In the 1960s, women in Germany and England swallowed tranquilizerpills of a drug named thalidomide, to assuage menstrual cramps andthe nausea of pregnancy. Several thousand of them gave birth tobabies with stubby, flipper-like limbs and other gross birth defects.

Thalidomide proved to be a teratogen as well as a tranquilizerbecause, paradoxically, it consisted of a mixture containing twochemically identical compounds, one levo-rotary, the other dextro-rotary.

It was Louis Pasteur who in the mid-1800s discovered that lightbeamed through crystals of organic chemicals would rotate either tothe left (levo-) or the right (dextro-). Subsequent scientistsdetermined peptides produced by nature turned light left (levo-),while it took organic chemistry to rotate the light to the right (dextro-). One molecule is the mirror-image of the other.

Thalidomide's therapeutic effect came from its natural, levo, peptide;its toxic action from the synthetic dextro.

"Peptides in general are not very good drugs," observed structuralbiologist Peter Kim, of the Massachusetts Institute of Technology'sWhitehead Institute. "The reason is that they are degraded bynaturally occurring enzymes, namely proteases," he told BioWorldToday. "But," he added, "not all peptides _ for example,cyclosporin _ are lousy drugs."

Kim is senior author of a report in today's Science that describes anovel method of drug discovery, based on overcoming the dextro(D)/levo (L) biochemical dichotomy between natural and lab-madecompounds. Its title: "Identification of D-peptide ligands throughmirror-image phage display."

His technology exploits "the wonderful explosion that's occurred ingenetically encoded combinatorial libraries, namely phage displaylibraries. We create phage [bacterial viruses]," Kim explained, "inwhich we insert random DNA sequences. This leads to the expressionon the surface of the phage of random peptide sequences, via thetranslation of the genetic code."

Then comes the dilemma: "The phage-display library is going toexpress peptides made out of L-rotary amino acids. What we want isphage that display D amino acids on their surface, but we can't dothat. It would mean changing the genetic code and making a wholenew strain of virus."

So what Kim and his co-authors did, in effect, was to follow Alicethrough the looking glass. That is, "we chemically synthesized ourprotein of interest out of D amino acids, to give us the drug target,and used that normal L phage to express random sequences. Andthose sequences, made in the D version, bound to our natural Ltarget."

To test this approach, Kim, a Howard Hughes Medical Researchinvestigator, sought to obtain D-peptide ligands for the Src homology3 domain of c-Src tyrosine kinase.

"We picked that protein," he explained, "because it has a smalldomain we could chemically synthesize, and also because that c-Srcactivity is thought to be essential for osteoclast-mediated boneresorption. Some people, for example, at Ariad [Pharmaceuticals Inc.in Cambridge, Mass.] think that inhibition of Src activity could beuseful as a treatment for osteoporosis."

The team synthesized that Src domain out of D amino acids, thenused normal L phage to select for sequences that bound to that Dversion.

"In the long run," Kim observed, "it is possible that some of these D-amino-acid-based peptides could serve as initial leads fordevelopment of drugs."

At present, he and his team are "working on trying to develop thegenerality of the method, by using other target proteins of potentialtherapeutic interest."

Kim and the Science article's first author, Ton Schumacher, are co-inventors of a patent application on the mirror-imaging techniquefiled by MIT. Through the institute's licensing office, "somediscussions have been started, at this point with biotechnologycompanies," he said. Kim himself is a co-founder of ScriptgenPharmaceuticals Inc., of Medford, Mass. (See BioWorld Today, May22, 1995, p. 1.) n

-- David N. Leff Science Editor

(c) 1997 American Health Consultants. All rights reserved.