The notion that excessive production of cytokines, especially of TNF, can be central to the pathogenesis of numerous acute and chronic diseases is now widely accepted and documented. The first publication clearly demonstrating the causative role of TNF in pathology is featured in the current issue of The Journal of Immunology as part of the “Pillars of Immunology” series (1). In it, Bruce Beutler and colleagues, then working at the Rockefeller University in New York City, showed that passive immunization with rabbit antiserum or purified Ig raised against murine TNF protected mice from the lethal effect of bacterial LPS.
TNF (now also known as TNF-α) owes its name to the work of Lloyd Old and colleagues at the Memorial Sloan-Kettering Cancer Center in New York City, who showed in the 1970s that tumor necrotizing activity in the sera of Bacillus-Calmette Guérin-sensitized mice injected with LPS was due to a host factor, which they named tumor necrosis factor or TNF (2). Anthony Cerami and colleagues at Rockefeller University (located across the street from Lloyd Old’s lab) had independently studied a factor dubbed “cachectin” suspected of causing wasting and a hypertriglyceridemic state in rabbits infected with Trypanosoma brucei (3, 4). The amino acid sequence of human TNF was established in the mid-1980s (5, 6). When, shortly thereafter, mouse cachectin was purified to homogeneity, its amino-terminal sequence turned out to be highly homologous to that of human TNF, thus establishing the identity of TNF and cachectin (7). It is interesting, though not unique in the history of cytokine research, that its two originally postulated functions, mediation of tumor necrosis or cachexia, are generally not considered to be among the most important physiological and pathophysiological activities of TNF.
The main significance of the featured article by Beutler et al. (1) is that it clearly and directly implicated TNF as a mediator of the pathology produced by LPS. Based on their findings, the authors postulated that Abs to TNF and other agents that block TNF synthesis or action may be useful in the treatment of shock induced by septicemia and possibly other causes. The authors did note that protection from the lethal effect of LPS was strongest when antiserum to TNF was given 6 or 3 h before the introduction of LPS but was minimal or completely absent when antiserum was administered 3 or 6 h after LPS introduction. The authors also noted that anti-TNF serum, while preventing LPS-induced lethality, had not protected animals from all adverse effects of LPS, because these mice did become febrile and appeared ill and distressed. Beutler et al. (1) concluded that in eliciting the many LPS-induced pathological actions, TNF acts in concert with other mediators, including IL-1, IFNs, and lymphotoxin. The same group subsequently showed that injection of rats with recombinant human TNF caused hypotension, metabolic acidosis, hemoconcentration, and rapid death, indicating that TNF alone can induce at least some of the deleterious effects of LPS (8).
That TNF and other monocyte/macrophage-derived cytokines may be the cause of pathology in endotoxemia as well as in some forms of malaria had been postulated earlier (9, 10), but complete evidence could not be provided due to the lack of pure TNF or Abs to TNF in those days. It is also appropriate to mention that the first demonstration of the adverse role of a cytokine in pathogenesis and of the effectiveness of anti-cytokine therapy was provided in a study showing that Ab to type I IFN prevented liver cell necrosis and death in suckling mice infected with lymphocytic choriomeningitis virus (11).
The introduction of anti-TNF therapies to human medicine in the late 1990s revolutionized the treatment of many chronic inflammatory disorders, including rheumatoid arthritis, ankylosing spondylitis, Crohn’s disease, ulcerative colitis, and severe plaque psoriasis (12). Three therapeutic agents — two mAbs and one soluble receptor construct that binds TNF in the human body, thereby preventing TNF from triggering receptor activity — have been licensed in the United States and most other parts of the world. Notably absent from the list of their approved therapeutic targets is sepsis and septic shock. Even though the work of Beutler et al. (1), along with later evidence that Abs to TNF can prevent Escherichia coli-induced sepsis in baboons (13), sparked a keen interest in the use of TNF antagonists for the treatment of sepsis, numerous clinical trials have failed to demonstrate a statistically significant beneficial effect (14). The likely main reason for the failure of anti-TNF therapies in sepsis can be found in the results published in the original milestone paper by Beutler et al. (1): Abs to TNF were efficient in preventing death in mice only when administered before or a very short time after the introduction of LPS. Another probable reason, also documented by Beutler et al. (1), is that not all LPS-induced pathology is TNF mediated. Eventually, the effort invested in the development of TNF antagonists, originally intended for the treatment of sepsis, produced rich dividends in the form of therapies for chronic inflammatory diseases.