Superantigen-Dependent Accelerated Death of Bovine Neutrophilic Granulocytes in vitro is Mediated by Blood Mononuclear Cells

Immunobiology (2000) 202, pp. 493- 507

© 2000 Urban & Fischer Verlag http://www.urbanfischer.de/journals/immunobiol

1Immunology

Unit, School of Veterinary Medicine, Hannover, Germany; 2Dermatology Unit, Department of Small Animal Medicin.-e and Surgery, Royal Veterinary College, London, UK

Superantigen-Dependent Accelerated Death of Bovine Neutrophilic Granulocytes in vitro is Mediated by Blood Mononuclear Cells HANS-JOACHIM SCHUBERTH l , CORINNA KRUEGER l , ANKE HENDRICKS 2 , DIANE BIMCZOK l , and WOLFGANG LEIBOLD l Received August 25, 1999 . Accepted October 10, 2000

Abstract While classical interactions of bacterial superantigens (SAgs) with antigen presenting cells and T cells have been studied intensively, the potential interactions of SAgs with granulocytes (PMNs) have gained much less attention. We investigated if in the bovine system SAgs have any direct or indirect influence on the fate of granulocytes, which are among those cells primarily responsible for the elimination of superantigen-producing bacteria. The tested SAgs (SEA, SEB) had no apparent direct effect on PMN viability (neutrophils and eosinophils). However, in the presence of blood mononuclear cells (MNCs), SAgs led to an accelerated death of neutrophils but not of eosinophils. Compared to medium controls, in SAg-stimulated cultures only about 20-50 % of the neutrophils survived after 24 hours in vitro. Accelerated death of neutrophils required the presence of at least 100/0 MNC and started between 2.5-24 h after initiation of the co-culture between MNC and PMN. Minimal effective SEA concentrations ranged between 10-100 pg/l (SEB 0.1-10 ng/l). The effect could be mimicked by culture supernatants of SAg-stimulated MNCs, suggesting that direct cell-cell interactions are not required for the killing. In the human system, where we tested the role ofTNF-a, an antibody specific for this cytokine was not able to abolish the death of human neutrophils. Brefeldin A, an inhibitor of golgi transport and cytokine secretion, which blocked the SAg-induced activation of bovine MNC did not abolish the killing of neutrophils. Blocking of nitric oxide generation or PGE2 synthesis also could not alter the SAg-induced killing of bovine neutrophils. The observed indirect negative effects of SAgs on neutrophils may provide new insights in mechanisms by which superantigens modulate the hosts immune response.

Introduction Bacterial superantigens (SAgs) bind to MHC class II molecules on antigen presenting cells and subsequently to certain V~ regions ofT cell receptors (1). The consequences of this type of interaction are manyfold. On the T cell side this leads to polyclonal proliferation, induction of apoptosis and/or anergy of the responding cells (2, 3) Similar Abbreviations: PGE 2 = prostaglandin E2 ; MNC = blood mononuclear cells; PMN = polymorphonuclear cells; SEA and SEB = staphylococcal enterotoxin A and B; SAg = superantigen; cMed, cSEA = conditioned supernatants of mononuclear cells cultured in the presence of medium or SEA. 0171-2985/00/202/03-493 $ 15.00/0

494 . H.-J.

SCHUBERTH et al.

responses are observed with SAg-presenting cells (4). In summary, the effects ofsuperantigens (SAg) on lymphoid cells, especially T and B lymphocytes are well recognized. Effects on other cells, particularly granulocytes, have gained considerably less attention. However, these cells are amongst the earliest that appear after infection with SAgproducing staphylococcus aureus under natural conditions or after in vivo challenge with superantigens (5). Teleologically, a local modulatory effect of SAg on innate immune cells would seem even more important for the survival of SAg-producing bacteria compared to the perturbation of cells of the adaptive branch of the immune system in the periphery. This type of alteration of the host immune response may also playa role in diseases of veterinary importance, where SAg-producing staphylococci (5. aureus, S. intermedius) are the causative agents for severe forms of mastitis in cattle (6) or pyoderma in dogs (7). Reports of direct interactions of SAg with neutrophils are scarce. Toxic shock syndrome toxin-l (TSST-l), but not SEA or SEB decreased human neutrophil bactericidal activity in vitro (8). TSST-l was also able to induce heat shock proteins (hsp70, hsp72) in isolated human PMN (9) and could potentiate the FMLP-induced LTB 4 generation of human PMN (10). Thus, at least TSST-l may significantly affect signal transduction pathways of human PMNs. Recently it was reported that SEA, SEB and TSST-l had no direct effect on human neutrophil apoptosis (11). Evidence for SAg effects on functional properties of neutrophils is mainly indirect. The migratory-inhibiting effect was demonstrated for TSST-l and was mediated by stimulated mononuclear cells (12). On the other hand, intraveneously administered SEA acted pro-migratory (13). Other reports stated increased granulocyte sensitivities to TNF-amediated signals after systemic T cell-dependent responses to SEB or phasic alterations in neutrophil CD 11 b/c expression after induction of lethal sepsis with enterotoxin-producing staphylococci (14). It is well established that human neutrophilic and eosinophilic granulocytes can express MHC class II molecules, especially after activation by IFN-y, GM-CSF, IL-3 and IL-5 (15-17). MHC class II expression was rather low, but, such granulocytes were shown to be capable of SAg presentation to T cells resulting in activation and proliferation (1). Thus, MHC class II expression may represent a rational mechanism, by which T cells and granulocytes get into close physical contact. One consequence of this type of interaction may lead to an enhanced killing of granulocytes, a phenomenon known as SDCC (superantigen-dependent cellular cytotoxicity) which was described for the elimination of autologous B cells, monocytes and activated MHC class II-positive T cells (19). We therefore asked whether SAg directly or indirectly modulate the vitality of neutrophils and/or eosinophils and we will present evidence that Staphylococcus aureus superantigens lead to a selective accelerated death of bovine neutrophils but not eosinophils in vitro. This death is mediated by SAg-activated mononuclear cells and does not require close physical contact between neutrophils and MNC.

Materials and Methods Cell separation

Blood from healthy cattle (n = 7) was obtained by venipuncture of the vena jugularis externa into heparinized vacutainer tubes (Becton Dickinson, Heidelberg, Germany) or tubes containing EDTA. The blood was layered on Ficoll-Isopaque and centrifuged at 10°C for 30 min at 1200 g. The interphase

Superantigen-induced, MNC-mediated accelerated death of bovine neutrophils in vitro .

495

contaning mononuclear cells was washed 3 times in PBS (200 g, 100 g, 80 g) and finally resuspended in culture medium (see below) or PBS. The purity of the cell population regarding platelet or PMN contamination was checked by flow cytometry. The packed cells below the Ficoll-Isopaque were lysed with distilled water. This was repeated (usually twice) until complete erythrolysis. The bovine granulocyte population contained substantial numbers of eosinophils as determined by their characteristic higher autofluorescence after flow cytometric evaluation and staining of the cells with a monoclonal antibody specific for neutrophils (see below). Human MNC and PMN from venous blood of 11 volunteers were separated the same way. To obtain highly purified bovine neutrophils, the cell suspension containing neutrophils and eosionophils was further separated over a discontinuos Percoll® gradient (Biochrom, Berlin, Germany). Percoll® was made isotonic by adding 9 gil NaCl. This 1000/0 solution was diluted to 70% with PBS over which the cells separated in eosinophils (in the interphase) and neutrophils (below the Percoll®). Subsequently, cells were washed 3 times in PBS (200 g, 100 g, 60 g) and finally resuspended in culture medium or PBS (4 x 106 cellsl ml). In vitro stimulation assays

Media used fur culturing single cell suspensions were based on Iscove (Biochrom, Berlin, Germany) supplemented with 100/0 fetal calf serum, 1mmol/l sodium pyruvate, 1 x 10 5 U II penicillin, 100 mg/l streptomycin, 10 mmol/l Hepes). Mercaptoethanol (2-ME) was omitted from the culture medium. MNC, MNC+PMN (at various ratios) or PMN alone (always 2 x 10 5 total cells per well) were plated in roundbottom microtiter plates in 175 1I1 culture medium. Cultures containing PMN were incubated between 30 min and 24 h. Superantigens, Staphylococcus aureus enterotoxins A and B (SEA, SEB, Boehringer Ingelheim, Heidelberg, Germany) were used at final concentrations between 1pg/l and 100 f..Lg/l. Cultures without superantigens served as controls. All cultures were made in triplicate and incubated in a humidified atmosphere (37°C, 5% CO 2 ). SAg- (cSEA) and nil-conditioned supernatants from MNC (cMed) were obtained after various times (30min to 24h) in vitro. They were kept frozen (-20°C) until further use. Each experiment was repeated at least twice with cells of at least three unrelated animals or human beings. Additives to in vitro cultures

Where indicated, the cell cultures contained the following substances in final concentrations; Brefeldin A (Sigma, Deisenhofen, Germany), an inhibitor of golgi transport and cytokine secretion (5mg/l); brefeldin A was primarily dissolved in methanol; NG-monomethyl-L-arginine (L-NMMA, Calbiochem, Bad Soden, Germany), an inhibitor of nitric oxide synthetase (lmmolll dissolved in culture medium), indomethacin, an inhibitor of cyclooxigenase-2 (5mmolll), PGE 2 (10- 6 mollI) (both Sigma, Deisenhofen, Germany). Indomethacin and PGE 2 were primarily dissolved in ethanol (96%). Final concentrations of methanol ranged between 0.035-0.0050/0 (v/v), ethanol was present at 0.1 % (v/v). Control cultures proved that neither solvent had any influence on the observed effects (data not shown). Recombinant human tumor necrosis factor a (rhTNF-a; Alexis, Grunberg, Germany) was prediluted in culture medium and used at a final concentration of 5 J..Lg/l. A monoclonal antibody specific for human TNF-a (mAb TNF-D, Alexis, Grunberg, Germany) was used at Img/l (final concentration). Membrane immunofluorescence

Labelling of neutrophils was performed with mAb Bol16 (mouse IgM (20)). MAb BolI6 detects bovine neutrophils and a subpopulation of monocytic cells. MHC class II-positive cells were stained with mAb Bo 139 (21) specific for BoLA-DR molecules. Staining was performed according to standard procedures (21) using a FITC-labelled secondary antibody (goat anti mouse IgG (H&L), Dianova, Hamburg, Germany). Samples were evaluated by flow cytometry after gating on viable cells which excluded the dye propidium iodide (PI, final 2mg/l). Flow cytometric determination of viable cell numbers

Absolute numbers of viable cells were determined by the standard cell dilution assay (SCDA (22)) after flow cytometric acquisition (FacScan ®, Becton Dickinson, Heidelberg, Germany) of cultured cells. The

496 .

H.-]. SCHUBERTH et al.

procedure was adapted for the bovine system with slight modifications and was described previously (21).

Results SAg have no effect on purified polymorphonuclear granulocytes

When highly purified neutrophils (containing ~0.1 % MNC; ~20/0 eosinophils) were incubated between 30 min and 24 h in vitro, between 20% and 600/0 of the neutrophils died (data not shown). The fraction of surviving neutrophils was individually different and showed also considerable day to day variation. The decrease in viable neutrophil numbers was not due to increased adherence to the plastic surface (which was controlled microscopically, data not shown). The presence of SEA or SEB in a wide concentration range had no apparent influence on the kinetics of neutrophil loss in vitro (Fig. 2A). This could also be observed with less pure PMN preparations after Ficoll-isopaque-separation which contained 3-5% MNC and 11-300/0 eosinophils (data not shown).

medium

SEA

3h

..

24h

mAb Bol16 Figure 1. Depletion of neutrophils in cocultures ofboMNC and boPMN occurs after stimulation with SEA. A mixture of mononuclear cells and ficoll-separated PMNs were cultured for 3 h or 24 h in the presence or absence of SEA (1 ~g/l). Bovine neutrophils were labelled with a mAb specific for this cell type (upper right qu~drant). The lower right quadrant contains MNC and low propidium iodide-positive cells (PlIo). Plh1gh cells were gated out.

Superantigen-induced, MNC-mediated accelerated death of bovine neutrophils in vitro .

160 120

Ii

I-I'i-_

497

A)

Ii

D-g-Q-D

purified neutrophils

80 40 O-L.--...,----r-~~--,--..---....---r-

50

B)

40 30

neutrophils

20 10

30 25 eosinophils

20 15

90 75

60 45

. i-.-.-.-.-·,. g D-D-D-D-D-Q-Q - -4 -3 -2 -1 0

SEB

log (Jig/I)

1 2

•

• -.-.- Ii

·-Ii ...· '

g D-D-D-Q-D-D-~

mononuclear

cells

-.- cattle #1 -0 - cattle #2

- -6 -5 -4 -3 -2 -1 0

SEA

log (Jig/I)

Figure 2. Sensitive and neutrophil-selective death-inducing effect of SAgs. Variable concentrations of SEA (lpg/l- ltlg/l) or SEB (IOOpg/l- IOOtlg/l) were tested on A) purified neutrophils or B) cocultures of MNC and PMN (containing neutrophils and eosinophils) for 24h in vitro. Eosinophils were identified flow cytometrically in FLI/SSC dotplots based on their characteristic higher autofluorescence. Numbers of vital cells (means and standard deviation of triplicates) were determined by a quantitative flow cytometric procedure. Data are presented for 2 representative animals of 4 tested cattle. Note the differences in the threshold concentrations of SEA and SEB above which the accelererated death of neotrophils occur.

498 .

H.-J. SCHUBERTH et al.

SAg lead to an accerelated death of neutrophils in co-culture with MNC

When culturing MNC together with PMNs (containing neutrophils and eosinophils) in the presence of SEA (at a ratio MNC:PMN = 50:50) we initially saw an apparently selective loss of neutrophils (Fig. 1) after 24 h. This effect started between 10-200/0 MNC in the mixed cultures and showed a plateau above 20% MNC (data not shown). The SAg concentrations needed for this neutrophil killing were rather low. They ranged between 10-100 pg/l for SEA and 0.1-10 ng/l for SEB (Fig. 2B). Numbers of vital eosinophils and vital MNC remained nearly unchanged after 24 h of co-incubation of PMN with MNC (Fig. 2B). SAg-mediated killing of nPMN is due to mediators/cytokines released from MNC

To test, whether the SAg-dependent killing of nPMN requires physical contact between MNC and nPMN, we produced supernatants of MNC stimulated in vitro for 30 min, 2.5 h, 15 hand 24 h in the presence or absence of SEA (0.1 and 10 ~g/l) or SEB (1 and 100~g/I). This was performed with MNC from 3 cattle and the supernatants harvested at the same time points were pooled. When highly purified neutrophils were incubated

cattle #1 180 160 140 120 ~

~

100

=

80

~

60

I

~

~

Z ~

180

=

160

'5=

140

~

~ .....

120 100 80 60

cattle #2

:;.~, ~~'" ~/ "\~\

~

supernatants o nil-conditioned

SEA-conditioned • 0.1 Jlg/I • 10 Jlg/I

.~.\

i~!-\

\/

0==1\ ~ ~~~==·1

&\~

~

..L

•

o nil-conditioned

SED-conditioned • 0.1 Jlg/I • 10 Jlg/I

0.5 2.5 15 24 120 0.5 2.5 15 24 120

hours Figure 3. Accelerated death of neutrophils can be induced with supernatants of SAg-stimulated MNC. MNC of one cattle were stimulated in vitro for the times indicated (X-axes) with two concentrations of either SEA (upper panels) or SEB (lower panels) (filled symbols). Parallel cultures with 2x 10 5 MNC in triplicates were set up for each time point, SAg and SAg concentrations. Supernatants of MNC cultured in medium alone served as controls (open symbols). Highly purified bovine neutrophils of2 representative cattle (out of 4 tested) were incubated for 24 h in the presence of the supernatants after which absolute numbers of vital neutrophils were determined by flow cytometry (means and standard deviations of triplicates). Note the difference between the animals regarding the sensitivities for SAginduced MNC supernatants taken at different times.

Superantigen-induced, MNC-mediated accelerated death of bovine neutrophils in vitro . 499 Table 1. The effect of TNF-a and anti-TNF-a on the superantigen-mediated accelerated death of human PMN. vital PMN (x 10-3)

1

in percent

n

medium + TNF-a + anti-TNF-a + TNF-a+anti-TNF-a

180±25 a 105 ±43 b 182±35 a 169±49b

100 59±24 102± 5 94± 18

11 11 7 7

cMed + TNF-a + anti-TNF-a

214±26 a 136±44 b 190±49a

120 ± 16 77±26 109±35

11 7 8

cSEA + TNF-a + anti-TNF-a

102±31 a 88±27 b 125 ±51 a

59±24 51 ±20 71 ±29

11 7 11

Human PMN of 11 donors were incubated for 24 h in medium or in media supplemented with conditioned MNC supernatant from one donor (cMed) or with SEA-induced MNC supernatant from the same single donor (cSEA). Parallel cultures contained either TNF-a (5 J.1g/I), anti-TNF-a (1 mg/l), or the combination ofTNF-a and anti-TNF-a. The numbers of vital PMN cultured for 24h in medium alone were set as 100 % for each indiviual tested. The numbers ofvital PMN present in each tested combination were also expressed as percentage of vital PMNs present in the medium controls of each tested individual. 1) For each block (medium, cMed, cSEA), significances of the differences to media without TNF-a or anti-TNF-a were calculated using the paired t-test. Same small letters indicate no difference (p>0.05), different letters indicate significant differences (p
for 24 h with these pooled supernatants, an accelerated killing could be observed with supernatants of SAg-stimulated MNC taken between 2.5 hand 24 h after the initiation of the in vitro culture (Fig. 3). This was dependent on the individual from which neutrophils were separated. Whereas neutrophils from cattle #1 began to die in SAg-conditioned culture supernatants harvested between 2.5 hand 15 h (e.g. Fig. 3, cattle #1, SEA 10 flg/l), the neutrophils from cattle #2 began to die in supernatants harvested between 15 hand 24 h. This individual feature was seen with both SEA- and SEB-induced MNC supernatants and was independent of whether supernatants contained cytotoxic factors derived from autologous or allogeneic MNC (data not shown). The effects of such supernatants were dependent on the SAg and its concentration. Supernatants which were taken early (30 min, up to 15 h) after the initiation of the in vitro culture (e.g. Fig. 3, cattle #2, SEB 0.1 J.1g/l) even delayed the killing of neutrophils. Overall, the results suggest, that the accelerated, indirectly mediated death of neutrophils does not depend on direct cell-cell interactions between MNCs and neutrophils.

TNF-a

is not the responsible cytokine

Bovine-specific reagents were not available at the time of this study. Therefore, the role of induced TNF-a was tested in the human system. As with bovine cells, SAg-induced MNC supernatants (cSEA) resulted in an accelerated neutrophil death; about 59% (±24%) remained vital compared to the' medium controls (Tab. 1). TNF-a addition instead of cSEA produced similar effects (Tab. 1, 59%±24%). However, effects ofTNF-a

500 . H.-J.

SCHUBERTH et al.

Supernatants from MNCs stimulated with nil

SEA(1 Ilg/I) II

100 ~

C/lO

]~

80

0..11

8- 60

..... 0

;::ll-<

~§

a

40

~]

20

d .t::

U

;:> ;:l

S

0 +

+ +

+

+

+ +

presence of brefeldin A dialysation of supernatants

Figure 4. Brefeldin A is not able to inhibit the release of SEA-induced release of MNC mediators responsible for the accelerated death of bovine neutrophils. MNC were cultured in vitro for 24 h without (nil) or with addition of SEA. Some cultures contained brefeldin A (5 mgll) as indicated below the bars. Since brefeldin A was toxic for neutrophils (nil, middle bar) some MNC supernatants of parallel cultures were dialysed to remove brefeldin A (indicated below the bars). Highly purifed neutrophils were incubated in the presence of the various MNC supernatants for 24 h after which the total numbers of viable neutrophils were determined (means and standard deviation of triplicates). Neutrophil numbers were related to the numbers present in the medium control (nil, left bar = 100%). Data are shown for one representative animal (out of 3 tested). The differences between the dark bars were not significant (paired t-test, p >0.2).

and cSEA addition to PMN cultures of different individuals were not correlated (r = 0.42, p > 0.2). Although addition of an antibody specific for TNF-a could partially restore the cSEA effect (from 590/0 to 71 0/0, Tab. 1) this was not significant (p > 0.07) and could not be seen in every individual tested (data not shown). Thus, TNF-a did not seem to be the responsible superantigen-induced cytokine for the accelerated death of neutrophils. The role of brefeldin

A sensitive

cytokines

Another approach to test whether secreted cytokines are responsible for the accelerated death of neutrophils was to block the cytokine secretion of SAg-activated MNCs with brefeldin A. Supernatants of SEA-stimulated MNC were produced in the presence and absence of brefeldin A. Subsequently, brefeldin A in the supernatants was removed by sephadex G 10 gel dialysation since it induced an accelerated neutrophil death when introduced in cultures of purified neutrophils: compared to the medium control (= 1000/0) only about 61 % neutrophils remained vital after 24 h in vitro. This reduction was nearly the same as seen with SAg-stimulated MNC supernatants (Fig. 4). Dialyzed supernatants of SAg-stimulated MNC in the presence of brefeldin A still induced an accelerated killing of neutrophils (Fig. 4) although the effects were slightly less pronounced. Thus, brefeldin A, although inhibiting the SAg-induced MNC blastogenesis (proving the inhibition of cytokine secretion, data not shown) was not able to inhibit the SAg-mediated release of MNC mediators which are relevant for the killing of neutrophils.

Superantigen-induced, MNC-mediated accelerated death of bovine neutrophils in vitro . 501

~

M

cattle #2

cattle #1

70 60

b 50

R 40 ~

30

3 .;;

20

~

SEA (~g/l)

10 0

-

-

~~

~~

V

~~

~~

~c~

0 0.01

0«;'V

~

.~

Figure 5. Effects of inhibition of nitric oxide synthesis (by L-NMMA), PGE2 synthesis (by indomethacin) and PGE2 supplementation in cocultures of bovine MNC and PMN. Mixtures of MNC and PMN (ratio 50:50) with and without stimulation by SEA were incubated for 24 h in the presence or absence of L-NMMA (Immol/l), indomethacin (5 mmol/l) or PGE2 (IO-6 molll). Numbers of viable neutrophils were determined flow cytometrically. This test was made with cells of 2 animals. One animal (cattle #2) was tested twice at different times with identical results. Significances were not determined.

Neiter

PGE 2 nor induced NO mediate the SAg-induced neutrophil death

In search for possible non-protein mediators responsible for the accelerated neutrophil death we examined the role of prostaglandin E2 (PGE2) and nitric oxide (NO). Figure 5 summarizes the effects of inhibitors of the inducible NO synthetase (iNOS) or cyclooxigenase 2 (COX-2) and the effect ofPGE2 addition to cocultures ofPMN and MNC in the presence or absence of SEA. Inhibition of the iNOS with L-NMMA in cultures without SEA either increased the numbers of vital nPMN slightly (e.g. from 38 x 103 to 50 X 103 , Fig. 5, cattle #1) or had no effect (Fig. 5, cattle #2). Control experiments proved, that SEA-stimulation of the MNC indeed induced NO release (data not shown), however, in no case did iNOS inhibition result in a modulation of the SEA-induced accelerated death of neutrophils. In contrast, PGE2 even increased the numbers of viable neutrophils in unstimulated cocultures of PMN and MNC in every case tested (Fig. 5). This was seen after addition ofPGE2 and, inversely after the inhihibition of endogenous synthesis ofPGE2 with indomethacin which decreased the numbers of vital nPMN. But again, neither addition of PGE 2 nor the inhibition of its synthesis could reverse the SEA-induced accelerated death of neutrophils. Inhibition of cyclooxigenase led to an even more accelerated decline in viable neutrophil numbers (Fig. 5). In summary, the soluble mediators of SAg-stimulated MNC, responsible for the accelerated killing of bovine neutrophils in vitro have not been identified yet.

Discussion The effects of bacterial superantigens on neutrophils have so far not been a focus of investigation, presumably because they do not primarily express the classical ligands (MHC

502 . H.-J.

SCHUBERTH et al.

class II, T cell receptor) to which these type of bacterial exotoxins bind. The recent observation that neutrophils and eosinophils can express MHC class II molecules upon stimulation (15 -17) directed our attention to this group of leukocytes. The rationale being the observation that, during an infection of the bovine udder with the potentially SAgproducing bacterium Staphylococcus aureus, significantly less granulocytes in the milk of affected mammary complexes are found, compared to infections with other gram-positive pathogens (Zerbe, personal communication). One reasonable explanation for this observation could be the SAg-induced accelerated death of neutrophils by superantigens. Therefore we tested whether SAgs mediate any influence on the survival rate of neutrophils in vitro and we presented evidence, that SAgs induce an accelerated killing selectively of neutrophils via stimulated mononuclear cells. Neutrophils are believed to be rather short-lived cells which undergo rapid apoptosis unless they are in a microenvironment supporting their survival (23). In vitro, the situation is more complicated in that apoptosis is initially induced but cells die secondary via necrosis or both mechanisms are induced in parallel (24). Even in cases where absolute numbers of vital PMNs considerably dropped in our system within the first 24 h, the flow cytom~trically detectable fractions of propidium iodide positive (PI+) PMNs (either PI low or PIh1gh) were not prominently different from those cultures where PMN numbers did not drop. This may indicate, that cellular desintegration, and hence, low numbers of PI+ cells due to rapid induction of secondary necrosis may have taken place in our in vitro system. For this reason we make no statement, whether the death of neutrophils was due to induction of apoptosis, necrosis or both mechanisms. On the other hand, the rapid desintegration of killed neutrophils in vitro after appropriate stimuli stresses the importance of quantitative detection systems of neutrophil death. In this regard, the SCDA method (22) proved to be of great value. This may also help to explain why our data are in striking contrast to findings by MOULDING et al. (11). They reported on an apotosis inhibiting effect ofSEB for human neutrophils in the presence ofMNC which they measured microscopically, determining the fraction of visible cells with apototic morphology. SEA or SEB did not affect the death kinetics of purified bovine PMNs. Even high concentrations had no influence on the total numbers ofvital neutrophils or eosinophils after 24 h in vitro (Fig. 2AB). Enhanced killing of neutrophils could only be observed either when co-culturing PMN and MNC in the presence of SAgs or by incubating purified neutrophils with SAg-conditioned supernates from MNC. The effectiveness of SAg-conditioned supernatants ruled out the possibility that neutrophils were killed via superantigen-dependent cellular cytotoxicity (SDCC) mediated by T cells (19). We observed this type of cellular cytotoxicity with bovine B cells (25) but this effect required much higher SAg concentrations. In contrast, SAg-dependent accerelated killing of neutrophils was exquisitely sensitive, starting between 10-100 pg/l (SEA) or 0.1-10 ng/l (SEB) (Fig. 2B). In addition, we did not observe any MHC class II expression on bovine neutrophils, either after preparation or after 24 h in vitro (data not shown). However, this expression could have been too low to be detected and therefore it remains possible, that the killing effect of SAg-conditioned MNC supernatants was due to the putative released MNC mediators/cytokines acting in concert with the still present SAgs. We therefore stimulated MNCs in the presence of agarose-immobilized SEA thereby eliminating free SEA from the conditioned supernatants. Such superantants still mediated the accelerated killing of bovine neutrophils (data not shown). Hence it follows that merely SAg-induced mediators/cytokines could be responsible for the observed effects.

Superantigen-induced, MNC-mediated accelerated death of bovine neutrophils in vitro·

503

The nature of the killing enhancing factors is still unclear. Several approaches to their characterization have failed so far. Surprisingly, most of the reported cytokines, mediators or substances which modulate the vitality of neutrophils are anti-apoptotic. This includes GM-CSF (26), G-CSF (27), IL-4 (28), IL-6 (29), IL-8 (30), IL-I0 (31), IL-15 (32), sodium butyrate (33), adenosine triphosphate and diadenosine polyphosphates (34), LPS (35), ICAM-l and CDl8-mediated adhesion (36), dexamethasone (37), leukotriene B4 (38), and antioxidants (39). Pro-apoptotic effects were shown for cycloheximide (27), exogenously applied nitric oxide (40), formyl-methionyl-Ieucyl-phenylalanine (FMLP) (38,41) and in part for the reactive oxygen species generated in neutrophils during the oxidative burst (42). For TNF-a and PGE 2 both pro- and antiapoptotic effects were reported (see below). The role ofTNF-a in our system was tested in the human system, where appropriate tools were right at hand (compared to the bovine system). It should be noted, that with human cells the same accelerated death of neutrophils could be observed as in the bovine system. Thus, the described phenomenon is not a species-specific one. The addition of rhTNF-a to human PMN resulted in an enhanced death of human PMN. This could be inhibited with an antibody specific for huTNF-a. However, this antibody was only partially able to revert the SAg-induced effect on human PMN (Table 1). This argued against TNF-a as the major responsible mediator for the death of PMN. In addition, the SAg stimulation of bovine MNC in the presence ofbrefeldin A had no effect on the release of death-accelerating factors (Fig. 4). Brefeldin A, an inhibitor of vesicularization and golgi transport repeatedly was reported to inhibit the release of several cytokines including TNF-a (43,44). Taken together, although we did not measure the released cytokines, we conclude that TNF-a is not the responsible cytokine for the accelerated death of bovine neutrophils. These findings are in agreement with several reports stating either no effect (27) or even an anti-apoptotic one ofTNF-a on human neutrophils (45-47). We cannot rule out the possibility that in vivo, SAg-induced TNF-a might contribute to an accelerated killing of bovine neutrophils as shown in the human system (46,48). Previously, we could demonstrate that SAgs induce PGE2 in bovine mononuclear cells (25). Regarding the role ofPGE 2 for neutrophil survival the available data are limited and still controversial despite numerous and pleiotropic effects of arachidonic acid metabolites on several immune mechanisms (49). Studies of KOLLER et al. (50) provided evidence for the involvement of arachidonic acid and distinct arachidonic acid metabolites in the regulation of apoptosis in human PMNs. The addition of arachidonic acid led to a significant increase in apoptosis. Leukotrienes seemed to be antiapoptotic, whereas prostaglandin synthesis inhibition with indomethacin inhibited apoptosis. In our hands, however, the opposite occured. PGE 2 when added exogeneously to purfied neutrophils clearly enhanced the numbers of viable cells after 24 h in vitro. Conversely, the inhibition of endogenous PGE2 synthesis with indomethacin enhanced the killing (Fig. 5), supporting the positive effect of PGE 2 for bovine neutrophils. Interestingly, these findings are in accordance with those obtained by WALKER et al. (51) who demonstrated an anti-apoptotic effect of PGE2 for human neutrophils. Taken together, although PGE2 synthesis and secretion is induced by SAgs in bovine cells (25), this mediator is clearly not reponsible for the observed accelerated killing of bovine neutrophils due to its anti-apoptotic effects.

The same probably holds true for another mediator, nitric oxide, whose synthesis inhibition in bovine MNCs was unable to block the SAg-induced indirect killing of bovine

504 . H.-]. SCHUBERTH et al. neutrophils (Fig. 5). We knew from earlier studies, that all tested SAgs are able to induce NO in bovine MNCs in considerable amounts (21). However, in contrast to other species, induced NO proved to be unable to modulate the proliferative reaction of bovine MNC. These differential effects in distinct species (52) may explain why others described NO as pro-apoptotic for human granulocytes (40) or found evidence for NO as a proapoptotic mediator for neutrophils in the development of skin lesions of patients with anaphylactoid purpura (53). Taken together, the observed indirect negative effects of SAg on the viability of bovine neutrophilic granulocytes may provide new insights in important pathogenetic mechanisms by which SAgs modulate the immune response of the host. This may add to some earlier in vivo observations where infections of surgical wounds with TSST-1 producing Staphylococcus aureus not usually elicited a purulent response from the host (12). It may be speculated that in addition to a proposed inhibition of PMN migration to sites of infection, the induction of accelerated neutrophil death also accounts for the poor elimination of exotoxin-producing staphylococci in some cases.

References 1. GASCOIGNE, N. R.]. 1993. Interaction of the T cell receptor with bacterial superantigens. Semin. Immunol. 5: 13. 2. HERMAN, A.,]. W KApPLER, ~ MARRACK, and A. M. PULLEN. 1991. Superantigens: Mechanism

ofT-cell stimulation and role in immune responses. Annu. Rev. Immunol. 9: 745. 3. HUANG, L., and 1. N. CRISPE. 1993. Superantigen-driven peripheral deletion ofT cells. Apoptosis

occurs in cells that have lost the alpha/ beta T cell receptor. ]. Immunol. 151: 1844. 4. AOUDJIT, E, R. AL-DACCAK, C. LEVEILLE, and W MOURAD. 1995. Effects and mechanisms underlying the interaction of bacterial superantigens with MHC class II-positive cells. In: THIBODEAU,], and R. ~ SEKALY (eds). Bacterial superantigens: Structure, function and therapeutic potential. R. G. Landes, p 147-160. 5. HERZ, U., R. RUCKERT, K. WOLLENHAUPT, T. TSCHERNIG, U. NEUHAUS-STEINMETZ, R. PABST, and H. RENZ. 1999. Airway exposure to bacterial superantigen (SEB) induces lymphocyte-dependent airway inflammation associated with increased airway responsiveness - a model for non-allergic asthma. Eur. ]. Immunol. 29: 1021. 6. MATSUNAGA, T., S.1. KAMATA, N. KAKIICHI, and K. UCHIDA. 1993. Characteristics of Staphylococcus aureus isolated from peracute, acute and chronic bovine mastitis. ]. Vet. Med. Sci.

55: 297. 7. BURKETT, G., and L. A. FRANK. 1998. Comparison of production of Staphylococcus intermedius 8. 9. 10. 11. 12.

exotoxin among clinically normal dogs, atopic dogs with recurrent pyoderma,and dogs with a single episode of pyoderma.]. Am. Vet. Med. Assoc. 213: 232. BERGER, E. M., G. A. SHIBAO, S. N. BROWN, and ]. E. REPINE. 1988. Variable effect of toxic shock toxins from different sources on neutrophil function in vitro. Inflammation 12: 447. HENSLER, T., M. KOLLER, ]. E. ALOUF, and W KONIG. 1991. Bacterial toxins induce heat shock proteins in human neutrophils. Biochem. Biophys. Res. Commun. 179: 872. HENSLER, T., M. KOLLER, C. GEOFFROY, ]. E. ALOUF, and W KONIG. 1993. Staphylococcus aureus toxic shock syndrome toxin 1 and Streptococcus pyogenes erythrogenic toxin A modulate inflammatory mediator release from human neutrophils. Infect. Immun. 61: 1055. MOULDING, D. A., C. WALTER, C. A. HART, and S. W EDWARDS. 1999. Effects of staphylococcal enterotoxins on humn neutrophil functions and apoptosis. Infect. Immun. 76: 2312. FAST, D.]., ~ M. SCHLIEVERT, and R. D. NELSON. 1988. Nonpurulent response to toxic shock syndrome toxin I-producing Staphylococcus aureus. Relationship to toxin-stimulated production of tumor necrosis factor. ]. Immunol. 140: 949.

Superantigen-induced, MNC-mediated accelerated death of bovine neutrophils in vitro·

505

13. MILLER, E. J., A. B. COHEN, and B. T. PETERSON. 1996. Peptide inhibitor ofinterleukin-8 (IL-8) reduces staphylococcal enterotoxin-A (SEA) induced neutrophil trafficking to the lung. Inflamm. Res. 45: 393. 14. NEUMANN, B., B. ENGELHARDT, H. WAGNER, and B. HOLZMANN. 1997. Induction of acute inflammatory lung injury by staphylococcal enterotoxin B. ]. Immunol. 158: 1862. 15. GUIDA, L., R. E. O'HEHIR, and C. M. HAWRYLOWICZ. 1994. Synergy between dexamethasone and interleukin-5 for the induction of major histocompatibility complex class II expression by human peripheral blood eosinophils. Blood 84: 2733. 16. WELLER, ~ E, T. H. RAND, T. BARRETT, A. ELOVIC, D. T. WONG, and R. W FINBERG. 1993. Accessory cell function of human eosinophils. HLA-DR-dependent, MHC-restricted antigen-presentation and IL-1 alpha expression. J. Immunol. 150: 2554. 17. GOSSELIN, E. J., K. WARDWELL, W E RIGBY, and ~ M. GUYRE. 1993. Induction of MHC class lIon human polymorphonuclear neutrophils by granulocyte/macrophage colony-stimulating factor, IFN-gamma, and IL-3. J. Immunol. 151: 1482. 18. FANGER, N. A., C. LIU, ~ M. GUYRE, K. WARDWELL, J. O'NEIL, T. L. Guo, T. ~ CHRISTIAN, S. ~ MUDZINSKI, and E. J. GOSSELIN. 1997. Activation of human T cells by major histocompatability complex class II expressing neutrophils: proliferation in the presence of superantigen, but not tetanus toxoid. Blood 89: 4128. 19. HEDLUND, G., M. DOHLSTEN, ~ A. LANDO, and T. I
506 .

H.-]. SCHUBERTH et ai.

32. GIRARD, D., M. E. PAQUET, R. PAQUIN, andA. D. BEAULIEU. 1996. Differential effects of interleukin-15 (IL-15) and IL-2 on human neutrophils: modulation of phagocytosis, cytoskeleton rearrangement, gene expression, and apoptosis by IL-15. Blood 88: 3176. 33. MOULDING, D., ]. A. QUAYLE, R. E. STRINGER, C. A. HART, and S. W EDWARDS. 1996. Regulation of neutrophil apoptosis by sodium butyrate. Biologicals 24: 301. 34. GASMI, L., A. G. McLENNAN, and S. W EDWARDS. 1996. The diadenosine polyphosphates Ap3A and Ap4A and adenosine triphosphate interact with granulocyte-macrophage colony-stimulating factor to delay neutrophil apoptosis: implications for neutrophil: platelet interactions during inflammation. Blood 87: 3442. 35. WILLIAM, R., G. WATSON, H. ~ REDMOND, ]. H. WANG, and D. BOUCHIER-HAYES. 1996. Bacterial ingestion, tumor necrosis factor-alpha, and heat induce programmed cell death in activated neutrophils. Shock 5: 47. 36. WATSON, R. W, O. D. ROTSTEIN, A. B. NATHENS, ]. PARODO, and ]. C. MARSHALL. 1997. Neutrophil apoptosis is modulated by endothelial transmigration and adhesion molecule engagement. ]. Immunoi. 158: 945. 37. MEAGHER, L. C.,]. M. COUSIN,]. R. SECKL, and C. HASLETT. 1996. Opposing effects of glucocorticoids on the rate of apoptosis in neutrophilic and eosinophilic granulocytes. ]. Immunoi. 156: 4422. 38. HEBERT, M. ]., T. TAKANO, H. HOLTHOFER, and H. R. BRADY. 1996. Sequential morphologic events during apoptosis of human neutrophils. Modulation by lipoxygenase-derived eicosanoids. J.lmmunol. 157: 3105. 39. WATSON, R. W, H. ~ REDMOND,]. H. WANG, C. CONDRON, and D. BOUCHIER-HAYES. 1996. Neutrophils undergo apoptosis following ingestion of Escherichia coli. J. Immunol. 156: 3986. 40. FORTENBERRY, J. D., M. L. OWENS, M. R. BROWN, D. ATKINSON, and L. A. BROWN. 1998. Exogenous nitric oxide enhances neutrophil cell death and DNA fragmentation. Am. ]. Respir. Cell. Mol. BioI. 18: 421. 41. KETTRITZ, R., R.]. FALK, ]. C. JENNETTE, and M. L. GAIDO. 1997. Neutrophil superoxide release is required for spontaneous and FMLP-mediated but not for TNF alpha-mediated apoptosis. J. Am. Soc. Nephrol. 8: 1091. 42. NARAYANAN, ~ K., K. RAGHEB, G. LAWLER, and]. ROBINSON. 1997. Defects in intracellular oxidative metabolism of neutrophils undergoing apoptosis. J. Leukoc. BioI. 61: 481. 43. BAUMGARTNER, R. A., V. A. DERAMO, and M. A. BEAVEN. 1996. Constitutive and inducible mechanisms for synthesis and release of cytokines in immune cell lines. J. Immunol. 157: 4087. 44. SOLOMON, K. A., M. B. COVINGTON, C. ~ DECICCO, and R. C. NEWTON. 1997. The fate of pro-TNF-alpha following inhibition of metalloprotease-dependent processing to soluble TNFalpha in human monocytes. J. Immunol. 159: 4524. 45. LILES, W C., A. KIENER, J. A. LEDBETTER, A. ARUFFO, and S. J. KLEBANOFF. 1996. Differential expression of Fas (CD95) and Fas ligand on normal human phagocytes: implications for the regulation of apoptosis in neutrophils. J. Exp. Med. 184: 429. 46. MURRAY,].,]. A. BARBARA, S. A. DUNKLEY, A. E LOPEZ, X. VAN-OSTADE, A. M. CONDLIFFE, I. DRANSFIELD, C. HASLETT, and E. R. CHILVERS. 1997. Regulation of neutrophil apoptosis by tumor necrosis factor-alpha: requirement for TNFR55 and TNFR75 for induction of apoptosis in vitro. Blood 90: 2772. 47. SWEENEY,]. E, ~ K. NGUYEN, G. M. OMANN, and D. B. HINSHAW. 1998. Autocrine/paracrine modulation of polymorphonuclear leukocyte survival after exposure to Candida albicans. Shock 9: 146. 48. GON, S., T. GATANAGA, and E SENDO. 1996. Involvement of two types ofTNF receptor in TNFalpha induced neutrophil apoptosis. Microbiol. Immunol. 40: 463. 49. SNIJDEWINT, F. G. M., KALINSKI, E. A. WIERENGA,]. D. Bas, and M. L. KApSENBERG. 1993. Prostaglandin E2 differentially modulates cytokine secretion profiles of human T helper lymphocytes. J. Immunol. 150: 5321. 50. KOLLER, M., WACHTLER, A. DAVID, G. MUHR, and W KONIG. 1997. Arachidonic acid induces DNA-fragmentation in human polymorphonuclear neutrophil granulocytes. Inflammation 21: 463.

:r

:r

:r

:r

Superantigen-induced, MNC-mediated accelerated death of bovine neutrophils in vitro .

507

51. WALKER, B. A., C. ROCCHINI, B. H. BOONE, S. Ip, and M. A. JACOBSON. 1997. Adenosine A2a receptor activation delays apoptosis in human neutrophils. J. Immunol. 158: 2926. 52. ADLER, H., B. FRECH, M. THONY, H. PFISTER, E. PETERHANS, and T. W JUNGI. 1995. Inducible nitric oxide synthase in cattle. Differential cytokine regulation of nitric oxide synthase in bovine and murine macrophages. J. Immunol. 154: 4710. 53. BANNO, S., Y. TAMADA, Y. MATSUMOTO, and M. OHASHI. 1997. Apoptotic cell death of neutrophils in development of skin lesions of patients with anaphylactoid purpura. J. Dermatol. 24: 94. PO Dr. H.-JOACHIM SCHUBERTH, Immunology Unit, School of Veterinary Medicine, Bischofsholer Damm 15, 0-30173 Hannover, Germany. Tel.: ++49 (0) 511 856 7267, Fax: ++49 (0) 511 856 7682, e-mail: [email protected]