By David N. Leff

In song, story and film, the legendary Robin Hood robs the rich and feeds the poor. So, too, one might say, does an enzyme in the body called matrix metalloproteinase (MMP).

This mighty molecule is infamous for causing many diseases, by destroying the extracellular matrices surrounding and supporting their cells. On MMP's "most wanted" list are osteoarthritis, cancer, bladder infection and gingivitis.

In arthritis, the enzyme presumably breaks down cartilage in joints. In cancer, tumor cells recruit it to tear down matrix, thus gaining access to the bloodstream, and spreading throughout the body. Drug designers have MMP on their hit list.

But like Robin Hood's two-hatted malefactor/benefactor rep, metalloproteinases also do good deeds, such as promoting normal growth and wound healing. In the small intestine of mammals, surprisingly, they unleash molecules that act as bouncers against gut-invading bacteria.

Today's Science, dated Oct. 1, 1999, breaks this news in an article titled: "Regulation of intestinal a-defensin activation by the metalloproteinase matrilysin in innate host defense." The paper's lead author is postdoctoral assistant research professor Carole Wilson, at Washington University in St. Louis.

"Its major finding," she told BioWorld Today, "is a new role for a metalloproteinase. People have been finding that these enzymes, which were thought to just degrade matrix, process other cell surface proteins. But this is the first time," Wilson added, "that a metalloproteinase has been associated with host defense. What we found is that the metalloproteinase we're studying, matrilysin, can activate defense molecules against microorganisms. The specific molecules we were looking at are made in the small intestine. They are very small, positively charged peptides, which kill bacteria by disrupting their membrane."

Cleavage Unlocks Gates Leading To Mayhem

These antibacterial molecules are made with long precursor sequences, which keeps them inactive while lying in wait for a microbial target of opportunity. What the paper's co-authors found was that matrilysin can cleave off that pro segment, thus galvanizing the defensive molecules, appropriately called "defensins." What triggers that cleavage, they showed, is exposure to bacteria.

The co-authors also demonstrated that matrilysin protects the intestine's epithelial-cell lining by regulating defensin activity.

How can defensins distinguish between harmful bacteria, such as Salmonella typhimurium, which causes food poisoning, and benign resident flora that permanently inhabit the gut? "This is something that people are still trying to understand," Wilson pointed out, "how the resident microflora manage to establish a niche, and avoid these antimicrobial defenses set up by the host. We have some data showing that germ-free mice, which have no endogenous microflora, don't seem to make matrilysin. Their guts' proliferative crypt cells are by definition devoid of bacteria. So we conclude that when a pathogen, or a new kind of bacterium, is encountered, that's what triggers the cleavage release."

The team's main experimental ally is a knockout mouse that lacks the gene for matrilysin. It resides on human chromosome 11, along with a cluster of other metalloproteinase genes. "I noticed that matrilysin in particular," Wilson recalled, "would show up in normal, uninjured epithelium. So I made the knockout about five years ago to see what the molecule was doing normally, and originally studied its role in cancer." Then she joined the Washington University group, focused on its intestinal role.

As reported in Science, the co-authors challenged matrilysin-minus knockouts (KO) and normal wild-type (WT) mice in vivo with two virulent micro-organisms that wreak food poisoning in humans, and sometimes death in mice. "This involved infecting the animals with E. coli, or Salmonella typhimurium," Wilson recounted, "which in mice causes typhoid fever and death, but in humans only a gastrointestinal illness.

"What we did first was orally infect the animals with E. coli. A couple of hours later, we took out their small intestines, and counted how many bacteria were remaining alive in the KOs vs. the WTs. Then, with the Salmonella, we just tested different doses of bacteria we administered, and measured how long it took the mice to get sick. The KOs sickened earlier and died sooner than did the WTs. The E. coli just stayed in the intestine, but the Salmonella actually invaded the rest of the body tissues. It's a systemic infection. Depending on the assay we used," Wilson pointed out, "there was a 10-fold to 100-fold difference in the tissue's ability to kill bacteria in both mouse strains.

"That allowed us to make certain conclusions about the role of matrilysin," Wilson observed. "By biochemical assays, seeing that the KO mice don't have these defensins, we could then say that defensins are important for meeting a microbial challenge, at least in the small intestine where we tested."

As for human lifestyles, Wilson pointed out, "we're interested now in trying to look in other tissues - upper airways, the reproductive system - where in both mice and people we see expression of matrilysin. So we'll try to somehow do the same type of experiments. I'll start by testing some human tissues, to see if there's a relationship between matrilysin and defensin processing, say, in the lungs. It could have importance in diseases like cystic fibrosis."

Caveat: Buffer Inhibitor Drugs With Antibacterials

Meanwhile, she noted that a number of companies are clinically testing MMP inhibitors, for treating cancer and arthritis. The idea being that these agents would block the ability of tumor cells to metastasize, or cartilage to erode.

"There is sort of a problem here with matrilysin," Wilson cautioned, "because it is showing up in a lot of tumors. But since the body makes it normally, it seems to have this protective function, too. You have to weigh the consequences of developing inhibitors. Maybe if you want to inhibit the enzyme, you have to be more careful about monitoring the patient for microbial infections.

"The problem is that a lot if inhibitors now target the whole family of metalloproteinases, including some that are beneficial. If you're going to inhibit matrilysin," she concluded, "you'd perhaps want to engineer something - either as part of the drug or part of the treatment - that would increase the bacteria-fighting ability of the patient, which could be compromised."