he relentless conveyor belt of new publications sometimes forces older papers out of the Top Ten before they have received due attention. One (#3) that is about to vanish was briefly mentioned two issues ago but deserves better. We suggested that Bruce Beutler's team (then at the University of Texas Southwestern Medical Center, Dallas) independently confirmed the work of others who had identified the so-called TLR-4 gene. The reverse is true. Beutler's group was first and, what is more, their publication represented a prodigious amount of work.

TLRs are a crucial component of the innate immune response, which early on in an infection mounts a non-specific defense against a generalized bacterial threat. It had long been known that some mice seemed to be resistant to LPS, the lipopolysaccharide produced by gram-negative bacteria, which triggers septic shock. Beutler, now at the Scripps Institute in La Jolla, California, mapped the mouse mutation but the closest he could pin it down by genetic means was to a stretch of DNA about 2.6 million base-pairs long. The mutation and the gene itself were finally identified in what was probably the largest such target yet attempted. "We solved it," Beutler tells Science Watch, "by a combination of overwhelming sequencing power and brute-force bioinformatic work."

The gene, TLR-4, proved homologous to the Toll-like receptors of Drosophila, which have a role in development. Mutations in Toll render the fly hyper-susceptible to fungal attack, which seems paradoxical given that mice with a similar mutation are resistant to LPS—they do not become sick, no matter how much LPS they are given. "Probably the mouse fails to recognize it is infected," Beutler explains. It does not mount an immune response, the bacteria multiply, and eventually the mouse goes into shock and dies.

Once Beutler's group had shown that TLR-4 responds to LPS and triggers the release of the cytokines that underly shock, other groups—notably Shizuo Akira's at the Hyogo College of Medicine, Japan—found similar systems. TLR-9 signals the unmethylated DNA that is typical of bacteria, while TLR-2 senses the peptidoglycan component of gram-positive and gram-negative bacterial cell walls. The therapeutic possibilities are obvious, but there had been early setbacks. One of the cytokines released in sepsis is tumour necrosis factor (TNF), but early attempts to treat sepsis by blocking TNF proved a failure. That could be because there are so many other signalling pathways that get around the block on TNF. "If you blocked all the TLRs," observes Beutler, "I suspect that the course of infection would be very different."

Might that do more harm than good? After all, the patient still has to deal with the bacterial load. There are those who say that given good antibiotic drugs, it is always possible to deal with the infection—provided shock does not enter the equation. Others argue that the lack of a good immune response may make it impossible to clear the infection with drugs. Beutler tells Science Watch that he believes that blocking the TLR-4 receptor will prove an effective approach to treatment, but the question, he says, can only be answered by clinical trials. "Looking a few years out, " Beutler adds, "it might be used in cases where sepsis is just suspected. You might treat with antibiotics plus a blocker of TLR signals."

Other explorations are also under way. It may be that at least some of the people who develop gram-negative infections have mutations in their TLRs. That is still a matter of speculation, and while Beutler has found polymorphisms in human TLR-4, he is reluctant to discuss them because the work is at such an early stage. But with at least 10 TLRs now identified, all of them believed to participate in microbial sensing, defects in TLRs might collectively be responsible for susceptibility to many forms of infection in humans.

Science writer Dr. Jeremy Cherfas
works with the Biotechnology and Biological Sciences
Research Council of the U.K., Swindon.