Prevention and treatment of staphylococcus aureus infections with recombinant cytokines

PREVENTION AND TREATMENT OF STAPHYLOCOCCUS AUREUS INFECTIONS WITH RECOMBINANT CYTOKINES Michael

Daley, 1 Thomas Williams, Patricia Coyle, Gary Furda, Ruth Dougherty, Phil Hayes

Antibiotic therapy is only moderately efficacious for bovine Staphylococcus aureus mastitis. We have used recombinant bovine cytokines to activate the natural host defenses, to prevent and treat bovine mastitis. Uninfected mammary glands infused with GM-CSF or IL-2 increased the percentage of phagocytic cells in the milk by 2-3 fold. IL-1 increased the number of polymorphonuclear cells in the milk, enhanced their inducible oxygen radical formation, and had no effect on phagocytosis. Treatment with IL-2 increased the number of polymorphonuclear cells in the milk, enhanced their inducible oxygen radical formation, and enhanced their phagocytosis. GM-CSF had no effect on the number polymorphonuclear cells in the milk but enhanced their inducible oxygen radical formation, and enhanced their phagocytosis. All cytokines were effective in preventing S. aureus infections (20-100 %). 52% of all chronically infected mammary gland quarters treated with three doses of IL-2 responded to therapy and 32% of the treated quarters remained cured. 75% of all mammary glands treated with three doses of IL-ls responded to therapy by clearing the infection and 22% of the treated glands remained cured. These studies demonstrate that recombinant bovine cytokines can be used effectively to prevent infections as well as treat established chronic infection.

Primary and secondary clinical infections due to Staphylococcus aureus are a ubiquitous problem especially in certain patient populations.lJ It has been suggested that resistance to classical antibiotic therapy3 can in part be accounted for by intracellular survival of bacteria.4 Although polymorphonuclear neutrophils (PMNs) clearly play a central role in the control of most S. aureus infections, PMNs may also contribute to the inefficiency of classical antibiotic therapy because intracellular bacteria are usually not affected by antibiotics. Cytokines have been reported to be effective in the treatment of some bacterial infection+’ presumably through activation of host defense

From the Immunology Group, Agricultural Research Division, American Cyanamid@ Company, Princeton, NJ 08543-0400, USA. *Address correspondence and reprint requests to: Dr Michael J Dalev. Immunologv. American Cvanamid Co, PO Box 400, Princkton, NJ 08543-0400. Received 28 April 1992; accepted for publication 5 October 1992 @ 1993 Academic Press Limited 1043-4666/93/030276+9 $08.00/O KEY

276

WORDS:

cytokines/infections/mastitis/prevention/treatment

mechanisms, but pleiotropic activities and side effects have precluded their wide clinical use. Clinical and sub-clinical bovine mastitis is a major cause of economic loss to the dairy industry with estimates of 2-4 billion dollars annually.8 One of the dominant etiological contributors to this economic loss is S. aureus. Although the current antibiotic therapy can be effective in the control of mastitis due to some organisms, therapy for S. aureuS is only moderately efficacious and requires 3-5 day milk withdrawal due to antibiotic residues.gJO S. uureus mastitis infections are characterized by viable bacteria in the milk and an elevated number of PMNs in the mammary tissue and milk.11 The initial development and/or continuation of S. aurezu mastitis depends upon the interaction between invading bacteria and the cells of the host’s defense system. Inefficiency of host PMNs may actually contribute to the reinfection of the gland and chronic cyclic nature of the infection.12 Because of these striking similarities to a variety of human clinical infectious diseases, bovine mastitis can provide a valuable experimental model for a variety of infectious diseases. The migration and activation of phagocytic cells can be regulated by a variety of host products,‘@7 CYTOKINE,

Vol.

5, No.

3 (May),

1993: pp 276284

Treatment of infections with Cytokines / 277

some of which are cytokines. In the bovine species, the genes for bovine IL-l, and IL-l, (r-BOIL-ll3), IL-2 (r-BOIL-2) and GM-CSF (r-BoGM-CSF), have been cloned and their products expressed.is20 IL-l, although somewhat pleiotropic in its effects, has been reported to have specific chemoattractant and activating properties on cells of the phagocytic series as well as other cell types. IL-2 can also have similar activities, but the effects on PMNs would probably have to be mediated through a secondary cascade of lymphocyte derived cytokines, although PMNs have been reported to express the IL-2 receptor.21 GM-CSF stimulates proliferation of early precursors of the myeloid series, accelerates their differentiation, and clinically is being used to accelerate the repopulation of bone marrow progenitors and their subsequent differentiation in treating genetic disorders** and after chemotherapy or radiation therapy.23 Polymorphonuclear cells and their state of biological activity play a central role in the pathogenesis of S. aureus infections. Cytokines can precisely regulate, directly or indirectly, PMN biological activity, and the bovine mammary gland represents an excellent closed organ model system to evaluate the local administration of these immunomodifiers. We therefore evaluated the in vivo application of targeted cytokine administration in the prevention and therapy of S. aureus infections.

Biological Effects of Local Cytokine Administration Both r-BOIL-l@ and r-BOIL-2 can induce an infiltration of PMNs into the milk in a dose related manner, while r-BoGM-CSF demonstrates no significant chemoattractant activity, Fig. 2. Infusion of another recombinantly derived protein, lysostaphin, has also failed to induce a significant somatic cell influx with as much as 100 mg of protein (unpublished datal3), supporting the specific nature of this influx of cells even in the mg/ml range. The minimal effective dose and maximal tolerable dose for r-BOIL-la and r-BOIL-2 is 0.5 pg-1000 pg and 0.5 mg-30 mg respectively, while intramammary infusion of up to 5 mg of r-BoGM-CSF failed to elicit

120

804 LR 2 2 ;

Cycling of Staphylococcus

aureus Infections

In experimental S. aweus mastitis a cyclic rise and fall in bacterial CFUs per ml of milk (viable S. aureus per ml) is observed, Fig. 1. This cycling of CFUs is also followed by cycling of the somatic cells (primarily PMNs). The cycling of the total number of PMNs is asynchronous with the cycling of the bacteria and the amplitude and/or frequency occurs independently in each quarter of the mammary gland, Fig. 1, panel A, B, and C. Bacterial counts can often fall below the levels of detection for as long as 1-2 days, although true long term spontaneous cures with daily sampling have been rare (< 2%). Concomitant with this quantitative cycling, the biological activity of these cells also cycles, such that when there are few PMNs in the gland they are also less able to ingest and kill bacteria.12 The inefficiency of these host cells in the elimination of bacteria has led to a possible explanation for relapsing, cyclic nature of the mastitis, as well as the lack of susceptibility to antibiotic therapy.

B

loo80 -

8 2

RESULTS

1 6QOLF

7z

60 40

- -

-

x 20. z 5 39 2 120 802 RR

-

-o-

_

-.-

2

3

Day

1. Cycling Mammary

O----y, 4

-

-

_

5 C

2

Figure Infected

-

3 sampling

of Somatic Glands.

Cells

4

(AM

and

9 5

Milking)

Staphylococcus

nureus

in

Three infected quarters from three mid-lactation Holstein dairy cows infected with S. aureuS were followed for 5 consecutive days. Both somatic cell counts (O---O) and colony forming units (O-O) of S. aureus were quantitated on each day. Data is expressed as the mean somatic cell count x lo-* per ml or number of CFU per ml x 101 of milk plated on each respective day. Differential counts on cells in the milk of infected glands demonstrated that > 95% PMNs.

CYTOKINE,

278 I Daleyetal.

any significant influx of cells into the gland. The specific in vivo biological activity for r-BOIL-l, as a chemoattractant is 10 000 fold greater than r-BOIL-2 on a mass basis, although their in vitro specific bioactivities are equivalent (see Materials and Methads). Inhibition of IL-2 bioactivity , soluble receptors, decreased bioavailability or decreased biostability in milk might explain this discrepancy. The peak number of cells after infusion with a maximal tolerable dose was 4-5 fold greater for IL-l than r-BOIL-2. The kinetics of PMN migration into the mammary gland generally occurred 8-16 hrs sooner for r-BOIL-l, infusions as compared to r-BOIL-2, depending upon the dose (data not shown). The dominant cell type for both r-BOIL-l, and r-BOIL-2 infusion was PMNs. In addition to quantitative changes in milk PMNs after cytokine infusion, qualitative changes (activation) was also observed. r-BOIL-lp induced an initial 150-fold increase in inducible super oxide production over the pre-treatment PMNs at 16 hours, decreasing to 20-fold at 40 hours, and back to normal levels by 112 hours, Fig. 3A. The inducible superoxide upon phorbol ester stimulation was significantly increased over the resident PMNs, Fig. 3C. The inducible superoxide from resident milk PMNs is significantly inhibited and can often be only 5% of that from peripheral blood PMNs from the same animal (unpublished observations). Phagocytosis was unaffected by intramammary r-BOIL-l, administration, Fig. 3B, perhaps suggesting that the newly recruited PMNs were only partially activated or primed. r-BOIL-2 administration also elicited a quantitative increase in milk PMNs, Fig. 3D, which returned within normal limits within 72 hours to control levels at the 2 mg dose. As with r-BOIL-lp administration, inducible superoxide was significantly increased over the resident PMNs, Fig. 3F, and remained so throughout the 96 hours of sampling. Phagocytosis was increased 2-3 fold over resident PMNs within 40-64

r

5 cc i

g y 5 e E 5 .E .-5 5 3 .=F z

Vol.

Prevention of Staphylococcus Infections by Cytokines

/O-O

10

i

1Z 0.001

o~o/o~o~o

0

0

P

/

I .-,Lp0.01 Cytokine

0.1

1.0

concentration

10 (kg per single

100 infusion)

1993: 276284)

aureus

The potential benefits of exogenous cytokine administration to the host, as it related to the differential induction of quantitative and qualitative changes in milk PMNs, was further explored in applied preventative and therapeutic experiments. r-BOIL-lp, r-BOIL-2 and r-BoGM-CSF were all capable of protecting the host from subsequent S. aweuS challenge, Table 1. The protective effect of r-BOIL-l, was most efficacious if administered 24 hours prior to S. aureuS challenge, but good protection was also seen if the cytokine was co-administered or infused 48 hours prior to challenge. This maximal protection appears to correlate with the peak quantitative biological response for this cytokine. The protective effect of r-BOIL-2 was most efficacious if administered 24-48 hours prior to S. aureus challenge, while less protection was seen if the cytokine was co-administered with challenge. The optimal protection also appeared to correlate

10000

100 :

3 (May

hours after infusion. Although the percent phagocytosis tapered off by 112 hours, it was still above the pre-treatment level, Fig. 3E. The average number of ingested beads per phagocytic cell was also increased upon activation. The delay of peak phagocytic activity, as compared to somatic cell counts, suggests that activation and chemoattraction can be distinctly regulated in vivo. r-BoGM-CSF failed to elicit any significant quantitative increase in milk PMNs, Fig. 3G, even with a 5 mg intramammary dose. However, the biological activity, as measured by inducible superoxide production was increased over the resident PMNs, Fig. 31, as well as the percent of cells which were phagocytic, was enhanced, Fig. 3H. The increase in phagocytosis was maintained above preinfusion levels for up to 112 hours.

Figure 2. Dose Response Gland to Intramammary Cytokines.

1000 :

5, No.

1000

IO 000

of the Bovine Mammary Infusion of Recombinant

Five fold serial dilutions of r-BOIL-1 from 1000.0 pg to 0.5 pg (0-o); r-BOIL-2 from ‘# 0 mg to 0.1 mg (0-n); and r-BoGM-CSF from 5.0 to 0.2 mg (O-O) total dose were administered at time 0 into 3-4 mammary glands from normal lactating Holstein dairy cattle. A total of 54 mammary gland quarters, 3-4 per group, were infused through the teat canal with various concentrations of recombinant bovine cytokines after the morning milking. Milk samples were collected 24 hours later and the number of viable somatic cells counted on a Coulter@ ZM with a C256 Channelyzer. The data is expressed as the Stimulation Index of the somatic cell count/ml after infusion divided by the somatic cell count/ml of the same gland prior to infusion.

Treatment

of infections

with

Cytokines

r-BoGM-CSF

rBoIL-2 D

I

l-dl-l cn__[I, H

6 5 3 2 I 04 ,[,

24

48

72

96

Hours Figure

3.

Quantity

F

C

0

and Quality

of Milk

0

after

recombinant

PMNs

from

24

48

72

cytokine

Mammary

I

96

0

intramammary

Glands

Infused

24

48

~-BOIL-~@

or

72

infusion

with

r-BOIL-2.

A total of 12 mammary glands from normal lactating Holstein dairy cattle were infused through the teat canal with either 200 pg of r-BOIL-la; 2 mg of r-BOIL-2; or 1 mg of r-BoGM-CSF. Milk samples were collected and the number of viable somatic cells counted on a Coulter@ ZM with a C256 Channelyzer. The data is expressed as the mean stimulation index +/SEM from three separate quarters sampled at 16, 40, 64 and 88 hours after infusion for r-BOIL-l@ (A); r-BOIL-2 (D); or r-BoGM-CSF (G). The Stimulation Index is equal to the Post-Treatment SCC divided by the PreTreatment SCC. The total percentage of phagocytosis by milk PMNs was quantitated by flow cytometry by measuring the mean percentage of cells ingesting 2~ fluorescent beads based upon triplicate analysis for of the mean of individual milk samples +/- SD, r-BOIL-2 (E), or the mean of pooled milk samples of 3-4 treated quarters, r-BOIL-l, or r-BoGM-CSF (H). The total population was first gated by forward angle light scatter, FALS, and 90” light scatter to only consider viable polymorphonuclear cells. The induction of superoxide after PMA (520ng) stimulation was also monitored using a whole cell cytochrome-c reduction assay, for r-BOIL-l@ (C); r-BOIL-1 (F); or r-BoGM-CSF (I). This data represents the mean of triplicate assays performed on pooled samples of the 3-4 treated quarters in the respective groups. The standard deviation for the superoxide and phagocytosis assay was within lO-15% of the mean.

TABLE

1.

Treatment’

of S. aureus Mastitis by Recombinant

#Glands*

Control7 IL-l

IL-2

Control GM-CSF

Prevention

27

12 15 14 14 13 9 8 11 10 10

/ 279

DW5

Bovine Cytokines.

Administration4

1 l4s 1 Pg 1 Pg lmg lmg lmg

0 hrs -24 hrs -48 hrs 0 hrs -24 hrs -48 hrs

lmg lmg

-24 -48

hrs hrs

% Infected5

% Control6

100 13 0 22 77 11 31 82 20 20

87 100 78 23 89 63 0 76 76

0

1 Single intramammary infusion administered after the afternoon milking. A total of 100 mammary glands from 22 different lactating Holstein dairy cattle were treated. 2 All glands were free of bacteria on five successive samples after the morning milking prior to the initial infusion. 3 The total dose in mass of protein is indicated. The biological activity of r-BOIL-l,, r-BOIL-2 and r-BoGM-CSF was 32 OKI Units/pg, 22 000 Unitsipg and 38 000 Units&g respectively. 4 Infusions of cytokines were administered simultaneously with S. Aureur (0); 24 or 48 hours prior to challenge with 30-200 CFUs of S. aureus. 5 Individual quarters were considered cured if the milk samples remained free of S. ~UXUF for 14 days after intramammary challenge with of S. aureur. 6 The % Control = (1 - (% Infected After Therapy divided by the % Infected in Control Quarters}) x 100. 7 Untreated quarters in Control 1 were challenged with 250 CFUs of S. aureur. Untreated quarters in Control 2 were challenged with 30 CFUs of S. aureu~.

280 I Daley

et al.

CYTOKINE,

with the peak quantitative biological response for this cytokine, but also marked the qualitative increase in phagocytosis and inducible superoxide, Fig. 3D & 3E. Unlike r-BOIL-l, and r-BOIL-2, r-BoGM-CSF elicited only minor quantitative changes in the milk PMNs. However, good protection from S. uureL(s was observed when r-BoGM-CSF was administered both 24 and 48 hours prior to challenge. Both these time periods were shown to increase phagocytic ability of the cells in the milk of treated glands, Fig. 3H.

Therapy ofEstablished Staphylococcusaureus Infections Established infections of mammary glands were also effectively treated with intramammary infusion of r-BOIL-lp and r-BOIL-2, Table 2. Both cytokines showed a dose response therapeutic efficacy. r-BOIL1, and r-BOIL-2 were most efficacious if doses were administered over 3 consecutive milkings, rather than in a single bolus. The percentage of quarters clearing the infection for > 1 milking (RESPONDING), as compared to the percentage of quarters which actually cure their infections (CURE) is a measure of the failure of the host cells to kill the bacteria. The relapse rate was the greatest for r-BOIL-lp (combined relapse rate of 73% for the 200 pg dose X’s 3) as compared

TABLE Treatment’ IL-l

(1 x)

IL-l

(3 x)

IL-2

(1 x)

IL-2

(3 x)

Na-cephapirin Cefa-Lak@

2

Efficacy of Recombinant #Glands 9 10 10 13 13 14 14 10 10 10 11 12 12 12 14 58

Vol.

5, No.

3 (May

to r-BOIL-2 (combined relapse rate of 40% for the 10 mg dose X’s 3). The high relapse rate for r-BOIL-$ suggests that although many host cells may indeed ingest S. aureus, these cells are inefficient at killing the bacteria. The relapse rate for r-BOIL-l, (54%) is comparable to the Na cephapirin treatment (57%), while the relapse rate for r-BOIL-2 (23%) was more closely approximated by the Cefa-Lak@ treatment (16%), albeit the overall cure rates are not as efficacious as Cefa-Lak@. This would suggest that activation of phagocytic cells in the milk would be an essential component of effective S. aweus therapy. Furthermore, if activation of the host cells is effective, as indicated by the r-BOIL-2 and r-BoGM-CSF, then a combination therapy with r-BOIL-l, or antibiotics may be synergistic.

DISCUSSION Although S. aureuS infections often undergo a waxing and waning of clinical signs, this bovine mastitis model system suggests that some of this variation may be due to a cyclic rise and fall of colony forming units, CFU (i.e. viable S. aureus) within the infected organ. This dynamic equilibrium between bacteria is

Bovine Cytokines as Mastitis Therapeutics. Dose2

Interval

% Response3

1ol-e

-

11 20 10 77 39 79 64 0 20 10 55 33 58 33 86 93

200 I.4 600 c1g 10 IQ 10 I% 200 I% 200 I% 2mg 10 mg 30 mg 2mg 2mg 10mg 10mg 200 mg 200 mg

1993: 276-284)

24 48 24 48

hrs hrs hrs hrs -

24 48 24 48 12 12

hrs hrs hrs hrs hrs hrs

% Cures4 11 0 0 8 0 7 29 0 0 10 9 8 25 25 42 22

1 S. aurew infections were established 2-3 weeks prior to treatment. Treatments were administered after the morning or afternoon milking by 1 (1 X) or 3 (3 X) successive intramammary infusions at the indicated times. A total of 120 mammary glands from 28 different lactating Holstein dairy cattle were used. Each cytokine treatment group contained 9-14 infected quarters. Recombinant bovine interleukin-la (IL-l) and interleukin-2 (IL-2) were obtained from IMMUNEX Corp. Seattle, WA. Na-cephapirin was prepared in PBS; and Cefa-Lak@(Bristol-Meyers) was used according to package insert instructions. 2 Cytokines were infused intramammary through the teat canal after the morning milking at the indicated intervals. The total dose was delivered in a total volume of 10 ml of sterile PBS. The biological activity of r-BOIL-l, and I-BOIL-~ was 32 000 UnitslDg and 22 000 Unitslpg respectively. A total dose per infusion of 200 mg of Na-cephapirin or Cefa-Lak@ was administered after consecutive afternoon and morning milkings, according to described procedures. 3 The % RESPONDING = quarters which transiently (> 1 milking) were free of S. LII(T~US divided by the total number of quarters treated. 4 The % CURED quarters = # of quarters which cleared their infection and remained free of S. LII(TPUS divided by the total number of quarters treated. Infections were considered to have relapsed if there were > 1 Cl- La mun three consecutive morning milkings, or > 20 CFUs for two consecutive milkings. Less than 2% of infected spontaneously cure; and > 97% of all quarters which relapse after therapy do so within the 14 day monitoring period.

Treatment of infections with Cytokines / 281

also followed by cycling of the host cells, primarily PM&, and the cycling of host cells is asynchronous with bacteria cycling. The qualitative changes in the host’s PMNs have been shown to contribute to this cycling infection. 12~4 Although the host may be as much as 99.99% efficient (see Fig. 1) at eliminating S. uureus, the remaining inefficient host cells with viable intracellular bacteria can permit a reinfection as well as protect the bacteria from antibiotic therapy.+*5,*6 Furthermore, antibiotic therapy can also inhibit the normal defense mechanisms of the host by decreasing phagocytic function27 and thereby further exacerbate the inefficiency of the phagocytic cell and subsequent relapse of the infection. The biological activity of resident PMNs in milk and the mammary gland have been shown to be significantly suppressed .*ssO Furthermore, the biological activity of phagocytosis and inducible superoxide production of PMNs from mastitic gland is quite variable. Recent reports have also suggested that some Holstein cattle may suffer from an adherence deficiency31 similar to Leukocyte Adherence Deficiencies (LAD) described in humans.32 We have demonstrated that the quantity and quality of PMNs within the mammary gland can be precisely modulated by various biological response modifiers, such as homologous cytokines. Although an activation of phagocytosis is necessary, it is insufficient without proper activation of the bactericidal components of the cell to efficiently eliminate bacteria.4,25 It is this dissociation of activation of phagocytosis without activation of the bactericidal activity that can in fact contribute to a source of reinfection as stated earlier. From the present data, it appears that the activation of superoxide may indeed precede the activation of phagocytosis upon cytokine therapy. In relative terms, this phenotype may effectively kill bacteria that is ingested, but their lack of phagocytic activation minimizes their contribution in eliminating an infection. Both of these activities are significantly inhibited or depressed in normal milk PMNs, as compared to circulating peripheral blood PMNs. This would suggest that phagocytosis may be the limiting biological activity for the PMN in the case of cytokine activation in this in vivo model system. We are presently exploring the role of intracellular adhesion molecules and integrins in the pathophysiology of mastitis, as well as secondary cytokine induction. These results emphasise that the approach to controlling mastitis, as well as other bacterial infectious diseases, should include a therapy which not only increases the quantity, but also the quality of PMNs in the target organ. An ideal therapeutic would perhaps have no effect on PMN numbers and optimally activate the resident cells in the gland. Clearly the host is efficient at dealing with the majority of S. uureus

infections, and it is likely that glands exposed to S. aureus, as with most bacterial infections, can be effectively eliminated during the initial infection. An immunocompromised state may play a role in the progression to an established infection.33 The role of the clinician in prevention is to prevent exposure or maximize the natural immunity perhaps to acute infection. This later approach to maximize natural immunity has been attempted through a variety of methods with respect to bovine mastitis.30,34J5 Although these approaches clearly increase the quantity of PMNs in the normal mammary gland, there is no information on the quality of these cells, and their effectiveness has at best been modest. We predict from the present study that an approach which concomitantly activates quantity and quality of cellular defense mechanism, with particular emphasis on the quality or activation, would be the most effective preventative. These potential beneficial effects to the host must be balanced with the potential local and systemic side effects that cytokines have been known to elicit. Although other investigators have suggested that IL-2 can have adverse short term effects on the morphology of the mammary gland,36 we have not observed any long term detrimental effects on milk production after cytokine treatment of hundreds of dairy cows with doses lOO-fold greater than those reported. The source of the recombinant cytokine, possible endotoxin content, and short term versus long term effects may explain these discrepancies. We have shown that chemoattraction and chemoactivation can be distinctly regulated in vivo. In protecting the mammary gland from challenge with S. uureus, activation of the resident PMN population appears to be sufficient as evidenced by the data with GM-CSF. Similarly, in the treatment of established infections, a therapeutic which concomitantly induces a qualitative, as well as quantitative change in the phagocytic cells of the gland tended to be the most efficacious (IL-l versus IL-2). Activation of phagocytic cells without a significant cellular influx may be sufficient to prevent disease because of the initial low bacterial burden, as evidenced by the data with r-BoGM-CSF. Presumably, a cytokine which concomitantly serves as a chemoattractant and chemoactivator, either alone or in conjunction with a conventional antibiotic therapy, would be the most effective therapeutic.36,37 While many of the therapeutic applications of systemic administration of cytokines have yielded only modest results,38-40 approaches which have focused on local or targeted cellular activation have seemed to be more efficacious.41 An analogous approach for infectious diseases, in which the specifically affected organ can be targeted may also provide a more efficacious modality for cytokine therapy. These observations have a wide

282

/ Daley

CYTOKINE,

et al.

application for the prevention and therapy of any infectious disease, both in domestic animals as well as human health, and perhaps suggest that targeted organ systems may provide a new practical application for cytokines in human medicine.

MATERIALS AND METHODS Animals

and Experimental

Model

A total of 32 early to mid-lactation Holstein-Fresian dairy cattle were used. Experimental infections were established with S. aweuS (strain Newbould 305)” as previously described.12 All mammary glands to be inoculated were free of bacteria on 5 consecutive morning milkings prior to infection and had individual quarter somatic cell counts (SCC) less than 200 000 cells/ml of milk. Inoculum were prepared from a four hour culture grown in trypticase soy broth and diluted in phosphate buffered saline, pH 7.0, (PBS) to contain 50-300 colony forming units (CFU) per ml. In vivo inoculations were made by infusion of 1.O ml diluted culture of S. uur. a into the teat cistern of a milked-out mammary quarter. Infections were followed daily for the presence of S. aureus. All infections were established for a minimum of three weeks before being used for analysis. Quarters were considered to respond to therapy if the morning milk sample was free of S. aureus for one milking or more after the last therapy. Quarters were monitored for fourteen days after the last therapy and were considered cured if samples continued to be negative throughout this monitoring period. Even with antibiotic therapy, we have observed with approximately 1000 experimental infections monitored on a daily basis for 30 days that the > 98% of all quarters that relapse will do so within this 1Cday period. Similarly, with this model system using daily samplings and the aforementioned criteria for ascribing an infection, there is less than 2% false negatives.

Microbiology Individual quarter milk samples were cultured by streaking 0.10 ml on a petri dish containing trypticase soy agar supplemented with 5.0% whole ovine blood. Plates were incubated aerobically at 37°C and examined after 24 and 48 hours incubation. Staphylococcus uureus isolates were counted and identified on the basis of colony morphology, hemolytic patterns and confirmed by positive coagulase test.42

Somatic Cell Determination Somatic cell counts were determined by CottIter@ counting and performed on 1 ml aliquots of fresh milk samples according to an adaptation of a previously described procedure.i2+3 Cells were pelleted initially at 2200 X’s g for 10 minutes and the resulting cellular pellet washed twice in PBS at 1200 X’s g for 5 minutes. The cellular pellet was diluted 1:50 or 1:500 in Isoton (Coulterm Diagnostics) and the fresh cell preparation was counted and gated at 7.4~1 diameter on a Coulter ZM (Coulter@ Electronics) based on

Vol.

5, No. 3 (May

1993: 276284)

the cell population profiled on a Coulter@ C256 Channelyzer. Only viable somatic cells were counted using this method.

Quantitti’on

of Phagocytosis-Flow

Cytometry

Ten to 250 ml of milk was collected and passed through cheese cloth to remove clumped PMNs. Milk was then centrifuged at 4°C 1200 x g for 30 min. The fat and skim milk removed, and cell pellets were washed twice by resuspending in PBS and counted on a Coulter@ counter. Cells were brought to a final concentration of 1 x 106 cells/ml with cold RPMI-1640 (GIBCO Grand Island, NY) supplemented with 5% fetal calf serum (GIBCO). Two micron fluorescent beads (Polysciences, Warrington, PA) were added to a concentration of 5.52 x 109 beads per 106 cells41 and then incubated at 37°C for 30 minutes. The samples were analyzed by flow cytometry (Coulter EPICS@’753, Princeton University, Princeton, NJ) as previously described.12

Whole Cell NADPH-Dependent

Superoxide Assay

The NADPH-dependent 0, generation was assayed as an indication of PMN activation.45 Briefly, milk from quarters was centrifuged at 1200 x g for 15 min to pellet milk PMN. Cell pellets were washed thrice in PBS, resuspended and counted. Fifty pl of 5 x 106 milk PMN/ml of PBS supplemented with glucose, magnesium, and calcium (PBS-G) were added to each microtiter well containing 200 pl of PBS with glucose and a final concentration of 75 PM 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 rig/ml to activate 0, production. The specific rate of reduction of cytochrome-c was measured using a narrow band 550nm filter on a ThermoMax@ kinetic plate reader (Molecular Devices Corporation, Menlo Pk, CA) as an indirect measure of 0, formation. Data was expressed as the mean rate of 0, generation in nM O,/min per 107cells.

Recombinant

Cytokines

Recombinant cytokines were prepared by IMMUNEX corporation and assayed as previously reported.is20 The biological activity of r-BOIL-la, r-BOIL-2 and r-BoGMCSF was 32 000 Unitsipg, 22 000 Units/pg and 38 000 Units/pg respectively. The r-BOIL-l, was assayed using the co-stimulator assay with homologous bovine thymocytes.18 The r-BOIL-2 was assayed using the bovine IL-2 dependent cell line, BT-2.19 The r-BoGM-CSF was assayed using the proliferation of non-adherent bone marrow cells.20 All in vitro biological activities were defined by one unit being equal to the amount requiring 50% of the maximal biological response. Cytokines were prepared for in vivo use by diluting them to concentrations such that the desired total dose was contained within 10 ml of phosphate buffered saline. Doses were administered in syringes fitted with teat cannulas by direct infusion through the teat cistern. A minimal biological response was defined as the lowest dose of cytokine administered into the mammary gland which elicited a significant influx of somatic cells as compared to the pretreatment SCC. A maximal tolerable dose was defined as a dose where significant increases in SCC were observed often plateauing in the dose response. In addition,

Treatment of infections with Cytokines I 283 sustained and severe clinical signs (fever, diarrhea, dyspnea, loss of milk production, external glandular swelling, loss of appetite, etc.) were observed in at least 60% of the animals.

Acknowledgement The authors thank Dr Karl Simkins, Linda James, Robert VanMater, Daniel Smith, and Raymond Brazina of the Clinical Laboratory, and MS Debbie Search of the Immunology Group, American Cyanamid Company for their assistance in herd management, sampling and bacteriology; Drs Paul Baker, Charles Maliszewski and David Urdall, and MS Kathy Picha of IMMUNEX Corporation for their assistance in supply of cytokine, scientific discussions and bioassays; and Mr Jerome Zawadzki of the Flow Cytometry Facility, Princeton University for use of the flow cytometer.

ABBREVIATIONS Abbreviations used in this manuscript: r-BOIL-l,, recombinant bovine interleukin 1 beta; r-BOIL-2, recombinant bovine interleukin 2; r-BoGM-CSF, recombinant bovine granulocyte-macrophage colony stimulating factor; PMNs, polymorphonuclear cells; SCC, somatic cell count (milk); CFU, colony forming unit (bacterial); PBS, phosphate buffered saline (pH 7.0); Cefa-lak@, commercial mastitis therapeutic containing Na cephapirin formulated in a peanut oil base.

REFERENCES 1. Wilson R, Hamburger M (1957) Fifteen years experience with staphylococcus septicemia in a large city hospital: Analysis of 55 cases in the Cincinnati General Hosoital 1940 to 1954. Am J Med 221437-457. 2. Wolleman OJ, Finland M (1942) Pathology of staphylococcal pneumonia complicating clinical influenza. Amer J Path 19:23-37. 3. Craven N, Anderson JC, Wilson CD (1983) Penicillin and (cloxacillin)-tolerant Staphylococcus aureus from bovine mastitis: Infection and lack of correlation between tolerance in vitro and response to therapy in vivo. Res Vet Sci 34:266271. 4 Li IW, Mudd S, Kapral FA (1963) Dissociation of phagocytosis and intracellular killing of Staphylococcus aureus by human blood leukocytes. J Immunol90:804-809. 5. Goronzy J, Weyand C, Quan J, Fathman CG, O’Hanley P (1989) Enhanced cell-mediated protection against fatal Eschericia coli septicemia induced by treatment with recombinant IL-2. J Immunol 142:1134-1138. 6. Cross AS, Sadoff, Kelly N, Brenton E, Gemski P (1989) Pretreatment with recombinant murine tumor necrosis factoraicachectin and murine interleukin-1, protects mice from letohal bacterial infection. JEM 169:2021-2127.

7. Van der Meer JW, Helle A, Aarden L (1989) Comparison of the effects of recombinant interleukin-6 and recombinant interleukin-1 on non-specific resistance to infection. Eur J Immunol 19:413-416. 8. Christ WL, Harmon RJ, Newman LE (1983) How much does mastitis cost? Udder Topics 6:69. 9. FDA Office of the Federal Register, National Archives and Records Administration (1989) Codd of Federal Regulations, Publication #21CFR, US Government Printing Office. 10. Aungier SPM, Austin FH (1987) A study of the efficacy of intramammary antibiotics in the treatment of clinical mastitis. Br Vet J 143:8&90. 11. Newbould FHS, Neave FK (1965) The response of the bovine mammary gland to an infusion of staphylococci. J Dairy Res 32:163-170. 12. Daley MJ, Oldham ER, Williams TJ, Coyle PA (1991) Bovine Mastitis. I. Quantitative and qualitative properties of host polymorphonuclear cells during Staphylococcus aureus mastitis infection. Am J Vet Res 52~474-479. 13. Oldham ER, D’Angelo D, Daley MJ (1991) LYSOSTAPHIN: Use of a recombinant bactericidal enzyme as a mastitis therapeutic. J Dairy Sci 74:4175-4182. 14. Springer TA (1990) Adhesion receptors of the immune system. Nature 346:425-434. 15. Piper PJ (1983) Pharmacology of leukotrienes. Br Med Bull 39~255-259. 16. Davies P, Bailey PJ, Goldenberg MM (1984) The role of arachidonic acid oxygenation products in pain and inflammation. Ann Rev Immunol2:335-357. 17. Fleischmann J, Golde DW, Weisbart RH, Gasson JC (1986) Granulocyte-macrophage colony-stimulating factor enhances phagocytosis of bacteria by human neutrophils. Blood 68:708-711. 18. Maliszewski CR, Baker PE, Schoenborn MA, Davis BS, Cosman D, Gillis S, Cerretti DP (1988) Cloning, sequence and expression of bovine interleukin 1, and interleukin lp complementary DNAs. Molecular Immunology 25:429-435. 19. Cerretti DP, McKereghan K, Larsen A, Cantrell MA, Anderson D, Gillis S, Cosman D, Baker P (1986) Cloning sequence and expression of bovine interleukin-2. PNAS 83:3223-3227. 20. Maliszewski CR, Schoenborn MA, Cerretti DB, Wignall JM, Picha KS, Cosman D, Tushinski RJ, Gillis S, Baker PE (1988) Bovine GM-CSF: Molecular cloning and biological activity of the recombinant protein. Molecular Immunol25:843-850. 21. Djeu JY, Li JH, Wei S, Rul H, Pearson CA, Leonard WJ and Blanchard D (1993) Function associated with IL-2 receptor-p on human neutrophils. J Immunol 150:96@-970. 22. Hammond WP, Price TH, Souza LM, Dale DC (1989) Treatment of cyclic neutropenia with granulocyte colonystimulating factor. NEJM 320:1306-1311. 23. Nemunaitis J (1990) Use of recombinant human granulocyte-macrophage colony-stimulating factor in graft failure after bone marrow transplantation. Blood 76: 246-253. 24. Sears PM, Smith BS, English PB, Herer PS, Gonzales RN (1990) Shedding pattern of Staphylococcus aureus from bovine intramammary infections. J Dairy Sci 73:2785-2789. 25. Marodi L, Leijh PCJ, Van Furth R (1983) A micromethod for separate evaluation of phagocytosis and intracellular killing of Staphylococcus aureus by human monocytes and granulocytes. J Immunological Methods. 57:353-361. 26. Craven N, Anderson JC (1984) Phagocytosis of Staphylococcus aureus by bovine mammary gland macrophages and intracellular protection from antibiotic action in vitro and in vivo. J Dairy Res 51:513-520. 27. Ziv G, Paape MJ, Dulin AM (1983) Influence of antibiotics and intramammary antibiotic products on phagocytosis of Staphylococcus aureus by bovine leukocytes. Am J Vet Res 44:385-388.

284 I Daley et al. 28. Niemialtowski M, Nonnecke BJ, Targowski SP (1987) Phagocytic activity of milk leukocytes during chronic staphylococcus mastitis. J Dairy Sci 71:78&787. 29. Paape MJ, Pearson RE, Schultze WD (1978) Variation among cows in the ability of milk to support phagocytosis and in the ability of polymorphonuclear leukocytes to phagocytose Staphylococcus aureus. Am J Vet Res 39: 1907-1910. 30. Craven N, Williams MR (1981) Defenses of the bovine mammary gland against infection and prospects for their enhancement. Vet Immunol and Immunopath 10:71-127. 31. Kerhli ME, Schmalstieg FC, Anderson DC, Van Der Maaten MJ, Hughes BJ, Ackermann MR, Wilhelmsen CL, Brown GB, Stevens MG, Whetstone CA (1990) Molecular definition of the bovine granulocytopathy syndrome: identification of deficiency of the Mac-l (CDllWCD18) glycoprotein. Am J Vet Res 51:1826-1836. 32. Anderson DC, Springer TA (1987) Leukocyte adhesion deficiency: an inherited defect in the Mac-l, LFA-1 and p 150.95 glycoproteins. Ann Rev Med 38:175-194. 33. Kehrli ME, Goff JP, Nonnecke BJ (1988) Effects of immunosuppression in Holstein cows on Staphylococcus aureus shedding in bovine milk. J Dairy Sci 71 (Suppl):166. 34. Paape MJ, Schultze WD, Guidry AJ, Kortum WM, Weinland BT (1981) Effect of intramammary polyethylene device on the concentration of leukocytes and immunoglobulins in milk on the leukocyte response to Eschericia coli endotoxin and challenge exposure with Staphylococcus aureus. Am J Vet Res 421774-782. 35. Smith KL (1982) Mastitis Control: A Discussion. J Dairy Sci 66: 1790-1794. 36. Reddy P, Reddy D, Pruiett S, Daley M, Shirley J, Chengappa M, Blecha F (1992) Immunomodulation of mastitis with bovine recombinant interleukin-2. Cytokine 4:227-231.

CYTOKINE,

Vol. 5, No. 3 (May 1993: 276-284)

37. Daley MJ, Dougherty R, Coyle P, Furda G, Johnston P (1992) Potentiation of antibiotic therapy of Staphylococcus aureus mastitis by recombinant bovine interleukin-2. J Dairy Sci 75:3330-3338. 38. Winkelhake JL, Gauny SS (1990) Human recombinant interleukin-2 as an experimental therapeutic. Pharmacological Rev 42:1-28. 39. Taguchi T (1984) Clinical effects of interferon on malignancies. Japan J Cancer Chemo 11:194-206. 40. Dinarello CA, Cannon JG, Mier JW, Bernheim HA, LoPreste G, Love DL, Webb AC, Auron PE, Reuben RC (1986) Multiple biological activities of human recombinant interleukin 1. J Clin Invest 77:1734-1739. 41. Kradin R, Yamin R, Kurnick J (1988) Immunological effects of adoptive immunotherapy with IL-2: an overview. Path01 Immunopathol Res 7:434. 42. National Mastitis Council Inc (1981) Microbiological procedures for use in the diagnosis of bovine mastitis. 2nd ed, Carter Press Inc, Ames, Iowa. 43. Phipps LW, Newbould FHS (1966) Determination of leukocyte concentrations in cows’ milk with a Coulter@ counter. J Dairy Res 33:51-54. 44. Steinkamp JA, Wilson JS, Saunders GC, Stewart CC (1982) Phagocytosis: Flow cytometric quantitation with fluorescent microscpheres. Science 215:64-66. 45. Mayo LA, Curnutte JT, (1990) Kinetic microplate assayfor superoxide production by neutrophils and other phagocytic cells. Meth in Enzymol186:567-579.