Inhibition of CD44, CD45 and LFA-3 mediated cytokine release from human monocytes by SK&F 86002 and pentoxifylline

0192-0561/93 $6.00 + .00 Pergamon Press Ltd. 01993 International Society for Immunopharmacology.

Int. J. lmmunopharmac., Vol. 15, No. 2, pp. 205-209, 1993. Printed in Great Britain.

INHIBITION OF CD44, CD45 A N D LFA-3 M E D I A T E D CYTOKINE RELEASE FROM H U M A N M O N O C Y T E S BY SK&F 86002 A N D PENTOXIFYLLINE UMA PRABHAKAR,* DAVID LIPSHUTZ and ALEMSEGED TRUNEH Department of Cellular Biochemistry and Immunology, SmithKline Beecham Pharmaceuticals, 709 Swedeland Road, King of Prussia, PA 19406, U.S.A. (Received 4 June 1992 and in final form 23 September 1992)

Abstract -- Compounds from two distinct pharmacological classes namely, SK&F 86002 and pentoxifylline,

were examined for their effects on TNFa and IL-1/3release by human monocytes stimulated with LPS or monoclonal antibodies to three cell surface glycoproteins, CD44, CD45 and LFA-3 (LFA-3 is also known as CD58). SK&F 86002, an inhibitor of 5-LO and CO in arachidonic acid metabolism, inhibited LPS-induced release of TNFa and IL-1/3 with an Ic~0of 1 /aM. At this dose, it also inhibited by >50°70, release of both cytokines induced by the three monoclonal antibodies. Pentoxifylline, a methylxanthine derivative with phosphodiesterase inhibitory activity, selectively inhibited LPS-induced TNFa release with an Its0 of 100/aM. TNFa and IL-1/3release mediated by the monoclonal antibodies were inhibited by less than 30070in the presence of 100/aM pentoxifylline. These results suggest that (a) LPS induced cytokine release shares a common step with the physiologically relevant stimuli (involving cross-linking of cell surface receptors), and that this pathway is sensitive to inhibition by SK&F 86002 and, (b) SK&F 86002 is more potent than pentoxifylline in inhibiting TNFcr and IL-I~ release induced by both stimuli.

TNFa and IL-I/3 are two cytokines which are important mediators in the initiation and regulation of immune and inflammatory responses, and also in a variety of inflammatory diseases including septic shock (Tracey et al., 1986). They can be produced by stimulating monocytes and macrophages with lipopolysaccharide (LPS), phorbol myristate acetate or gram-positive bacterial components (Le and Vilcek, 1987; Gerrard, Seigel, Dyer & Zoon, 1987). Recently, three monocyte surface glycoproteins were described as putative receptors responsible for the physiological triggering of TNFa and IL-1/3 (Webb, Shimizu, Seventer, Shaw & Gerrard, 1990). In that study, monoclonal antibodies to CD44, CD45 and LFA-3 induced TNFa and IL-1/J production by human peripheral blood monocytes. SK&F 86002 is a member of a novel class of nonsteroidal anti-inflammatory compounds that inhibit both the lipoxygenase and cycloxygenase pathways of arachidonic acid metabolism (Griswold, Webb, Schwartz & Hanna, 1987). This compound also inhibits the release of both TNFa and IL-1/J by LPSstimulated monocytes (Lee, Rebar & Laydon, 1989;

Lee, Votta, Dalton, Griswold, Bender & Hanna, 1990). There is a large, but conflicting body of literature on the effects of lipoxygenase inhibitors on TNFa and IL-I(J synthesis, and the exact mechanism by which SK&F 86002 and other such agents inhibit TNFa and IL-1/3 release remains largely unknown. In contrast pentoxifylline, a phosphodiesterase inhibitor, is a more selective inhibitor of LPS-induced TNFa release with no effects on IL-1/3 release. The observed effect of pentoxifylline on T N F a release appears to be mediated by the accumulation of cAMP (Endres et al., 1991). The aim of this study was to examine the effects of cytokine synthesis inhibitors against physiologically relevant stimuli, in order to determine whether or not activation of monocytes by such stimuli to release TNFa and IL-1/3 share a common step. We compared the effects of SK&F 86002 and pentoxifylline on TNFa and IL-1/3 release by human monocytes stimulated either with LPS or with monoclonal antibodies to three monocyte cell surface glycoproteins, CD44, CD45 and LFA-3 (CD58). In this report we demonstrate that the two stimuli used,

*Author to whom correspondence should be addressed. 205

206

U. PRABHAKARet al.

do indeed share a common SK&F 86002 sensitive step for the induction of TNFa and IL-lfl. Furthermore, SK&F 86002 is more potent than pentoxifylline in inhibiting T N F a and IL-lfl release induced by LPS or by physiologically relevant activators through crosslinking the CD44, CD45 and LFA-3 glycoproteins. EXPERIMENTAL PROCEDURES

Reagents

RPMI 1640 medium was purchased from M A Bioproducts (Walkersville, MD), and fetal calf serum (FCS) was obtained from Hyclone Laboratories Inc. (Logan, UT). Bacterial lipopolysaccharide (LPS, E. coil 055:B5) and pentoxifylline (Trental) were purchased from Sigma (St Louis, MO). Monoclonal antibodies to CD44, CD45 and LFA-3 were purchased from Amac Inc. (Westbrook, ME). The Limulus Amoebocyte Lysate (LAL) kit was purchased from Associates of Cape Cod (Woods Hole, MA). Peroxidase conjugated goat anti-rabbit antibody was obtained from Pierce (Rockford, IL). Human IL-lfl ELISA kit was purchased from R&D Systems (Minneapolis, MN). SK&F 86002 [5-(4-pyridyl)-6(4-fluorophenyl)-2,3-dihydroimidazo (2,l-b)diazole] was synthesized at SmithKline Beecham Pharmaceuticals. The test compounds were dissolved in DMSO/ethanol (50/50; v/v) and the final concentration of the two solvents was maintained at 0.5%. DMSO/ethanol by itself, at these concentrations had no effect on LPSstimulated cytokine production. Monocyte &olation

Peripheral blood monocytes obtained from normal human volunteers by plateletpheresis (Red Cross, Philadelphia, PA) were purified according to Collata, Peri, Villa & Mantovani (1984), to yield monocytes (average purity of monocytes in these preparations was >85 °7o as determined by differential count analysis, and viability was >95o7o based on trypan blue exclusion). For experiments using antibodies as the cytokine stimulus, monoclonal antibodies (10/ag/ml) to the monocyte glycoproteins in phosphate-buffered saline (PBS pH 7.4, 0.5 ml/well of 24-well Nunc multidishes) were incubated for 6 h at 4°C and unbound M A b removed by washing twice with PBS. The wells were then filled with 1 ml control medium (RPMI 1640 containing 1°70 FCS, penicillin and streptomycin at l0 U/ml), or medium containing compound. Purified monocytes were added at a

concentration of 1 x 106 cells/well and the plates were incubated overnight ( 1 2 - 14 h) at 37°C in 5°70 C02. For experiments using LPS as stimulus, adherent monocytes were pretreated with or without test compound in 1 ml medium for 1 h, and then stimulated with or without 0.I /ag/ml LPS for 1 2 - 1 4 h. Supernatants were collected, centrifuged at 1000 × g to remove cell debris, and then assayed for TNFa and IL-lfl by ELISA. All components of the growth media as well as the MAb used in this study, were evaluated for the presence of endotoxin using the LAL test, and were found to contain <10 pg/ml. TNFa assay

A standard "Sandwich E L I S A " protocol was adapted for use in quantitation of human TNFa (K. Esser, SmithKline Beecham Pharmaceuticals, personal communication). Immunoassay plates (Nunc Immunolon-4 96-F, Dynatech Laboratories, Alexandria, VA) were coated for 2 h at room temperature with a murine anti-human TNFc~ antibody (16 gg/ml in 50 mM PBS pH 7.5). Plates were blocked with PBS containing 0.5°70 caesin, 0.01°70 thimersol, 0.001% phenol red and 0.25°70 Tween-20 (Block buffer) for 1 h at 37°C. After three washings with wash buffer (consisting of PBS and 0.05°70 Tween-20), test samples were added to the plate and incubated overnight at 4°C. A standard titration curve was obtained by making serial dilutions of a known sample of recombinant human TNFa in block buffer identical to the test samples. Next, the plates were washed five times with wash buffer and incubated with rabbit anti-human TNF~ antibody (1 : 1000 dilution in block buffer) for 2 h at 37°C. Plates were washed five times with wash buffer and incubated with peroxidase conjugated goat anti-rabbit antibody (1 : 5000 dilution in block buffer) for 2 h at 37°C. Following five washes with wash buffer, substrate (1 mg/ml ortho-phenylenediamine in 0.1M citrate buffer pH 4.5 containing 0.1 °7o urea peroxide) was added to the plates for 20 min, and the color reaction stopped by addition of 0.1M sodium fluoride. Spectroscopy (450 nM) was performed using a micro-ELISA autoreader (Titertek Multiscan MC). The ELISA had a lower detection limit of 50 pg/ml and was linear up to I000 pg/ml. 1I.- lfl assay IL-lfl levels in test samples were determined by ELISA kits as per the manufacturers instructions. The lower detection limit of this assay was 10 pg/ml.

207

Human Monocytes and Cytokine Release

A

A

120-

8200

100~ 300

~_

so.

~

40200

0 Control

CD44

CD45

LFA-3

LPS

0

0.01

0.1

1

10

100

1000

Drug Concentration (pM)

B

B

~00].

120100

j.o 100

o

6O

5- - ~

40-

200

0 Control

CD44

CD45

LFA-3

LPS

Fig. 1, Stimulation of human monocytes by antibodies to CD44, CD45 and LFA-3 (CD58) to release TNFa and IL-1/L Monocytes were treated with 0.1 lag/ml LPS or 10/ag/ml immobilized antibodies to CD44, CD45 and LFA3 for 12-14 h. Culture supernatants were assayed for TNFa (A) or IL-1/3 (B) as described in Experimental Procedures. Unstimulated cultures (C) expressed TNFa or IL-lp at levels that were below the detection limits of the assay. Data are representative of 2 separate experiments.

RESULTS Figure I shows that monoclonal antibodies to CD44, CD45 and LFA-3 (CD58) when coated on to plastic dishes at 10 pg/ml, induced release of TNFa (up to 228 pg/ml) and IL-1/~ (up to 87 pg/ml) by human monocytes. A dose of 10/ag/ml of each monoclonal antibody was chosen for subsequent studies since it produced the maximal release of both T N F a and IL-1/3 in a dose range of 0.1 - I00/ag/ml (data not shown). The maximal cytokine release by monocytes was induced by 0.1/ag/ml of LPS, and was 8 _+ 1.2 n g / m l for T N F a and 3.3 +_ 0.1 ng/ml for IL-1/J. The inhibitory effects of SK&F 86002 and pentoxifylline on LPS-induced T N F a and IL-1/3 release are shown in Fig. 2. SK&F 86002 inhibited

0.01

0.1

1

10

100

1000

Drug Concentration (pM)

Fig. 2. The effect of SK&F 86002 and pentoxifylline on LPS-induced cytokine release. Monocytes were pretreated with increasing concentrations (as shown in figure) of SK&F 86002 ([D) or pentoxifylline (Q) for 1 h and then treated with 0.1 vg/ml of LPS (optimal concentration) for 12-14 h. TNFa (A) and IL-1/~ (B) levels in culture supernatants were measured by ELISA as described in Experimental Procedures and are expressed as percent control. The control value is the maximum amount of cytokine produced by LPS alone (8 ± 1.2 ng/ml TNFa, 3.3 ± 0.1 ng/ml IL-1/'J). Data are representative of 2 separate experiments. TNFa and IL-1/] release in a dose-dependent manner with an approximate Ics0 of 1/aM. The release of IL-I/3 was similarly inhibited in response to 1 ng/ml LPS (data not shown). Pentoxifylline inhibited LPSinduced TNFa release with an ICs0 of 100/aM, whereas IL-I/3 release was inhibited less than 20% at the highest concentration (1 mM) used. Both of the cytokine synthesis inhibitors did not appear to be cytotoxic at the concentrations tested based on trypan blue exclusion. The lack of adverse effects on cell viability and metabolism by SK&F 86002 has also been demonstrated in an earlier study (Lee et al., 1990).

208

U. PRABHAKARet aL

Table 1. The effect of SK&F 86002 and pentoxifylline on monoclonal antibody induced TNFo release MAb

Control

TNFa pg/ml SK&F 86002

Pentoxifylline

CD44 CD45 LFA-3

228 _+_+0.7 102 -4- 15 147 +__12

68 _+ l0 51 _.+1 72 +__8

175 + 13 68 _ 3 126 -e 10

Monocytes were incubated with 1/aM SK&F86002 or 100/aM pentoxifylline and 10/~g/ml of monoclonal antibodies for 12 - 14 h. TNFa levels in monocyte supernatants were measured by ELISA. Data are representative of 2 separate experiments.

Table 2. The effect of SK&F 86002 and pentoxifylline on monoclonal antibody induced IL-I/~ release MAb

Control

IL-1/3 pg/ml S K & F86002

Pentoxifylline

CD44 CD45 LFA-3

87 +__4 85 +__3 45 _.+7

~<10 ~<10 ~<10

65 +__5 63 +_3 37 + 2

Monocytes were treated as described in the legend to Table 1. IL-lp levelsin culture supernatants were measured by ELISA. Data are representative of 2 separate experiments. Table 1 shows the effects of SK&F 86002 and pentoxifylline at 1 and 100/aM, respectively, on TNFa release induced by the three monoclonal antibodies. SK&F 86002 at 1/aM inhibited the release of TNFa by >50%0, in response to all the three antibodies tested. At a dose of 10gM, SK&F 86002 completely inhibited the release of T N F a to levels below the detection limits of the assay (data not shown). Pentoxifylline at 100 gM inhibited the release of TNFa by less than 30% under these conditions. As shown in Table 2, SK&F 86002 at 1 pM, completely inhibited the release of IL-lfl to levels below the detection limits of the assay, in response to the three monoclonal antibodies, whereas pentoxifylline at 100/~M, caused less than 30°7o inhibition of IL-1i3 release.

DISCUSSION Monocytes/macrophages produce TNFa and IL-lfl in response to a variety of biological stimuli. These cytokines appear to be detrimental to the host when released in inappropriate amounts and may

result in cachexia, cell damage, high fever or shock symptoms (Cerami & Beutler, 1988; Waage, Halstensen & Espevik, 1987; Le & Vilcek, 1987). The nature, concentration and duration of the exogenous stimuli determines the degree of TNFa and IL-lfl synthesis and release (Nain et al., 1990). This has been amply demonstrated with LPS of gram negative bacteria (Waage et al., 1987; Girardin, Grau, Dayer, Roux-Lombard, J5 Study Group & Lambert, 1988; Beutler, Krochin, Milsark, Leudke & Cerami, 1986). One potentially important physiological signal that triggers cytokine release is via c e l l - cell contact. For example, TNFa is released when monocytes contact tumor cells (Webb & Gerrard, 1990) and IL-I/3 is released during antigen presentation to T-cells (Le & Vilcek, 1987). More recently, immobilized monoclonal antibodies to CD44, CD45 and LFA-3 (monocyte surface receptors) were shown to induce cytokine production in vitro (Webb et al., 1990). The question that we address in this report is whether or not the activation of monocytes to release cytokines by LPS or putative physiological receptors share a common step. We used as probes two cytokine synthesis inhibitors, SK&F86002 and pentoxifylline, to examine their effects on TNFa and IL-1/3 release induced by LPS or monoclonal antibodies to three monocyte glycoproteins, CD44, CD45 and LFA-3 (CD58). We have demonstrated that SK&F 86002 can effectively inhibit release of both TNFt~ and IL-1/J when either monoclonal antibodies or LPS are used to stimulate the monocytes. In contrast, pentoxifylline is at least a 100-fold less potent than SK&F 86002 in inhibiting LPS induced TNFt~ release. Thus, we show that SK&F 86002 blocks the release of TNFa and IL-I/3 induced by both sets of stimuli whereas both stimuli are much less sensitive to the inhibitory action of pentoxifyUine. The levels of TNFa and IL-1/3 induced by the monoclonal antibodies in our studies are somewhat lower from those reported by Webb et al. (1990). These differences may be attributable to (a) donor to donor variability in the monocyte response and/or (b) differences in the epitopes to which the monoclonal antibodies are directed. Although monoclonal antibodies induce low levels of TNFa and IL-lfl in comparison to LPS, there is evidence in the literature that such levels of plasma TNFa, do indeed contribute to pathological states (Waage et al., 1987; Leroux-Roeis et al., 1988). However, one must exercise caution in correlating in vitro data with in vivo conditions. Furthermore, whether or not receptor molecules involved in such cell-cell interactions would act independently or in

Human Monocytes and Cytokine Release combination with other stimuli to produce higher and cumulative levels of T N F a and IL-1/3 remains to be investigated. Interestingly, the suppression of monoclonal antibody-induced T N F a release by pentoxifylline was much less than when LPS was used as the stimulus. The reasons for this finding are not clear at this time, since the levels o f T N F a produced with Mab stimulation were also m u c h lower than the amounts produced with LPS stimulation. However, one possible explanation could be that the two stimuli may exhibit subtle differences in the mechanism by which they regulate T N F a release, with L P S being more sensitive to the inhibitory effects of agents such as pentoxifylline. Based on the overall results presented in this study,

209

we conclude that LPS and the putative physiological receptors share a c o m m o n step in the activation of cytokine production by h u m a n monocytes. We are currently investigating whether or not other stimuli that trigger production of T N F a and IL-I{J by monocytes, would also utilize an activation step that is similar to LPS. A n understanding of these mechanisms/pathways will aid in the design of selective inhibitors o f cytokine production that m a y be beneficial in prophylaxis as well as treatment of diseases, where there is evidence that exaggerated levels of T N F a and IL-lfl contribute to pathology.

Acknowledgements - - We thank Dr Klaus Esser for providing the TNFa antibodies.

REFERENCES

BEUTLER, B., KROCHIN,N., M1LSARK)W., LEUDKE,C. 8/: CERAMI,A. (1986). Control of cachectin (tumor necrosis factor) synthesis: mechanism of endotoxin resistance. Science, 232, 977- 980. CERAMI,A. & BEUTLER,B. (1988). The role of cachectin/TNF in endotoxic shock and cachexia. Irnrnun. Today, 9, 28 - 31. COLLATA,F., PERI, G., VILLA,k. & MANTOVANI,A. (1984). Rapid killing of actinomycin-D treated cells tumor cells by human mononuclear cells. J. Irnrnun., 132, 936-944. ENDRES, S., FULLE, H. J., SINHA, B., STOLL, D., DINARELLO, C. A., GREZER, R. & WEBER, P. C. (1991). Cyclic nucleotides differentially regulate the synthesis of tumor necrosis factor a and interleukin-1 production. Immunology, 72, 5 6 - 60. GERRARD, T. L., SEIGEL, J. P., DYER, D. & ZOON, K. C. (1987). Differential effects of interferon-a on interleukin-1 secretion by monocytes. J. Imrnun., 138, 2535- 2540. GIRARDIN, E., GRAU, G. E., DAYER, J. M., ROux-LOMBARD, P., J5 STUDYGROUP & LAMBERT,P. H. (1988). Tumor necrosis factor and interleukin-I in the serum of children with severe infectious purpura. New Engl. J. Med., 319, 397 - 400. GRISWOLD, D. E., WEBB, E., SCHWARTZ,L. & HANNA,N. (1987). Arachidonic acid-induced inflammation: inhibition by dual inhibitors of arachidonic acid metabolism, SK&F 86002. Inflammation, 11, 189- 199. LE, J. & VILCEK, J. (1987). Tumor necrosis factor and interleukin-l: cytokines with multiple overlapping activity. Lab. Invest., 56, 234 - 248. LEE, J. C., REBAR, L. & LAYDON, J. T. (1989). Effect of SK&F 86002 on cytokine production by human monocytes. Agents Actions, 27, 277 - 279. LEE, J. C., VOTTA, B., DALTON, B. J., GRISWOLD, D. E., BENDER, P. E. & HANNA, N. (1990). Inhibition of human monocyte IL-1 production by SK&F 86002. Int. J. Irnmunother., VI, 1 - 12. LEROUX-ROElS, G,, OFFNER, F., PHILIPPE, J. & VERMEUIEN, A. (1988). Influence of blood-collecting systems on concentrations of tumor necrosis factor in serum and plasma. Clin. Chem., 2373 - 2374. NA1N, M., HINDER, F., GONG, J. H., SCHMIDY, A., BENDER, A., SPRENGER,H. & GEMSA, D. (1990). Tumor necrosis factor-a production of influenza virus-infected macrophages and potentiating effect of lipopolysaccharide. J. Irnrnun., 145, 1921 - 1928. TRACEY, K. J., BEUTLER, B., LOWRY, S. F., MERRYWHEATHER,J., WOLPE, S., MILSARK, I. W., HARIR, R. J., FAHEY, T. J. III., ALBERT, J. D., SHIRES, G. T. & CERAMI, A. (1986). Shock and tissue injury induced by recombinant human cachectin. Science, 234, 470- 474. WAAGE, A., HALSTENSEN,A. & ESPEVIK,T. (1987). Association between tumor necrosis factor in serum and fatal outcome in patients with meningococcal disease. Lancet, 1, 355. WEBB, D. S. A. & GERRARD, T. L. (1990) IFNa and IFNy can affect both monocytes and tumor cells to modulate monocyte-mediated cytoxicity. J. Irnmun., 144, 3643- 3648. WEBB, D. S. m., SHIMIZU,Y., SEVENTER,G. m. V., SHAW, S. & GERRARD,T. L. (1990). Physiological triggers of human monocyte TNF and IL-1 release. Science, 249, 1295- 1297.