 30 2

2

40 72

48 72

5

Z

>Tj

0 2 [6 0 22

0

...... tTl

[)

>-l ......

0 Z

rFJ

......

Z

[)

0

~

Vl

IV

01

c..o

}J

O'l

"'"

II

TABLE II THE CHANGES IN CELL COUNTS DURING LACTATION ASSOCIATED WITH INFECTION

Mean cell counts per mi. throughout lactation Period of lactation (week numbers)

Uninfected Log mean

±

SE

S. aureus

C. bovis

Geometric mean

±

SE

Geometric mean

Log mean

X 10- 3

Log mean

±

to

S. epidermidis

Geometric

Log mean

±

::0 ....

Geometric meal!

SE

mean

SE

2'755 ± 0 01026

568 (20)

3° 06 7 ± 0 01801

116 7 (7)

2 °7 16 ± 0°079°

520

3° 152 0°7'7

±

14 18 (32)

3° 189 ± 0° '426

' 546

1-4

2'347 ± 0°07 16

222 ( 18)':'

5-8

2 0290 ± 0°04 14

195 (51 )

2 0622 ± 0 00543

4 18 (63)

9-12

2 °4 22 ± 0°0638

26 4 (22 )

2 0612 ± 000284

4 09 ( 150)

3°211 ± 0' 1047

162 7 (24)

3° 047 0° ' 49

±

Intermediate

25 6 2 ± 0°0590

36 5 (20)

2 °9 10 ± 0 00181

81 3 (43 2)

3° 176 ± 0°05 66

1500 (56)

2 °9 13

±

- I I to -8

2 0602 ± 0°08 19

400 (4)

3°022 ± 0°0220

1053 ( 19 1)

3° 19 1 ± 0°0842

' 553 (24)

0°05 24

-7 to -4

2808 ± 0"5 20

642 (4)

3° 0 76 ± 0' 02 4 2

119 1 (224)

3°18 7 ± 0°0788

'537 (24)

3°068 ± 0 °08 70

3° 080 ± 0 °1359

1203 (4)

3° 17' ± 0' 026 4

1484 (230)

3° 2 79 ± 0006 99

1899 (20)

3° 1962 0 010 ' 9

...,

.....

en

::r:

<:

...,t'1 t'1

'11 5

818

::0 .... Z

~

><

c..... 0 C

::0

~

t' ~

-3 to end

* The numbers of values are given in parentheses.

}J

±

117 0

~O'l

U1

145 1

INTRA MAMMARY INFECTIONS IN COWS

infected with C. bovis and those infected with Staph. aureus. Because of the small numbers of quarters infected with Staph. au reus this group includes quarters which had a concurrent infection with C. bovis during all or part of the infected period. Uninfected quarters. As only one quarter remained uninfected throughout the lactation the number of values for cell counts for the end of the lactation relate to this one quarter. The mean counts rose steadily throughout the lactation. The rise which was apparent during the first 8 weeks continued and became significantly higher by the intermediate period (P < 0'05). The further rise after this period again became significant by the end of the lactation (P < 0'05). Only in the last 2 months of the lactation did the mean values exceed 50 x 10 3 cells per ml. Q,uarters infected with C. bovis. The mean count for the intermediate period was significantly higher than those for the first 12 weeks (P < 0'01). A further significant rise occurred during weeks - I I to - 8, and also from weeks - I to -4 to the last period of the lactation (P < 0'01). Q,uarters irifected with Staph. aureus. The cell counts in this group were high throughout the lactation and no significant changes occurred throughout the period. All the mean counts were above one million cells per ml.

Comparison of cell counts in relation to the type of irifection present The mean levels of cell counts in the quarters infected with C. bovis or Staph. aureus were both significantly higher than the level in uninfected quarters up to the weeks - I I to - 8 (P < 0'0 I ). After this the cell levels of infected quarters were higher but not significantly so. The mean cell count from quarters infected with Staph. au reus was significantly higher than quarters infected with C. bovis up to and including the intermediate period. Thereafter the levels remained higher but were not significantly so although it should be noted that the mean levels from quarters infected with C. bovis exceeded one million cells per ml. throughout this period. Staph. epidermidis irifections. Because of the small numbers of quarters infected with Staph. epidermidis and the fact that some followed previous infections which had necessitated treatment, a full progression of cell counts throughout lactation was not possible. The available figures are shown in Table II. The levels are similar to those from quarters infected with C. bovis. DISCUSSION

The main fact that emerges from this study of intramammary infections is the invasiveness ofC. bovis. It is then associated with mean cell counts higher than the mean counts from uninfected quarters. This work was done on the same herd as a previous report (Forbes & Hebert, 1968) in which staphylococci were the main intramammary bacteria, the only difference being that four cows had been culled. C. bovis was introduced inadvertently into the herd between these lactations by another cow. Although she was also culled before the start of the lactation reported here and despite reasonable hygienic pre-

BRITISH VETERINARY JOURNAL, 126, 5

cautions during milking the organism quickly spread throughout the herd. The marked difference in the incidence of staphylococcal infections between this and the previous lactation is probably associated with the introduction of C. bovis. This could not be proved experimentally as possible control animals quickly acquired the corynebacterial infection but the incidence of staphylococcal teat canal infections throughout the lactation period described here was high (Forbes, 1970). Presumably the staphylococci invaded the gland but could not establish as intramammary infections due to the presence of C. bovis. Newbould & Neave ( I965) showed that Staph. aureus when inoculated into glands infected with corynebacterium was not recovered 10 minutes later with the frequency that it was from uninfected glands. They attributed this to a higher number of leucocytes in the milk rather than to the presence of the corynebacterium itself. In the work reported here the mean cell levels in quarters infected with C. bovis were higher than in uninfected quarters and it is probable that the leucocytes were responsible for the failure of staphylococcal organisms to establish in the udder parenchyma. In addition the disappearance of existing staphylococcal infections was often associated with a new infection of corynebacteria and it is possible that the inflammatory response of the gland alters in some way to cause expulsion of staphylococci. It is of interest that those types of Staph. epidermidis infections which did occur all provoked an obviously high cell count and there was not any infection by a non-pathogenic type described earlier (Forbes & Hebert, 1968). This is probably due to such types having a greater susceptibility to the presence of leucocytes than pathogenic types. The only bacteria to cause clinical mastitis in this work were recognized pathogens, i.e. Staph. aureus and Str. uberis but most quarters had intramammary infections with other bacteria. This would seem to confirm that the facility with which bacteria invade the mammary gland is unrelated to their pathogenicity within it. The passage of bacteria through the teat canal is not fully understood (Forbes, 1968), but it is probably dependent upon the weight and type of infection on the teat skin and in the teat canal. The subsequent establishment of infection within the gland is further dependent upon the inherent pathogenicity of the organism and the susceptibility of the host. This in turn may be influenced by the numbers of leucocytes in the milk as a result of an existing or recent intramammary infection. In investigating the mechanisms by which pathogens colonize the udder parenchyma the role of "commensal" bacteria cannot be ignored. REFERENCES BLACKBURN, P. S. (1966).J. Dairy Res., 33,1 93. CULLEN, G. A. (1967). Vet. Rec., 80, 188. EDWARDS, S. J . & JONES, G. W. (1966) . Dairy Res., 33, 261. FORBES, D. (1968).J. Dairy Res., 35, 399. FORBES, D. (1970). To be published. FORBES, D. & HEBERT (1968). Vet. Rec., 82, 69. JAIN, N. C., LAsMANIS, J . & SCHALM, 0. W. (1967). Am. vet. MURPHy,J. M. & STUART, 0. W. (1954) , Cornell Vet., 440 501.

J.

J.

Res., 28, 1243.

INTRAMAMMARY INFECTIONS IN COWS NEWBOULD, SCHALM, O. SCHALM, O. SCHALM, O.

F. H. S . & NEAVE, F. K. (1965) .]. Dairy R es., 3 2, 157. W., CARROL, E.]. & LASMANIS, J. (1964). Am.]. vet. Res., 25, 104. W., LASMANIS,]. & CARROL, E.]. (1966). Am.]. vet. Res., 27, 1537. W., LASMANIS,]. & CARROL, E.]. (1967). Am.]. vet. Res., 2 8 , 685. (Accepted for publication 28 November 1969)

La signification p athogenique de diverses infections intraIIlaIIlellaires (Forbes) R esUDle. On a determine la presence d'infections intramamallaires dans tous les quartiers de dix-huit vaches en prelevant a la seringue des echantillons de lait dans Ie sinus de la tetine a travers la paroi de la tetine. Les prelevements etaient effectues au cours de la semaine precedant l'accouchement, 8 semaines apres l'accouchement, et ensuite a un mois d'intervalle pendant toute la lactation. Les numerations cellulaires du lait du debut de la traite preleve chaque semaine etaient faites par un compteur electronique. Toutes les infections intramamellaires sont indiquees et Ie type est rapporte aux taux moyens des numerations cellulaires. C. bovis infectait tous les quartiers, souvent concurremment avec d'autres infections. C. bovis et Staph. epidermidis provoquaient des numerations cellulaires plus elevees que dans les quartiers non infectes, mais sans causer de mastite clinique; celle-ci etait associee seulement a la presence de Staph. aureus et Str. uberis. On discute la signification des numerations cellulaires elevees dans les quartiers infectes par des bacteries relativement nonpathogenes par rapport a la susceptibilite a d'autres infections pathogenes. Pathogene BedeutsaDlkeit verschiedener intraIIlaIIlIIlar er Infektionen (F orbes) Zusaullnenfassung. Die in allen vier Quadranten vorhandene Infektion in achtzehn Klihen wurde durch Milchproben bestimmt, die mit einer Spritze von dem Zitzenkanal durch die Zitzenwand entnommen wurden. Die Proben wurden in der Woche vor dem Kalben, 8 Wochen nach dem Kalben und danach in monatlichen Zeitabstanden wah rend der Melkzeit gemacht. Die Zellzahlung der in jeder Woche entnommenen Vormilch wurde mit einem elektronischen Zahler durchgefiihrt. Es wird liber aIle intramammaren Infektionen berichtet, und eine Beziehung zwischen der Art der Infektion und der durchschnittlichen Zellzahlung hergestellt. Abgesehen von einer Zitze infizierte C. bovis alle anderen, mehrfach gleichzeitig mit anderen Infektionen. C. bovis und Staph. epidermidis riefen betrachtlich h6here Zellzahlungen hervor als in den gesunden Zitzen gefunden wurden, jedoch klinische Mastitis konnte nur in Verbindung mit Staph. aureus und Str. uberis festgestellt werden. Man berichtet liber die Rolle der hohen Zellzahlungen in Quadranten, die mit Bakterien von verhaltnismassig schwacher pathogener Wirkung infiziert wurden, sowie der Beziehung zu der Empfanglichkeit flir andere pathogene Infektionen. Signific ado patogenico de varias infecciones intraIIla lDarias (Forbes) ResUDlen. Se determin6 la presencia de infecciones intramamarias en todos los cuartos de dieciocho vacas compilando muestras de leche del seno de la ubre sacandolas con ayuda de una jer inguilla a traves de las paredes de la ubre. Las muestras se obtuvieron dentro de la semana anterior al parto, 8 semanas despu es del par to y desde entonces a intervalos mensuales dura n te toda la lactaci6n. Los calculos celulares del calestro que, se efectuaron todas las semanas, se hicieron con una computadora electr6nica. Se presenta un informe sobre todas las infecciones intramamarias asi como el tipo relacionado a los niveles medios del calculo de celulas. C. bovis infect6 a todos excepto un cuarto, a menudo concurrentemente con otras infecciones. C. bovis y Staph. epidermidis causaron calculos de celulas significativamente mas elevados que los obtenidos en cuartos no infectados, pero no causar on mastitis clinica q u e estuvo asociada solamente con Staph. aureus y Sir. uberis. Se discu te el papel de calculos de celulas elevados en cuartos infectados con bacteria relativamente no patogenica, en relaci6n can la susceptibilidad a otras infecciones patogenicas.