With bacterial resistance rising, the need for new antibiotics is not exactly a secret. The question is where to get them.

Most antibiotics are made by what's known as semisynthesis. Basically, a natural precursor is made by fermentation of microorganisms, which has the advantage in that it makes large quantities. That precursor then is modified by synthetic chemistry to obtain derivatives with maximal antibacterial activity.

There are limits with that approach, mainly that molecules with a large number of functional groups can make synthetic chemistry difficult.

"You can only do what you can do and hope that the products that come out of that are effective," said Andrew Myers, professor of chemistry and chemical biology at Harvard University. "And a lot of great products have come out of that, but at the same time, there are limits to what is possible."

Myers is senior author of a paper in the April 13, 2005, issue of Science describing a fully synthetic route to one group of antibiotics: tetracyclines.

Structurally, tetracyclines consist of a row of four fused six-carbon rings, named A to D from right to left. They have what's known to chemists as a high density of polar functionality; many of their carbons have nitrogen or oxygen groups attached, and the exact orientation of those groups is important to molecular function. The polar functionality makes tetracycline very difficult to synthesize; such synthesis, in fact, has eluded chemists for more than 50 years.

Functionally, tetracyclines prevent the translation of bacterial messenger RNA into proteins by binding to the small subunit of bacterial ribosomes, blocking protein synthesis. Crystal structure studies of tetracycline bound to its bacterial target have shown that one place in which structural variation can influence antibacterial activity is the leftmost D ring. Achieving variability in the D ring can be done to some extent by fermentation, but again, that approach is more limited than a synthetic chemistry approach.

"For example, I don't know of any way, and I'm not aware that anybody else does, to substitute an oxygen or a nitrogen for one of the carbons on the D ring by fermentation," Myers told BioWorld Today.

Myers and his colleagues managed to tackle tetracycline synthesis by breaking it into two parts. They separately synthesized an AB and a D precursor, which then were fused.

"What we have shown is that the tetracycline problem can be reduced to the AB precursor," Myers said. "AB plus D gives ABCD, and D can be widely varied."

Myers and his colleagues actually used bacterial fermentation as the first step of making the AB precursor, which allowed them to produce that precursor in large quantities. Because the fermentation was used in the first step in the synthesis process, it did not greatly reduce the options for making different variants of the AB precursor.

"We still had a lot of latitude," Myers said.

The AB precursor then was further modified by synthetic chemistry and fused to a D ring, which had been made synthetically. Fusing the two precursors late in the synthesis game allowed the researchers to create many different D rings.

The fusion of the AB and D precursors is what creates the C ring. Like the D ring, the orientation of the C rings strongly influences the biological activity of the finished molecule. "It was very important, and lucky for us, that [the C ring's polar groups] were formed with the right orientations" in the synthesis process, Myers said. He added that his group had some theories on why this was the case, but no definitive explanation yet.

Using their approach, the scientists were able to synthesize several promising tetracycline derivatives that showed activity against a variety of different bacteria. One of them, which sports an added E ring, is active against multidrug resistant Staphylococcus aureus.

The next problem is how to make sufficient quantities of any promising new tetracyclines. "We can make multigram amounts of the AB precursor, and we can easily make 20 milligrams or so of final product, which is enough to answer all sorts of biochemical questions," Myers said. "But we clearly cannot make tons. There's a lot of work left before this can be used in the clinic."

That could prove another formidable problem. Part of the penicillin lore is that scientists were not able to synthesize it in sufficient quantities when it was first being used clinically; at one point, half the existing supply was used for the treatment of just one patient, and the drug was so scarce that it was recycled from patients' urine for repeated administration.