Intelligent Design, Science Literacy and the Case of Antibiotics

Science literacy is increasingly seen as important for people to be effective participants in a society where the pace of scientific discovery and translation to products and processes is increased. Effective engagement with GM crops, stem cell therapies, burgeoning healthcare costs and global warming require a reasonable degree of science literacy.

The need for science literacy is emphasised by a recent post at Uncommon Descent by commentator Gil Dodgen, who makes this remarkable statement:

In the meantime, medical doctors should prescribe multiple antibiotics for all infections, since this will decrease the likelihood that infectious agents can develop resistance through stochastic processes. Had the nature of the limits of Darwinian processes been understood at the outset, the medical community would not have replaced one antibiotic with another in a serial fashion, but would have prescribed them in parallel.

As I’ve said before in a different context, to paraphrase Mr. Babbage[1], I cannot apprehend the confusion of mind that would result in the above statement. One would have to be ignorant of 60 years of biomedical research and medical practice to say that. Both Humble Monkey and The Sandwalk have already commented on this post, and it’s got a big helping of Respectful Insolence, but I want to discuss this from a pharmacological perspective.

Combination therapies to combat resistance have been in use since the 1940’s to prevent the rapid emergence of streptomycin resistance during therapy for tuberculosis (streptomycin and p-aminosalicylic acid was the drug combination, Kucers, 1997)). The understanding of the need for antibiotic combinations was firmly grounded in evolutionary biology then as now (see also Allan et al 1985, Millatovic et al, 1987).

A brief look at Goodman and Gilman’s Manual of Pharmacology and Therapeutics shows combination therapy indicated to reduce the development of resistance (page 712); it’s also in the Medicines Australia Handbook. G&G is hardly an obscure publication and is one of the mainstays of medical education, so it is hard to be unaware of combination therapy. Combination therapies for TB and leprosy are standard practice. In the world of viruses, triple therapy for HIV, is the standard, and has resulted in significant long term survival and improved quality of life. With protozoa, dual therapies for malaria are recommended by the WHO. This is not an exhaustive list of micro-organisms where combination drugs are given to slow the rise of resistance, merely the better known ones. While Mr. Dodgen is not expected to have a medical therapy reference handy, even a brief search on the internet would acquaint him with the long history of the use of combination antibiotic therapies in medicine.

Now, while we may with justification roll our eyes at Mr. Dodgen’s ignorance, he does ask an interesting question. Given that we know (and have known for some time, based on our understanding of evolution), that combination therapies slow the appearance of resistance, why are not all anti-microbial therapies combination therapies? The reasons are manyfold, and not related to the medical research communities firm understanding of the rise of resistance (a recent Medical Journal of Australia article carrying the plaintive title “Why do we not yet have combination chemotherapy for chronic hepatitis B?” gives some good insights into this question)

One of the reasons is simple; to have a combination therapy, one must have multiple drugs to combine. In the case of HIV, before we could have triple therapy, we need to develop several drugs to deliver. In the early days of HIV, there was only one drug AZT, which bound to the enzyme reverse transcriptase. It was only after several years of intense research that we were able to provide several different drugs to combine.

Another problem is that the targets for these drugs must be different; a drug variant that binds to the same site on a target protein will be affected by the same or similar mutations that produced resistance in the first place. But in drug development, variants of a successful structure are easiest to make, so these will be first to be made. It took a while, but drugs were developed that could inhibit reverse transcriptase through other binding sites (the non-nucleoside analogs) and finally drugs were developed that could inhibit HIV protease, the viral enzyme required for processing viral proteins into their functional forms. Now triple therapy, a combination of two reverse transcriptase inhibitors (both AZT-like and non-nucleoside) and a HIV protease inhibitor, is standard. But it took a while to get there.

The principle applies to combination antibacterials. We have a wider range of cellular targets to hit in bacteria, compared to viruses (which hijack our own cellular machinery, making them difficult to target), but that range of targets is not immense. Many of the antibiotics we have are variants on the same basic structure, and while these have very important therapeutic properties, they are not good candidates for combination therapies. As I mentioned above, it is generally easier in drug development to modify an existing drug to have improved properties then to come up with a completely new drug that acts at a different target (but not always, drug design can have all sorts of surprises in store). Despite what people may think, it takes quite a bit of work to make a new drug, and drugs that may have wonderful properties in the test tube fail utterly in humans (one classic was at the turn of the century when a promising new group of antibiotics failed because they were insoluble in water). As well, given the costs of drug development, there is little incentive for drug companies to develop new antibiotics acting at unique targets. In The Pipeline has some interesting ruminations on drug development that may help you keep all this in perspective.

Once you have new drugs, all is not plain sailing. All drugs have some degree of toxicity, adding drugs together can increase the risk of toxicity, for example the combination of beta-lactam antibiotics and aminoglycosides produce renal toxicity which outweighs the benefits of reduced resistance development. As with this case, not all antibiotics are suitable for combination because of their toxicity. As well, and this may seem surprising to you, some antibiotics interfere with each other to produce less of an effect. For example, bacteriostatic antibiotics (those that slow the growth of bacteria, rather than kill them) interfere with bactericidal antibiotics; this limits our choice of drugs even more.

As you can see, setting up antibiotic combinations is not straight forward. To quote from the above Cochrane review:

The disadvantages of combination therapy may include additional costs, enhanced drug toxicity, the possible induction of resistance caused by the broader antibiotic spectrum (Manian 1996; Weinstein 1985), and possible antagonism between specific drug combinations (Moellering 1986).

So, despite GilDodgen’s claims, we do have combination antibiotic therapies. The failure to develop universal combination therapies for all bacterial infections is not due to blinkered Darwinism, but due to the limitations of the drugs that we humans develop. Limitations that are entirely understandable given the fallible information we have about organisms and their interaction with chemicals, and the finite supply of resources that we have available to develop drugs.

This brings me back to science literacy. I just have to turn 45 degrees to my left to pick up Goodman and Gilman (most physicians will have it or something like it in similar reach). This sort of resource is not available to most people, and I would be surprised if GilDodgen has Goodman and Gilman or Martindales within easy reach. But he probably has read reports in newspapers or seen items on TV about antibiotic resistance. It would be easy then to simply assume that medical science is just rolling out the same old sausage line drugs with little regard to actually thinking about resistance. However, a science literate prson would assume that newspaper reports are limited, and go looking for more in depth information. Google is the modern encyclopaedia, and typing in “combination therapy” antibiotic retrieves a number of hits. The very first discusses combination therapy to reduce the rise of resistance. Then there are others such as this one and the Cochrane review , which clearly shows that combination therapy is used, and it’s limitations. Even typing “combination therapy” antibiotic into Wikipedia will retrieve this reference , which clearly shows the use of combination therapy.

Clearly, scientific literacy is not just having a passing understanding of the facts of science, but also the ability to use the methods and resources of Science. Sadly, the denizens of Uncommon Descent do not have scientific literacy.

Now let’s look at some more of GilDodgen’s claims:

Here’s a prediction and a potential medical application from ID theory: Design a chemical or protein which would require a triple CCC to defeat its toxic effects on a bacterium, and it will exhaust the probabilistic resources of blind-watchmaker mechanisms to counteract the toxic effects.

A “CCC” is a chloroquine resistance cluster, which is not a cluster at all and requires considerably less than a population of 1020 organisms to evolve it (see Nick Matzke’s review of this for more detail). Designing such a drug would be difficult, even with the protein structures we have, designing de novo a drug that requires multiple mutations to displace it is a non-trivial task. Even here, evolutionary biology is required, in order to find which enzymes are critical targets, and which have mutated the least since remote common ancestors, making their structures highly intractable to mutation.

Such a success could and will only come from engineering and reverse-engineering efforts, not from Darwinian theory.

Sadly for GilDodgen and Uncommon descent, no. It will require a large input from evolution and evolutionary theory to design such drugs. We will need evolutionary theory to locate suitable targets first. Then drugs will be developed either from screening natural products, where evolution has crafted many antimicrobial compounds vai paths we aren’t clever enough to understand yet, or developed using drug synthesis approaches that mimic natural selection . Some complete de novo design is done, building drugs from first principles, but this has not been more successful than evolutionary based approaches.

In summary, we do have combinatorial antibiotic therapies, where they are possible and appropriate. New antibiotic therapies with reduced risk of development of resistance will come from the drug discovery process, but only by utilizing our understanding of and the principles of evolutionary biology. The commentators at Uncommon Descent could use some pointers in science literacy.


Allan JD,Moellering RC.Management of infections caused by gramnegative bacilli: the role of antimicrobial combinations. Reviews of Infectious Diseases 1985;7 Suppl 4:559–71.

Kucers A. Evolution of antituberculous therapy. In: Kucers A, Crowe S, Grayson M, Hoy J, editors. The use of antibiotics. 5th ed. Oxford: Butterworth-Heinemann, 1997: 1194.

Milatovic D, Braveny I. Development of resistance during antibiotic therapy. European Journal of Clinical Microbiology 1987;6(3):234– 44.

[1] Charles Babbage was a computer pioneer who had to deal with staggering incomprehension.

On two occasions I have been asked, – “Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?” In one case a member of the Upper, and in the other a member of the Lower, House put this question. I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question.