A while ago, I mentioned a serious problem in the attempt to combat drug resistance:

…I think the greatest challenge is medicinal chemistry. By way of prelude, in days of yore, before the FDA approved virtually everything it was sent, people used to grouse about how ‘the FDA wouldn’t approve aspirin.’ While that might have been true, the reality is that if you gave a medicinal chemist a list of twenty compounds slated for further development that included aspirin, aspirin would probably be at the bottom of the list. Aspirin, arguably one of the most commercial successful drugs, is an awful molecule. No medicinal chemist in her right mind would choose aspirin (I’m not going to go into details; either you’re a chemist/know enough chemistry to understand this, or I would have to provide a very, very long explanation that I would probably screw up).

I remember a policy-oriented meeting I attended that had the chiefs of drug development from the large drug companies. A very famous medicinal chemist was giving an overview of the principles used to choose ‘good’ molecules. At the end of his talk, one of the VPs asked a one word question, “Aspirin?” Said famous medicinal chemist had to acknowledge that these ‘principles’ would have ruled out aspirin based on every single criterion.

The problem is that, in terms of chemical diversity, we’re looking for our keys under the streetlight. The molecules from which drug development begins (sometimes referred to as ‘scaffolds’) are pretty limited. They often have the advantage that they’re easy to alter, so it’s possible to generate thousands of related compounds cheaply and quickly that can then be screened for antibiotic activity (or any other drug activity, such as killing cancerous cells).

We need new sources of compounds.

…these molecules are often less than ideal (‘bad’), so there will have to be more research into how to modify these compounds (which might explain some of the reluctance to pursue these compounds).

On the policy front, anything that can close this gap would be very useful. We also need to support the basic research that will make working with and altering these non-traditional compounds easier.

So this seems promising (boldface mine):

A 64-year-old class of antibiotics that has been a cornerstone of medical treatment has been dramatically refreshed by dogged chemists searching for ways to overcome antibiotic-resistant bacteria.

In work described today in Nature, a team of chemists built molecules similar to the drug erythromycin, a key member of the macrolide class, from scratch. In doing so, they were able to generate more than 300 variations on erythromycin that would not have been feasible by merely modifying the original drug — the way that scientists would normally search for new variants of existing antibiotics.

The process generated several variants on erythromycin that can kill bacteria that are resistant to the antibiotic. Although much testing remains before any of the molecules could be used in people, many of them show promise, says chemist Phil Baran of the Scripps Research Institute in La Jolla, California. He adds that the work holds potential for the future of antibiotics: “The fact that you can now make deeply modified analogues in a practical way with chemical synthesis opens the door to a whole host of derivatives that could never have been dreamt of before.”

This is an example of what needs to be done, though it remains to be seen if this particular approach will pan out. Hopefully, as chemists get better at doing this sort of thing, the venture capitalists will provide money, otherwise this probably won’t get very far.