Surprising science is still being highly cited, but only in the outer reaches of the Hot Topics, where two papers of potential medical interest arouse attention. Hyeryun Choe and Michael Farzan at Harvard University School of Medicine led a group that investigated how the virus responsible for severe acute respiratory syndrome (SARS) interacts with the cells of its host. Their paper, at #10, demonstrates that angiotensin-converting enzyme 2 (ICE-2), which is involved in the regulation of blood pressure, is a functional receptor for the SARS coronavirus. This has all sorts of implications, as the authors note. It hints that other molecules that bind to ICE-2, or even a soluble form of the ICE-2 receptor, could block the course of SARS infection. The structure of the binding region of ICE-2 could also point towards a suitable vaccine against SARS. And a cell line already approved for vaccine production could be enlisted to make SARS for an attenuated or killed virus vaccine by giving it the ability to express ICE-2, which is also essential for viral replication.

At #13 (S. Rakoff-Nahoum, et al., Cell, 118[2]: 229-41, 23 July 2004; 27 citations) is an entirely fresh insight into toll-like receptors (TLRs). These are the basis of the innate immune system. They recognize certain bacterial products such as lipopolysaccharides and mount a non-specific attack. The problem is that the mammalian body is host to bacteria that are not merely not pathogenic, but positively benign. The lower intestine and colon house an estimated 10¹³ bacteria of diverse species that confer multiple benefits. Why don’t the TLRs recognize and attack them? It had been thought that the TLRs were kept away from the benign bacteria by the epithelia of the gut, while pathogenic bacteria had virulence factors that enabled them to slip through the gut wall, and that brought them into contact with TLRs. Gut microflora have been implicated in inflammatory conditions such as Crohn’s diseases, and so there is considerable interest in the study of commensal bacteria.

Ruslan Medzhitov and his colleagues at Yale University School of Medicine thought that TLRs might be involved, and so studied mice deficient in MyD88, an adaptor molecule essential for TLRs to induce inflammatory cytokines. They dosed MyD88-negative mice with a toxin that destroys the colon epithelium, allowing commensal bacteria to come into contact with TLRs on white blood cells and elsewhere. Medzhitov reasoned that as these mice could not mount an effective TLR-mediated attack, they would not show as severe a response to the toxin as mice with an intact TLR pathway. To the team’s surprise, MyD88-deficient mice showed much higher morbidity and mortality than wild-type mice, which were essentially unharmed.

A painstaking series of experiments eliminated all kinds of alternative explanations and came up with several new ideas about the role and operation of TLRs, not related to immune functions. Interactions between commensal bacteria and the host’s TLRs prevent epithelial injury, perhaps by stimulating the normal production of protective compounds. TLRs also influence cell reproduction in the lining of the gut, maintaining a balance between cell death and proliferation and thus maintaining the integrity of the epithelium. This is of particular interest because medical interventions that involve intestinal damage, such as radiotherapy, chemotherapy, and colonic surgery, are often preceded by antibiotic treatment to reduce the risk of infections. Medzhitov’s work suggests this may not be such a good idea, and that deliberately stimulating TLRs might improve tissue repair and healing.