My post the other day on Phil Baran’s public-private funding ideas brought in a lot of comment. As usual, I encourage a look a the comments section if you’re interested in the topic, or any topic that comes up around here – a lot of people who know whereof they speak show up. But I wanted to follow up on a response that I got through email, from an experienced organic chemistry professor. He says that it’s gotten harder and harder for him to do any such partnerships with pharma, so I asked him for details. Here (with permission) is most of his email:
* In the past, pharmaceutical companies had more projects related to families of natural products (corticosteroids, betalactams, macrolides, prostaglandines) and the complex chemistry involved encouraged long term collaborations with academic groups. It seems now that companies want academics to just make new scaffolds or molecules for testing and this is less conducive to the discovery of new chemistry.
* Automation has limited the number of chemical reactions that are of use to medicinal chemists. Furthermore, the “designers” in the West are dumping more and more of the synthetic burden (e. g. synthesis of a new scaffold to feed the robots) on the shoulders of chemists in China and India. Interestingly, however, automation has at the same time created a demand for late stage modifications and spurred a revival of “C-H functionalisations” and redox radical chemistries building on the Minisci reaction. So perhaps more long term collaborations will arise on this front.
* I have personally found that problems from “process and development” were more focused and often pin-pointed to gaps in synthetic methodology (hence a source of interesting research problems). Medicinal chemists in “discovery” tend to design compounds they can make; if they can’t, they think of another structure, an option not available to process chemists. There is therefore less pressure to come up with new chemistry or to ask people in academia to solve a particular problem (exception: the late stage modification discussed in the previous bullet).
* I very much regret the disappearance of “Central Research” from large diversified companies such as Dupont. I do not understand these decisions. Funding such laboratories only consumes a small fraction of R&D budgets, but the long term benefits are enormous. Furthermore, chemists in such “laboratories” had a broader view of chemistry and would scout the whole of academic research programmes in search of ideas and inspiration. This resulted in tighter (and healthier) contacts between industrial and academic chemists and more long term collaborations. . .
* There is also the question of the diminishing importance of “small molecules” as drugs and whether this is diminishing the interest of pharmaceutical companies in collaborating with hard core organic chemists.
“Nostalgia isn’t what it used to be…” I suppose that the pendulum will swing back at some point. . .
I have a few comments on these, and I’m sure that readers here will have many more. The point about fewer complex natural product scaffolds is probably true, but at the same time, I think that there’s plenty of good organic chemistry to be done by trying to make useful and interesting new scaffolds
. Natural products are certainly the most complex things that synthetic organic chemistry usually takes on, and I appreciate that working on these structures can lead (in many cases, has to lead) to new chemistry. (It doesn’t always, of course, which gives you the worst of both worlds!)
One could draw similarly complicated systems all day, though, just out of the imagination. But the key thing about the complexity of natural products is that (1) you know that such a structure can exist and (2) you know that it has interesting features like biological activity. These are crucial points, because (dodecahedrane aside) no one’s going to put the sort of time, money, and effort needed to make a big complex structure unless that structure has those things going for it. No one’s going to give you money to make one, either. You can go in and modify a structure like vancomycin
if you have reason to believe that your new molecules are going to be useful, but no one’s going to be able to do that just for the heck of it, so natural products do have unique things going for them in organic synthesis.
At the same time, there are a lot of smaller and (apparently) simpler scaffolds that no one knows how to make
– whole classes of heterocycles, bicyclic and tricyclic compounds, small and medium-sized rings, that are almost totally unrepresented in medicinal chemistry. I think that biopharma should be funding people who have good ideas about how to generate such intermediates. The point about late-stage functionalization is a good one, too, and I think that biopharma should be putting up money for that kind of thing as well. There’s a lot of potential there, and it’s hard chemistry to realize.
But I’m not sure if automation itself has (so far) limited the number of reactions of use to medicinal chemists. It’s true that we run a whole lot of metal-catalyzed couplings and amide formations, but the machines didn’t force us into that – we did it to ourselves. Thus far, automation has made a much bigger mark in compound handling and compound purification, and much less of one in actual synthesis, although I think that’s going to be changing. (As an aside, I’m going to be speaking at the European Federation for Medicinal Chemistry meeting
in Manchester later this year, and I think my talk is going to be focusing on automation and AI in future med chem).
I can very much agree, though, with the regrets about the loss of Central Research departments. The comments to the Baran post went into some of the reasons why this has happened. Those reasons (tax law, near-monopoly positions of some companies, a focus on shorter-term returns to shareholders) are all comprehensible, but that doesn’t mean that they’re all desirable. The next time we mess around with corporate tax law, I wouldn’t mind seeing at attempt at making this sort of long-term research more attractive, although it would be tricky to set that up without encouraging mere window dressing. I remember while I worked in New Jersey in the early 1990s, the whole R&D site got renamed as a “Research Institute”, which as far as I could tell involved only changing the signs, business cards, and letterheads. The reason was apparently a New Jersey tax law that was more favorable to “research institutes”, so there things went: ya want it, ya got it. One of the big problems in messing around with the tax code is that you tend to get that sort of thing first, and any desirable effects second.
Now, to the question of the diminishing importance of small molecules – there’s a topic that’s come up here before, and in the macro sense, it’s probably true. Look at the number of antibodies and proteins that are in the list of top-selling drugs, and look at how money has poured over the years into various biological platforms (antisense, RNAi, CRISPR) that promise to tackle disease pathways that small molecules have had a hard time reaching. At the same time, the small molecule med-chem world is very much alive. I think it’s more that the pie has gotten bigger, rather than the small-molecule slice has been shrinking on its own. The effect of this on academic collaboration is another question, and a good one.
Finally, that pendulum. I’m less and less of a fan of the pendulum metaphor in predicting where scientific fields will go, because I think it presupposes too much. There’s no such law of motion for us: some fields just go away and don’t come back, just like some businesses. I first heard that the pendulum would swing back for organic synthesis in 1986, from my PhD advisor, so I’m not waiting for the return swing. On the other hand, if you toss the pendulum metaphor, then things aren’t so deterministic: we can, then, start to make the field as relevant as possible right now. You don’t get the (spurious?) hope that things necessarily have to come back, but you get, perhaps, a little more mental freedom to operate by not waiting and not, perhaps, thinking on some level that waiting is necessary. We shall see!