The anti-inflammatory activities of cannabinoid receptor ligands in mouse peritonitis models.

In this report, we describe experiments in which cannabinoid receptor ligands were evaluated for effects on the development of a peritoneal inflammation when elicited in mice with thioglycollate broth or staphylococcus enterotoxin A. The cannabinoid receptor agonists [(-)-11-hydoxy-Delta(8) tetrahydrocannabinol-dimethylheptyl] (HU-210) and [(R)-(+)-[2,3-dihydro-5-methyl-3-[(4-morpholinyl)methyl[pyrrolo[1,2,3-de]1,4-benzoxazin-6-yl](1-naphthalenyl) methanone] (WIN 55212-2) blocked the migration of neutrophils into the peritoneal cavity in response to these inflammatory stimuli. This effect was caused by a delay in the production of the neutrophil chemoattractants, KC and macrophage inflammatory protein-2. HU-210 and WIN 55212-2 blocked neutrophil chemokines and neutrophil migration whether administered subcutaneously (s.c.) or intracerebroventricularly (i.c.v.). Their modulatory effects on the inflammation were antagonized by centrally administered [N-(piperdin-1-yl)-5-(4-chloropheny)-1-(2,4-dichloropheny)-4-methyl-1H-pyrazole-3-carboxamide hydrochloride] (SR141716A), a selective cannabinoid CB(1) receptor antagonist. This latter observation, and the ability of the cannabinoid receptor agonists to suppress the peritoneal inflammation at relatively low doses when administered i.c.v., indicated a role for central cannabinoid CB(1) receptors in the anti-inflammatory activities of HU-210 and WIN 55212-2. The cannabinoid receptor agonists had no effect on monocyte migration elicited by thioglycollate, despite their ability to suppress monocyte chemotactic protein-1 levels in lavage fluids. The cannabinoid CB(2) receptor antagonist, [N-[(1S)-endo-1,3,3-trimethylbicyclo[2.2.1]heptan-2-yl]5-(4-choro-3 methylphenyl)-1-(4-methylbenzyl)pyrazole-3-carboxamide] (SR144528) inhibited the peritoneal inflammation in a manner analogous to that of HU-210 and WIN 55212-2 when administered i.c.v., but it did not appear to act through central cannabinoid CB(1) receptors. The present results add to the body of literature indicating that cannabinoid receptor ligands have diverse anti-inflammatory properties.

Modulation of cytokine responses in Corynebacterium parvum-primed endotoxemic mice by centrally administered cannabinoid ligands.

The cannabinoid receptor agonists [(-)-11-hydoxy-Delta(8)tetrahydrocannabinol-dimethylheptyl] (HU-210) and [(R)-(+)-[2,3-dihydro-5-methyl-3-[(4-morpholinyl)methyl[pyrrolo[1,2,3-de]1,4-benzoxazin-6-yl](1-naphthalenyl) methanone] (WIN 55212-2) were previously shown to downregulate inflammatory cytokines (tumor necrosis factor alpha and interleukin-12) and to upregulate antiinflammatory interleukin-10 when administered intraperitoneally (i.p.) to mice before an endotoxin challenge. Cytokine modulation coincided with the onset of behavioral changes that are associated with cannabinoid agonist activated central cannabinoid CB(1) receptors. Both effects were antagonized by [N-(piperdin-1-yl)-5-(4-chloropheny)-1-(2,4-dichloropheny)-4-methyl-1H-pyrazole-3-carboxamide hydrochloride] (SR141716A) a selective cannabinoid CB(1) receptor antagonist. In the present study, we have investigated further the apparent role of central CB(1) cannabinoid receptors in cytokine modulation by HU-210 and WIN 55212-2. When administered intracerebroventricularly (i.c.v.), the drugs modulated cytokine responses at doses that were threefold to fourfold lower than those found effective by the i.p. route. SR141716A blocked cytokine modulation when coadministered centrally with the agonists, while a selective cannabinoid CB(2) receptor antagonist, (N-[(1S)-endo-1,3,3-trimethylbicyclo[2.2.1]heptan-2-yl]5-(4-choro-3 methylphenyl)-1-(4-methylbenzyl)pyrazole-3-carboxamide) (SR144528) had no effect. Surprisingly, SR144528 was found to modulate cytokines itself when injected i.c.v.

Effects of cannabinoid receptor agonist and antagonist ligands on production of inflammatory cytokines and anti-inflammatory interleukin-10 in endotoxemic mice.

Previous studies have shown that mice primed with Corynebacterium parvum produce higher levels of inflammatory cytokines than unprimed mice upon challenge with lipopolysaccharide (LPS). Herein, we describe experiments in which two cannabinoid (CB) agonists, WIN 55212-2 [(R)-(+)-[2, 3-dihydro-5-methyl-3-[(4-morpholinyl)methyl]pyrrolo[1,2,3-de]1, 4-benzoxazin-6-yl](1-naphthyl)methanone) and HU-210 [(-)-11-hydroxy-delta(8) tetrahydrocannabinol-dimethylheptyl], were examined for their effects on LPS-induced cytokines in C. parvum-primed and unprimed mice. These agonists have been reported to bind selectively to the CB2 and CB1 receptor subtypes, respectively. WIN 55212-2 (3.1-50 mg/kg i.p.) and HU-210 (0.05-0.4 mg/kg i.p.) decreased serum tumor necrosis factor-alpha and interleukin-12 (IL-12) and increased IL-10 when administered to mice before LPS. The drugs also protected C. parvum mice (but not unprimed mice) against the lethal effects of LPS. The protection afforded to C. parvum mice could not be attributed to the higher levels of IL-10 present in these mice after agonist treatment. The WIN 55212-2- and HU-210-mediated changes in the responsiveness of mice to LPS were antagonized by SR141716A [N-(piperdin-1-yl)-5-(4-chloropheny)-1-(2, 4-dichloropheny)-4-methyl-1H-pyrazole-3-carboxamide hydrochloride], a selective CB1 receptor antagonist, but not by SR144528 [N-[(1S)-endo-1,3,3-trimethylbicyclo[2.2. 1]heptan-2-yl]5-(4-choro-3-methylphenyl)-1-(4-methylbenzyl)p yrazole-3 -carboxamide], a selective antagonist at the CB2 receptor. Therefore, both CB agonists modulated LPS responses through the CB1 receptor. Surprisingly, SR141716A itself modulated cytokine responses in a manner identical with that of WIN 55212-2 and HU-210 when administered alone to mice. The agonist-like effects of SR141716A, which were more striking in unprimed than in primed mice, suggested that the antagonist also could function as a partial agonist at the CB1 receptor. Our findings indicate a role for the CB1 receptor subtype in cytokine modulation by CB ligands.

Endogenous mouse interleukin-10 is up-regulated by exogenously administered recombinant human interleukin-10, but does not contribute to the efficacy of the human protein in mouse models of endotoxemia.

In murine models of experimental endotoxemia, inflammatory cytokines as well as antiinflammatory interleukin-10 (IL-10) appear in the circulation after the injection of lipopolysaccharide (LPS). There is considerable experimental evidence to suggest that the major function of endogenously produced IL-10 is to down-regulate inflammatory cytokine production. Indeed, the protective effects of exogenously administered IL-10 against murine endotoxin lethality have been shown to correlate with its ability to inhibit the LPS-induced production of tumor necrosis factor-alpha (TNF-alpha) and interferon-gamma (IFN-gamma). While mouse IL-10 (mIL-10) has been used in the majority of studies in murine endotoxemia, we have found the human homolog to be equally effective in suppressing inflammatory cytokine production and in protecting mice from endotoxin lethality. However, we have recently observed that the LPS-induced endogenous IL-10 response is enhanced when mice are treated with recombinant human IL-10 (rhuIL-10). The upregulation of endogenous IL-10 by exogenously administered rhuIL-10 is particularly evident in mice that are primed with Corynebacterium partum (Proprionibacterium acnes). In the present study, we have examined the potential contributions of the increased circulating levels of mouse IL-10 to the inhibitory effects seen with rhuIL-10 on inflammatory cytokine production and endotoxin lethality. We show that pretreatment with a neutralizing anti-mouse IL-10 monoclonal antibody (mAb) has no effect on the ability of rhuIL-10 to suppress an LPS-induced inflammatory cytokine response in these mice. In contrast, the suppressive effects of the human protein on inflammatory cytokine responses are blocked completely by pretreating the animals with an anti-huIL-10 mAb. These data show that despite the up-regulated endogenous IL-10 response, it is the exogenously administered rhuIL-10 that is directly responsible for the suppressed inflammatory cytokine responses that are observed when the human protein is given to endotoxemic mice.

Elevated levels of NO in both unchallenged and LPS-challenged C. parvum-primed mice are attributable to the activity of a cytokine-inducible isoform of iNOS.

Elevated levels of nitric oxide (NO2-/NO3-) were detected in the serum of mice 3-7 days after priming with Corynebacterium parvum (Propionibacterium acnes). The serum NO2-/NO3- response was completely inhibited when C. parvum-primed (C. parrum) mice were treated with N(G)-monomethyl-L-arginine (L-NMMA) or aminoguanidine (AG) on days 6 and 7 post priming. The response was also inhibited when the mice were treated with interleukin-10 (IL-10) and the cytokine was most effective when given in multiple doses beginning on the day of priming. In contrast to L-NMMA and AG, IL-10 had no effect on the serum NO2-/NO3- response when administered to the mice on days 6 and 7 post priming. The inducible isoform of NOS (iNOS) appeared to be responsible for the elevated NO2-/NO3- response in C. parvum mice because iNOS transcripts were readily detected in their livers. Moreover, these transcripts as well as the circulating levels of NO2-/NO3- were dramatically reduced when the mice were treated with anti-tumor necrosis factor alpha (anti-TNF-alpha) or anti-interferon-gamma (anti-IFN-gamma) monoclonal antibodies (mAbs) during the priming interval. There was a modest increase (less than twofold) in the serum NO2-/NO3- response following a lipopolysaccharide (LPS) challenge to C. parvum mice (C. parvum/LPS mice). LPS had a more dramatic stimulatory effect if the levels of NO2-/NO3- preexisting in C. parvum/LPS mice were reduced by treatment with L-NMMA, AG, or IL-10 before the challenge. Thus the levels of NO2-/NO3- that preexisted in C. parvum/LPS mice appeared to influence their ability to mount a NO2-/NO3- response subsequent to the LPS challenge. The NO2-/NO3- response did not contribute to lethality in C. parvum/LPS mice because anti-TNF-alpha and anti-IFN-gamma mAbs were protective but had no effect on serum NO2-/NO3- levels when administered to mice 24 h before the LPS challenge.

Administration of interleukin-10 at the time of priming protects Corynebacterium parvum-primed mice against LPS- and TNF-alpha-induced lethality.

Several laboratories have described the protective effects of interleukin-10 (IL-10) in mouse models of lethal endotoxemia. In most of these experiments, protection was observed in normal mice that were given a lethal dose of LPS. However, we failed to observe protection with IL-10 in LPS-challenged mice that had been primed with Corynebacterium parvum (Proprionibacterium acnes). We have extended our studies with IL-10 in C. parvum-primed mice and in some cases have observed protection that appears to depend on the strength of the sensitization to C. parvum. When IL-10 was administered to mice at the time of priming, it was particularly effective in blocking sensitization, as evidenced by the inability of treated mice to mount a strong inflammatory cytokine response when subsequently challenged with LPS. Following such treatment with IL-10, C. parvum-primed mice were also protected from a subsequent lethal challenge with rMuTNF-alpha. In addition, the mice were protected against LPS- and TNF-alpha-induced lethality with a single dose of an anti-TNF-alpha or anti-IFN-gamma mAb given at the time of priming. Our results suggest that TNF-alpha and IFN-gamma are produced early after priming with C. parvum and are at least partly responsible for the enhanced sensitivity of the mice to LPS and TNF-alpha. IL-10 affords protection to the mice because of its ability to block the C. parvum-induced TNF-alpha and IFN-gamma responses.

Inhibitory effects of recombinant human interleukin 10 on disease manifestations in a P-->F1 model of acute graft versus host disease.

Interleukin 10 (IL-10) is a cytokine with both antiinflammatory and immunosuppressive properties. In the present study, we have examined the effects of recombinant human IL-10 (rHuIL-10) on the development of acute graft-vs.-host disease (GVHD) in unirradiated (C57B1/6JxA/J) F1 recipients of parental A/J lymphocytes. rHuIL-10 (2.5 to 100 micrograms/mouse administered subcutaneously) caused a significant reduction in splenomegaly in GVH mice. GVH splenocytes exhibited an augmented capacity to produce IFN-gamma when stimulated in culture with Con A or LPS. The IFN gamma produced in response to LPS stimulation was found to be derived from CD4+ and CD8+ T cells with little or no contribution from the NK1.1+ subpopulation of the GVH spleen. Treatment with IL-10 in vivo was found to diminish the capacity of splenocytes to produce IFN gamma when stimulated with LPS but not with Con A. IL-10 did not protect GVH mice from a lethal dose of LPS but caused a marked reduction in the serum TNF alpha response triggered by the LPS challenge. We conclude that IL-10 may be useful in controlling those clinical manifestations of acute GVHD that arise as a result of the activities of proinflammatory cytokines such as IFN gamma and TNF alpha.

The cooperative effects of TNF-alpha and IFN-gamma are determining factors in the ability of IL-10 to protect mice from lethal endotoxemia.

Recent studies have demonstrated that interleukin-10 (IL-10) has the capacity to protect mice from the lethal effects of endotoxin. In this investigation, we have examined the ability of IL-10 to protect both normal mice and Corynebacterium parvum-primed mice against endotoxin lethality. In the overwhelming majority of experiments, recombinant murine IL-10 (rMuIL-10) and recombinant human IL-10 (rHuIL-10) did not protect normal BALB/cJ mice from lipopolysaccharide (LPS)-induced lethality at doses up to 10 micrograms/mouse. Despite their inability to protect, both IL-10 preparations were highly effective in preventing the increase in serum tumor necrosis factor alpha (TNF-alpha) that occurred in response to the lethal dose of LPS. Moreover, a neutralizing antibody against TNF-alpha gave only partial protection when administered alone to BALB/cJ mice. Treatment with a combination of neutralizing antibodies against TNF-alpha and interferon-gamma (IFN-gamma) resulted in complete protection. In contrast to BALB/cJ mice, normal BDF1 mice were protected from lethal endotoxemia by treatment with both rMuIL-10 and rHuIL-10. However, IL-10 did not protect C. parvum-primed BDF1 against LPS lethality even though it caused a reduction in the LPS-induced serum TNF-alpha response in C. parvum-primed mice as well as in normal BDF1 mice. Neutralizing antibodies against TNF-alpha and IFN-gamma were protective when administered alone to normal BDF1 mice, as previously demonstrated in C. parvum-primed mice. These findings suggest that lethal endotoxemia is a result of the cooperative activities of TNF-alpha and IFN-gamma in normal mice of the BALB/cJ and BDF1 strains as well as in C. parvum-primed BDF1 mice. IL-10 appears to be less effective in protecting mice from lethal endotoxemia when cooperation between IFN-gamma and TNF-alpha is facilitated by high-level production of the cytokines as in C. parvum-primed mice or when there is evidence of strong synergy between them as in normal BALB/cJ mice.

Lipopolysaccharide-induced cytokine production and mortality in mice treated with Corynebacterium parvum.

Tumor necrosis factor alpha (TNF-alpha) has been shown to be an important mediator of the lethal effects of endotoxin in several experimental models of septic shock. However, studies with a recombinant human interleukin-1 (IL-1) receptor antagonist protein (IL-1ra) suggest a role for IL-1 as a mediator of septic shock as well. In the present study, we show that mice treated in vivo with Corynebacterium parvum are primed for the production of interferon-gamma (IFN-gamma) and exhibit an enhanced capacity to produce serum IL-1 alpha, TNF-alpha, and IL-6 when challenged intravenously with lipopolysaccharide (LPS). The majority of C. parvum-treated mice die within 24 h of an LPS challenge. Pretreatment with a rat antimouse TNF-alpha monoclonal antibody (mAb) protected 90% of the animals against the lethal endotoxin challenge, while an anti-IFN-gamma mAb gave approximately 75% protection. The anti-IFN-gamma mAb also caused a reduction in LPS-induced serum TNF-alpha and IL-1 alpha. Anti-IL-1 alpha, anti-IL-1 beta, and anti-IL-6 neutralizing mAb did not protect against lethality when administered to mice prior to the LPS challenge. These results indicate that TNF-alpha and IFN-gamma are major mediators of endotoxin shock in C. parvum-treated mice. The results further suggest that the IFN-gamma produced by C. parvum-primed mice in response to an LPS challenge serves as a stimulus for enhanced production of TNF-alpha and IL-1 alpha. These findings are consistent with an increasing body of evidence suggesting a major role for IFN-gamma in lethal endotoxemia.

A study of cytokine production in acute graft-vs-host disease.

The role of cytokines in the development of acute graft-vs-host disease (GVHD) was investigated in B6AF1 mice that were injected with parental A/J lymphocytes. Splenocytes from GVH mice exhibited an increased capacity to produce interleukin (IL)-1, IL-6, and TNF-a when stimulated in culture with lipopolysaccharide (LPS). This enhanced capacity was diminished following in vivo treatment with immunosuppressive drugs. Concanavalin A-stimulated GVH spleen cells produced significantly lower levels of IL-2 but higher levels of interferon-gamma (IFN-gamma) than did syngeneic spleen cells. Immunosuppressive therapy in vivo increased the capacity of GVH spleen cells to produce IL-2. However, immunosuppressants differed in their effects on IFN-gamma production. Sch 24937 (6-bromo-5-chloro-2-[1-(methylsulfonyl)acetyl] 3-(2-pyridyl)indole) enhanced or had no effect while cyclosporin A consistently decreased the capacity of splenocytes to produce this lymphokine. These results indicate that the capacity of GVH splenocytes for cytokine production can be differentially affected by the actions of some pharmacological agents. The data also indicate that there may be differential regulation of the production of IL-2 and IFN-gamma by the Th1 subset in the GVH spleen.

Inhibition of collagen II-induced arthritis in mice--a comparison of the effects of Sch 24937, immunosuppressants and nonsteroidal antiinflammatory drugs on the clinical expression of disease.

Previous studies from this laboratory have described Sch 24937 as a potent immunosuppressive agent that is particularly effective in suppressing humoral immune responses in mice. These findings prompted an evaluation of the effects of Sch 24937 in type II collagen-induced arthritis in mice where disease manifestations include the development of a strong humoral response to the collagen antigen. Sch 24937 reduced the incidence and severity of arthritis in collagen sensitized mice which appeared to be directly related to the immunosuppressive properties of the drug. However in contrast to the steroid betamethasone which also exhibited immunosuppressive activity, Sch 24937 did not prevent the changes occurring in the lymphocyte population of the draining lymph nodes of mice immunized with type II collagen. While the exact mechanism of the immunosuppressive activity of Sch 24937 remains to be elucidated, its mode of action in suppressing arthritis differs at least to some extent from that of a steroid.

Inhibition of graft-vs-host induced immunodeficiency with immunosuppressive therapy.

The pathologic features of the acute graft-vs-host disease occurring in unirradiated (C57Bl/6 X A/J)F1 mice injected intravenously with lymphocytes from the C57Bl/6 parent are similar to those reported for other parental----F1 hybrid combinations. When stimulated in culture with concanavalin A, lipopolysaccharide or alloantigen, spleen cells from B6AF1 mice that had been injected 11 days previously with B6 lymphocytes exhibited proliferative responses that were drastically reduced in comparison to the responses of spleen cells from F1 hosts injected with syngeneic lymphocytes. IL2 production in GVH spleen cell cultures was also diminished. Proliferative responses and IL2 production were partially restored in mice given immunosuppressive therapy with azathioprine, cyclosporin A or Sch 24937 a drug whose inhibitory effects on cellular and humoral immune responses in mice have recently been described. Phenotypic analyses by flow cytometry of the GVH splenocyte population indicated that the most consistent change in the GVH spleen was the appearance of an Lyt2+ L3T4+ T cell subset which in the majority of experiments was accompanied by an increase in cells expressing only the Lyt2 antigen. Both subpopulations were reduced in mice that had recovered immunological responsiveness following immunosuppressive therapy. The results suggest that in this GVH model the development of an immunodeficient state is directly related to the induction of an active T suppressor cell population and that such cells are effectively eliminated from the splenocyte population following treatment with some immunosuppressive drugs.

A study of the immunosuppressive properties of Sch 24937.

Sch 24937 (6-bromo-5-chloro-2-1[(methylsulfonyl) acetyl] 3-(2-pyridyl indole) was previously shown to be an immunosuppressant with potent inhibitory effects on B lymphocyte mediated immune responses. The present investigation was primarily designed to compare the immunopharmacological profile of Sch 24937 with that of cyclosporin A (CSA) a well-known immunosuppressive drug that has selective effects on T-lymphocyte-mediated immune responses. The results show that while the immunosuppressive activity of CSA in vitro is superior to that of Sch 24937, in general the latter agent is a more potent inhibitor of immune responses in vivo. The activity of Sch 24937 in rat models of adjuvant arthritis and experimental allergic encephalomyelitis is also described. While Sch 24937 exhibits the type of immunopharmacological profile that is likely to yield a good therapeutic effect in the treatment of immune-mediated chronic inflammatory diseases, hepatotoxicity associated with the compound precludes its development for the treatment of non-life threatening human autoimmune conditions.

A study of the mechanism of Con A-induced immunosuppression in vivo.

The plaque-forming cell (PFC) response to sheep erythrocytes (SRBC) is suppressed in a dose-related manner when concanavalin A (Con A) is administered intravenously to mice prior to or after immunization with antigen. The magnitude of suppression as well as the duration of the Con A effect greatly depends on the concentration of antigen used for immunization. Although profound suppression of the anti-SRBC PFC response is observed in intact mice pretreated with Con A for 4-24 hr, spleen cells from these mice do not exhibit suppressive activity when transferred into normal recipients or when cotransferred with normal spleen cells into irradiated recipients. Moreover, the cells from Con A-treated mice respond as normal spleen cells to SRBC when transferred alone into irradiated hosts. Suppression of the anti-SRBC PFC is only observed when adoptive hosts of cells from Con A-treated mice are also injected with Con A within 48 hr (but not 72 hr) of cell transfer and immunization. This time course of responsiveness to the suppressive effects of Con A is similar to that observed in normal mice and in irradiated recipients of normal spleen cells. The immune response to SRBC is also suppressed in adoptive hosts of normal spleen cells that are pretreated with Con A 4-24 hr prior to irradiation and cell transfer. Although functionally inactive when transferred into adoptive hosts, spleen cells from mice pretreated with Con A for 4-24 hr can suppress a primary antibody response to SRBC in vitro. The suppressive activity, which cannot be detected in the spleens of mice when the interval between pretreatment and assay is longer than 24 hr, is present in a subpopulation that bears the Thy 1.2 and Lyt 2 phenotype. Taken together the results obtained in in vivo and in vitro functional assays suggest that a suppressor cell population is activated following in vivo treatment with Con A, but that the cells rapidly lose their state of activation when removed from a Con A environment. This phenomenon is in all probability responsible for the failure to demonstrate suppressive activity in the spleens of Con A-treated mice using in vivo functional assays.

Comparative effects of azathioprine, cyclophosphamide and frentizole on cellular immunity in mice.

This report extends previous observations on the immunosuppressive properties of cyclophosphamide (CPA), azathioprine and frentizole (15) to include murine models of cellular immunity. Systemic and local graft vs host reactions (GVHR) were most effectively suppressed by CPA. In contrast to frentizole, both CPA and azathioprine were found to inhibit the proliferation of parental T-cells in a systemic GVHR. However, CPA was the only agent capable of inhibiting the proliferation of T-cells following contact sensitization with oxazolone. Mice pretreated with a high dose of CPA or frentizole prior to inoculation with the murine sarcoma virus exhibited accelerated tumor growth. However, there was no accelerated growth of murine sarcoma virus induced tumors or an SaI spindle cell fibrosarcoma during rather prolonged therapy with immunosuppressive doses of CPA, azathioprine or frentizole. Normal mice treated with CPA showed a more drastic reduction in lymphoid elements of the spleen and thymus than mice treated with azathioprine or frentizole. Studies on the mitogenic responsiveness of spleen cells obtained from normal mice after an eight day course of therapy suggested that CPA has some selectivity of action on B-cells and azathioprine on T-cells.

Comparative effects of azathioprine, cyclophosphamide and frentizole on humoral immunity in mice.

Data is presented comparing the activities of three immunosuppressive agents, cyclosphosphamide, frentizole and azathioprine in models of humoral immunity in mice. Cyclophosphamide and frentizole suppressed the primary and secondary plaque forming cell responses to sheep erythrocytes at lower doses than did azathioprine. Prolonged suppression of serum antibody titers occurred following short-term therapy with cyclophosphamide or frentizole, but not azathioprine. Azathioprine was also the least effective agent in suppressing a primary response to the T-independent antigen, trinitrophenylated lipopolysaccharide. All three agents were found to inhibit the induction and activity of suppressor cells at immunosuppressive doses.