SummaryThis study investigates the molecular mechanisms underlying the induction of and protection from T cell activation-associated hepatic injury. When BALB/c mice were given a single intravenous injection of concanavalin A (Con A) (>10.3 rag/mouse), they developed acute hepatic injury as assessed by a striking increase in plasma transaminase levels within 24 h. Histopathologically, only the liver was injured while moderate infiltration of T cells and polymorphonuclear cells occurred in the portal areas and around the central veins. The induction of hepatic injury was dependent on the existence as well as the activation of T cells, as untreated BALB/c nu/nu mice or BALB/c mice pretreated with a T cell-specific immunosuppressive drug, FK506, failed to develop disease. Significant increases in the levels of various cytokines in the plasma were detected before an increase in plasma transaminase levels. Within 1 h after Con A injection, tumor necrosis factor (TNF) levels peaked, this being followed by production of two other inflammatory cytokines, interleukin 6 (IL-6) and IL-1. Passive immunization with anti-TNF but not with anti-IL-1 or anti-IL-6 antibody, conferred significant levels of protection. Moreover, administration of rlL-6 before Con A injection resulted in an IL-6 dose-dependent protection. A single administration of a given dose of rlL-6 completely inhibited the release of transaminases, whereas the same regimen induced only 40-50% inhibition of TNF production. More than 80% inhibition of TNF production required four consecutive rlL-6 injections. These results indicate that: (a) TNFs are critical cytokines for inducing T cell activation-associated (Con A-induced) hepatitis; (b) the induction of hepatitis is almost completely controlled by rlL-6; and (c) rlL-6 exerts its protective effect through multiple mechanisms including the reduction of TNF production.

A single intravenous injection of concanavalin A (Con A) induces T-cell activation and an acute hepatitis in mice. This study investigated the role of interferon gamma (IFN-gamma) in the pathogenesis of this hepatitis model. Striking increases in the plasma levels of various cytokines, including tumor necrosis factor (TNF), interleukin-2 (IL-2), and IFN-gamma, were detected before the increase in plasma aminotransferase levels induced by Con A injection. TNF levels peaked within 2 hours, whereas IFN-gamma levels peaked at 6 hours after Con A injection. In contrast to a sharp peak of TNF levels, high IFN-gamma levels were detected for a more prolonged period. Passive immunization with anti-IFN-gamma monoclonal antibody (MAb) conferred a dose-dependent protection against liver injury in this model. This protection was observed when anti-IFN-gamma MAb was administered at least 30 minutes before Con A injection but not when given 1 hour after Con A injection. The protection from Con A-induced hepatitis was also induced by administration of rIL-6 before Con A injection. rIL-6 treatment induced significant albeit incomplete inhibition of IFN-gamma and TNF production, whereas this regimen did not affect IL-2 production. Despite striking protective effects of rIL-6 or anti-IFN-gamma MAb, comparable levels of cellular (both T cell and polymorphonuclear cell) infiltration were detected in liver sections from animals untreated, or treated with either rIL-6 or anti-IFN-gamma MAb. Moreover, electron microscopic examination showed that infiltrating T cells exhibited a blastoid appearance in all groups. These results indicate that IFN-gamma plays a critical role in the development of Con A-induced acute hepatitis and suggest that IL-6 administration can regulate the manifestation of hepatitis through mechanisms including the reduced production of inflammatory cytokines such as IFN-gamma.

Liver disease, including hepatitis and cirrhosis, repmodel. Striking increases in the plasma levels of various resent a major cause of morbidity and mortality. In cytokines, including tumor necrosis factor (TNF), in-chronic hepatitis, liver parenchymal cells are destroyed terleukin-2 (IL-2), and IFN-g, were detected before the by host lymphoid cells, especially by activated T cells increase in plasma aminotransferase levels induced by responding to viral antigens. 1 The cellular and molecuCon A injection. TNF levels peaked within 2 hours, lar mechanisms underlying the development of hepatowhereas IFN-g levels peaked at 6 hours after Con A injeccellular destruction have not been studied extensively tion. In contrast to a sharp peak of TNF levels, high IFNg levels were detected for a more prolonged period. because of the lack of availability of suitable animal Passive immunization with anti-IFN-g monoclonal anti-models. body (MAb) conferred a dose-dependent protectionRecently, a new hepatitis model was developed in against liver injury in this model. This protection was which liver-specific inflammatory lesions are induced observed when anti-IFN-g MAb was administered at by injection of concanavalin A (Con A). 2 Importantly, least 30 minutes before Con A injection but not when the hepatitis could be induced without pretreatment given 1 hour after Con A injection. The protection from with D-galactosamine, 3 which is an absolute requireCon A-induced hepatitis was also induced by adminis-ment for the induction of hepatitis by injection of endotration of rIL-6 before Con A injection. rIL-6 treatment toxin from gram-negative bacteria such as lipopolysacinduced significant albeit incomplete inhibition of IFNcharide (LPS). 4,5 Furthermore, the hepatic injury g and TNF production, whereas this regimen did not affect IL-2 production. Despite striking protective ef-appears to be a consequence of T-cell activation. 2,6 Our fects of rIL-6 or anti-IFN-g MAb, comparable levels of initial study 6 using this model showed that tumor necellular (both T cell and polymorphonuclear cell) infil-crosis factor (TNF) and interleukin-6 (IL-6) function as tration were detected in liver sections from animals un-positive and negative modulators, respectively, for the treated, or treated with either rIL-6 or anti-IFN-g MAb. development of Con A-induced hepatitis. Despite the Moreover, electron microscopic examination showed massive production of interferon gamma (IFN-g) by that infiltrating T cells exhibited a blastoid appearance Con A-stimulated T cells in vitro, the involvement of in all groups. These results indicate that IFN-g plays a this cytokine in the pathogenesis in Con A-induced critical role in the development of Con A-induced acute hepatitis has not been investigated. It is increasingly hepatitis and suggest that IL-6 administration can reguevident that IFN-g produced by T cells after stimula-

A single intravenous injection of concanavalin A (Con A) induces T-cell activation-associated inflammatory injury selectively in the liver. This study investigated the strain difference in the development of Con A-induced hepatic injury. Normal C57BL/6 and BALB/c spleen cells produced comparable levels of T-cell-derived lymphokines (interferon gamma [IFN-gamma], tumor necrosis factor alpha [TNF-alpha], and interleukin-2 [IL-2]) following in vitro stimulation with Con A. A single intravenous injection of Con A to C57BL/6 mice induced the plasma levels of TNF-alpha and IL-2 comparable with or slightly higher than those observed in BALB/c mice, whereas the same treatment resulted in an apparently lower level of IFN-gamma production in C57BL/6 mice. RNA from livers of Con A-treated C57BL/6 mice exhibited lower levels of IFN-gamma mRNA than RNA of BALB/c livers. Unexpectedly, a dramatic difference in the severity of hepatic injury was observed between C57BL/6 and BALB/c. Namely, the peak alanine transaminase (ALT) level was more than 15,000 U/L and inducible as early as 8 hours after injection of 0.2 mg Con A per mouse in the C57BL/6 strain, whereas the peak was approximately 3,000 U/L and induced as late as 24 hours after Con A injection in the BALB/c strain. The increase in plasma ALT levels was limited to less than 10% by injection of anti-IFN-gamma monoclonal antibody (mAb) in both strains. The C57BL/6 strain inducing lower levels of IFN-gamma exhibited higher IFN-gamma responsiveness as exemplified by the intrahepatic expression of an IFN-gamma-inducible gene, an inducible type of nitric oxide (NO) synthase (iNOS). These results indicate that, while IFN-gamma produced in vivo by activated T cells induces hepatic injury, there exists a striking strain difference in the induction of IFN-gamma-dependent hepatic injury.

We established an interleukin‐6 (IL‐6)–dependent cell line from murine plasmacytoma MOPC‐104E cells. This cell line (designated PIL‐6) was found to respond to murine and to human IL‐6, but not to any other cytokines. We used this cell line to investigate the involvement of IL‐6 production in type II collagen–induced arthritis in DBA/1 mice. Only marginal IL‐6 activity was detected in sera from DBA/1 mice inoculated with Freund's complete adjuvant (FCA) alone, with an unrelated protein (bovine serum albumin) plus FCA, or with type II collagen plus Freund's incomplete adjuvant. However, enhanced IL‐6 activity was observed in DBA/1 mice that had been injected with type II collagen plus FCA to induce arthritis. The elevated level of serum IL‐6 activity was associated with high levels of IL‐6 produced when lymph node cells from arthritic mice were stimulated in vitro with type II collagen. We also found that the L3T4+ T cell subset is responsible for the enhanced production of IL‐6 in arthritic mice. The results are discussed in the context of potential roles of IL‐6 in the induction and/or expression of chronic, progressive arthritis.