Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF NEW MEXICO on 11/29/14 For personal use only.
B R E F REPBRTSIWPORTS BREFS
in rat adrenal gland and bovine cerebellum using selective nonpeptide antagonists. J. Cardiovasc. Phamacol. f 7: 177 - 184. Wong, P. C., Hart, S. D., Zaspel, A. M . , et al. 1 9 9 0 ~Functional . studies of mnpeptide angiotensin I1 receptor subtype-specific ligands: DuP 753 (AII-1) and PI4123177 (AH-2). J. Phamacol. Exp. Ther. 255: 584-592. Wong, P. C., Price, W. A., Jr., Chiu, A. T., et al. 1998b. Mypotensive action of DuP753, an angiotensin I1 antagonist, in spontane-
769
ously hypertensive rats: non peptide angistensin II receptor antagonists: X. Hypertension (Dallas), 15: 459-468. Wong, P. C., Price, W. A., Jr., Chiu, A. T., et al. 1 9 9 8 ~Non . peptide angiotensin I1 receptor antagonists. IX. Antihypertensive activity in rats of DuP 753, an orally active antihypertensive agent. J. Phamacol. Exp. Ther. 252: 726-732.
Production of tumour necrosis factor by cells exposed to sulphonamide reactive metabolites MICHAEL J. RIBDER,~ MONICAMASK,AND INGRIDA. BIRD Division of Clinical Phamctplogy, Child Health Research Institute, Degartmen~sof Paediatrkcs and Pha~macology and Toxicology, University Q JWestern Ontario, London, Ont., Canada N6C 2V5 Received August 8, 1991 RIEDER,M. J., MASK,M., and BIRD,I. A. 1992. Production of tumour necrosis factor by cells exposed to sulphonarnide reactive metabolites. Can. J. Physiol. Pharmacol. 70: 7 19-722. Hypersensitivity reactions are the most common adverse events associated with therapy with the sulphonamide antibiotics. These reactions have been shown to occur among individuals with pharmacogenetically determined differences in the capacity of their cells to detoxify reactive products of oxidative metabolism of the sulphonamides. These reactions appear to be propagated by an inflammatory response by the immune system. To investigate the role of the cytokine tumour necrosis factor (TNF-a) in these reactions, we studied the production of TNF-a by peripheral blood mononuclear cells (PBMCs) that had been incubated with sulfamethoxazole and murine microsomes in the presence and absence of a microsomal-activating system and TNF-a production by PBMCs in the presence and absence of the hydroxylamine derivative of sulfamethoxazole. The PBMCs showed a time-related increase in the production of TNF-a. There was no increase in TNF-a production seen during incubation with sulphonamide reactive metabolites; rather, there was a decrease in TNF-CY elaboration that was most marked when PBMCs were incubated with the hydroxylamine of sulfamethoxazole. There is no evidence from these in vitro studies that TNF-a is involved as a mediator of the inflammatory response in sulphonamide hypersensitivity adverse dmg reactions. Key avords: sulphonamide, tumour necrosis factor, adverse drug reaction, hydroxylamine. RIEDER,M. J., MASK,M., et BIRD,I. A. 1992. Production of tunnour necrosis factor by cells exposed to sulphonamide reactive metabolites. Can. J. Physiol. Phamacol. 70 : 719-722. Les reactions d'hypersensibilitk constituent les effets nCgatifs les plus frCquemment relevCs suit h un traitement aux sulfamides. On a constate que ces reactions se produisaient chez les individus montrant des diffkrences phamcolsgique dans la capacitC de leurs cellules de dCtoxiquer les produits rkactifs du m6tabolisme oxydatif des sulfamides. Ces rbctions segnblent &re propagees par une rCponse inflamatoire du systkme imunitaire. Pour 6tudier le r8le du facteur de nCcrsse des tumeurs a (TNF-a) dans ces reactions, nous avons examink la production de TNF-cu par les cellules mononucl6aires anguines pCriphCriques (CMSP), incubCes avec du sulfamCthoxazole et des microsomes murins, en prCsence et en l'absence d9un systbme d'activation microsomiale, ainsi que la production de TNF-a par les CMPS, en prCsence et en l'absence du dCriv6 hydroxylamine du su8famCthoxazole. Les CMPS ont montr6 une augmentation, reliee au temps, de la production de TNF-a. La production de TNF-a n9a pas augment6 durant l'incubatiesn avec les mktabolites rkactifs aux sulfonamides; elle a phutdt diminui, et de facon tr&smarquee, lorsque les CMPS ont Ct6 incubCes avec le dCriv6 hydroxylamine du sulfamCthoxazole. Ces ktudes in vitro n'ont pu dtmontrer que le TNF-a est impliqui en tant que mCdiateur de la rCponse inflammatoire dans les reactions m~icamenteusesindksirables d9hypersensibilitClors du traitement aux sulfonamides. Mots ckks : sulfesnamide, facteur de nkcrose des tumeurs, rCactions mCdicamenteuses indCsirables, hydroxylamine. [Traduit par Ba rkdastion]
Introduction sdphonde first invoduced into practice more than 50 years ago, continue to be one of fie major classes of agents, for ambulatory therapy. Adverse reactions, the most serious of which are 'Author for correspondence at the following address: Clinical Pharmacology, Department of Paediatriss, Children's Hospital of Western Ontario, 800 Commissioner's Road East, London, Ont., Canada N6C 2V5. Printed in Canada I Imprim6 au Canada
hypersensitivity reactions characterized by fever and rash and, in many cases, other organ involvement, complicate 3 -5 % of all courses of sulphonamide therapy (Rieder et al. 1989). One of the most striking features of sulphonamide hypersensitivity reactions is the of a high fever? which is fie earliest and most common clinical manifestation of these reactions (Rieder et al. 1989). The pathogenesis of sulphonamide hypersensitiviv reactions appears to involve the Qetoxication of reactive metabolites of sdphonamide oxidative metabolism, with the peripheral blood
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF NEW MEXICO on 11/29/14 For personal use only.
720
CAN. J . PHYSIOL. PHARMACOL. VOL. 70, 1992
mononuclear cells (BBMCs) of patients who develop sulphonarnide hypersensitivity reactions showing markedly increased cell death when exposed to these compounds relative to the cells of control subjects (Shear and Spielberg 1985; Shear et a!. 1986; Park al. 1987; Rieder et al. 1988, 1989). Although compromised detoxication appears to be critical in the initiation of sulphonamide hypersensitivity reactions, atypical lymphocytes and eosimphilia are found in the blood of patients with hypersensitivity reactions. The persistence and continuing development of the symptoms of these reactions despite discontinuation of sulphonamide therapy and the clinical improvement seen after cor-ticosteroid therapy have suggested that these reactions have an immune component (Park et ak. 1987; Rieder et a!. 1989). The involvement of mediators of the immune response in the pathogenesis of sulphonamide hypersensitivity adverse drug reactions is unknown. Over the last decade we have come to a better understanding of the importance of immune mediators, including cytokines, in the control of the inflammatory response (Warren 1990). Tumor necrosis factor alpha (TNF-a) is a 1%-kDapeptide produced and secreted by msnocytes and macrophages that exerts a variety of effects across a considerable range of cell types (Sugarman et ak. 1985; DeForge et ak. 1990; Warren 1990). In Bow concentrations, TNF-a is a potent pyrogen (Warren 1990). In higher concentrations, it has been shown to produce a biphasic febrile response by a direct response as well as by triggering the release of interleukin-1 (Warren 1990). This suggests that the cytokines such as TNF-a are important candidates to consider in the search for immune mediators of sulphonamide hypersensitivity reactions. The objective of this study was to test for a correlation between TNF-a production and the cytotoxicity of sulphonamide reactive metabolites.
Materials and methods
balanced salt solution to stop the rnonoxygenase-catalyzed reaction (Spielberg 1980). The cells were then incubated for 18 h at 37°C in a water bath to allow expression of cytotoxicity (Spielberg 1980; Shear and Spielberg 1985). After the 18-h incubation, 8 ~4gof TB was added to each test tube and PBMCs from experimental and control tubes were scored visually for TB dye exclusion using a light microscope. A visual score of the number of viable (white) and nonviable (blue) cells was obtained and recorded manually. Cytotexicity was determined by calculating the ratio of nonviable to viable cells. This determination has been shown to have a small intraobserver variability but a large interobserver variability. To reduce the variability of this assay, all TB assays were scored by the same observer (IAB), who counted 108 cells from each test tube. For fluorescence assays, PBMCs were adjusted to a concentration of 2.0 X lo6 cellslml in 1% mM 4-(2-hydroxy1ethyl)-1-piperazine ethane sulfonic acid (HEPES) buffer and loaded into a purpose-built 96-well microtiter plate (Baxter Healthcare Corp., Mundelein, Ill.). A standard curve was prepared on each plate, with cells plated in concentrations from 0 to 50 000 cellslwell in increments of 10 000 cells/ well. For concentration -toxicjlty experiments, the N-hydroxylamine derivative of sulfamethoxazole (SMX-HA) was added to wells in increasing concentrations from 0 to 400 pM. Stock solutions of SMXHA were prepared daily just prior to plating with the cells. The cells were incubated with SMX-HA for 2 h in a humidified incubator at 37°C. The cells were washed and resuspended in 100 pL of HEPES containing 0.5 % ! bovine serum albumin and incubated in a humidified incubator at 37°C for 18 h. After incubation, the cells were washed and resuspended with HEPES buffer containing 50 j4g of BCECFImL. The cells were incubated for 30 min in a humidified incubator at 37OC. Fluorescence was determined using a Pandex Fluorescence Concentration Immunoassay Analyzer (Baxter Healthcare Cerp., Mundelein, Ill.). Linear regression analysis was used to derive an equation from the standard curve for each plate comparing cell number with fluorescence. This was then used t s calculate the number of viable cells remaining in each well. Reagents for the cytoxicity assay were obtained from Sigma Chemical Co. (St. Louis, Mo.) except for the hydroxylarnine derivative of sulfamethoxazole (SMX-HA), which was synthesized as previously described (Rieder ef al. 1988).
cybstoxicity assay in vitro The assays for cytstsxicity in vidro were performed as previously described using PBMCs collected from whole blood as the target cells (Spielberg 1980; Shear and Spielberg 198%;Shear ef a / . 1986; Wieder et (11. 1988, 1989). Cytotoxicity was assessed by trypan blue dye exclusion (TB assay) and by fluorescence using the dye 2',7'-bis(2-carboxyethy1)-5-(4)-carboxyfluorescein (BCEGF) as an indicator of cell viability by determining BCECF fluorescence of live cells (Molecular Probes Inc., Eugene, Oreg.). Fluorescence has been demonstrated to be a reproducible and accurate assay for the assessment of cytotoxicity produced by reactive drug metabolites (Leeder et al. 1989). Cytotoxicity assays using trypan blue dye exclusion were perfomed using a murine microsomal-activating system to generate sulphommide reactive metabolites as previously described (Shear and Spielberg 1985; Shear et al. 1986). Murine hepatic microsomes were prepared from National Institutes of Health General Purpose Swiss mice induced by treatment with phenobarbital. Microsomal protein (0.5 mg) was incubated with PBMCs, 8.6 mM nicotine adenine dinucleotide phosphate, 2.4 mM glucose 6-phosphate, 2 units of glucose-6-phosphate dehydrogenase, and sulfarnethoxazole in concentrations of 0.015, 0.15, or H .5 mM. Control incubations were prepared that contained cells alone or cells plus microsomes and the microsomal-activating system without the addition sf sulfamethoxazole. TB assays were done in test tubes using a cell concentration of lo6 cellltube; duplicate tubes were prepared for each experimental condition. During TB assays, the test tubes were incubated for 2 h in a water bath heated to 3TQC. After 2 h of incubatisn, the cells were centrifuged and resuspended with 0.5% albumin (wlv) in HEPES
aaBrF-a enzyme immunoassay For each cytotoxicity assay, duplicate samples were prepared for the collection of supernatant for TNF-a assays. Supernatants were stored at -70°C until their use in the TNF-a assay. TNF-a assays were performed using BIBKINE TNF-a test kits, which were obtained from T-cell Sciences, Inc. (Cambridge, Mass.). All reagents for the assay, including standards of lyophylized recombinant human TNF-a, were provided in the BIBKINE TNF-a test kit except for the stop solution (2 M H,SO,), which was obtained from BBH Inc. (Toronto, Ont.). The TNF-a kit is specific for bioactive, human TNF-a and is unaffected by the presence of denatured TNF-a, lyrnphotoxin, or interleuatins (Cuturi et al. 1987). On the day of assay, 50 pL of the supernatant was added to micrstitre plate wells to which a murine anti-TNF-a monocloml antibody had been adsorbed. The supernatant was incubated with the murine anti-TNF-a monoclonal antibody for 2 h at 37°C. The wells were washed and 100 pL of a solution of horseradish peroxidase conjugated murine anti-TNF-a was added to all wells. The plates were incubated for 2 h at 37°C. The wells were washed and 100 pL of a solution of o-phenylendiamine was added to each well. The plates were incubated for a further 2 h in an incubator at 37OC. One hundred microlitres of H202 was added to each well as a substrate for o-phenylenediawnine and incubation was continued for a further 30 min. The basis for the assay is the formation of a coloured product formed from o-phenylenediarnine in proportion to the amount of TNF-a bound to the antibody adsorbed to the plate. The absorbance of the wells was read at 490 nm with a Multiscan Titretek Microplate Reader (Flow Laboratories, McLean, Va.). The absorbance of test
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF NEW MEXICO on 11/29/14 For personal use only.
BHEF REPORTS / RAPPORTS BRBFS
0.0
1.0 Sultarnethsxazole Concentration (mM)
Sulfamethoxazoke Hydroxylamine Concentration (uM)
FIG. 1. Changes in TNF-or production in PBMCs incubated with murine microsomes, a microsoml-activating system, and increasing concentrations of sulfamethoxazole. TNF-or production was determind after 18 h incubation. Data represent the mean f SE of experiments performed in duplicate using the BBMCs of five volunteers. Fifty thousand cells were incubated per well. TNF-a! production by PBMCs incubated with B .0 mM sulfametRoxazole was 218 20 pg after 18 h incubation.
+
wells was compared to the absorbance of a standard curve prepared with serial dilutions of recombinant human TNF-er. Data analysis Linear regression was performed using a STAWiew 5 12 and a Microsoft Excel program with an Apple Macintosh Plus computer (Apple Computers, Cupertino, Calif.).
+
Ethical approval The protocol for donation of PBMCs by volunteers was approved by the Review Board for Health Research Involving Human Subjects, University of Western Ontario.
Results We first investigated the production of TNF-(r by PBMCs under the conditions used for our cytotoxicity assays as described above. Under our usual assay conditions, the PBMCs demonstrated a concentration-dependent production of TNF-a (data not shown). The production of TNF-a! by PBMCs under these conditions was time dependent, with minimal amounts of TNF-a detected after 1- or 2-h incubations, and with a marked increase in TNF-a concentration detected after 18 h of incubation (data not shown). Incubation of PBMCs with sulfamethoxazole, murine microsomes, and a microsomal-activating system or with the hydroxylamine derivative of sulfamethoxazoleproduced doserelated toxicity consistent with results described previously (data not shown; Shear and Spielberg 1985; Rieder et ad. 1988, 1989). We studied TNF-(r production under these conditions. TNF-(r production by PBMCs in the presence of murine microsomes, a microsomal-activating system, and increasing concentrations of sulfmethoxa.zole was found to be less than that observed when cells were incubated without the addition of a murine microsomal-activating system (Fig. 1). This reduction in TNF-(r production was found when cells were incubated for 1, 2, or 18 h. There was no correlation between the toxicity associated with these concentrations of sulfamethoxazole (for 8.15 mM, 18 & 3%) and the decline observed in TNF-a concentrations. We then determined TNF-a production in PBMCs incubated with the chemically synthesized hydroxylamine derivative of sulfamethsxazole. A similar decline in TNF-cr production was
FIG.2. Changes in TNF-a! production associated with incubation of BBMCs with increasing concentrations of the hydroxylamine derivative of sulfamethoxazole. TNF-er production was determined after 18 h incubation. Data represent the mean f SE of experiments performed in duplicate using the PBMCs of five volunteers. Fifty thousand cells were incubated in each well.
observed when PBMCs were incubated with increasing concentrations of SMX-HA (Fig. 2). Again, the reduced production of TNF-a! was present after 1, 2, and 18 h incubation. In the case of experiments performed with SMX-HA, the decline in TNF-a production was correlated with the number of viable cells present after 18 h incubation (R2 = 0.994).
The pathogenesis of sulphonamide hypersensitivity reactions appears to involve initial events involving dmg metabolism with the subsequent evolution of the reactions propagated by the immune system (Shear and Spielberg 1985; Shear et al. 1986; Park et al. 1987; Rieder et al. 1989; Spielberg et sk. 1989). TNF-a! is believed to be involved in the acute-phase response, as well as having effects on neutrophil activation, endothelial adhesion, calcium release from bone, and stimulation of tissue enzymes such as collagenase (Warren 1990). TNF-(r is believed to be a proximal mediator in septic shock in vivo, a disease state characterized by fever, malaise, and hypotension (DeForge et al. 1990). In concert, these factors suggested that TNF-a may play a role in the evolution of sulphonamide hypersensitivity adverse drug reactions. Although TNF-(r was considered previously to be only a product of monocytes and macrophages, Cuturi et ad. (1987) have found that lymphocytes are capable of rapidly producing high levels of TNF-a. Previous research into the mechanism(~)of sulphonamide hypersensitivity adverse drug reactions has used PBMCs, which are primarily lymphocytes, as the target tissue (Spielberg 1980; Shear and Spielberg 1985; Shear et ak. 1986; Rieder et ad. 1988, 1989). The advantage of PBMCs for this research is that they possess many of the main metabolite detoxication pathways but do not activate the sulphonamides to any significant degree (Spielberg 1980). The study of TNF-(r production in vitro is problematic, because of the difficulties in isolating TNF-a from biological samples collected in vivo. Thus, we elected to study the production of TNF-(r in response to sulphonamide reactive metabolites in vitro. We have demonstrated that TNF-a is elaborated by PBMCs under the conditions of incubation used for our cytotoxicity
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF NEW MEXICO on 11/29/14 For personal use only.
722
CAN. J. PHYSIOL. PHARMACOL. X7OL. 70, 1992
assays. Our results are in agreement with those s f Cuturi et ul. $1987),indicating h a t peripheral blood lymphocytes are capable of producing and releasing TNF-a in vktra? in the range of 150-250 pg/mE of culture medium. The time-related increase we observed in TNF-a release occurred over the time course previovsly used for our cytotoxicity assays, suggesting that increases in TNF-a production associated with exposure to reactive drug metabolites might be expected at this time. Incubation of cells with reactive derivatives of the sulphonamides, either generated in vitro by microsomal oxidation of sulfamethoxazole or chemically synthesized, was not associated with an increase in TNF-a release. Rather, a decrease in TNF-a concentration into the medium was demonstrated. As noted above, the decrease in TNF-a release occurred whether the reactive derivatives were provided in a synthetically pure form or generated in vitro by microsomal metabolism. This suggests that the cytofine TNF-a is not a mediator of the inflammatory response associated with sulphonamide hypersensitivity adverse drug reactions. The decline in TNF-a release when PBMCs were incubated with murine microssmes, a microsomal activating system, and sulfamethoxmole was noted at concentrations of sulfamethoxazole only associated with a modest degree of toxicity, suggesting that the mechanism of reduced TNF-a release may involve a direct effect of sulphonamide reactive metabolites on the immune system. Eeeder et al. (1991) have found that sulphonamide reactive derivatives appear to have an inhibitory effect on the function of natural killer cdls. In contrast, the dose-related decline in TNF-a release noted with SMX-HA appeared to correlate with the decline in viable cell number demonstrated by eytotsxicity assays. There was a m c h more pronsuncd decline in both toxicity and TNF-a release noted when PBMCs were incubated with the hydroxylarnine derivative of sulfarnethoxuole. This suggests that a decline in TNF-a concentration into the medium may be related to both a general toxic effect of the compound as well as effects of sulphonamide reactive metabolites on the h n c tioning s f the immune system. Mediators of the immune response that appear to be involved in the propagation of sulphsnamide hypersensitivity adverse drug reactions remain unhown. We are currently pursuing further studies with respect to the role of other immune mediators, including interleukin- 1 , in the pathogenesis of sulphonarnide hypersensitivity adverse drug reactions as well as mechanisms of alterations in immune function associated with reactive dmg mehbolites.
Acknowledgements This work was supported by grant MT 10641 from the Medical Research Council of Canada (MRC) and a start-up grant from the Child Health Research Institute. Dr. Rieder holds a Pharmaceutical Manufacturers' Association of Canada1 MRC Career Award. We wish to thank Drs. Neil Shear, W. Alrnawi, J. Bend, and K. Borkowski for their input into the preparation of this manuscript. Cutbari, M. C., Murphy, M . , Costa-Giorni, M. P., and Weinrnann, R. 1987. Independent regulation of tumor necrosis factor and Iymphotoxin production by human peripheral blood lymphocytes. J. Exp. Med. 165: 1581 -1594. DeForge, L. E., Nguyen, D. T., Igunkel, S. L., and Remick, D. G. 1990. Regulation of the pathophysiology of tumor necrosis factor. J. Lab. Clin. Med. 116: 429-438. Leeder, %. S., Dosch, H.-M., Harper, B. A . , et a&. 1989. FIu(ssrescence-based viability assay for studies of reactive drug metabolites. Anal. Biochem. 17'7: 364-372. Leeder, J. S., Nakhooda, A., Spielberg, S. P., and Dosch, H.-M. 1991. Cellular toxicity of sulfarnethoxazole reactive metabolites. II. Inhibition of natural killer activity in human peripheral blood mononuclear cells. Biochem. Pharmacol. 41: 575 -583. Bark, B. K . , Coleman, J. W., and Kitteringham, N. 8.1987. Drug disposition and drug hypersensitivity. Biochem. Pharmacol. 36: 581 -589.
Rieder, M. B., Uetrecht, J., Shear, W. H., and Spielberg, S. P. 1988. Synthesis and in vitro toxicity of hydroxylamine metabolites of sulfonarnides. J . Pharmacol . Exp . Tl-aer. 244: 724 -728. Rieder, M. J . , Uetrecht, J . , Shear, N. H., eta&.1989. Diagnosis of sulfonarnide hypersensitivity reactions by in vitro "rechallenge" with hydroxylarnine metabolites. Am. Intern. Med. 110: 286289.
Shear, N. H . , and Spielberg, S. P. 1985. ln vitro evaluation of a toxic metabolite of sulfadiazine. Can. J. Phy siol. Pharrnacol. 63: 13766- 1372.
Shear, N. H., Spielberg, S. P., Grant, D. M., et a&.1986. Differences in metabolism of sulfomamides predisposing to idiosyncratic toxicity. Ann. Intern. Med. 105: 179-184. Spielberg, S. P. 1980. Acetaminophen toxicity in human lympho9. Pharmacol. Exp. Ther. 213: 395 -398. cytes in \!&'fro. Spielberg, S. P., keder, 9. S., Cribb, A. E., and Dosch, H.-M. 1989. Is sulfamethoxazole hydrsxylamine (SMX-HA) the proximal toxin for s~lfamethoxazoLe(SMX) toxicity? Eur. J. Clin. Pharmacol. 36: 04.37 (Abstr. A173). Sugarman, B. J., Aggarwal, B. B., Maas, P. E., et al. 1985. Recombinant human tumor necrosis factor-alpha: effects on proliferation of nom~aland transformed cells in vitro. Science (Washington, D.C.), 238: 943 -945. Warren, 9. S. 19%. Interleukins and tumor necrosis factor in inflammation. Crit. Rev. Clin. Lab. Sci. 28: 37-59,
B R E F REPBRTSIWPORTS BREFS
in rat adrenal gland and bovine cerebellum using selective nonpeptide antagonists. J. Cardiovasc. Phamacol. f 7: 177 - 184. Wong, P. C., Hart, S. D., Zaspel, A. M . , et al. 1 9 9 0 ~Functional . studies of mnpeptide angiotensin I1 receptor subtype-specific ligands: DuP 753 (AII-1) and PI4123177 (AH-2). J. Phamacol. Exp. Ther. 255: 584-592. Wong, P. C., Price, W. A., Jr., Chiu, A. T., et al. 1998b. Mypotensive action of DuP753, an angiotensin I1 antagonist, in spontane-
769
ously hypertensive rats: non peptide angistensin II receptor antagonists: X. Hypertension (Dallas), 15: 459-468. Wong, P. C., Price, W. A., Jr., Chiu, A. T., et al. 1 9 9 8 ~Non . peptide angiotensin I1 receptor antagonists. IX. Antihypertensive activity in rats of DuP 753, an orally active antihypertensive agent. J. Phamacol. Exp. Ther. 252: 726-732.
Production of tumour necrosis factor by cells exposed to sulphonamide reactive metabolites MICHAEL J. RIBDER,~ MONICAMASK,AND INGRIDA. BIRD Division of Clinical Phamctplogy, Child Health Research Institute, Degartmen~sof Paediatrkcs and Pha~macology and Toxicology, University Q JWestern Ontario, London, Ont., Canada N6C 2V5 Received August 8, 1991 RIEDER,M. J., MASK,M., and BIRD,I. A. 1992. Production of tumour necrosis factor by cells exposed to sulphonarnide reactive metabolites. Can. J. Physiol. Pharmacol. 70: 7 19-722. Hypersensitivity reactions are the most common adverse events associated with therapy with the sulphonamide antibiotics. These reactions have been shown to occur among individuals with pharmacogenetically determined differences in the capacity of their cells to detoxify reactive products of oxidative metabolism of the sulphonamides. These reactions appear to be propagated by an inflammatory response by the immune system. To investigate the role of the cytokine tumour necrosis factor (TNF-a) in these reactions, we studied the production of TNF-a by peripheral blood mononuclear cells (PBMCs) that had been incubated with sulfamethoxazole and murine microsomes in the presence and absence of a microsomal-activating system and TNF-a production by PBMCs in the presence and absence of the hydroxylamine derivative of sulfamethoxazole. The PBMCs showed a time-related increase in the production of TNF-a. There was no increase in TNF-a production seen during incubation with sulphonamide reactive metabolites; rather, there was a decrease in TNF-CY elaboration that was most marked when PBMCs were incubated with the hydroxylamine of sulfamethoxazole. There is no evidence from these in vitro studies that TNF-a is involved as a mediator of the inflammatory response in sulphonamide hypersensitivity adverse dmg reactions. Key avords: sulphonamide, tumour necrosis factor, adverse drug reaction, hydroxylamine. RIEDER,M. J., MASK,M., et BIRD,I. A. 1992. Production of tunnour necrosis factor by cells exposed to sulphonamide reactive metabolites. Can. J. Physiol. Phamacol. 70 : 719-722. Les reactions d'hypersensibilitk constituent les effets nCgatifs les plus frCquemment relevCs suit h un traitement aux sulfamides. On a constate que ces reactions se produisaient chez les individus montrant des diffkrences phamcolsgique dans la capacitC de leurs cellules de dCtoxiquer les produits rkactifs du m6tabolisme oxydatif des sulfamides. Ces rbctions segnblent &re propagees par une rCponse inflamatoire du systkme imunitaire. Pour 6tudier le r8le du facteur de nCcrsse des tumeurs a (TNF-a) dans ces reactions, nous avons examink la production de TNF-cu par les cellules mononucl6aires anguines pCriphCriques (CMSP), incubCes avec du sulfamCthoxazole et des microsomes murins, en prCsence et en l'absence d9un systbme d'activation microsomiale, ainsi que la production de TNF-a par les CMPS, en prCsence et en l'absence du dCriv6 hydroxylamine du su8famCthoxazole. Les CMPS ont montr6 une augmentation, reliee au temps, de la production de TNF-a. La production de TNF-a n9a pas augment6 durant l'incubatiesn avec les mktabolites rkactifs aux sulfonamides; elle a phutdt diminui, et de facon tr&smarquee, lorsque les CMPS ont Ct6 incubCes avec le dCriv6 hydroxylamine du sulfamCthoxazole. Ces ktudes in vitro n'ont pu dtmontrer que le TNF-a est impliqui en tant que mCdiateur de la rCponse inflammatoire dans les reactions m~icamenteusesindksirables d9hypersensibilitClors du traitement aux sulfonamides. Mots ckks : sulfesnamide, facteur de nkcrose des tumeurs, rCactions mCdicamenteuses indCsirables, hydroxylamine. [Traduit par Ba rkdastion]
Introduction sdphonde first invoduced into practice more than 50 years ago, continue to be one of fie major classes of agents, for ambulatory therapy. Adverse reactions, the most serious of which are 'Author for correspondence at the following address: Clinical Pharmacology, Department of Paediatriss, Children's Hospital of Western Ontario, 800 Commissioner's Road East, London, Ont., Canada N6C 2V5. Printed in Canada I Imprim6 au Canada
hypersensitivity reactions characterized by fever and rash and, in many cases, other organ involvement, complicate 3 -5 % of all courses of sulphonamide therapy (Rieder et al. 1989). One of the most striking features of sulphonamide hypersensitivity reactions is the of a high fever? which is fie earliest and most common clinical manifestation of these reactions (Rieder et al. 1989). The pathogenesis of sulphonamide hypersensitiviv reactions appears to involve the Qetoxication of reactive metabolites of sdphonamide oxidative metabolism, with the peripheral blood
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF NEW MEXICO on 11/29/14 For personal use only.
720
CAN. J . PHYSIOL. PHARMACOL. VOL. 70, 1992
mononuclear cells (BBMCs) of patients who develop sulphonarnide hypersensitivity reactions showing markedly increased cell death when exposed to these compounds relative to the cells of control subjects (Shear and Spielberg 1985; Shear et a!. 1986; Park al. 1987; Rieder et al. 1988, 1989). Although compromised detoxication appears to be critical in the initiation of sulphonamide hypersensitivity reactions, atypical lymphocytes and eosimphilia are found in the blood of patients with hypersensitivity reactions. The persistence and continuing development of the symptoms of these reactions despite discontinuation of sulphonamide therapy and the clinical improvement seen after cor-ticosteroid therapy have suggested that these reactions have an immune component (Park et ak. 1987; Rieder et a!. 1989). The involvement of mediators of the immune response in the pathogenesis of sulphonamide hypersensitivity adverse drug reactions is unknown. Over the last decade we have come to a better understanding of the importance of immune mediators, including cytokines, in the control of the inflammatory response (Warren 1990). Tumor necrosis factor alpha (TNF-a) is a 1%-kDapeptide produced and secreted by msnocytes and macrophages that exerts a variety of effects across a considerable range of cell types (Sugarman et ak. 1985; DeForge et ak. 1990; Warren 1990). In Bow concentrations, TNF-a is a potent pyrogen (Warren 1990). In higher concentrations, it has been shown to produce a biphasic febrile response by a direct response as well as by triggering the release of interleukin-1 (Warren 1990). This suggests that the cytokines such as TNF-a are important candidates to consider in the search for immune mediators of sulphonamide hypersensitivity reactions. The objective of this study was to test for a correlation between TNF-a production and the cytotoxicity of sulphonamide reactive metabolites.
Materials and methods
balanced salt solution to stop the rnonoxygenase-catalyzed reaction (Spielberg 1980). The cells were then incubated for 18 h at 37°C in a water bath to allow expression of cytotoxicity (Spielberg 1980; Shear and Spielberg 1985). After the 18-h incubation, 8 ~4gof TB was added to each test tube and PBMCs from experimental and control tubes were scored visually for TB dye exclusion using a light microscope. A visual score of the number of viable (white) and nonviable (blue) cells was obtained and recorded manually. Cytotexicity was determined by calculating the ratio of nonviable to viable cells. This determination has been shown to have a small intraobserver variability but a large interobserver variability. To reduce the variability of this assay, all TB assays were scored by the same observer (IAB), who counted 108 cells from each test tube. For fluorescence assays, PBMCs were adjusted to a concentration of 2.0 X lo6 cellslml in 1% mM 4-(2-hydroxy1ethyl)-1-piperazine ethane sulfonic acid (HEPES) buffer and loaded into a purpose-built 96-well microtiter plate (Baxter Healthcare Corp., Mundelein, Ill.). A standard curve was prepared on each plate, with cells plated in concentrations from 0 to 50 000 cellslwell in increments of 10 000 cells/ well. For concentration -toxicjlty experiments, the N-hydroxylamine derivative of sulfamethoxazole (SMX-HA) was added to wells in increasing concentrations from 0 to 400 pM. Stock solutions of SMXHA were prepared daily just prior to plating with the cells. The cells were incubated with SMX-HA for 2 h in a humidified incubator at 37°C. The cells were washed and resuspended in 100 pL of HEPES containing 0.5 % ! bovine serum albumin and incubated in a humidified incubator at 37°C for 18 h. After incubation, the cells were washed and resuspended with HEPES buffer containing 50 j4g of BCECFImL. The cells were incubated for 30 min in a humidified incubator at 37OC. Fluorescence was determined using a Pandex Fluorescence Concentration Immunoassay Analyzer (Baxter Healthcare Cerp., Mundelein, Ill.). Linear regression analysis was used to derive an equation from the standard curve for each plate comparing cell number with fluorescence. This was then used t s calculate the number of viable cells remaining in each well. Reagents for the cytoxicity assay were obtained from Sigma Chemical Co. (St. Louis, Mo.) except for the hydroxylarnine derivative of sulfamethoxazole (SMX-HA), which was synthesized as previously described (Rieder ef al. 1988).
cybstoxicity assay in vitro The assays for cytstsxicity in vidro were performed as previously described using PBMCs collected from whole blood as the target cells (Spielberg 1980; Shear and Spielberg 198%;Shear ef a / . 1986; Wieder et (11. 1988, 1989). Cytotoxicity was assessed by trypan blue dye exclusion (TB assay) and by fluorescence using the dye 2',7'-bis(2-carboxyethy1)-5-(4)-carboxyfluorescein (BCEGF) as an indicator of cell viability by determining BCECF fluorescence of live cells (Molecular Probes Inc., Eugene, Oreg.). Fluorescence has been demonstrated to be a reproducible and accurate assay for the assessment of cytotoxicity produced by reactive drug metabolites (Leeder et al. 1989). Cytotoxicity assays using trypan blue dye exclusion were perfomed using a murine microsomal-activating system to generate sulphommide reactive metabolites as previously described (Shear and Spielberg 1985; Shear et al. 1986). Murine hepatic microsomes were prepared from National Institutes of Health General Purpose Swiss mice induced by treatment with phenobarbital. Microsomal protein (0.5 mg) was incubated with PBMCs, 8.6 mM nicotine adenine dinucleotide phosphate, 2.4 mM glucose 6-phosphate, 2 units of glucose-6-phosphate dehydrogenase, and sulfarnethoxazole in concentrations of 0.015, 0.15, or H .5 mM. Control incubations were prepared that contained cells alone or cells plus microsomes and the microsomal-activating system without the addition sf sulfamethoxazole. TB assays were done in test tubes using a cell concentration of lo6 cellltube; duplicate tubes were prepared for each experimental condition. During TB assays, the test tubes were incubated for 2 h in a water bath heated to 3TQC. After 2 h of incubatisn, the cells were centrifuged and resuspended with 0.5% albumin (wlv) in HEPES
aaBrF-a enzyme immunoassay For each cytotoxicity assay, duplicate samples were prepared for the collection of supernatant for TNF-a assays. Supernatants were stored at -70°C until their use in the TNF-a assay. TNF-a assays were performed using BIBKINE TNF-a test kits, which were obtained from T-cell Sciences, Inc. (Cambridge, Mass.). All reagents for the assay, including standards of lyophylized recombinant human TNF-a, were provided in the BIBKINE TNF-a test kit except for the stop solution (2 M H,SO,), which was obtained from BBH Inc. (Toronto, Ont.). The TNF-a kit is specific for bioactive, human TNF-a and is unaffected by the presence of denatured TNF-a, lyrnphotoxin, or interleuatins (Cuturi et al. 1987). On the day of assay, 50 pL of the supernatant was added to micrstitre plate wells to which a murine anti-TNF-a monocloml antibody had been adsorbed. The supernatant was incubated with the murine anti-TNF-a monoclonal antibody for 2 h at 37°C. The wells were washed and 100 pL of a solution of horseradish peroxidase conjugated murine anti-TNF-a was added to all wells. The plates were incubated for 2 h at 37°C. The wells were washed and 100 pL of a solution of o-phenylendiamine was added to each well. The plates were incubated for a further 2 h in an incubator at 37OC. One hundred microlitres of H202 was added to each well as a substrate for o-phenylenediawnine and incubation was continued for a further 30 min. The basis for the assay is the formation of a coloured product formed from o-phenylenediarnine in proportion to the amount of TNF-a bound to the antibody adsorbed to the plate. The absorbance of the wells was read at 490 nm with a Multiscan Titretek Microplate Reader (Flow Laboratories, McLean, Va.). The absorbance of test
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF NEW MEXICO on 11/29/14 For personal use only.
BHEF REPORTS / RAPPORTS BRBFS
0.0
1.0 Sultarnethsxazole Concentration (mM)
Sulfamethoxazoke Hydroxylamine Concentration (uM)
FIG. 1. Changes in TNF-or production in PBMCs incubated with murine microsomes, a microsoml-activating system, and increasing concentrations of sulfamethoxazole. TNF-or production was determind after 18 h incubation. Data represent the mean f SE of experiments performed in duplicate using the BBMCs of five volunteers. Fifty thousand cells were incubated per well. TNF-a! production by PBMCs incubated with B .0 mM sulfametRoxazole was 218 20 pg after 18 h incubation.
+
wells was compared to the absorbance of a standard curve prepared with serial dilutions of recombinant human TNF-er. Data analysis Linear regression was performed using a STAWiew 5 12 and a Microsoft Excel program with an Apple Macintosh Plus computer (Apple Computers, Cupertino, Calif.).
+
Ethical approval The protocol for donation of PBMCs by volunteers was approved by the Review Board for Health Research Involving Human Subjects, University of Western Ontario.
Results We first investigated the production of TNF-(r by PBMCs under the conditions used for our cytotoxicity assays as described above. Under our usual assay conditions, the PBMCs demonstrated a concentration-dependent production of TNF-a (data not shown). The production of TNF-a! by PBMCs under these conditions was time dependent, with minimal amounts of TNF-a detected after 1- or 2-h incubations, and with a marked increase in TNF-a concentration detected after 18 h of incubation (data not shown). Incubation of PBMCs with sulfamethoxazole, murine microsomes, and a microsomal-activating system or with the hydroxylamine derivative of sulfamethoxazoleproduced doserelated toxicity consistent with results described previously (data not shown; Shear and Spielberg 1985; Rieder et ad. 1988, 1989). We studied TNF-(r production under these conditions. TNF-(r production by PBMCs in the presence of murine microsomes, a microsomal-activating system, and increasing concentrations of sulfmethoxa.zole was found to be less than that observed when cells were incubated without the addition of a murine microsomal-activating system (Fig. 1). This reduction in TNF-(r production was found when cells were incubated for 1, 2, or 18 h. There was no correlation between the toxicity associated with these concentrations of sulfamethoxazole (for 8.15 mM, 18 & 3%) and the decline observed in TNF-a concentrations. We then determined TNF-a production in PBMCs incubated with the chemically synthesized hydroxylamine derivative of sulfamethsxazole. A similar decline in TNF-cr production was
FIG.2. Changes in TNF-a! production associated with incubation of BBMCs with increasing concentrations of the hydroxylamine derivative of sulfamethoxazole. TNF-er production was determined after 18 h incubation. Data represent the mean f SE of experiments performed in duplicate using the PBMCs of five volunteers. Fifty thousand cells were incubated in each well.
observed when PBMCs were incubated with increasing concentrations of SMX-HA (Fig. 2). Again, the reduced production of TNF-a! was present after 1, 2, and 18 h incubation. In the case of experiments performed with SMX-HA, the decline in TNF-a production was correlated with the number of viable cells present after 18 h incubation (R2 = 0.994).
The pathogenesis of sulphonamide hypersensitivity reactions appears to involve initial events involving dmg metabolism with the subsequent evolution of the reactions propagated by the immune system (Shear and Spielberg 1985; Shear et al. 1986; Park et al. 1987; Rieder et al. 1989; Spielberg et sk. 1989). TNF-a! is believed to be involved in the acute-phase response, as well as having effects on neutrophil activation, endothelial adhesion, calcium release from bone, and stimulation of tissue enzymes such as collagenase (Warren 1990). TNF-(r is believed to be a proximal mediator in septic shock in vivo, a disease state characterized by fever, malaise, and hypotension (DeForge et al. 1990). In concert, these factors suggested that TNF-a may play a role in the evolution of sulphonamide hypersensitivity adverse drug reactions. Although TNF-(r was considered previously to be only a product of monocytes and macrophages, Cuturi et ad. (1987) have found that lymphocytes are capable of rapidly producing high levels of TNF-a. Previous research into the mechanism(~)of sulphonamide hypersensitivity adverse drug reactions has used PBMCs, which are primarily lymphocytes, as the target tissue (Spielberg 1980; Shear and Spielberg 1985; Shear et ak. 1986; Rieder et ad. 1988, 1989). The advantage of PBMCs for this research is that they possess many of the main metabolite detoxication pathways but do not activate the sulphonamides to any significant degree (Spielberg 1980). The study of TNF-(r production in vitro is problematic, because of the difficulties in isolating TNF-a from biological samples collected in vivo. Thus, we elected to study the production of TNF-(r in response to sulphonamide reactive metabolites in vitro. We have demonstrated that TNF-a is elaborated by PBMCs under the conditions of incubation used for our cytotoxicity
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF NEW MEXICO on 11/29/14 For personal use only.
722
CAN. J. PHYSIOL. PHARMACOL. X7OL. 70, 1992
assays. Our results are in agreement with those s f Cuturi et ul. $1987),indicating h a t peripheral blood lymphocytes are capable of producing and releasing TNF-a in vktra? in the range of 150-250 pg/mE of culture medium. The time-related increase we observed in TNF-a release occurred over the time course previovsly used for our cytotoxicity assays, suggesting that increases in TNF-a production associated with exposure to reactive drug metabolites might be expected at this time. Incubation of cells with reactive derivatives of the sulphonamides, either generated in vitro by microsomal oxidation of sulfamethoxazole or chemically synthesized, was not associated with an increase in TNF-a release. Rather, a decrease in TNF-a concentration into the medium was demonstrated. As noted above, the decrease in TNF-a release occurred whether the reactive derivatives were provided in a synthetically pure form or generated in vitro by microsomal metabolism. This suggests that the cytofine TNF-a is not a mediator of the inflammatory response associated with sulphonamide hypersensitivity adverse drug reactions. The decline in TNF-a release when PBMCs were incubated with murine microssmes, a microsomal activating system, and sulfamethoxmole was noted at concentrations of sulfamethoxazole only associated with a modest degree of toxicity, suggesting that the mechanism of reduced TNF-a release may involve a direct effect of sulphonamide reactive metabolites on the immune system. Eeeder et al. (1991) have found that sulphonamide reactive derivatives appear to have an inhibitory effect on the function of natural killer cdls. In contrast, the dose-related decline in TNF-a release noted with SMX-HA appeared to correlate with the decline in viable cell number demonstrated by eytotsxicity assays. There was a m c h more pronsuncd decline in both toxicity and TNF-a release noted when PBMCs were incubated with the hydroxylarnine derivative of sulfarnethoxuole. This suggests that a decline in TNF-a concentration into the medium may be related to both a general toxic effect of the compound as well as effects of sulphonamide reactive metabolites on the h n c tioning s f the immune system. Mediators of the immune response that appear to be involved in the propagation of sulphsnamide hypersensitivity adverse drug reactions remain unhown. We are currently pursuing further studies with respect to the role of other immune mediators, including interleukin- 1 , in the pathogenesis of sulphonarnide hypersensitivity adverse drug reactions as well as mechanisms of alterations in immune function associated with reactive dmg mehbolites.
Acknowledgements This work was supported by grant MT 10641 from the Medical Research Council of Canada (MRC) and a start-up grant from the Child Health Research Institute. Dr. Rieder holds a Pharmaceutical Manufacturers' Association of Canada1 MRC Career Award. We wish to thank Drs. Neil Shear, W. Alrnawi, J. Bend, and K. Borkowski for their input into the preparation of this manuscript. Cutbari, M. C., Murphy, M . , Costa-Giorni, M. P., and Weinrnann, R. 1987. Independent regulation of tumor necrosis factor and Iymphotoxin production by human peripheral blood lymphocytes. J. Exp. Med. 165: 1581 -1594. DeForge, L. E., Nguyen, D. T., Igunkel, S. L., and Remick, D. G. 1990. Regulation of the pathophysiology of tumor necrosis factor. J. Lab. Clin. Med. 116: 429-438. Leeder, %. S., Dosch, H.-M., Harper, B. A . , et a&. 1989. FIu(ssrescence-based viability assay for studies of reactive drug metabolites. Anal. Biochem. 17'7: 364-372. Leeder, J. S., Nakhooda, A., Spielberg, S. P., and Dosch, H.-M. 1991. Cellular toxicity of sulfarnethoxazole reactive metabolites. II. Inhibition of natural killer activity in human peripheral blood mononuclear cells. Biochem. Pharmacol. 41: 575 -583. Bark, B. K . , Coleman, J. W., and Kitteringham, N. 8.1987. Drug disposition and drug hypersensitivity. Biochem. Pharmacol. 36: 581 -589.
Rieder, M. B., Uetrecht, J., Shear, W. H., and Spielberg, S. P. 1988. Synthesis and in vitro toxicity of hydroxylamine metabolites of sulfonarnides. J . Pharmacol . Exp . Tl-aer. 244: 724 -728. Rieder, M. J . , Uetrecht, J . , Shear, N. H., eta&.1989. Diagnosis of sulfonarnide hypersensitivity reactions by in vitro "rechallenge" with hydroxylarnine metabolites. Am. Intern. Med. 110: 286289.
Shear, N. H . , and Spielberg, S. P. 1985. ln vitro evaluation of a toxic metabolite of sulfadiazine. Can. J. Phy siol. Pharrnacol. 63: 13766- 1372.
Shear, N. H., Spielberg, S. P., Grant, D. M., et a&.1986. Differences in metabolism of sulfomamides predisposing to idiosyncratic toxicity. Ann. Intern. Med. 105: 179-184. Spielberg, S. P. 1980. Acetaminophen toxicity in human lympho9. Pharmacol. Exp. Ther. 213: 395 -398. cytes in \!&'fro. Spielberg, S. P., keder, 9. S., Cribb, A. E., and Dosch, H.-M. 1989. Is sulfamethoxazole hydrsxylamine (SMX-HA) the proximal toxin for s~lfamethoxazoLe(SMX) toxicity? Eur. J. Clin. Pharmacol. 36: 04.37 (Abstr. A173). Sugarman, B. J., Aggarwal, B. B., Maas, P. E., et al. 1985. Recombinant human tumor necrosis factor-alpha: effects on proliferation of nom~aland transformed cells in vitro. Science (Washington, D.C.), 238: 943 -945. Warren, 9. S. 19%. Interleukins and tumor necrosis factor in inflammation. Crit. Rev. Clin. Lab. Sci. 28: 37-59,
