Science Writing Experiment: In Fleming’s Footsteps

The world is facing an invisible onslaught. Year by year, mutation by mutation, the fateful clock of microbial evolution ticks away, leading us towards a world unlike anything in living memory. It isn’t Blade Runner or Zombieland, though it does have more than a passing resemblance to an alien invasion. Call it the bacterial Second Coming.

Where does it all lead? If we’re not very careful, to a world where almost every family knows firsthand the misery of losing a child to infectious disease. A world where common scratches from everyday objects could lead to death or painful infirmity. A world where epidemics leap from livestock to humans and back, or catch rides on international trade routes, with little more to stop their progress than basic sanitation measures. A world where life-saving surgeries we currently take for granted get swamped by the risk of post-operative infection. A world, in short, that we thought consigned to ages long past. In the words of Margaret Chan, director of the World Health Organization, it is “an end to modern medicine as we know it. Things as common as strep throat or a child’s scratched knee could once again kill.’”

Over the past several years, many of us have heard frightening stories: the spread of MRSA (Methicillin-Resistant Staphylococcus Aureus) infections in hospitals, the over-prescription of antibiotics, or of exceptional individual cases of bacterial rapacity, like that of Aimee Copeland, who in 2012 underwent a quadruple amputation after a zip-lining accident led to a catastrophic Aeromonas infection. But, for the most part, we still take for granted that the arsenal of antibiotics will be there when we need them, quietly underwriting our way of life, insulating us from that strange and invisible realm of bacterial life that, in past centuries, always threatened behind the veneer of daily routine.

It’s not like we weren’t warned that that realm was far from vanquished–and warned well in advance, too. In fact, the warnings were loud and clear from the very beginning of the antibiotic age–and not exactly from your garden-variety doomsday theorist, either. They instead came, literally, from the source. In 1945, in the closing remarks of his Nobel prize lecture, Alexander Fleming, discoverer of penicillin, surveying the the revolution that he had almost singlehandedly launched, already foresaw what we now know to be an uncomfortably  reality:

“It is not difficult to make microbes resistant to penicillin in the laboratory by exposing them to concentrations not sufficient to kill them, and the same thing has occasionally happened in the body… The time may come when penicillin can be bought by anyone in the shops. Then there is the danger that the ignorant man may easily underdose himself and by exposing his microbes to non-lethal quantities of the drug make them resistant.”

The story of Fleming’s discovery of penicillin is one of the supreme examples of the interplay of sheer serendipity and tenacious directed effort in scientific discovery. Fleming had been searching and testing for antibacterial substances since shortly after the end of the First World War, and had already built himself a sizable reputation. For instance, Fleming’s pre-penicillin work had not been without some promising leads. In 1921, for instance, he had discovered that a protein found in human tears and saliva, which he dubbed “lysozyme”, has potent antibacterial activity. But lysozyme is a large and fragile molecule–too easily destroyed either in storage or in the digestive system of the patients, and too difficult to distribute to the location of the infection. For a practical medication, the ideal is something small, sturdy, and easily diffusible.

The great moment of discovery, by Fleming’s account, seemed almost banal at the time, and even tinged with annoyance. He had gone away for a month’s summer vacation in 1928, leaving his studies of the Streptococcus bacterium temporarily aside. Upon his return at the beginning of September, he found that the agar plates housing his prized Streptococcus had had unwelcome guests: spores of mold–which are plentiful in many indoor environments, including the notably untidy Fleming lab–had germinated in the middle of many of the plates, creating “a white fluffy mass which rapidly increases in size and after a few days sporulates, the centre becoming dark green and later in old cultures darkens to almost black”. No doubt this would require a re-do of his planned experiment.

Thinking the plates ruined, Fleming was about to throw them out, when he noticed something odd: the mold had created a visibly clear zone extending around it, very different from the cloudy appearance of the surrounding bacterial culture. In this zone, the bacteria had died (or “lysed”), becoming transparent. Something was being given off by this particular mold–a relatively common strain, which Fleming later identified as “Penicillium notatum”–that could mow down deadly bacteria even at a distance. Thinking again, Fleming decided these tainted plates might prove a decent subject of study after all.

The bright light of hindsight tends to paint many pivotal moments in science with a significance and purposiveness that was far from evident at the time, and to downplay the enormous number of false leads, weird artifacts, and almost-breakthroughs that litter scientists’ day-to-day efforts. On the one hand, this observation of the “kill zone” created by this moldy interloper was certainly worthy of a little more attention. On the other hand, at the time it was but one in a long line of “promising” leads being investigated by Fleming’s laboratory, and the initial observation of bacterial lysis was not even necessarily that remarkable. As Fleming wrote in 1945:

“To my generation of bacteriologists the inhibition of one microbe by another was commonplace. We were all taught about these inhibitions and indeed it is seldom that an observant clinical bacteriologist can pass a week without seeing in the course of his ordinary work very definite instances of bacterial antagonism… It seems likely that this fact that bacterial antagonisms were so common and well-known hindered rather than helped the initiation of the study of antibiotics as we know it today.”

Indeed, others before Fleming had observed remarkable effects of molds on bacterial growth, even in ancient times. In 1877, Louis Pasteur reported that anthrax bacteria were inhibited by certain molds. Later, an Italian scientist, Vincenzo Tiberio, had been able to conclude that molds from a well near his house could produce bacteria-killing substances; a couple years later, in 1897, the French physician Ernest Duchesne even observed such behavior in a Penicillium mold, Penicillium glaucum, commonly found in blue cheeses–but his research was ignored.

Here we see a remarkable example of the odd, Kuhnian inconsistency of scientific progress: a phenomenon of great importance may be stumbled upon by many investigators before one of them finally is able to assign it its proper significance in the context of current paradigm. Fleming’s results, however, stood out by being at once too precise in their description, and far too general in their implications, to be passed over as merely another isolated example of the eons-old war between microorganisms. In his historic 1929 paper, Fleming summarized several key points about the curious bacteria-busting mold and, importantly, reported that the active component it produced could be concentrated by drying up broth the mold had grown in and extracting the residue with alcohol. Using a quite standard nomenclature based on the organism that produced it (and to avoid the unbearable title of “mould broth filtrate”), Fleming decided to name this active component “penicillin”.

Even at this early stage in the research, penicillin exhibited many remarkable properties that hinted at something above and beyond the long-known, garden-variety inhibition of one microbe by another. For one thing, it was remarkably potent. A “good sample” of this agent could be diluted up to 800-fold after growing for a week and still completely kill staphylococci; or, using instead the “small drops of bright yellow fluid which collect on the surface of the mould”, up to 20,000-fold. It was also easily diffusible, unlike lysozyme, “so that in the few hours before the microbes show visible growth it has spread out for a centimetre or more in sufficient concentration to inhibit growth of a sensitive microbe”, and it was quite stable unaffected by brief boiling, though alkaline solutions degraded it faster than neutral or acidic ones. It did not kill microbes instantaneously, as for instance bleach does, but instead required several hours for full effect.

Penicillin also showed an interesting pattern of efficacy. It did not kill every bacterial strain: it generally had far less effect on Gram-negative bacteria like E. coli and V. cholerae than on Gram-positive ones like Staphylococcus, Streptococcus, and diphtheria. (In hindsight this makes sense, as Gram negative bacteria have a more complex multi-layered cell wall, which makes it harder for antibiotics of any kind to get inside.) Besides the obvious medical interest, Fleming noticed this selectivity also allowed the isolation of bacterial strains that, though normally overwhelmed and outnumbered by faster-growing strains like Streptococci, were insensitive to the antibiotic; penicillin could simply be used to kill off the competition. Interestingly, this marks the first example of a now very widespread molecular biological technique, that of selection by antibiotic resistance. Scientists aiming to incorporate a desired DNA strand into a cell include an antibiotic resistance gene in the DNA strand; the cells are grown in antibiotic, and so only those that have incorporated the DNA (with resistance gene) survive.

But probably most important of all, penicillin was amazingly non-toxic for animals, even in “enormous” doses. “Half a c.c. injected intraperitoneally into a mouse weighing about 20 gm. induced no toxic symptoms,” Fleming wrote, while in humans, “irrigation of large infected surfaces […] was not accompanied by any toxic symptoms.” In stark contrast, early antibacterials in use at the time were quite terrible in this regard. Pyocyanase, an extract of the highly dangerous bacterium Pseudomonas aeruginosa, often contained leftover toxins from the bacteria that made it hardly better than the infection it was supposed to treat. Another class of pre-penicillin antibiotics, including arsphenamine and novarsenobillon, were more effective, especially for syphilis; but they too posed a not inconsiderable health problem–they were, after all, based on arsenic. Sulfa drugs, developed in the 1930’s by Gerhard Domagk and colleagues, block the bacterial ability to make folic acid (vitamin B9); they were a considerable improvement, but still posed serious risks of toxicity.

Through these and other experiments, Fleming offered the world the first inkling of the fateful role a humble mold was to play in the future of humanity. The world, for its part, largely ignored the 1929 paper–for the time being. Fleming had found something that is extraordinary in any field, but in medicine especially: a real-life magic bullet. But even after the potential of the compound was fully appreciated, it had to wait for the development of new methods of mass production. The original mold strain from Fleming’s lab had been easy to grow and the extract was certainly potent, but it proved far from productive enough for widespread application. Nine years lapsed before Walter Florey and Ernst Chain took up the subject again, isolating the exact chemical properties of penicillin and reviving the search for a more productive strain. Astonishingly, the most crucial advance in the bringing of penicillin to wide use involved the discovery of a penicillin-producing strain hundreds of times more productive, now known as Penicillium chrysogenum, on a moldy cantaloupe in a trash can in Peoria, Illinois.

New mutant versions of the chrysogenum strain were soon developed, allowing still higher yields. Finally entering mass production by the latter half of World War II (about the same time that Fleming, along with Florey and Chain, shared the Nobel Prize in Medicine), penicillin rapidly overtook the sulfa drugs and became the treatment of choice for wounded Allied soldiers in World War II, offering complete cures for diseases ranging from strep throat to syphilis.


Antibiotics in general act as very specific molecular saboteurs or traps, targeting one or another of the bacterial cell’s critical systems–either systems that have no equivalent in the patient’s cells, or equivalents that are so different that they are not affected by the drug. The mechanism of penicillin is illustrative of this, for the target of the saboteur in this case is the bacterial cell wall, a tough casing of cross-linked proteins and sugars that holds the contents of the bacterium together. Animal cells don’t have cell walls at all, instead relying on their own internal scaffolding, a “cytoskeleton”, to hold their shape. Bacteria, however, must continually maintain their cell wall in order to be able to divide, and even to avoid exploding from the pressure of all the substances dissolved inside them. This upkeep, in turn, depends on an enzymes called transpeptidases.

Penicillin turns out to be one of a vast class of antibiotics that work on the same general principle, one that is exceptionally clever yet also simple. All of them contain a molecular structure known as a “beta-lactam” ring, which is what chemists would call “strained”–its bond angles are too tight, which makes reactive and under tension, yet perfectly stable when left by itself. This makes it something like the biochemical equivalent of a spring-loaded bear trap, for in the presence of the right trigger, the beta-lactam readily snaps forth and locks on to whatever hapless molecule happens to be nearby.

Moreover, beta-lactam antibiotics’ shapes resemble an ingredient of the bacterial cell wall, allowing them to bind tightly to the transpeptidase. Then the trap is sprung: the unstable lactam ring breaks, attacking the transpeptidase and forming an irreversible bond that effectively ruins it. With its transpeptidases out of commission, the bacteria’s cell wall soon becomes like string cheese, ready to peel apart. It’s great fun to watch the result through a microscope: one by one, the single-celled beasties burst like water balloons, leaving behind harmless, ghostly clouds of cellular goo.


For a while following the debut of penicillin as a mass commodity, progress in the development of new beta-lactam antibiotics accelerated. The cephalosporins–more potent relatives of penicillin–were discovered in the 1940s, also from a mold, becoming available in the mid ’60s. Then, beginning in the 1970s, came the carbapenems, with a tougher versions of the beta-lactam structure; a strong carbon-carbon bond replaces a more vulnerable sulfur bond, making these drugs harder for bacteria to degrade and effective against more strains. The industriousness of medicinal chemists was set to work finding detours around the selectivity Fleming had noticed in the original penicillin compounds, each of these sub-classes soon was populated with dozens of derivatives, each optimally deadly for a specific germ or type of infection that the parent compounds might hardly touch–a broad and impressive armamentarium indeed.

But these improvements on the beta-lactam idea, it seems, have now been pushed out nearly as far as they can go, while nature continues to prove a most inventive adversary. One by one, despite the best tricks and tweaks of the medicinal chemists, the beta-lactams have fallen victim to natural (or in this case, artificial) selection’s arms race. Even at the beginning of penicillin’s splendid success a family of bacterial enzymes known as beta-lactamases was waiting in the wings.

If penicillin is like a bear-trap, then think of the beta-lactamases as like a branch the bear uses to spring the trap so that he can proceed unhurt. There are many kinds of beta-lactamases, but all work by reacting with the beta-lactam ring, popping it open before it can ever reach the transpeptidase. The sabotage is thwarted, and the bacterial wall stays intact.

What’s more, a world of widespread antibiotic use means there is pressure of selection, so that only those bacteria that can resist the antibiotics will survive. Either by mutation, or by getting hold of the DNA needed to produce beta-lactamases, bacteria evolve the ability to disable (or otherwise evade) the antibiotic and soon replace their more vulnerable competitors. A kind of freewheeling genetic stock market is also in play between microbes, where DNA fragments of at least 20,000 resistance genes are passed along from cell to cell, making still more bacteria resistant, all in the raw Darwinian pursuit of maximum growth and fitness.

Much has been said about the folly of prescribing antibiotics frivolously or carelessly–for instance as a kind of placebo for flus and colds, which are viral, not bacterial, in origin–and there has been plenty of contempt more or less deservedly heaped on patients who abandon their treatment courses of antibiotics as soon as they feel better, thus giving the germs another round of opportunity to acquire resistance. Agriculture, however, may be the worst culprit of all, as one of the lesser-known consequences of Fleming’s revolution is that antibiotics have in applied to far greater quantities to protect animals than people. Industrial farmers soon found that by lacing their livestock’s food with antibiotics they could avoid most of the risk of transmissible infections decimating their herds, making it possible to raise them in much more confined spaces. At the same time, animals raised with a constant regimen of cheap, plentiful antibiotics grow larger, faster. The result has been a truly industrial-scale overuse of antibiotics with them commonly found in sewage from farms–it’s hard to imagine designing a more comprehensive way to maximize the amount of bacteria getting a chance to develop resistance, whether through DNA-swapping or outright mutation. The result has been an evolutionary overdrive.

First to fall was penicillin itself, which for some strains is now down to about 10% effectiveness. Then the cephalosporins began to fade as well, with the first “extended spectrum” resistant cases appearing in the 1980s. More recently, the carbapenems, which are kept as last-line treatments and whose structure was thought to be nearly immune to the development of resistance, have begun to appear vulnerable; imagine scientists’ unpleasant surprise when, in the late 1990s, they first found strains of Enterobacteriaceae harboring both new types of beta-lactamases–such as NDM-1, first found in New Delhi in 2008 and now worldwide–that specialize in chopping the ostensibly-invulnerable carbon-carbon bond of the carbapenem, as well as tiny “efflux” pumps that drive the antibiotics back out of the cell faster than they can be absorbed. (An excellent timeline of resistance developments can be found here.) Step by step, the antibiotics revolution was being chipped away; terms like MRSA and “superbug” slipped into our lexicon.

While it is astonishing how much has been achieved using variations on the single beta-lactam theme found in penicillin, it might also seem in retrospect like an epic case of putting too many eggs in one basket, an invitation to widespread resistance. After all, the beta-lactams were far from the only class of antibiotics discovered, but they were the first and the most widespread, accounting for 65% of the world’s market for antibiotics as of late 2003. Alas, the story of resistance was hardly confined to the beta-lactams. Most other classes–such as quinolones, polymyxins, tetracyclines, aminoglycosides, drugs that sabotage diverse targets from the cell membrane to protein synthesis to DNA copying–gradually fell prey to the molecular arms-race as well, as drug developers’ new tricks and modifications to old prototypes were outpaced by the constant evolutionary experimentation of the bacteria themselves.

In a manner reminiscent of a phased tactical retreat, medical providers have developed a tier system, retaining a few especially strong and relatively unscathed antibiotics for the most problematic cases, like the inner walls of a fortified city. Many of these last-line antibiotics were natural choices for the job not only because they had not yet evoked resistance, but because they were not so easy to administer carelessly, and their side-effects made them unpalatable, thus retarding the development of resistance in the first place. Vancomycin, for example, has been available since the 1950s, but was long considered an unappealing option because it has to be given intravenously and carries a risk of rashes and kidney damage. But as resistance claimed other safer and more effective antibiotics throughout the 1980s, vancomycin use shot up more than a hundredfold. Nowaydays it is often the last line of defense for highly resistant cases of MRSA and intestinal infections.

Even with this tiered approach of keeping in reserve rarer, little-known and even undesirable antibiotics that had not had much chance to induce resistance, the retreat has continued. The carbapenems, as already discussed, have given ground on many fronts. Vancomycin too has been beaten by some bacteria, with the first resistant cases recorded in 1988; though such cases remained thankfully rare until around 2006, they now can be found “in hospitals in most countries“. Yet, for all the defeats suffered against the onslaught of resistance, there remained nearly always some antibiotic that could take care of a given infection–if the right combination could be found in time. Also, there were still some antibiotics that were untouched by resistance entirely.


Late in 2015, that final wall came down; and not surprisingly, it happened in the agricultural sector. Last November, using genetic sequencing tools to scan DNA in the antibiotic-laden stew that is the Chinese livestock industry, a group of scientists reported on a new E. coli strain carrying a never-before-seen gene, called MCR-1, that confers resistance for colistin, an antibiotic of the polymyxin class.

Rather like the case of vancomycin, colistin had been neglected for many years due to its kidney toxicity but, as the front lines of bacterial resistance conquered one standby drug after another, it had increasingly become a critical fall-back option for the treatment of many kinds of difficult Gram-negative infections. In fact, until the Chinese discovery, it was the only remaining antibiotic for which there was no known microbial resistance mechanism. This had made it, for some infections, literally the last line of defense–though this unfortunately did not suffice to prevent its wide-scale use in agriculture.

Following the news of MCR-1’s emergence, articles with doomful titles such as “Antibiotics are dead–it’s time to find another cure” filled the media for a time, along with somewhat forcedly optimistic claims that this or that tentative new discovery in the biomedical pipeline would ensure the uninterrupted supremacy of antibiotics. Then the news cycle subsided, and both the optimists and the pessimists alike turned to other matters. But in a world of rapid and comprehensive global trade, combined with the genetic stock market where resistance genes are traded among bacteria with alarming ease, MCR-1 was bound to travel just as quickly as the genes that laid low the carbapenems.

Sure enough, just in the past week or so, news broke that MCR-1 has been detected in the United States for the very first time, in a 49-year old Pennsylvania woman–bringing the issue back to a boil. In the words of this latest report,  from the Walter Reed Army Institute of Research, the rise of MCR-1 “heralds the emergence of truly pan-drug resistant bacteria”. The rise of multi-drug resistant bacteria that cannot be killed by any known medicine seems to have jumped the final hurdle.

But we have heard plenty by this point about the scary advances in resistance by this point. What’s somewhat less talked about is the inverse question–the long-running dearth of new antibiotics to combat these new superbugs. What has caused this dearth, and how can we move past it?


Antibiotics were never exactly easy to discover and develop, but for a long time it was nevertheless a challenge with reasonable chance of success and a reasonable payoff. In the so-called “golden age” of antibiotic discovery, not only was medicinal chemistry busily tweaking old antibiotics to yield useful variants with new effects but, beginning roughly from the time of Fleming’s Nobel lecture into the late 1960s, whole new classes of antibiotics beyond the beta-lactams were being discovered every few years.

How did scientists of that bygone era manage such productivity? By looking through dirt. Not only is dirt easy enough to find, but it turned out that soil bacteria (and fungi) had developed quite a number of chemical trump cards in the billion-year-old competition that is microbial life. Penicillium, it soon turned out, was just one of the star players in this competition; an invisible kingdom of previously un-investigated soil bacteria beckoned. Following the lead of the Russian-American microbiologist Selman Waksman, who discovered a slew of antibiotics from soil bacteria (and coined the term “antibiotics”), an army of scientists fanned out over the globe in search of new microbes, culturing them in the lab and testing them, laboriously, for antibacterial prowess and then, for those that showed such prowess, passing them on to the biochemists to isolate and characterize the active compounds responsible. Out of this collection, the vast majority of lead compounds, as usual, led precisely nowhere, being either too toxic, too hard to make, or too ineffective. But enough of them proved to be magic bullets to sustain a wave of progress, to keep well ahead of resistance–and to make the enterprise profitable.

But then the well of soil-borne chemicals began to dry up. In fact, after 1987, not a single new antibiotic class (see p. 31) was discovered. As this drought deepened, much hope was placed in new strategies, such as high-throughput screens and “rational” antibiotic design. In the former case, scientists proposed to use large numbers to increase their odds: screen massive libraries of compounds for new leads using cutting-edge high-fidelity assays until, like a gold panner who swirls enough river sand, a few new golden lead compounds pop out of the rabble. In the latter case, the goal is to begin with the exact structure of some vulnerable bacterial molecule (often from X-ray crystallography or NMR), and then build up, atom by atom, the structure of the perfect biochemical saboteur, like deducing the shape of the perfect key for a strange lock. This would be done either with a combination of the know-how of structural chemists, and raw computational power.

With few exceptions, both kinds of discovery largely proved a flop, with ever more expensive techniques pressed into service to screen ever more lead compounds meeting with ever-diminishing returns. As antibiotic resistance made more and more news, investments in R&D rose steadily in the 1990s, but results were severely discouraging and expensive. In a typical example, 14 high-throughput runs using state-of-the art methods, at a cost of $1 million per run, produced only one new lead compound–which, like so many promising leads, did not make it through clinical trials. And as if to taunt the already-dwindling ranks of the antibiotic searchers, these expensive combinatorial and rational approaches often turned up none other than the same classes of antibiotics that had been discovered decades before.

This last result, in itself, is rather astounding. If one tries to enumerate all the possible molecules small enough to be readily synthesized and easily absorbed as a drug, one is soon confronted with a mind-boggling range of possibilities. For every atom in the molecule, another substituent may take its place, in turn to have new substituents added on to it, and so on in a great combinatorial explosion. This imaginary realm of all possible molecules is known as “chemical space”, and even when limited just to smallish, antibiotic-sized molecules it contains, by some estimates, more possible compounds than there are atoms in the Earth. (Imagine digging all over and through the planet for new atomsonly to find precisely the same few atoms that you threw out years before.)

So either combinatorial chemistry and rational design suffers from a dismal lack of flexibility and imagination, so that it only explores a very narrow and antibiotic-poor nether-realm of chemical space, or we are faced with the uncomfortably awe-evoking possibility that nature, in billions of years of evolution, only hit upon a handful of really good solutions to the problem of killing bacteria without killing everything else around them. It may be that the new antibiotics are out there, somewhere, in that hypothetical chemical space, but involve shapes or elements so strange or so hard to synthesize that they don’t occur in nature and never get touched by high-throughput screens.

After a while, drug companies turned to more tractable and lucrative problems, ones that offer more repeat business–unlike antibiotics, which are often fully curative after just a dozen or so doses. Antibiotic development is now looked on as a kind of white elephant by most companies, with a few passionate advocates bowing to profit motives and disappointing results. Even if we take the hopeful and far from unreasonable view that there do remain novel antibiotic classes out there to be found out in the wilds of chemical space, we seem destined to tread over the same ground over and over–unless we can tap into new and promising lead compounds from a very novel, unexplored source.


The good news is that we are just now encountering a very large unexplored potential source of lead compounds, one the soil-searchers of the antibiotic golden age had no clue about. From comprehensive analyses of DNA diversity in soil samples, we now know that of the myriads of bacterial species to be found in a typical clod of healthy dirt, only about 1% can actually be grown in the lab using conventional techniques, like agar plates or nutrient broths. If only we can find a way to catch hold of this other 99%, our pool of candidates for new antibiotic discovery would grow a hundredfold.

There’s something deeply wondrous and humbling about the realization that our understanding of practically an entire kingdom of life may only reflect the limitations of our own experiments, and the propensity of a few unusually audacious and self-sufficient species to define our world view, like the class clown who steals all the attention. The vast majority of soil bacteria, it turns out, are more quiet personalities–slow growing, and so intensely connected and adapted with their own micro-ecosystem that they literally cannot exist apart from it. In this twist, the classic story of the microbial war of all against all has turned out to be just as much a story of cooperation and mutual dependence.

Yet for all the talk about “microbial dark matter” (huge numbers of microbial strains thought to be un-culturable or otherwise undiscovered), most of the inroads into the subject have been confined to genetic sequencing, a bit of proteomics, and much less to biochemistry or drug discovery, even though the latter are usually the main rationales given. But there have been striking counterexamples, of which none might be so promising as the extremely recent discovery, in a species of bacterium no one previously knew existed, of the first truly novel antibiotic in decades–one that may offer a unique new approach in the battle against resistance.

In November of 2015, a large team of researchers, mainly from the small biotech company NovoBiotics in collaboration with Northeastern University and the University of Bonn, announced in the journal Nature that they had used a novel proprietary technology, called the iChip, to successfully culture the previously un-culturable. The bacterium in question was dubbed Eleftheria terrae, and its near-wondrous product was an unusual peptide molecule, which they named teixobactin.

To sum up the secret of their success: if you can’t take the bacteria into the lab, take the lab to the bacteria. The iChip itself is surprisingly low-tech: a thin piece of plastic with numerous rows of tiny holes drilled in it, each containing a tiny plug of agar plus a diluted soil sample, and walled off by membranes on each side. To allow the un-culturable specimens to grow as naturally as possible, the iChip is then simply stuck back into the soil, and after a while the bacteria within can be taken to the lab for analysis. This method, claim the inventors, boosts the proportion of recoverable species from a lowly 1% to a respectable 50%.

Back in the lab, researchers were able to isolate teixobactin from the surrounding bacteria, and assign its 3-D chemical structure using NMR. They describe this molecule as “an unusual depsipeptide which contains enduracididine, methylphenylalanine, and four D-amino acids”. Such a bout of rebarbative terminology is rare even for biochemists, but it basically means that the molecule is made with modified amino acids, ones outside the usual 20 found in the genetic code, and that its backbone contains some ester bonds (oxygen) where a typical protein would have only amide bonds (nitrogen).

Though it little affects Gram-negative bacteria, teixobactin proves a potent foe against many heavy-hitting Gram-positive bacteria, such as Bacillus anthracis (anthrax), S. aureus (MRSA), and M. tuberculosis (TB), killing them at much smaller doses than vancomycin requires. It sabotages the same general machinery as vancomycin–by destabilizing precursors in the building of the cell wall, so that the bacteria burst–but does so in a different way. Teixobactin does not act on a protein or an amino acid at all, but rather to the fatty-sugary part of two molecules known as “Lipid II” and “Lipid III”, which are critical precursors both for building the bacterial cell wall, and for regulating powerful enzymes called “autolysins”. When the autolysins are cut free by teixobactin, the already weakened cell wall gets digested entirely.

Even more exciting than this potency, however, is the hope that this lipid-based mechanism may represent a whole new paradigm in antibiotic design, one which has a built-in resistance to, well, resistance. Nearly all antibiotics discovered prior to teixobactin bind to specific proteins, like penicillin with transpeptidase. Yet since proteins are made of amino acids that are directly encoded in a cell’s DNA, they are also easy to modify by direct mutations; evolution can readily jigger with the DNA until a new mutant protein arises that can shrug off the antibiotic. But since lipids are not coded in DNA and are also fairly hard to chemically modify, there is strong reason to believe that evolution may be stymied in its search for a resistance mechanism to teixobactin–perhaps even for good.

In order to test this hypothesis, the scientists did under controlled conditions exactly what you’re not supposed to do when taking an antibiotic: expose bacteria to just sub-lethal amounts of it for an extended time. For most known antibiotics, many natural resistance genes, such as the beta-lactamase gene, are always latent in a tiny subset of the population, or can be evolved quickly–so in the presence of a sub-lethal dose of antibiotic, the pressure of artificial selection takes over, and after a few generations the population is almost completely resistant. But while S. aureus exposed to sub-lethal doses of ofloxacin, a common fluoroquinolone antibiotic, both showed clear signs of resistance after a few days, the teixobactin lost none of its potency even after 27 days.

There has already been some success in producing derivatives of the new drug synthetically. Of course the wider hope is that once scientists further investigate this quiet 99% of the soil’s inhabitants, teixobactin will prove to be the tip of the iceberg–just the first entry in a successful re-boot of the heyday of the 1940s-1960s.

But if there’s one thing to remember whenever a new “wonder drug” or “promising lead compound” is trumpeted, it’s that they usually don’t make it very far–and if they do, the process to approval is long and tortuous, and dotted with unintended consequences. Much remains to be seen regarding teixobactin’s potency and its toxicity–although it has already been shown to cure mice of a 90% lethal dose of MRSA without apparently harming them. As for the notion that teixobactin may be the first totally “resistance resistant” antibiotic, there have been similar claims before. Vancomycin itself was once thought to be nearly immune to resistance, until strains turned up in which a single strategic modification that rendered the drug unable to bind to its target. (Scientists, continuing the arms race, soon found that changing a single atom in the antibiotic could restore its potency.) And while binding to a lipid instead of a protein is a clever trick for avoiding resistance, it still leaves open the emergence of efflux pumps that remove it, or an enzyme that cleaves it.

But even if teixobactin research is too much in its infancy to judge its promise, there are plenty of other approaches getting more attention as the struggle against resistance heats up. Bacteriocins, highly specific proteins made by bacteria themselves to kill off their competitors, are another hope some people have begun to swear by, though somewhat exotic and uncharted territory at this point. Being proteins, the majority of them are likely to be less stable, harder to make, and harder to deliver, somewhat like Fleming’s disappointment with lysozyme. Phage therapy, which makes use of ancient viral species that prey very specifically on bacteria, is another approach, and appealing in part because the risk of side-effects is very low, and the viruses can themselves evolve to counter resistance. And quorum quenching/anti-virulence agents, though little heard of, offer another approach, by disabling the bacterial production of dangerous toxins to make even resistant infections less deadly.

These new ideas and new molecules may yet turn the tide of resistance, and reinvigorate the human side in the battle against microbes. But they just as well may fall short. In the search for antibiotics, after all, we are perusing the tree of life–and in so doing, we are just as likely to end up coming face to face with assumptions we never acknowledged before, as we are to find the next blockbuster therapy.

In particular, we must face up to our assumption that the normal state of affairs is a never-ending ratchet of just-in-time breakthroughs, perfectly compensating for the threats built up by human short-sightedness and greed. The antibiotic age represents, to paraphrase Eugene Wigner on another subject, “a wonderful gift we neither understand nor deserve”, a gift of a type that science only very rarely affords us. If that age and that gift is to be preserved, we must learn to be good stewards of it, as of the earth itself.

The human race survived without antibiotics for most of its history–but life was often nasty, brutish, and short. What will life be like in a future world of ten or fifteen billion human beings, closely packed in cities, linked by lightning-fast transport, and all jockeying for the necessities of decent life, if and when the lucky discoveries peter out? How far will we manage to keep walking, however laboriously, in Fleming’s footsteps?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s