"People think sepsis and think infection, but infection is not the problem. The problem is an overwhelming inflammatory response, which you can get from many things. Infection, yes, but also inflammation or even trauma. In sepsis, proinflammatory cytokines are what will kill you. So if you want to treat sepsis, forget infection. Instead, treat the cytokines."

So said Luis Ulloa, assistant investigator at the Institute for Medical Research at North Shore-LIJ in Manhasset, N.Y., and senior author of a study that sheds light on just how one might deal with those proinflammatory cytokines. The paper, titled "Cholinergic agonists inhibit HMGB1 release and improve survival in experimental sepsis," appears in the November 2004 issue of Nature Medicine.

So far, treating the cytokines - or anything else in sepsis - is easier said than done. In a "News and Views" article accompanying the paper, authors Michael Mathay and Lorraine Ware recapitulated the sad story of developing sepsis treatments: more than 30 unsuccessful randomized clinical trials to date, using cytokines as well as a plethora of other approaches, including neutralization of endotoxin, inhibition of reactive oxygen species, nitric oxide and arachidonic acid metabolites. The authors dryly summarized that with the exception of activated protein C, progress in translating preclinical studies to effective clinical treatments "has been poor."

Among the failed trials are several investigating possible cytokine inhibitors, specifically TNF-a and IL-1b. However, TNF-a and IL-1b are released early in the development of sepsis; they appear before sepsis can be diagnosed in a clinical setting, and, by the time sepsis is diagnosed, they have often returned to normal. For that reason, Ulloa's group focused on later mediators of sepsis instead.

One such late mediator, the protein HMGB1, is released by macrophages about 20 hours after the induction of sepsis, which could make it a more promising candidate for therapeutic intervention. Additionally, prior research showed that HMGB1 leads to sepsis-like symptoms when administered, and that anti-HMGB antibodies were successful in treating sepsis symptoms.

From Receptors To Mechanisms To Symptoms

The researchers previously demonstrated that vagus nerve stimulation, which leads to acetylcholine release, could reduce blood levels of TNF-a in an animal model of sepsis. There, they began by investigating which acetylcholine receptor subtype is responsible for that effect. Pharmacology studies indicated that a nicotinic acetylcholine receptor subtype, known as a7nAchR, inhibits the release of HMGB1, and furthermore, that nicotine was more effective in inhibiting the release than acetylcholine itself. Protein and mRNA levels of HMGB1 were unaffected by nicotinic stimulation; instead, nicotine appeared to work by preventing HMGB1 from being ferried from the nucleus to the cytoplasm, which usually is the prelude to its extracellular secretion.

The scientists went on to investigate the effects of nicotine in two animal models: endotoxemia, a disease state that mimics some features of sepsis and is induced by administration of endotoxin, and outright experimental sepsis induced by a polymicrobial infection.

When given before and after endotoxin administration, nicotine reduced the clinical symptoms of endotoxemia and reduced mortality from 56 percent to about 19 percent of animals at the most-effective dose. Most encouragingly, both serum HMGB levels and mortality were reduced when nicotine was administered 24 hours after the induction of sepsis via a polymicrobial infection - a time point at which, the paper showed, mice had clear symptoms of sepsis.

Ulloa and colleagues also investigated the intracellular pathways leading to macrophage activation. They found that nicotine administration could reverse endotoxin's activation of the NFkB pathway, and that inhibition in turn prevented macrophage activation and HMGB1 release.

Ulloa drew a parallel between the research in Nature Medicine and earlier work on the neuro-immune axis that identified glucocorticoids as anti-inflammatory compounds that also are released after vagus nerve stimulation. Glucocorticoids work via the bloodstream, which makes their endogenous release slower-acting than the neurotransmitter-based mechanism identified by Ulloa and his colleagues.

Collaborators Wanted More Specific Agonists'

One of the co-authors of the Nature Medicine paper is a co-founder of and consultant to Critical Therapeutics Inc., of Lexington, Mass. That company announced in October that it concluded a Phase I trial of a small-molecule synthesis inhibitor of HMGB1 and said it planned to enter Phase II by the end of 2004. However, Ulloa noted that his group is interested in additional collaborations to develop more specific nicotinic agonists.

"No one is looking to use nicotine to treat sepsis, because it has both good and bad effects," he said, although he also noted that nicotine is in clinical use to treat ulcerative colitis, a chronic inflammatory condition, and that the work points to a possible mechanism for nicotine's effectiveness in that disease. "Our study has identified a new receptor and a mechanism. Now we are looking to develop more specific nicotinic agonists to control systemic inflammation," he said.