Staphylococcus aureus Mastitis: Pathogenesis and Treatment with Bovine Interleukin-1β and Interleukin-2

SfaphyrOcoccus aureus Mastitis: Pathogenesis and Treatment with Bovine Interleukin-1f3 and Interleukin-2 MICHAEL J. DALEY,’~ PATRICIA A. COYLE THOMAS J. WILLIAMS,’ GARY FURDA,’ RUTH DOUGHERTY! and PHILLIP W. HAYES2 Agricultural Research Division American Cyanamid Co. Princeton, NJ 08543WX

immunomodulators that augment natural defensive mechanisms similar to the normal response to pathogens and may prove to be suitable alternatives to, or may be used in combination with, antibiotics as effective mastitis therapeutic agents. (Key words: cytokines, therapy, mastitis)

ABSTRACT

Polymorphonuclear leukocytes play a central role in the pathogenesis of bovine mastitis. Intramammary challenge with Staphylococcus aureus was shown to induce both quantitative and qualitative changes in mammary gland polymorphonuclear leukocytes. Intramammary infusion of recombinant bovine interleukin18 and interleukin-2 elicited a similar cellular response. Staphylococcus aureus, interleukin-lp, and interleukin-2 all increased the number of somatic cells after intramammary infusion and activated the inducible superoxide production in milk polymorphonuclear leukocytes. Interleukin-2 also activated phagocytosis of these cells, and their activation was maintained for 3 to 5 d after intramammary administration. Inter- leukin-lp and interleukin-2 were moderately effective in the therapy of experimental S. aureus mastitis. Approximately 54% of the glands treated with interleukin-lp responded to therapy by transiently clearing the milk of s. aureus, 30% of which relapsed, and a total of 38% of the treated glands remained cured. In contrast, 83% of glands treated with interleukin-lp responded to therapy, but 50% of these quarters relapsed. A total of 42% of the quarters treated with interleukin-lp remained cured. Homologous recombinant cytokines are effective

Abbreviation key: BRM = biological K sponse modifiers, IL = interleukin. 0;= superoxide anion, PBS = phosphate-buffered saline, PMA = phorbol 12-myristate 13-acetate, PMN = polymorphonuclear cell, rb = recombinant bovine. INTRODUCTION

Received August 30, 1990. Accepted June 28, 1991. ‘Immunology Group. 2Biochemistry and Physiology Group. 3Correspondcnce: Michael Daley, Immunology Group, Agricultural Research Division, American Cyammid Co., PO Box 400, Princeton, NJ 08543-0400. 1991 J Dairy Sci 74:44134424

Subclinical and clinical mastitis is a major cause of loss in milk production to the dairy industry; economic losses are estimated at 2 to 4 billion dollars annually in the US (4). One of the dominant etiological pathogens contributing to this economic loss is Staphylococcus aureus (23). Although antibiotic therapy can be effective in the control of mastitis from a variety organisms, it is only moderately efficacious for S. aureus infections and requires a 3- to 5d milk withdrawal (2, 5). This milk discard is the major m e of economic loss to the dairy farmer and is a cause of some resistance to therapy in lactating cows, especially in light of the often poor response to therapy. Improved mastitis therapeutics remain a major unfilled need of the daiq industry. Staphylococcus aureus mastitis infection is characterized by viable bacteria in the milk and an elevated number of somatic cells in the mammary tissue and milk (1, 7, 9, 14). In an infected gland, these host cells are >95% polymorphonuclear leukocytes (PMN). The development of chronic nature of s. aureus

4413

4414

DALEY ET AL.

mastitis depends on the interaction between invading bacteria and these cells of the defense system (10, 28). Although the defense system of the mammary gland can be effective in eliminating as much as 3 to 4 logs of bacteria, it has been suggested that the less activated PMN may in fact contribute to the relapse of an existing infection and chronic cyclic nature of S. aureus mastitis (9). Furthermore, once organisms have been phagocytosed, they may be protected from the effects of antibiotic therapy (7, 19. 34). The defense mechanisms of the host can be stimulated by pathogens and their by-products (16) and various biological response modifiers (BRM) (38), including cytokines (26, 35). Cytokines regulate the activity of the host’s defense system in response to pathogens. For example, interleukin (1L)-1, although pleiotropic in many of its effects, can activate B cells, T cells, macrophages, and other primary and secondary cells of the defense system (27, 36). Interleukin-2 is a lymphocytederived glycoprotein that primarily induces the clonal expansion of activated T and B cells (15), but, in so doing, it can also initiate the secondary secretion of a variety of other cytokines by the activated lymphocytes. Both bovine IL-1 and L 2 have recently been cloned and expressed (3, 20) and have been shown to potentiate the biological activity of vaccines as well as a natural killer-like activity (32, 33). Given that normal regulation of PMN in response to infections can be mediated by cytokines (8, 11, 12, 13), the availability of recombinant bovine (rb) L - l p and rbIL-2 allows for the fist time an evaluation of the effects of exogenous administration of cytokines on the potentiation of normal defense mechanisms of the mammary gland. Enhancement of natural defense mechanisms might minimize or eliminate the need for antibiotics in the treatment of mastitis infections. However, a key to utilizing these recombinant cytokines optimally for the prevention and therapy of bovine mastitis (because these cytokines are part of a highly integrated regulatory network), lies in a thorough understanding of their role in the normal host response to a particular pathogen. The following Journal of Dairy Science Vol. 74, No. 12, 1991

study was initiated to identlfy the biological response of the mammary gland to the initiation of S. aureus infection and then to attempt to mimic or potentiate this response through the administration of exogenous cytokines. MATERIALS AND METHODS

Animals and Experimental Model

Three normal uninfected mammary quarters from three separate dairy cows were used to examine the normal host response to challenge with 100 to 200 cfu of S. aureus. At 16, 40, 64,112, and 160 h after intramammary infusion, the number of somatic cells and the ability of these cells to produce superoxide (03after induction with phorbol 12-myristate 13-acetate (PMA) was measured. For trials studying the response of cytokines to chronic S. uureus infection, a total of 32 lactating Holstein-Friesian dairy cows were used. Experimental infections were established with S. aureus strain Newbould 305 (25). All mammary glands to be inocuIated were free of bacteria on five consecutive am. milkings prior to infection and had individual quarter SCC less than 200,000 ceIIs/ml of milk. Inocula were prepared from a 4-h culture grown in trypticase soy broth and diluted in sterile distilled water to contain 50 to 300 cfu/ml. In vivo inoculations were made by infusion of 1 ml of diluted culture of S. aureus into the teat cistern of a millred out mammary quarter. Experimental mammary quarters were cultured daily to determine the presence of S. aureus. All infections were established for a minimum of 3 wk before being used for therapeutic treatment. Quarters were considered to respond to therapy if the a.m. milk sampIe was h e of S. aureus for a minimum of one milking immediately following the last intramammary infusion. Infections were considered cured if milk samples remained free of S. aureus for 14 successive daily samplings of milk, Fourteen days of continuous sampling were sufficient to eliminate most false negatives for cures, because even with antibiotic therapy >95% of all quarters that relapse do so within 14 d [(9), unpublished observations on over 1200 treated experimental infections].

SYMPOSIUM: IMMUNOBlOLOGY OF CYTOKINES

Mlcrobloiogy

Individual quarter milk samples were collected prior to each a.m. milking and cultured by swirling .1 ml in duplicate petri dishes containing trypticase soy agar supplemented with 5 % whole ovine blood. Plates were incubated aerobically at 37'C and examined after 24 h. StaphyIococcus aureus isolates were counted and identified by colony morphology and hemolytic patterns. Periodically, isolates were also tested for coagulase activity (24) and antibiotic sensitivity.

4415

analyzed at the Flow Cytometry Laboratory, Princeton University, using a Coulter EPICS 753 (Coulter Electronics) as previously described (9). Greater than 90% of the cells from infected glands were PMN, and a total of 5000 PMN were analyzed by flow cytometry for each individual phagocytosis determination. Whole Cell Superoxide Assay

Somatic cell counts were determined by Coulter counting (9, 30) and performed on 1-ml aliquots of fresh milk samples collected immediately prior to the a.m. or p.m. milking. Cells were pelleted initially at 2200 x g for 10 min, and the resulting cellular pellet was washed twice in phosphate-buffered saline (PBS) at 1200 x g for 5 min. The cellular pellet was diluted 150 or 1500 in ISOTON@ (Coulter Diagnostics, Hialeah, FL). The fresh cell preparation was counted using a gate for exclusion of less than 6.8 p in diameter on a Coulter ZM (Coulter Electronics, Hialeah, FL). The cell population was also profiied on a Coulter C256 Channelyzer (Coulter Electronics). Only viable somatic cells, which would be the only cells retaining biological activity, were counted using this criterion.

The NADPHdependent 0;generation was assayed as an indication of PMh' activation (22). Briefly, milk from quarters were centrifuged at 1200 x g for 15 min to pellet milk PMN. Cell pellets were washed thrice in PBS, resuspended, and counted. Fifty microliters of 5 x 106 milk PMN/ml of PBS supplemented with glucose, magnesium, and calcium were added to each microtiter well containing 200 pl of PBS with glucose and a final concentration of 75 p.M cytochrome c. Plates were incubated for 10 to 15 min at 37'C. To each well, PMA was added to a final concentration of 100 ndml to activate 0;production. The specific rate of reduction of cytochrome c was measured using a narrow band 550-MY filter on a ThennoMax@ kinetic plate reader (Molecular Devices Corporation, Menlo Park, CA) as an indirect measure of 0;formation. Data were expressed as the mean rate of 0generation in nanomolar 0;per minute per 10 cells.

Quantitatlon of Phagocytosis by Flow Cytometry

Recombinant Cytoklnes and lntramammary infusions

Ten- to 250-ml samples of milk were collected immediately prior to the am. or p.m. milking and passed through cheesecloth to remove clumped leukocytes. Milk was then centrifuged at 4'C, 1200 x g for 30 min. Fat and skim milk were removed, and cell pellets were washed twice and resuspended in PBS and counted on a Coulter counter. Cells were brought to a final concentration of 1 x 106 viable cellslml with cold RpMI-lW (Gibco, Grand Island, NY) supplemented with 5% heat-inactivated fetal calf serum (Gibco). Then 1.6- to 2-p fluorescent beads (Polysciences, Warrington, PA) were added to a concentration of 5.52 x lo7 beads/lo6 cells (37). This mixture was incubated at 37°C for 30 min and then

Recombinant cytokines ( r b e l p and &E2) were prepared and tested for in vitro biological activity by IMMUNEX Corporation, Seattle, WA (20, 31). The specific activity of rbIL2 was 20,000 U/pg of biological activity and 1.4 ng of endotoxidmg of protein. The specific activity of rbIL-lj3 was 30,000 U/pg of biological activity with 43 ng of endotoxh/mg of protein. Endotoxin levels at doses administered were 100-fold below the dose required to elicit any biological response in vivo with purified endotoxin. Cytokines were diluted to give the required dose to be delivered in 10 ml of sterile PBS. Doses were administered in syringes fitted with single use teat cannulas, and cytokines were infused through the teat

Somatlc Cell Determinatlon

3

Journal of Dairy Science Vol. 74. No. 12. 1991

4416

DALEY ET AL.

canal of the treated mammary quarter. For all therapeutic trials, rbIL-lp and rbL-2 were administered three times after the p.m. milking at either 24- or 48-h intervals into quarters that had been infected for a minimum of 3 wk. A total dose of 10 mg of rbIL-2 or 200 pg of rbIL-lp in 10 ml of sterile PBS was administered in each of the three infusions into each infected quarter. Sodium cephapirin (Sigma Chemical Co.,St. Louis, MO) was administered as a 200-mg dose in 10 ml of PBS after consecutive p.m. and am. milkings. Similarly, Cefa-Lak@(Bristol-Myers, Evansville, IN> was administered after consecutive p.m. and am. milkings according to package insert.

Cycllc Nature of Host Cell Actlvatlon In Experimental Staphylococcus aun?us Mastltls lnfectlons

In the experimental S. aweus mastitis model, a cyclic rise and fall in colony-forming units per milliliter of milk (or viable S. aureus per milliliter) is observed (Figure 2A). Cycling of colony-forming units concomitantly occurzed with an asynchronous cycling of the somatic cells ( S O to 95% PMN by flow cytometry). In addition to the SCC, biological activity, as measured by PMA-induced 0;production, was also measured (Figure 2B). The ability to induce 0;fluctuated almost 20-fold over the 5d sampling period and at each sampling was greater than the activity of PMN from uninfected quarters. Statistically, d 1 and Analysls 4 of the cycling infection were different (P < All data were analyzed by a two-tailed Stu- .05) from d 2, 3, and 5 by Student’s t test. dent’s t test, except for efficacy data for which the Mantel-Haenzsel-corrected chi-square Effect of rblL-lp and rblL-2 on Actlvatlon of Milk PMN from Mastltlc Dairy Cows analysis was used. as Measured by Superoxlde Induction and Phagocytosls RESULTS Actlvatlon of Host Polymorphonuclear Activity After lntramammary Challenge with Staphylococcus aureus

Normal uninfected quarters were first challenged with S. aureus to examine the normal host response to a pathogen. Figure 1A represents the total number of somatic cells and the rate of 0;generation (Figure 1B) at various time points after S. aureus challenge. At time 0, no significant 0;could be generated upon addition of PMA, suggesting that milk PMN are significantly suppressed in function. By 40 h after S. aureus challenge, the SCC increased 20-fold over the prechallenge SCC. At the same time, PMA-induced 0; increased to over 15 nM O$nin per lo7 milk PMN, a level approximately equal to the activity of peripheral blood PMN in the bovine (5 to 30 nM O-;/min per lo7 PMN, unpublished observations). The number of cells (as indicated by asterisk) and the generation of 0;(all time points) after S. aureus challenge were significantly different from prechallenge levels (P e .05) by Student’s t test. Journal of Dairy Science Vol. 74. No. 12. 1991

Nine mastitic quarters fkom four different animals were untreated or infused with 200 pg of rbIL-lp or 2 mg of rbLL-2 based upon previous dose-response data (Daley et al., unpublished data). Superoxide formation by milk PMN was monitored at 0. 16 and 24,40 and 48,64 and 72, and 112 and 120 h postinfusion. By 48 h, the SCC increased 170-fold for the rbIL-lp infusion (Figure 3A) and 4OO-fold for the rbIL-2 infusion (Figure 3D) over the pretreatment SCC (P e .01 and P e .005, respectively). Although administration to normal uninfected glands has been shown to enhance inducible 0;production and phagocytosis paley et al., unpublished data), stimulation of PMN from infected quarters was only moderate and quite variable after cytokine therapy compared with pretreatment levels. Polymorphonuclear cells fkom mastitic glands treated with rbIL-lf3 or rbIL-2 demonstrated two- to fourfold stimulation (P < .01 for L - l p for time 0 vs. other time points, and P e .01 for L-2 for time 0 vs. 112 h) or 0;production over the pretreatment-infected gland control cells (Figure 3, B and E). Increased stimulation of 0;production was seen within 16 h postinfusion and continued through 88 h for the rbIL-lp treatment.

SYMPOSNM: IMMUNOBIOLOGY OF CrrOMNEs

The glands treated with rbIL-2 demonstrated a significant induction of Oi (P < .@) over the PMN from infected, untreated quarters (time 0) at 112 h after the last intramammary infusion of rbIL-2. After rbL2 administration, inducible 0; production initially decreased and then increased to over 14 nM 0; /min per lo7 milk PMN. Unlike r b L l p administration, which showed a suppression of phagocytosis after infusion (P < .01 for time 0 vs. 18 h), phagocytosis remained enhanced (no statistical difference) at all times in the r b L 2-treated group (Figure 3, C and D). In Vivo Efflcacy of rblL-lp and rblL-2 Formulations

A total of 12 to 13 quarters per treatment were infused with rbIL-lD, rbIL-2, Na cephapirin in saline, or Cefa-Lak*. The number of infected quarters that cleared their infection for a minimum of one milking after the last therapy (response) and the number that remained infection b for 14 consecutive daily milkings after the last intramammary infusion (cures) were monitored flable 1). Although the overall cure rate was equivalent for the rbIL-2 treatment (38%) and rbIL-Ij3 treatment (42%), the percentage of quarters responding that went on to cure was different, 70% versus 51%, respectively (not statistically different). Therefore, the rbIL2 quarters that respond remained free of bacteria more frequently than the quarters treated with rbIL-lj3. Correlation of Superoxide Induction of Milk PMN and Therapeutic Efflcacy After rblL-2 Infusion

Somatic cells were collected from mastitic quarters after cytokine therapy, and their cells were assayed for PMA-induced 0;formation. Superoxide induction was also monitored at 16, 40, 64, and 136 h after the last infusion, and the quarters were separated into cured (Figure 4A) or relapsed (Figure 4B). Both groups had equivalent starting levels at 24 h posttreatment of inducible 0;(not statistically different). Cured quarters at the 136-h time point demonstrated a 60% decrease in inducible 0;formation (P< .05), but relapsed quarters only demonstrated a 7% decrease compared with the pretreatment levels. This would suggest that the endogenous PMN population

4417

of an infected gland requires a continuous source of pathogen to maintain the biological activation of the resident cells. DISCUSSION

In a study of the pathogenesis of experimental S. aweus bovine mastitis, it was shown that the host exhibits a cyclic rise and fall of host phagocytic cells (9) in the milk ( S O to 95% PMN). Concomitant with, but asynchronous to, this host cellular response is a cyclic rise and fall of detectable S. aureus in the milk of the infected gland. Not only does the quantity of PMN change during this cycling, but the quality, as measured by phagocytosis and intracellular bacterial killing efficiency, also changes (9). The present study has extended this observation to include cycling of oxidative metabolism as well as the activation of phagocytic cells on in vivo bacterial challenge. This clearly demonstrates that the host responds to a challenge fmm most pathogens by enhancing the quantity and quality of the phagocytic cells in the m a m m q g l a d Although partial cellular activation does occur after bacterial challenge and during a cycling infection, inefficient phagocytosis and killing may contribute to the relapse of the infection and failure of thempy by antibiotics (6). Furthermore, antibiotics can inhibit the normal defense mechanisms of the host by decreasing phagocytic function (40)and thereby further exacerbate the frequency of relapsing infections. Exogenous administration of rbIL-lp and r b L 2 appears to potentiate the normal response of the host to infection. Recombinant bovine interledcin-lf3 induced a quantitative influx of host PMN after an intramammary infusion. These cells were partially activated in that the inducible 0;formation was enhanced over resident mi& PMN from uninfected glands, but only modest effects of rbIL-lj3 administration were seen on PMN from infected glands. This might suggest that once PMN are activated by S. aureus, further activation by cytokines is only modest. However, maintenance of concomitant activation of the bactericidal components of the PMN, as well as phagocytosis by cytokine therapy, is necessary to eliminate bacteria efficiently (18, 21). Journal of Dairy Scienct VoI. 74, No. 12, 1991

441 8

DALJ3Y ET AL.

A

20 -15--

al

clr C

-9a

lo--

L 0

5 --

I

v

0

1

-24

16

40

64

112

160

16

8

14

f

12

10

8 6 4

I

W

2

a

I -24

16

40

64

112

160

Hours Postchallenge Figure 1. Induction of phagocytic cell activation after challenge with Sraphylococcus uureus-ni infusion. The polymorphonuclearcells from three mammqy glands from three cows challenged with S. w e u s were collected at the indicated time periods. Somatic cell counts and phorbol 12-myristate 13-acetate (PMA)-indacible superoxide &on (0; ) production werc assayed in triplicate. Data are expressed as the mean stimalation index for SCC (posttreatment SCC divided pretreatment SCC) (A), or the mean rate of PMA-induced 0;f SEM, exprcsscd as nan~molar0;per minute per 10 viable somafic cells, (J3). 'Ihe asterisk in panel A indicates statistically sig&hnt SCC from pretreatment levels (P< .OS). In panel B, time points 40,64,112, and 160 h are all statistically differmt by Student's I test (2' < .01)

from the pretreatment levels.

Journal of Dairy Science Vol. 74. No. 12, 1991

4419

SYMPOSIUM: IMMUNOBIOLOGY OF CrrOKINES

1

B

T 8 --

1

4

--

01 1

2

3

4

5

Day of Milk Sampling Figure 2. Phagocytic cell activation in established Staphylococcus aweus infections. The SCC, colony-foanits (Sraphylococcus oweus), and rate of superoxide anion (0;) of milk polymoqhonuclear were monitored from a cycling S. aweus mastitis infection for 5 consecutive d. The SCC per milliliter of milk (0)and colony-fOrming units per milliliter of milk (0)(A), and phorbol 12-myistate lSaCttate (PMA)-inducible 0; formation (B) are shown on each day. Data are expressed as the mean of duplicate samples for SCC x lfl,’ml and for colony-formingunits x lO-*/ml of milL; and mean rate of oxygen radical generation is expressed in nanomoles of 0 - eper 107 milk c e ~ after s PMA induction (* SD) for triplicate assays. The asterisks represent statistically significant difference (P < .05) compared with d 1 by Student’s t-test.

Journal of Dairy Science Vol. 74. No. 12, 1991

4420

DALEY ET AL.

Dissociation of activation of phagocytosis without activation of bactericidal activity could, in fact, contribute to a source of S. aureus for the relapsed infection. Consistent with this observation, treatment of infected glands with rbLlf3 transiently cleared the milk of bacteria in 83% of the mammary glands treat& only 42% of the treated quar-

ters remained free of infection after 14 d This high relapse rate is represented in the low ratio of cures to responding quarters, 51%. and is similar to that of Na cephapirin in saline treatment Treatment with rbIL2 resulted in a slightly lowered quantitative influx of cells, but these cells were activated for their inducible 0;at

IL-1 TRITED am= A

D I

n

r

0

la

40

64

112

i

16

40

u

112

lE4..

-_1111 100 10

0

4

h

0

is

io

a*

;12

I

I

.

.

40

u

111

1.

I

1 F

Time Posttreatment (Hours)

Time Posttreatment (Hours)

Pigure 3. Activation of milk polymorp~o~lear ctuS from mastitic glands after inh 'y infusion wilh recombinant bovine interleukin-1s (rbLlg) and intdcalrin-2 (rbIL-'2). Three to four iufected mannnar! glands from dairy cows were infused with -1s or rbE.-2 at each of three Wccessme am.or p s l milking. Somatlc cell counts. phagocytosis. and phorbol12-myristate 13-acttate (PMA)-indaced wpcroxide anion O;< ) production were monitored at 16 and 24,40 and 48,M and 72, and 112 and 120 h after the last W o n . Ihc gamk%ric mean nombaf SEM of MlmBtic cells per milliliter of milk is shown ( A -1s and D rblL2). The ram rate (k SEM) of EMA-induciblc 0; formation, exas naaomoltx of opio per 107 viable cells. is a l s ~s b o ~ n(B: -1s and E: MIA). he mean p-c index is shown (C: rb&@ and -2). The meauphagocytic indcx is expregstd as the meanparentage of cells pbagocytosiog 1 or Inole beads of dimc (x) divided by the mcan pgcentagc of cells PhaBocytosiog 1 or more beads at time 0 (prior to tnxtmeot). AU posaxeabnmt time points that werc significantly different from prctrcatmezlt tinac points are indicated by asterisks (*P < .05,**P < .all).

Journal of Dairy Science Vol. 74, No. 12, 1991

4421

SyMposIuM: IMMUNOBIOMGY OF CYTOKINES

TABLE 1. Therapeutic efficacy of ina~amammary infusion of recombinant bovine intaleukh (rblL)-lB and rbLL2.

Treamat

Dose'

rbE-lp

(mg) 2

rbE-2

Cefa-Lak Na

10

ig

Regimen 3x, am. 3x, pm. 2x pmJam. 2x. pmJam.

Numbex of g-

%Response*

%d

c~re:response~

12 13

83 54

425 385 425 775

.5 1 .70

12 13

86 92

.48 .83

'Treatments were administered after tbe a.m. milking (rbIL-lfI. Cda-Lak@, or Na cephapkin) or after the p s l milking (rblL-2, Ma-Lak@, or Na cephapirin) as indicated. Cytoldnes wcre administaed three times at 24-h intervals. and antibiotics were administacd on Consecutive pm. and am. milkhgs. *Number of lreated quarters with milk samples fiee of Srupltylococcus uureus for 1 d or more after the last tneament. divided by the totd number of qwutcrs treated. %umber of treated quarters 6ree of S. uwew for 1 4 4 comecntive milk samples after the k t treatment. divided by the total number of quarters treated. 4Cure:response ratio = Percentage of cured quarters divided by percentage of responding quarters. h o n e of the groups were shown to be statisticaly different from O n e another by Mantel-Haenszel Corrected chisquare analysis with P > .l.All groups were similarly compared with 50 mtreakd control qnarters followed over a 7-d period (0/50 = cared) and were shorn to be ~ignifimtlyd B m at P < .05.

several time points after administration (three to four greater than even the optimal time point for hIL-lf3). Unlike the rbLlf3 treatment, phagocytosis did not decrease but remained at an activated level following treatment of infected glands, thereby suggesting that the new cells being recruited into the gland continued to be activated. It is indeed surprising that intramammary administration of rbiL-2 elicited an influx of PMN, because PMN have not been described to express the IL-2 receptor (17). Presumably, r b L 2 must be initiating a secondary cytokine cascade that mediates the observed effects in vivo. The treatment of infected quarters with rbL-2 resulted in only 54% of the mammary glands transiently clearing their infection, but 38% of the treated quarters remained free of S. aurezu at 14 d. The ratio of cures to responding quarters was .7, a ratio that more closely resembles the therapeutic efficiency of CefaL&@, .&I, albeit the o v e d cure rate is less, 38 versus 77% for the Cefa-Lak@'. Although the ability to potentiate further the activation of milk P M N from infected quarters after cytokine therapy was variable, a decrease in inducible 0; was predictive of effective therapy. Inducible 05 levels from cured quarters demonstrated a two- to fourfold decrease over relapsed quarters (P < .05). This would suggest that the activation state of resident mammary gland PMN requires a continuous

source of pathogen to maintain the biological activation. Similarly, the activation by cytokines was shown to be transient and reversible. In mastitic glands, activation is maintained through bacterial components and secondary host signals such as cytokines. Exogenous cytokine administration merely mimics normal host reactions to infectious agents, and because phagocytic cell function is significantly depressed in normal milk PMN compared with peripheral blood PMN, overcoming this suppression or maintaining activation with cytokines may be important in the pvention and therapy of mastitis. If formulations of cytokines can be deve loped that are equivalent to or better than current therapies, they may totally or partially circumvent prolonged milk withdrawal because of antibiotic residues. Alternatively, cytokines as an adjunct to cment mastitis therapy may reduce the amount of antibiotic required. Although previous investigators have proposed an increase in the number of host cells as an approach to prevention (7, 29, 3 9 , we have further shown that the concomitant activation of these cells may be of equal importance in the treatment of bacterial infections. Clearly, this activation can be mimicked by exogenous administration of recombinant cytokines. These observations should be applicable to the therapy and prevention of a wide variety of acute and chronic infectious diseases in both domestic animals and humans. Journal of Dairy Science Vol. 74, No. 12. 1991

4422

DALEY ET AL.

14, 12-

3 al 0

lo--

h

8--

.E

6-

0

A

*

\ 7-

\

b

4--

f

2 -0 --

-2 !

0

I

16

40

64

"1

136

B

rn 10 al 0

h

0

\ T-

.-C

E

\

2

3 ,

-2 O 0

16

40

64

136

Hours After First Cytokine Infusion Figure 4. Correlation of milk polymorphonuclear cells activation and therapeutic efficacy from mastitic glands after in1mmmua1-y infusion with recombinant bovine intcrleakin-2 (-2). A total of 16 cycling Stuphylococcw aureus mastitis infections were treated with rbIL2 and monitored for cures. Each symbol represents the phorbol 12-myristate 13-acetate pMA)-inducible superoxide anion (03prodaction of individual guarters for 6 consecutive d after the last treatment with rbIL2. 'Ihc rate of PMA-induced superoxide induction is expressed as the mean rate of triplicate assays expressed as namrmoles of 0-@in per lo7 milk polymorphonwlearcells.Data are grouped into cured (n = 10) quarters (A) or relapsed (n = 6) guarters (B). The asterisks iodicate that the 136-h grouping was statistidy different (P < .01) from the IGh samples by Student's f test.

Journal of Dairy Science Vol. 74, No. 12, 1991

SYMPOSlW+t IMM"0BIOLOGY OF CYTOKDES

ACKNOWLEDGMENTS

'The authors wish to thank Jerome Zawadzki of the Princeton University, Flow Cytometry Laboratory, for technical assistance with flow cytometry and P. Baker, H. Sassenfeld, D. Urdal, C. Maliszewski, and K. Picha, of IMMUNEX Corporation, Seattle, WA, for assistance with production, supply, and analytical assays for recombinant bovine cytokines used in this study. The technical assistance of Debbie Search and support in our dairy herd management from K. Simkins, L. James, and R. VanMater, D. Smith, and R. Brazina of the Clinical Laboratory, American Cyanamid Co., were also instrumental in the completion of this study. REFERENCES 1 Anderson, J. C. 1982. Mechanism of staphylococcal virulence in relation to bovine mastitis. Br. Vet. J. 132:229. 2 Aungier. SP.M., and F. H.Austin. 1987. A study of the efficacy of infkmummary antibiotics in the treatment of clinical mastitis. Br. Vet. J. 143278. 3Cerretti. D. P.. K. McKereghan, A. Larsen, M. A. Canh-ell, D. Anderson, S. Gillis, D. Cosman, and P. Baker. 1986. Cloning sequence and expression of bovine interleukin-2. Roc. Natl. Acad. Sci. USA 83: 3223. 4 Christ, W. L., R J. Harmon, and L.E. Newman. 1983. How much does mastitis cost? Udder Topics 66. 5 Code of Federal Regulations. 1989. ?DA Office of the Federal Register, National Archives and Records Administration Publ. 21US Govt. Printing Office, Washingmq Dc. 6 Craven, N., and J. C. Anderson 1984. phagocytosis of Staphylococcus aureus by baine mammary gland macrophages and intracellular protection from antibiotic action in vitro and in vivo. J. Dairy Res. 51513. 7Craven, N., and M. R. Williams. 1981. Defenses of the bovine 7 gland against infection and prospects for theu enhancement. Vet Immunol. Immunopathol. 107. 8Czup1ymki, C. I., and J. F. Brown. 1987. Recombinant murine interleub-la enhancement of nonspecific antibacterial resistance. Infect. Immun 55: 2061. 9 Daley, M. J., E. R Oldham. T.J. Williams, and P. A. Coyle. 1991. Bovine mastitis. I. Quantitative and qualitative properties of host polymorphonuclear cells during Staphylococcus uureu mastitis infection. Am. J. Vet. Res. 52:474. IODerbyshire, J. B.. and D. T. Berman. 1968. Leulcocyte responses of the bovine udder to infusion of certain initatus. Am. J. Vet. Res. 291971. 11 Dexter, T. M, and E. Spooner. 1987. Growth and differentiation in the hematopoietic system. AMU. Rev. Cell Bioi. 3:423. 12 Dharello. C. A. 1988. Biology of interleukin-I. Fed.

4423

Am. Soc. Exp. Biol. J. 2:108. 13Faltynek, C. R, and J. J. OppenheimCr. 1988. Interferons in host defense. I. Natl. Cancer Inst. 80151. 14pOx, L. K.. and J. S. McDonald. 1988. Functional activity of neutrophils h m bovine mammary glands infected with Staphylococcu aureus. I. Dairy Sci. 71: 3521. 15 Gillis, S., and K. A. Smith. 1977. Long-term culture of tumor-specific cytotoxic T cells. Nature (Land.) 268: 154. 16 JacLson. I. A., D. E. Shuster, W. T. Silia, and R I. Harmon. 1990. physiological response to intramammary or intravenous treatment with endotoxin in lactating cows. J. Dairy Sci. 73:627. 17 Leonard,W. J.. M. Kamhisa, M.Kronke, N. I. Peffer, P. B. Svetlik, M.Sullivan, and W.C. Greene. 1985. Structure of the inkrleukb 2 receptor gene. Science 230633. 18 Li, I. W.,S. Mudd, and F. A. Kapral. 1963. Dissociation of phagocytosis and intracellular killing of Sraphylococcus aweus by human blood leukocytes. J. Immuwl. 9O:SW. 19 Madgwck, L., S. Maya, and P. Keen 1989. Penetration of antibiotics into bovine neutrophils and their activity against mimcdMa~Stuphylococcus aureus. J. Antimicrob. Chemother. W709. 20MaliszewsLi, C. R, P. E. Baker, M. A. Schoenbom, B. S. Davis. D. Cosman, S. Gillis, and D. P. Cemtti. 1988. Cloning, sequence and expression of bovine interleukin la and interleukin 18 complementary DNAs. Mol. Immuwl. 25:429. 21 Marodi, L., P.CJ. Lei& and R Van Furth. 1983. A micromethod for separate evaluation of phagocytosis and intracellular killing of Staphylococcus aureus by human monocytes and granulocytes. J. Imrnunol. Methods 57:353. 22IUiiyo. L. A., and J. T. Cumutte. 1990. Kinetic microplate assay for superoxide production by neumphils and other phagocytic cells. Methods Enzymol. 186:567. 23 McDonald, J. S. 1977. Streptococcal and staphylococcal mastitis. J. Am. Vet. Med. Assoc. 170:1157. 24 National Mastitis Council, Inc., 1981. Microbiological procedures for llse in the diagnosis of bovine mastitis. 2nd ed. carter press, Inc., Am=, IA. 25Newbould, F.H.S., and F. K. Nave. 1965. 'Ihc response of the bovine mammary gland to m iofusion of staphylococci. I. Dairy Res. 32:163. 26Nickerson, S. C., W. E.Owens, and J. L. Watts. 1989. Effects of recombinant granulocyte colony-stimulating factor on Staphylococcus aureus mastitis m lactating dairy cows. J. Dairy Sci. 723286. 27 Oppenheim, J. J., E.J. Kovacs, K. Matsushima. and S. K. Darams. 1986. There is more than one interleukin1. h u w l . Today 7:45. 28 F'aape, M. J., and W. P. Wergin 1977. The leukocyte as a defense mecham'a. I. Am. Vet Med. Assoc. 1701214. 29 Paape, M.I., W. P. Wergin, A. J. Guidry, and R E. Pearson. 1977. Lenkocytes--secorad line of defense against invading mastitis pathogens. J. Dairy Sci. 62: 135. 3OphiPps, L. W., and FH.S. Newbould. 1966. Determination of leukocyte concenfrations in cows' milk with

-

Journal of Dairy Science Vol. 74, No. 12, 1991

4424

DALEY ET AL.

a Coulter@ counter. I. Dairy Res. 33:51. 31 Picha, K.S.,and P.E.Baker. 1986. Bovine T l y m p h cyts. I. Gmration and maintaurnce of an i ~ t e r k ~ k i d~-t, ~ - 2 cytotoxic T lymphocyte ctll line. Imrmmology 52131. 32Reddy. D. N., P. G. Reddy, H. C. Mhocfm, B. W. Penwick, P. E. Baker, W.C. Davis, and P.Blecha. 1990. Adjuvaniciq of recombinantbovine k k u k i n 1s: influeme on immonity, infection and latency in bovine hapes virus-1 infcclion. Lymphokhe Res. 9: 295. 33Reddy, P. G., P. Blecha, H. C. hfinocba, G. A. Anderson, J. L.Morril, P.J. Fdorka-Cray. and P. E. Baker. 1989. Bovine recombinaot interlcalitt2 augments immunity and resistance to bovine herpes virus infection. Vet Immanol. Immumpathol. 23:61. 34 Russell, M.W..and B. Reiter. 1975. Phagocytic deficiency of bovine milk leukocytes: an effect of casein J. Reticuloendothel. Soc. 18:l. 35 Sayers, T.J., T.A. Wiley, C. A. Bull,A. C. Dew m, A. M. Pilaro, and B. Lokresh. 1988. Effect of cytokines on polymorphonuclear neutrophil infiltra-

Jaurnal of Daky Science Vol. 74, No. 12, 1991

tion in the mause: plustaglandin- and leukotricncindependem induction of infiltration by E-1 and tu-

mor IlcQosis factor. I. Inrmanol. 141:1670. 36 Stamch, M. J., and D. D. Wood.1983. The edjwanticity of intcrlenLin 1 in vivo. I. Immunol. 13a2191. 37 Steinkanp, J. A., J. S. Wilson, G.C. Samnlm, and C. c. stmvart. 1982. Phagocytosis:flow cytomtric quantitatiOn with fluomcent microsphats. Science 215: 64. 38 Woodward, L.P.. R. L.Jesman,D.0.Parrington, and K. E. J e n ~ e n . 1983. Enhsnctd antibody-dependent bactericidal activity of Imltrophils from calves tna!ed with lipid amhe am ti at or. Am. J. Vet Res. 44:389. 39 Ziv. G.,M.I. Paape, and W. S. Schulk. 1985. Effect ofabradd’ . 1 device on mitk somatic cell count and mftranunrrmpy infection. Bovine Pract 0(20):102. 40Ziv, G.,U I. Paape, and A. U Dulin. 1983. MUence of antiLliotics and intra 1 antibiotic products on phagocytosis of Staphylococcus aweus by bovine leukocytes. Am. J. Vet Res. 44.385.