By David N. Leff

A soil bacterium by the name of Deinococcus radiodurans is back in the news, nearly half a century after it turned up in a can of spoiled beef and pork. In 1956, food scientists were testing these meat containers for sterilization by gamma radiation. That treatment had killed off all the contaminating microbes save one. Only D. radiodurans survived the gamma rays and contaminated the meat. (See BioWorld Today, Nov. 19, 1999.)

¿This is an exotic sort of bacterium,¿ observed biochemist/structural biologist Frangois Franceschi, ¿because it is the most resistant organism against radiation found so far. D. radiodurans can withstand something like 3,000 times the radiation that will kill any other living being.¿

Franceschi, a research group leader at the Max-Planck Institute in Berlin, is senior author of the cover story in the current Nature, dated Oct. 25, 2001. Its title: ¿Structural basis for the interaction of antibiotics with the peptidyl transferase center in eubacteria.¿

¿What we did,¿ he told BioWorld Today, ¿was to analyze the structure of ribosomes at a molecular level, locating which parts of antibiotics are important for binding to the ribosome, and which parts of the ribosome bind to the antibiotic. This could help design new antibiotics. Until now, drug designers were synthesizing 10,000 compounds relatively close to the antibiotics they knew worked, then analyzed them, but without really knowing what their proposed chemical changes would do.¿

Franceschi continued: ¿With these ribosomal structures, we report, you could already make an educated guess as to which groups could improve interaction between the antibiotic and the ribosome. One of the things we found is that all these antibiotics interact only with RNA ¿ the ribonucleic acid. Ribosomes are a complex of proteins and RNA that translates genes into proteins. Its bigger of two subunits, which we analyzed structurally, consists mainly of some 33 different proteins. We found that a wide range of antibiotics interact only with ribosomal RNA. All the antibiotics we studied are not only clinically relevant, but also hamper the formation of the crucial peptide bond. There were previous reports on antibiotics that the bond binds to the small ribosomal unit, but so far no structural reports on antibiotics binding to the large ribosomal unit ¿ the place where the peptide bond is formed.¿

How To Build A Protein Chain

To elucidate the all-important peptide bond, Franceschi said: ¿Proteins are built from amino acid chains, and the task of the ribosome is to synthesize proteins in cells. The peptide bond happens when you couple two amino acids, and increase the size of the chain by one. That¿s the peptidyl-transferase activity. When the peptide bond forms between two amino acids ¿ one already in the ribosome, the other incoming ¿ the protein gets bigger by one amino acid. The genetic code is read by the ribosome, and this is what produces the protein.

¿The rationale behind our whole story,¿ he explained, ¿was that to study these ribosomal structures we needed to crystallize them. Then, because of the way ribosomes are made, we required a really strong radiation beam to reveal their atomic structure. This is normally done in synchrotrons ¿ particle accelerators that emit very high-intensity X-ray beams. We collect such data at both the CERN synchrotron in Grenoble, France, and the National Argonne Lab in Chicago, where we get beam time.

¿One of the problems we have with biological samples,¿ Franceschi pointed out, ¿is that the ribosomal crystals decay quite fast because of radiation damage. But because this Deinococcus bacterium was radiation-resistant, and we knew that ribosomes also resist radiation, we thought, OK, maybe crystals of this bacterium will withstand the radiation, so we could collect a lot more data with fewer crystals.¿¿

Franceschi and his co-authors beamed in on half a dozen frontline antibiotics to visualize their anti-bug interactions with the ribosomes in the cells of target pathogenic bacteria ¿ notably chloramphenicol, clindamicin and the large class of antibacterials called macrolides.

¿Erythromycin is very important clinically,¿ he observed, ¿because all these macrolide antibiotics are used a lot for human therapy, to treat everything from acne to stomach ulcers from Helicobacter pylori infection to syphilis. These are really wide-range antibiotics. They also have the advantage of relatively low toxicity, which means you can use them fairly safely. And in fact a lot of pharmaceutical companies are focusing their efforts on this macrolide class of antibiotics. Their results will also give a hand to people who are trying to design antibiotics that will be more efficient in a rational way.¿

Macrolides Perk Commercial Interest

¿One novel thing we did about these macrolides,¿ Franceschi said, ¿is that we showed exactly the way they work. In these structures we could see for the first time that these antibiotics are binding to the entrance of a tunnel. The ribosome is like a tunnel, and proteins being synthesized, once they leave this peptidyl transferase region at its entrance, go into the tunnel, travel all the way through it and come out at the back of this ribosomal exit. And this macrolide class of antibiotics, we reported, blocks the entrance of the tunnel. It¿s the first time the action of the antibiotic was visualized.

¿Two days ago,¿ he said, ¿some people from Pfizer [Inc., of New York] were really excited about using our models to design new antibiotics. People from [GlaxoSmithKline plc, of London], too, showed interest in our model because they are into macrolides.¿

Also, he noted, a small biotechnology company in New Haven, Conn., is starting up right now. It¿s called Ribex Pharmaceuticals Inc., and is affiliated with Yale University and its pioneer antibiotics structuralist Tom Steitz, with whom Franceschi is in touch.

Steitz, who holds an endowed chair of molecular biophysics and biochemistry at Yale, told BioWorld Today: ¿Knowing how these half-dozen or so antibiotics bind to the large ribosomal subunit, we started [Ribex] with the idea of structure-based drug design to modify existing antibiotics ¿ or designing new ones with a completely different chemical framework.¿ Yale has filed a patent covering Steitz¿s structural drug design, and licensed it to Ribex.

¿We¿ve been around as a company for six to eight months,¿ Susan Froshauer, CEO of Ribex, told BioWorld Today, adding, ¿We¿ve just closed a seed-money round.¿