Metabolic Magic

Streptomycetes are soil bacteria that could easily be mistaken for fungi, their cells snaking through the earth in long threads that resemble mycelial networks. To propagate when their survival is threatened, they break through the earth’s surface and then cannibalise themselves, using their last resources to build aerial platforms that release spores into the atmosphere to be carried away on the wind.

Streptomycetes also produce a bonanza of antibiotics. The mid-century ‘golden age’ of antibiotic discovery was largely thanks to them: by one count, 55 per cent of all antibiotics discovered between 1945 and 1978 came from streptomycetes. Why they should produce so many isn’t clear. Selman Waksman, the soil microbiologist who coined the term ‘antibiotic’ in the 1940s, suggested that streptomycetes don’t make antibiotics in nature. Only the ingenuity of the human experimenter, Waksman said, could coax these unruly bacteria into delivering the goods.

It was a characteristically ungenerous stance. Waksman was awarded a solo Nobel Prize in 1952 – for the discovery of streptomycin, the first antibiotic to be effective against tuberculosis – having cheated the student who did the crucial experiments, Albert Schatz, out of his share of the credit. Waksman had a vested interest in emphasising the human factor. Historically, US law had held that patents should not be awarded for ‘products of nature’. Thanks in part to Waksman’s efforts to patent natural antibiotics, the 1952 Patent Act amended that. (Some scientists and their lawyers argued further that bacteria should be patentable too. The current state of play in the US was established by a case in 1980: yes, but only if they’ve been genetically modified.)

Waksman was wrong that streptomycetes don’t naturally produce antibiotics. In soil they make only small amounts, but during their last metabolic gasp before sporulation, antibiotic production peaks – perhaps to coat their cells with a kind of anti-climb paint, preventing other hungry microbes from gaining entry at this critical moment. Waksman may even have got it backwards. Recent studies have provided evidence that streptomycetes are capable of making many more potential antibiotics than the ones recovered in usual laboratory conditions. The bonanza may not be over.

Unfortunately, knowing that a microbe makes an antibiotic is not the same as knowing how. This problem has always bedevilled antibiotic development. Penicillin was produced by the Penicillium fungus, but only in maddeningly small quantities that required the growth of vast amounts of sloppy mould in huge shallow vessels.

During the Second World War, American and British scientists tried to solve the problem by making penicillin synthetically in a test tube. John Sheehan, a chemist at Merck, thought the scale of their efforts was equalled only by the Manhattan Project. But they failed. The fungi were capable of far more advanced metabolic magic. Working on penicillin synthesis, Sheehan said in his memoir The Enchanted Ring (1982), was like ‘attempting to repair the mainspring of a fine watch with a blacksmith’s anvil, hammer and tongs’. His team at MIT eventually managed it in 1957.

We are now able to read microbial genomes, but elucidating how they produce antibiotics remains complicated. The workhorse of such investigations has traditionally been Streptomyces coelicolor (the sky-coloured streptomycete), whose genome was first sequenced in 2002. For a bacterium its genome is large and complex, stuffed with dense clusters of genes that work together to conjure up antibiotics.

In an age of AI-aided antibiotic discovery, experiments with S. coelicolor are rather passé. But last month, a team from Warwick and Monash universities reported research into how it makes a particular class of antibiotics called methylenomycins. Like saboteurs lobbing bricks over the cell’s factory wall, the scientists turned off various genes in the relevant cluster to see what happened, allowing them to reverse engineer the assembly line: molecule A is turned into B, which is turned into C, which becomes the final antibiotic.

They found that the antibiotic’s precursors were themselves antibiotics. That isn’t really a surprise. But some of them were between ten and a hundred times as effective as the final antibiotic. So why does S. coelicolor bother to turn them into something weaker? Perhaps, the researchers suggest, the molecular endpoint could have a function in nature other than killing. (The idea that many of the diverse molecules made by soil bacteria should be understood as signals between cells rather than weapons has been proposed before.)

In any case, it seems that the antibiotic diversity of the streptomycetes is not exhausted. But people need to go looking for it: the researchers started on this problem twenty years ago, with no particular goal in sight. Methylenomycins have been known about since the 1970s, but have never been approved for use in patients. Given that, the press release is surprisingly optimistic about the prospects for this ‘new’ intermediate antibiotic – yet to be tested in animals, let alone humans – as well as for finding other antibiotics using similar methods. We’ll see.

Looking at the intermediate steps before the final molecule has a long pedigree in antibiotic research. When German scientists at Bayer in the 1930s developed a synthetic red dye that killed bacteria, Prontosil, they didn’t bother to test a colourless intermediate molecule they used to make it. A group of rival French scientists did and found that it was just as effective. Prontosil’s reign was short-lived: since there was no patent on the colourless molecule, hundreds of near identical drugs flooded the market.

Then and now, our ignorance of what happens inside cells is often far more profound than is conveyed by the neat diagrams displayed in textbooks. One of the authors of the recent paper owns shares in a company that promises ‘chemistry from nature’. It’s a painful way to learn. Disentangling the biochemical cat’s cradles that power life can be nightmarish. But there are plenty of reasons to persist. As Sheehan said, ‘nature designed the penicillin molecule to teach organic chemists a little humility.’