I have an anniversary to celebrate this time of year: it’s now been thirty years since I started work in industrial drug discovery. Given the state of the industry over that time, just being able to say that at all has called for some luck and some flexibility along the way, but I’m very glad to still be at it.
What were the drug labs like in the fall of 1989? For equipment, I can think of two big differences: no column chromatography machines (prepacked columns, fraction detection), and no walkup LC/MS. There was nothing holding back the first of those, although if you’d shown me one during my graduate studies my mind would have reeled at such sybaritic luxury). But the lack of LC/MS is more understandable, considering that the electrospray technique had only been invented the year before by Fenn and co-workers. So yeah, what this means is that early on in my industrial career I still poured silica gel columns by hand and checked the fractions by TLC, but the clock was ticking on both of those tasks.
Another difference was that our NMRs didn’t even have sample handlers on them. Now, that was probably the company being cheap about it, because there was already one on the 300 Mhz machine in my graduate department by the time I left it. But we had the good old sign-up sheet, where you put your phone number down. And no, that was not a cell phone. Someone called you to tell you that they were finishing up their sample, so you went down and ran yours (you had a 15-minute time slot, in theory), and when you were done you checked the list and called the next lucky chemist. If I knew the next person well enough, I usually did that by saying “Get your worthless butt down here to the NMR, you hack” or something equally encouraging.
You got an automatic printout of the spectrum off an Apple LaserWriter at the end of your run; what you did not do was go back to your desk and work up the data on your computer. Because you didn’t have a computer. And that is the single biggest difference between the chemistry department of 1989 and the one of 2019, because at my first company the powers that be weren’t yet convinced that it was a good idea for people to have computers at all. The feeling apparently was that folks would sit at their desks and mess around with them rather than be out there in front of their fume hoods where they belonged – in fact, I heard that phrase almost word for word.
Nonetheless, within the next year or so, desktop computers began to make their appearance, at first shared between two or three people at a time. They were Mac IIci models; the head of our area was an Apple fan, and I was grateful for that, since it was Word and ChemDraw on a bunch of Mac SEs that had gotten me out of grad school a few years before. And that meant the introduction of e-mail, so I always have the option of appearing like a paleolithic hominid by telling people that I started working in the industry before we even used e-mail. We checked the darn mail slots instead – sheets of paper for memos, and those yellow envelopes with the little string closures for stuff that was more important or private. The string thingies saved wear and tear on our primitive flint hand axes, naturally (we used those to score the glass TLC plates instead). There was a manager who expressed a wish for pizza one day at a meeting, whereupon one of his reports walked down the street, bought him a slice, and then sent it to him in an interoffice mail envelope. But he was one of our more interesting specimens, admittedly, and the same dude who changed the network name, once we had a network, of the NMR room LaserWriter to “LaserWrither”, and no, it wasn’t me.
The whole lack-of-computers thing, and especially the lack-of-network thing, led to some situations. One of the biologists who’d been running assays for the first project I ever worked on retired in my first couple of years. It was only some time later that we caught up to the idea that he had (of course) been keeping up with all the assay data himself, on his own computer. With his own files, using his own FileMaker database, which now for some reason we could not find on his old machine, and there had been no particular provision for backing all this stuff up. So the call went out for everyone to search through their old hardcopy memos and meeting handouts to try to reconstruct as much of the information as we could. Even at the time I thought that this was clearly not the pinnacle of smooth data management.
The assays for the next project were cloned GPCR subtypes, and that was a pretty new thing, too. I came in just as the tissue-slice assays for those sorts of things were disappearing under the tide of molecular biology. Farewell possum trachea and rat vas deferens – some will recognize that this was a muscarinic project, and that neither of those examples are invented at all. The advent of pure cloned subtypes was greeted with some anticipation; we hoped that this would really settle things out, tell us key information about the new compounds, and make things move along faster and more surely. In reality, they did indeed help us to optimize against a particular subtype, to levels never seen before – but of course, the receptor assays did nothing to help out with most of the real problems with such a project. Such as the PK giving trouble as the compounds got decorated more and more baroquely to achieve that selectivity and potency (numbers that we really felt uncomfortable backtracking on). Or such as the fact that the whole project was aimed at Alzheimer’s in 1991 and was therefore doomed – that only became apparent later on. I also got to work on beta- and gamma-secretase before we even knew which enzymes those were, and you can guess how successfully that turned out.
Edit: of course, another big difference from that era was the way you searched the literature. I have waxed lyrical on this subject before, but don’t get the impression that I miss flipping through bound volumes of Chemical Abstracts, because yikes).
So that was the world of 1989 med-chem, and I hope I haven’t made it seem too distant. Because it wasn’t, in all the important ways: it was and is still a matter of skill and luck, of intuitive leaps and brute empirical experimentation, and the compounds still have to get made and get tested before you have any hope of knowing what’s going on. We use different equipment now, different instruments to play the music, but the song does indeed remain the same.