Science Editor

For CD23, it's location, location, location.

CD23 is an immunoglobulin E receptor; in its membrane-bound form, it sits on B cells, where it serves to keep a lid on IgE responses. As long as CD23 stays put in the membrane, all's well.

But once it gets cleaved and released into circulation, it exacerbates those same IgE responses. Allergies can be one result. Soluble CD23 also is found in the joints of mice with rheumatoid arthritis, and high levels of soluble CD23 are bad news for patients with B-cell chronic lymphocytic leukemia.

"In a very general sense, the membrane-bound form is protective against an IgE response, whereas the soluble form increases IgE levels," researcher Carl Blobel told BioWorld Today.

Because the CD23 receptor function depends on whether it is anchored in the safe haven of the membrane or at sea in the bloodstream, blocking it pharmacologically - which affects both kinds of receptors equally - is likely to cause as many problems as it solves. An alternate approach might be to block the enzyme that releases CD23 from the membrane.

But which enzyme?

The enzyme doing the cutting belonged to a family known as sheddases. Sheddases, also known as ADAMs, "function as signaling switches on the cell surface," Blobel explained. Incyte Corp., of Wilmington, Del., is testing sheddase inhibitors for targeting cancer. (See BioWorld Today, July 25, 2006.)

Previous papers had shown that a number of members of the ADAM family can cut CD23. But Blobel expressed a healthy skepticism toward in vitro results: "Oftentimes, people assume that if an enzyme can cleave something, it will do so in vivo - which is not always the case," he said. But "sometimes you get a different answer [in cells] than if you take a purified protein and feed it to a purified enzyme." ADAMs, as well as their substrates, are membrane anchored, which means that a given ADAM might never meet the substrate it theoretically could cut.

For this reason, Blobel, who is chairman of the arthritis and tissue degeneration program at the Hospital for Special Surgery, and professor of medicine at Cornell's Weill Medical College, and his colleagues used a smattering of approaches to pin down which ADAM is the one actually doing the dirty work under natural conditions. They reported their findings in the November 2006 issue of Nature Immunology.

The take-home message: Go for ADAM10.

They began by studying the relative abundance of membrane-bound and free-floating CD23 in cells that were either sheddase knockouts or overexpressed different sheddases.

Armed with a short list of ADAM candidates, they next went into knockout mice. That approach left two possible candidates: ADAM10 and ADAM33. However, studying ADAM10 in a knockout is literally not a viable option. Such animals "die so early that you can't work with them."

Blobel and his colleagues went to good old pharmacological inhibition, but that had its own challenges: Available ADAM10 inhibitors also will work on ADAM33, which was the distinction the researchers were trying to make. Cell culture helped here. The researchers were able to find a dose of the inhibitor G1 that inhibited ADAM10, but not ADAM33. "With that information, we were able to go into mouse primary cells - and into human primary cells, which is another jump," he said.

ADAM33, which had an effect on CD23 release in only some of the original cell studies, was not expressed on B cells. ADAM10, on the other hand, was not only expressed, but when the researchers treated B cells with an ADAM inhibitor at the dose that was selective for ADAM10, they found that pretty much stopped CD23 in its tracks; very little soluble CD23 was found from such cells. Taken together, the results "pretty much ruled out ADAM33," Blobel said, leaving ADAM10 as the last man standing.

The researchers' natural habitat was as varied as their methods. Participants hailed from the Hospital for Special Surgery at Weill Medical College of Cornell University, in New York; Virginia Commonwealth University in Richmond; units of GlaxoSmithKline plc; UCB Celltech; Christian-Albrechts University in Kiel and the Rhein-Westphalian Technical University in Aachen, both in Germany; Kyoto University in Japan; Schering-Plough Research Institute; and Amgen Inc.