Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases.
To contact Derek email him directly: derekb.lowe@gmail.com
Twitter: Dereklowe
Noted chem-blogger Milkshake seems to have had a close call with a fire started by a tiny potassium hydride residue. It looks like he made it through without serious injury, but that sort of thing will definitely shake a person up.
I hate potassium hydride. Its relative sodium hydride is a common reagent, but it's much tamer (and even so, can cause interesting fires - I knew someone who ignited a heap of it on the pan of a balance while he was weighing it out, which slowed things down a bit). Sodium hydride is usually sold as a 60% dispersion, a dark grey powder soaked with mineral oil to keep it from deteriorating too quickly (and to keep it from setting everything on fire). You can buy 95% sodium hydride, the dry stuff, and there are people who swear by it, but I tend to sweat at it. You never know if it's been stored properly; you may be adding a slug of sodium hydroxide to your reaction without knowing it. And there's the fire part. You'll want to move briskly if you're using the 95%, and I'd pick a day when the humidity is low.
But potassium hydride, that's another beast entirely. It makes the sodium compound look like corn meal, in terms of how forgiving it is. You can't get away with the clumpy oily powder form at all - traditionally, KH is sold as a gooey dispersion of grey powder sitting under a few inches of mineral oil. If it's well dispersed, it's supposed to be 35%. You shake the stuff up until you think it's even mixed, then pipet out the amount of gunk that corresponded to the KH contained therein. Sure you do. What actually happens is that you pipet out the stuff, noticing while you do that it's already settling out inside the pipet, thereby to clog it up when you try to transfer it. No fun.
It's becoming available now dispersed in a block of wax, which is not such a bad idea at all. Wax isn't any harder to get out of your reaction than oil is, and you can carve off chunks and weigh them without so many what-am-I-doing moments. But Milkshake worries that this ease of use will lead to more fires during workups (which is where his reaction ran into trouble), and he may well be right. If you're going to use KH, don't let your guard down.
Someone has been soaking up the atmosphere at a large pharma company, for sure. "Look, I'm a chemist. I thought you hired me to do chemistry. But so far, all I've heard is gibberish. . .don't you do chemistry here?".
Some of you may enjoy that, but for others, it might just be a bit too realistic to be amusing. . .
The same user has several other videos on YouTube, such as this one, which (in addition to a few four-letter words), features the phrase "Get off the Kool-Aid!" Clearly someone needs to go through some more training. (Thanks to Pharmalot for the original link).
There's a new paper out in Nature Chemistry called "Quantifying the Chemical Beauty of Drugs". The authors are proposing a new "desirability score" for chemical structures in drug discovery, one that's an amalgam of physical and structural scores. To their credit, they didn't decide up front which of these things should be the miost important. Rather, they took eight properties over 770 well-known oral drugs, and set about figuring how much to weight each of them. (This was done, for the info-geeks among the crowd, by calculating the Shannon entropy for each possibility to maximize the information contained in the final model). Interestingly, this approach tended to give zero weight to the number of hydrogen-bond acceptors and to the polar surface area, which suggests that those two measurements are already subsumed in the other factors.
And that's all fine, but what does the result give us? Or, more accurately, what does it give us that we haven't had before? After all, there have been a number of such compound-rating schemes proposed before (and the authors, again to their credit, compare their new proposal with the others head-to-head). But I don't see any great advantage. The Lipinski "Rule of 5" is a pretty simple metric - too simple for many tastes - and what this gives you is a Rule of 5 with both categories smeared out towards each other to give some continuous overlap. (See the figure below, which is taken from the paper). That's certainly more in line with the real world, but in that real world, will people be willing to make decisions based on this method, or not?
The authors go for a bigger splash with the title of the paper, which refers to an experiment they tried. They had chemists across AstraZeneca's organization assess some 17,000 compounds (200 or so for each) with a "Yes/No" answer to "Would you undertake chemistry on this compound if it were a hit?" Only about 30% of the list got a "Yes" vote, and the reasons for rejecting the others were mostly "Too complex", followed closely by "Too simple". (That last one really makes me wonder - doesn't AZ have a big fragment-based drug design effort?) Note also that this sort of experiment has been done before.
Applying their model, the mean score for the "Yes" compounds was 0.67 (s.d.0.16), and the mean score for the "No" compounds was 0.49 (s.d. 0.23, which they say was statistically significant, although that must have been a close call. Overall, I wouldn't say that this test has an especially strong correlation with medicinal chemists' ideas of structural attractiveness, but then, I'm not so sure of the usefulness of those ideas to start with. I think that the two ends of the scale are hard to argue with, but there's a great mass of compounds in the middle that people decide that they like or don't like, without being able to back up those statements with much data. (I'm as guilty as anyone here).
The last part of the paper tries to extend the model from hit compounds to the targets that they bind to - a druggability assessment. The authors looked through the ChEMBL database, and ranked the various target by the scores of the ligands that are associated with them. They found that their mean ligand score for all the targets in there is 0.478. For the targets of approved drugs, it's 0.492, and for the orally active ones it's 0.539 - so there seems to be a trend, although if those differences reached statistical significance, it isn't stated in the paper.
So overall, I find nothing really wrong with this paper, but nothing spectacularly right with it, either. I'd be interested in hearing other calls on it as it gets out into the community. . .
As a follow-up to that post on open offices (and the others referenced in it), I've had a letter from a reader who wonders the following:
(1) How many recent research buildings have been built with open offices, as opposed to cubicles or actual office space? Is this the wave-of-the-future, or is it just a few high-profile examples getting attention?
(2) Does anyone know of any examples where a research department has tried an open-office plan and moved back from it after the experience?
Just to clarify, I don't mean large, relatively open lab spaces (those are pretty common, and often seem to work just fine). What's in question are the wide-open no-walls office and desk areas, with the extreme being the ones where no one has any actual assigned space at all. Thoughts?
SOLITUDE is out of fashion. Our companies, our schools and our culture are in thrall to an idea I call the New Groupthink, which holds that creativity and achievement come from an oddly gregarious place. Most of us now work in teams, in offices without walls, for managers who prize people skills above all. Lone geniuses are out. Collaboration is in.
But there’s a problem with this view. Research strongly suggests that people are more creative when they enjoy privacy and freedom from interruption. And the most spectacularly creative people in many fields are often introverted, according to studies by the psychologists Mihaly Csikszentmihalyi and Gregory Feist. They’re extroverted enough to exchange and advance ideas, but see themselves as independent and individualistic. They’re not joiners by nature.
Well, I wish that I could describe myself as "spectacularly creative", but the rest of that last sentence sounds pretty much like me, anyway. I have no problem talking with people when I meet them. I speak up at meetings, and I really enjoy giving talks to audiences. At the same time. I find that my best thinking is done very much alone. Once I've got something worked out in my head, I'm fine with roaming up and down the halls telling people about it and hearing the reaction. But that working-out has to be done in silence. The phone rings, and my thoughts all take off like like a flock of pigeons. Getting to settle back into their assigned places is not the work of a moment.
For all I know, the new book addresses this problem, but we really need a wider spectrum of words other than "introvert" and "extrovert". There are people who absolutely need human company, human noises and chatter around them. Others would rather have a bit of that, but feel it can be overdone, or just need it in defined amounts, like a meal. And some people don't mind much one way or another, while others are irritated or even panicked by it. You can sort people out, in similar fashion, by their responses to solitude and silence. Given that any research organization is going to have a variety of types in it, you'd think that there would need to be some places where the quiet types could hang out, just as there should be some where the gregarious ones can find what they need.
Let's see here. . .145 messages in the work e-mail queue, but most of them are automated reminders that reminded me of the same thing every day of the break. Now to the lab bench. . .now, that was a good idea, making sure that everything was labeled before leaving. As I've said here before, too many times you come back to a bunch of stuff that you were just sure that you're remember every detail of, and feel like a moron as you look at the label on the vial or flask: "Second batch". "Mostly clean". "Large run". Fascinating! Large run of what, exactly? I have, in years past, been reduced to running NMR and LC/MS on my own reactions just to try to figure out what they were, and that's not right.
Reagents that I'd ordered back before the break have come in, and I do recall why I ordered them, at least. You don't want to put in a request for anything sensitive in late December, though, not if it's going to sit out on your bench at RT for a week or ten days. I'm glad I'm not a cell-culture person or a rodent-raiser; my stuff doesn't need to be fed, washed, or watered, and I have the luxury of just walking away from it.
Big pile of junk on the desk, though, some of which never should have stopped there on its way to the recycling bin. I saved this for this morning, since I thought clearing things off would be a good way to jump-start my brain into work mode again. It's different now than it used to be, though - paper's more ephemeral. I have the PDFs of these papers stored, so the hard copy's just a convenience, and if I can't figure out what use it is by looking it it, into the bin it goes without a worry. In the days of paper files, I had to spend a bit more time wondering if I'd regret tossing something that was hard to obtain and actually useful. No more: if it looks useless or unrecognizable, into the blue bin it goes.
And then, for those of us in industry, the company starts waking up. Meeting invitations begin to arrive, to fill out the new year's calendar. Looking at your own schedule, you see the first repeating meetings from last year starting to show up, although some of these will get canceled because there's nothing to talk about yet. People who wanted something from you back in December will start to remember what it was, at about the same time that you remember the people that you wanted something from.
Time, shortly, for the first reaction, the first LC/MS trace, the first NMR, the first lab assay result of the new year. And for some of us, the first blog post. Welcome back!
For those of you keeping count of how many elements you've used in your chemical careers, you now have another possibility. This paper suggests that uranyl anions are good for epoxide polymerization, so who knows, they may be good for something else as well. I don't anticipate adding this one to my life list, but there's at least a chance of it now. . .
Here's an interesting exercise carried out in the medicinal chemistry departments at J&J. The computational folks took all the molecules in the company's files, and then all the commercially available ones (over five million compounds), minus natural products, which were saved for another effort, and minus the obviously-nondruglike stuff (multiple nitro groups, solid hydrocarbons with no functionality, acid chlorides, etc.) They then clustered things down into (merely!) about 20,000 similarity clusters, and asked the chemists to rate them with up, down, or neutral votes.
What they found was that the opinions of the med-chem staff seemed to match known drug-like properties very closely. Molecular weights in the 300 to 400 range were most favorably received, while the likelihood of a downvote increased below 250 or above 425 or so. Similar trends held for rotatable bonds, hydrogen bond donors and acceptors, clogP, and other classic physical property descriptors. Even the ones that are hard to eyeball, like polar surface area, fell into line.
It's worth asking if that's a good thing, a bad thing, or nothing surprising at all. The authors themselves waffle a bit on that point:
The results of our experiment are fully consistent with prior literature on what confers drug- or lead-like characteristics to a chemical substance. Whether the strategy will yield the desired results in the long term with respect to quality, novelty, and number of hits/leads remains to be seen. It is also unclear whether this strategy can lead to sufficient differentiation from a competitive stand-point. In the meantime, the only undeniable benefits we can point to is that we harnessed our chemists’ opinions to select lead-like molecules that are totally within reasonable property ranges, that fill diversity holes, and that have been purchased for screening, and that we did so in a way that promoted greater transparency, greater awareness, greater collaboration, and a renewed sense of involvement and engagement of our employees.
I'll certainly give them the diversity-of-the-screening-deck point. But I'm not so sure about that renewed sense of involvement stuff. Apparently 145 chemists participated in total (this effort was open to everyone), but no mention is made of what fraction of the total staff that might be. People were advised to try to vote on at least 2,000 clusters (!), but fewer than half the participants even made it that far. Ten people made it halfway through the lot, and 6 lunatics actually voted on every single one of the 22,015 clusters, which makes me think that they had way too much time on their hands and/or have interesting and unusual personality features. A colleague's reaction to that figure was "Wow, they'll have to track those people down", to which my uncharitable reply was "Yeah, with a net".
So while this paper is interesting to read, I can't say that I would have been all that happy participating in it (although I've certainly had smaller-scale experiences of this type). And I'd like to know what the authors thought when they finally assembled all the votes and realized that they'd recapitulated a set of filters that they could have run in a few seconds, since they're surely already built into their software. And we all should reflect on how thoroughly we seem to have incorporated Lipinski's rules into our own software, between our ears. On balance, it's probably a good thing, but it's not without a price.
Talking about hydrogenation here the other day brought up another thought: there's a point where lab work becomes quite difficult, and there are not a lot of good options to help with that. I'm talking about scale-up work, the grey zone between benchtop synthesis and production.
The first of those is where I spend my time. I've often said that there are really only two yields to be calculated in medicinal chemistry: enough and not enough. For some cases, "enough" can be two milligrams, if all you need is one assay (although you're not adding to the screening collection that way). Twenty is plenty for a new compound; it'll be in stock for years at the usual rates of screening.
But later on, when you start to get interested in a particular molecule, those numbers inflate quickly. In vivo tests, PK, toxicology - all these can start to chew up much larger amounts of compound. Hundreds of mgs, then grams, tens of grams, hundreds of grams to get through preclinical because it turns out that you need another large-animal run - those of you in the labs will be familiar with the progression. And well downstream of people like me, there's pilot plant work and real commercial production, where things are measured in kilos and up.
Those folks have a tricky job, but they have one advantage: they know what compound they're making, and eventually they've settled on a route that works. (If you can't do that, well, you don't have a drug). And for a real money-making compound, it's worth investing in dedicated equipment, whole dedicated facilities if need be, just to make it correctly. Earlier in the process, though, you have to be ready for all kinds of chemistries to make all kinds of compounds.
The medium-scale is where those two worlds collide. A medicinal chemist can generally make things up to the tens of grams, maybe a hundred or two, using roughly the same techniques that are used on the smaller scale: round-bottom flasks, suction filtration, magnetic stirring, standard-sized rotary evaporators, silica gel columns. Everything gets bigger and more unwieldy, and it always takes more time (and more solvent) than you thought, but it can get done. Some of the more exotic small-scale chemistries do start to break down on you, which also also adds to the time needed when you have to come up with alternatives.
But if you're always having to work on roughly the hundred-gram scale, you're straddling two regimes. The size of the glassware gets hard to manage - things are heavy, they tip over, they crack - and you really have to have more serious capacity for things like solvent evaporation. But this is way too small for industrial-plant equipment, the kind of thing where you design the process to start on the third floor to take advantage of gravity as you pump the contents of the big batch reactor downstairs for the next step. And it's getting too big for scaled-up versions of standard equipment, but at the same time, you need the versatility that general-purpose labware provides.
Some kinds of gear help to bridge this gap - overhead mechanical stirrers, outboard circulating chillers, and large-capacity rota-vaps come to mind. But there are many other cases where something that's neither benchtop nor pilot plant is needed, and doesn't necessarily exist. Hydrogenation is a case in point. It's done on an industrial scale; that's where all that partially hydrogenated vegetable oil comes from, for one thing. And hydrogenation is common on the gram scale or below. But hydrogenating three hundred grams of something can be a real pain in many labs. The common solution is roll-your-eyes-and-split-it-into-batches, but that gets old fast. . .
We organic chemists have always liked the hydrogenation reaction. Take your compound up in a solvent, add a pinch of black catalyst powder, and put some hydrogen gas into the vessel. Come back a few hours later, filter off the catalyst, and there's your cleanly reduced compound, ready for the next step, often looking even better than it did before you ran the reaction.
For many decades, the standard ways to run these reactions have been to either take a balloon of hydrogen gas and attach it to the top of your round-bottom flask (as in this video clip), or run it on a "Parr shaker". That last piece of equipment has been with us, essentially unchanged, since the 1920s. It's simplicity itself: a thick-walled glass bottle for your reaction, a tube and stopper running into it (with a framework to hold it down under pressure), a hydrogen reservoir, and a motor to shake the bottle around. Its relentless dackadackadackadacka noise is one of the standard sounds of organic chemistry. These things are always off in separate hydrogenation rooms, and when you have several of them running in there at once the out-of-phase clatter makes sequential thought almost impossible. I wish that there were an audio file I could link to, but working organic chemists will all know the tune.
There are newer ways to run the reaction, and flow chemistry is the obvious choice. The "H-Cube" was an early entry into this space, and many of them are to be found around the chemistry world. Unfortunately, many of them are also found gathering dust. Uptake of the machine has been uneven, despite some obvious advantages. That's because the first-generation machine has some obvious disadvantages, too: you have to change the catalyst cartridge every time you want to try something different, because there's only one at a time. The cartridges themselves are not too large, so if your reaction isn't efficient enough, you can have a problem with not being able to run everything in one-time-through mode. And there's no liquid handling - you have to load your sample and collect it in whatever means you see fit. Various people have modified the machine over the years to get around these limitations, and the company now sells a machine incorporating many of these ideas. And there are competitors out there as well.
So here's my question for the chemical audience: has anyone had enough nerve to ditch the Parr shakers completely? I've heard of places that have done it, but when you inquire closely, you often find that there are still a couple around that do a disproportionate share of the hydrogenations. Are there any flow solutions that work well enough to get away with this? You'd think that there would be advantages to a walk-up instrument, if it were robust enough - put your starting solution in position A-3 on the rack, tell it what pressure and temperature you want, which catalyst to use, and add your run to the queue. Come back after lunch and there it is, eluted into another container, ready for you to pick up. NMR machines work this way, and so do microwave reactors. But do hydrogenators? Today, in the real world? Experiences with such things welcome in the comments. . .
I wanted to send people to this 50-year retrospective in J. Med. Chem.. It's one of those looks through the literature, trying to see what kinds of compounds have actually been produced by medicinal chemists. The proxy for that set is all the compounds that have appeared in J. Med. Chem. during that time, all 415, 284 of them.
The idea is to survey the field from a longer perspective than some of the other papers in this vein, and from a wider perspective than the papers that have looked at marketed drugs or structures reported as being in the clinic. I'm reproducing the plot for the molecular weights of the compounds, since it's an important measure and representative of one of the trends that shows up. The prominent line is the plot of mean values, and a blue square shows that the mean for that period was statistically different than the 5-year period before it (it's red if it wasn't). The lower dashed line is the median. The dotted line, however, is the mean for actual launched drugs in each period with a grey band for the 95% confidence interval around it.
As a whole, the mean molecular weight of a J. Med. Chem. has gone up by 25% over the 50-year period, with the steeped increase coming in 1990-1994. "Why, that was the golden age of combichem", some of you might be saying, and so it was. Since that period, though, molecular weights have just increased a small amount, and may now be leveling off. Several other measures show similar trends.
Some interesting variations show up: calculated logP, for example, was just sort of bouncing around until 1985 or so. Then from 1990 on, it started a steep increase, and it's hard to tell if that's leveling off or not even now. At any rate, the clogP of the literature compounds has been higher than that of the launched drugs since the mid-1980s. Another point of interest is the fraction of the molecules with tetrahedral carbons. What you find is that "flatness" in the literature compounds held steady until the early 1990s (by which point it was already disconnected from the launched drugs), but since then it's gotten even worse (and further away from the set of actual drugs). This, as the authors speculate, is surely due to metal-catalyzed couplings taking over the world - you can see the effect right in front of you, and so far, the end is not in sight.
Those two measures are the ones moving the most outside the range of marketed drugs. And despite my shot at early combichem molecules, it's also clear that publication delays mean that some of these things were already happening even before that technique became fashionable (although it certainly revved up the trends). Actually, if you want to know When It Changed in medicinal chemistry, you have to go earlier:
It is worth noting that these trends seemed to accelerate in the mid-1980s, indicating that some change took place in the early 1980s. The most likely explanations for an upward change in the early 1980s (before the age of combinatorial chemistry or high-throughput screening) seem to be advances in molecular biology, i.e., understanding of receptor subtypes leading to concerns about specificity; target-focused drug design and its corresponding one-property-at-a-time optimization paradigm (possibly exacerbated by structural biology); and improvements in technologies which enabled the synthesis and characterization of more complex molecules.
Target-based drug design, again. I'm really starting to wonder about this whole era. And if you'd told me back in, say, 1991 about these doubts that I'd be having, I'd have been completely dumbfounded. But boy, do I ever have them now. . .
Well, actually, this might not be an anomaly. Medicinal chemists will have heard of the "magic methyl" effect, where small changes can make a big difference in affinity for a drug candidate. This morning I heard an interesting talk by Phil Sanderson of Merck on allosteric Akt inhibitors for cancer. I won't go into all the kinase-ness, although it was definitely worth hearing about. What caught my eye was something he mentioned at the end of the talk. The first compound below was an early screening hit in their work, something that had been in Merck's files since the early 1970s. After a huge amount of work over many years, which you can follow though the literature if you like with a search for "allosteric" and "Akt", they found that four-membered rings were very useful in the structures. Going back to the original structure and adding that same modification to it improved its potency by roughly 100-fold.
One methylene group! You wonder what might have happened if they'd done that early in the project, but as Sanderson correctly noted, no one would have done that (it's synthetically tricky; no one would have put in the time). And they don't have any structural information that seems to explain this effect, he says. So if you're looking for an illustration of what makes medicinal chemistry the wild ride it is, you've got an excellent one here.
I've been in the lab all afternoon setting up reactions, and that prompts me to write about something that I've been noticing. Is it just me, or does Aldrich seem to be abandoning the practice of putting any useful information on their labels?
This has been creeping up for a while, but I worry that instead of an anomaly it's the way of the future. I just got in several bottles of reagents from Aldrich, and basically all they have on their labels are the names of the compounds. No molecular weight, no density, no melting or boiling point: nothing but a line of type surrounded by an Aldrich label. And while I can go look these things up, and while my electronic notebook often is able to provide the information, it would still be a lot more convenient to have it on the label as well. You know, like it used to be.
I assume that this is a cost savings. As a rule, I assume that the most likely answer to any question that starts out "I wonder how come they. . ." is "money". But it's a shame.
A conversation the other day got me to thinking: over the course of my career, I've worked in the following therapeutic areas (more or less in chronological order): CNS (dementia, then Alzheimer's), diabetes, osteoporosis, obesity, oncology, anti-bacterials, multiple sclerosis, and antivirals. That covers a fair amount of ground, but there are still areas I've never really touched on - not much that would qualify as cardiovascular and not much inflammation, for example. So I'm sure that there are readers out there who have seen more drug discovery territory than I have - anyone who thinks that they have the local record, feel free to leave details in the comments.
A second question is whether there are therapeutic areas that you'd always wanted to try but never have. (Anti-infectives would have been in that category for me until the last few years). The opposite of that is well worth asking, too: are there disease areas that you regret ever having touched on? For my part, I learned a lot doing my Alzheimer's work, but in retrospect, much of it was a ferocious waste of effort, considering the results, so I'd probably put that one at the top. Other candidates?
I was talking with some colleagues about underused synthetic chemistry technologies the other day, and one that came up was high pressure. Here's a new paper from JACS looking at pressure effects on a common reaction (Michael addition), and there are quite a few others like it scattered around the literature. In general, reactions that have a lot of steric congestion, or whose transition state occupies less volume than the starting complex, will show some effects as you go to higher pressure.
But no one ever does it. Well, not quite "no one", but pretty damned few people do. I think the problem is that you need special equipment, for the most part, and you also need to have the idea of using high pressure. Both of those are in short supply. But I wonder if someone were to make a lab-friendly high pressure reactor, if it might get taken up a bit more. (Note to equipment manufacturers: I am not promising to buy the thing if you make it. But it's a thought).
A colleague of mine is running a Diels-Alder reaction this morning, and turned out to have never run one before, despite many years of experience in chemistry. (I'd bet, though, that a fair number of chemists who have run the reaction did it in an undergraduate lab and never have since). I've run them - although it's been a while - and I've done the Claisen rearrangement (ditto), the Knoevenagel condensation, the Barbier reaction, and the Henry reaction. I've done plenty of Horner-Emmons-Wadsworth reactions (although not in the last few years), Jones oxidation, Birch reduction, the Arbuzov reaction, and a Chichibabin pyridine synthesis, many years ago. And I've done a Cannizzaro, the Gabriel synthesis, Ferrier rearrangements, the Shapiro reaction, Peterson olefination, and Lindlar reduction. I've run Sandmeyer reactions, the Prins, Staudinger reduction, Ullmann coupling, and Weinreb ketone synthesis. I've done the Wolff–Kishner reduction (once) and Wurtz coupling (once), a Dakin-West (once), a Darzens (once), and a Delepine reaction (once).
But I've never done a straight aldol condensation, at least, not on purpose. And I've never, as far as I can recall, actually done a Fischer indole synthesis, or the lovely Skraup reaction. I've never run a Bayliss-Hillman, a Ritter reaction, a Cope rearrangement, a Julia olefination, a Pictet-Spengler, a Nazarov cyclization, nor a pinacol, and I don't think I've ever set up an ene reaction.
So what's on your list? What's the most famous reaction you've never run? Is there some reaction you've always sort of wanted to do, but never had the reason?
Interesting post from Milkshake over at Org Prep Daily on solvents that don't get used as much as they might in synthetic chemistry. Among them: trifluoroethanol, methyl t-butyl ether, and 1-methoxy-2-propanol. Definitely worth a look for those of us who are trying to get things to work at the bench - other nominations welcomed in the comments.
And if you're looking for someone to do that, I believe that Milkshake himself is still looking for a position (unpaid advertisement!)
OK, here's how I understand the way that medicinal chemistry now works at Pfizer. This system has been coming on for quite a while now, and I don't know if it's been fully rolled out in every therapeutic area yet, but this seems to be The Future According to Groton:
Most compounds, and most actual chemistry bench work, is apparently going to be done at WuXi (or perhaps other contract houses?) Back here in the US, there will be a small group of experienced medicinal chemists at the bench, who will presumably be doing the stuff that can't be easily shipped out (time-critical, difficult chemistry, perhaps even IP-critical stuff, one wonders?) But these people are not, as far as I can tell, supposed to have ideas of their own.
No, ideas are for the Drug Designers, which is where the rest of Pfizer's remaining medicinal chemistry head count are to be found. These are the people who keep trac of the SAR, decided what needs to be made next, and tell the folks in China to make it. It's presumably their call, what to send away for and what to do in-house, but one gets the sense that they're strongly encouraged to ship as much stuff out as possible. Cheaper that way, right? And it's not like there's a whole lot of stateside capacity, anyway, at this point.
What if someone working in the lab has (against all odds) their own thoughts about where the chemistry should go next? I presume that they're going to have to go and consult a Drug Designer, thereby to get the official laying-on of hands. That process will probably work smoothly in some cases, but not so smoothly in others, depending on the personalities involved.
So we have one group of chemists that are supposed to be all hands and no head, and one group that's supposed to be all head and no hands. And although that seems to me to be carrying specialization one crucial step too far, well, it apparently doesn't seem that way to Pfizer's management, and they're putting a lot of money down on their convictions.
And what about the whole WuXi/China angle? The bench chemists there are certainly used to keeping their heads down and taking orders, for better or worse, so that won't be any different. But running entire projects outsourced can be a tricky business. You can end up in a situation where you feel as if you're in a car that only allows you to move the steering wheel every twenty minutes or so. Ah, a package has arrived, a big bunch of analogs that aren't so relevant any more, but what the heck. And that last order has to be modified, and fast, because we just got the assay numbers back, and the PK of the para substituted series now looks like it's not reproducing. And we're not sure if that nitrogen at the other end really needs to be modified any more at this point, but that's the chemistry that works, and we need to keep people busy over there, so another series of reductive aminations it is. . .
That's how I'm picturing it, anyway. It doesn't seem like a particularly attractive (or particularly efficient) picture to me, but it will at least appear to spend less money. What comes out the other end, though, we won't know for a few years. And who knows, someone may have changed their mind by then, anyway. . .
Here's a general question for all you lab types, prompted by some rearranging that I've been doing over at my bench: what piece of equipment do you get the least use out of for the space it takes up? Those dusty items that haven't been touched in a couple of years are obvious candidates, but feel free to add some instruments that work, but crowd out other useful items. . .
I had to use some potassium permanganate a little while back - first time in years I'd had any of it out in the lab, and I was reminded of just what a spectacular purple color the stuff has.
There's some of it dissolving in water, via Flickr, and it's hard to beat for sheer purplelosity. But the solid doesn't look as impressive; it's quite dark (which is probably how it makes such an intense color on dissolution). So what's the best purple solid in the lab?
I have to promote my personal favorite, chromium (III) chloride (image courtesy of the Wikipedia entry).
That's a pretty good shot, but it really should be experienced in person. The stuff is metallic purple flakes, weirdly reflective - it looks like it should be the color of a custom racer's hood, rather than anything you'd actually order from a chemical supply house. Now all I have to do is find a use for it in the lab. . .
Chemists who don't (or don't yet) work in drug discovery often wonder just what sort of chemistry we do over here. There are a lot of jokes about methyl-ethyl-butyl-futile, which have a bit of an edge to them for people just coming out of a big-deal total synthesis group in academia. They wonder if they're really setting themselves up for a yawn-inducing lab career of Suzuki couplings and amide formation, gradually becoming leery of anything that takes more than three steps to make.
Well, now there's some hard data on that topic. The authors took the combined publication output from their company, Pfizer, and GSK, as published in the Journal of Medicinal Chemistry, Bioorganic Med Chem Letters and Bioorganic and Medicinal Chemistry, starting in 2008. And they analyzed this set for what kinds of reactions were used, how long the synthetic routes were, and what kinds of compounds were produced. Their motivation?
. . .discussions with other chemists have revealed that many of our drug discovery colleagues outside the synthetic community perceive our syntheses to consist of typically six steps, predominantly composed of amine deprotections to facilitate amide formation reactions and Suzuki couplings to produce biaryl derivatives. These “typical” syntheses invariably result in large, flat, achiral derivatives destined for screening cascades. We believed these statements to be misconceptions, or at the very least exaggerations, but noted there was little if any hard evidence in the literature to support our case.
Six steps? You must really want those compounds, eh? At any rate, their data set ended up with about 7300 reactions and about 3600 compounds. And some clear trends showed up. For example, nearly half the reactions involved forming carbon-heteroatom bonds, with half of those (22% of the total) being acylations. mostly amide formation. But only about one tenth of the reactions were C-C bond-forming steps (40% of those were Suzuki-style couplings and 18% were Sonogoshira reactions). One-fifth were protecting group manipulations (almost entirely on COOH and amine groups), and eight per cent were heterocycle formation, and everything else was well down into the single digits.
There are some interesting trends in those other reactions, though. Reduction reactions are much more common than oxidations - the frequency of nitro-to-amine reductions is one factor behind that, followed by other groups down to amines (few of these are typically run in the other direction). Among those oxidations, alcohol-to-aldehyde is the favorite. Outside of changes in reduction state, alcohol-to-halide is the single most favorite functional group transformation, followed by acid to acid chloride, both of which make sense from their reactivity in later steps.
Overall, the single biggest reaction is. . .N-acylation to an amide. So that part of the stereotype is true. At the bottom of the list, with only one reaction apiece, were N-alkylation of an aniline, benzylic/allylic oxidation, and alkene oxidation. Sulfonation, nitration, and the Heck reaction were just barely represented as well.
Analyzing the compounds instead of the reactions, they found that 99% of the compounds contained at least one aromatic ring (with almost 40% showing an aryl-aryl linkage) and over half have an amide, which totals aren't going to do much to dispel the stereotypes, either. The most popular heteroaromatic ring is pyridine, followed by pyrimidine and then the most popular of the five-membered ones, pyrazole. 43% have an aliphatic amine, which I can well believe (in fact, I'm surprised that it's not even higher). Most of those are tertiary amines, and the most-represented of those are pyrrolidines, followed closely by piperazines.
In other functionality, about a third of the compounds have at least one fluorine atom in them, and 30% have an aryl chloride. In contrast to the amides, there are only about 10% of the compounds with sulfonamides. 35% have an aryl ether (mostly methoxy), 10% have an aliphatic alcohol (versus only 5% with a phenol). The least-represented functional groups (of the ones that show up at all!) are carbonate, sulfoxide, alkyl chloride, and aryl nitro, followed by amidines and thiols. There's not a single alkyl bromide or aliphatic nitro in the bunch.
The last part of the paper looks at synthetic complexity. About 3000 of the compounds were part of traceable synthetic schemes, and most of these were 3 and 4 steps long. (The distribution has a pretty long tail, though, going out past 10 steps). Molecular weights tend to peak at between 350 and 550, and clogP peaks at around 3.5 to 5. These all sound pretty plausible to me.
Now that we've got a reasonable med-chem snapshot, though, what does it tell us? I'm going to use a whole different post to go into that, but I think that my take-away was that, for the most part, we have a pretty accurate mental picture of the sorts of compounds we make. But is that a good picture, or not?
Well, I've been traveling this week, but have found a bit of time to blog. Today we have something new about SAR, courtesy of the BASF marketing department.
Medicinal chemists are all familiar with the "magic methyl group" effect - the phenomenon of a single methyl sending a compound over the top in terms of activity, selectivity, PK, or what have you. I've seen it several times myself. Usually you starting wondering why you didn't just put the thing in six months before, but that's rarely the way things work out.
Well, we're not the only people who notice such things. Check out this ag-chem ad from BASF for their Kixor herbicide (sent along by an alert reader). Scroll down to the bottom of the page, and you'll find:
Methyl groups serve as "metabolic handles" in crops delivering crop safety and giving you confidence knowing you have made the right choice"
There you have it! Methyl groups add confidence! Now that's worth knowing - and you have to wonder what secrets other functional groups hold. Do carboxylic acids put a spring in your step? Do para-fluoros freshen the breath? Will a sulfonamide help you make the big sale? Someone in the advertising department might believe it - as Dilbert put it, marketing people even believe marketing surveys, so what's the limit?
Well, this post needs updating. In it I mentioned never running a Prins reaction again since the 1980s, nor any photochemistry, and today what do I find myself doing? Both of them, although not at the same time.
I am, fortunately, not running the Prins this way. But even bringing it up at all recalls to me a key part of my education. When I first joined my graduate school research group, I was put to making some tetrahydropyran systems. I was handed a synthesis, drawn up before my arrival, of how to make the first one, and like most first-year grad students, I gamely dug and and started to work on it.
I should have devoted a bit more thought to it. I won't go into the details, but it was a steppy route that relied, in the final ring-closure step, on getting the cyclic ether to form where one of the partners was a neopentyl center. The organic chemists in the audience will immediately be able to guess just how well that went.
So I beat on it and whacked at it, getting nowhere as I used up my starting material, until I was finally driven to the library. In the spring of 1984, that was a different exercise than it is now, involving the 5-year Chemical Abstracts indices and an awful lot of page flipping. (I haven't so much as touched a bound volume of CA in I don't know how many years now). If you were a nomenclature whiz, you could try looking up your compound, or something like it, in the name index, but a higher-percentage move was often to look up the empirical formula. That gave you a better shot, because (if it was there at all) you could see how CA named your system and work from there.
To my great surprise, the second set of collective indices I checked (the good ol' 9th), yielded a direct hit on an empirical formula, and the name looked like exactly what I had been trying to make. The reference was in Tetrahedron, which we most certainly had on the shelf, and I zipped over to see if there was any detail on how to make the stuff.
There was indeed. A one-stepper Prins cyclization gave just the ring system I'd been trying to make, and that was one step from the intermediate I needed. I just stared at the page, though. I honestly couldn't believe that this was real (as I mentioned, I was in about my second month of grad school lab work). Surely the synthesis I'd been given was the way to make this stuff? Surely the people responsible for it had checked the literature before drawing it up? (After all, it had only taken my a few minutes to find the stuff myself). Surely I couldn't just make the ring in one afternoon using two starting materials I could buy cheaply from Aldrich?
Well, surely I could. And that's just what I did, and got my project moving along until the next interesting difficulty came up a couple of months later. But I still recall standing there in the Duke chemistry library, looking at that journal article "with a wild surmise" that perhaps I should check things out for myself next time instead of just taking everyone else's word. It took a couple more lessons for me to really grasp that principle (Nullius in verba!, but it's helped me out a great deal over the years. I have the 27-year-old photocopy I made that afternoon in front of me now. It's a good reminder.
Here's a question for all the organic chemists out there. A discussion with some colleagues the other day got me to thinking about the reactions that we all tend to underuse. The category I offered up was gaseous reagents. Outside of hydrogenation, I think that many of us sort of go "Ehh. . ." when we come across transformations that need lecture bottles, cylinder, regulators, and so on.
Add to that the unpleasant nature of many of the gases themselves, and it's easier to find something else to do. But there are a lot of good reactions and reagents in this category - metal-catalyzed CO insertions, reactions with ammonia, acetylene, sulfur dioxide, etc. There's just a bit of a higher activation barrier to getting around to running them.
I'd say that photochemistry and electrochemistry are in this "rather do something else" category as well. Other nominations welcome!
For the chemists out there in the crowd: have you been looking for a paper to read that's filled, beginning to end, with good, solid, old-fashioned medicinal chemistry? Look no further than this one, on recent reports of isosteres. This sort of thing is still the heat of med-chem as it's practiced in the real world - messing around with the structure of an active molecule to see what you can improve and what you can get away with.
If you're not a medicinal chemist, the idea of a bioisostere is some chemical group that can substitute for another one. Classic examples are things like swapping in a tetrazole ring for a carboxylic acid or an oxadiazole for an ester. Here are some examples - even if your organic chemistry is shaky, you can see the similarities across these structures. If it works, you can change the other properties of your molecule (solubility, stability, selectivity) for the better while still keeping the key features that made the original group valuable for activity. It's not something that just automatically comes through every time - sometimes there just is no substitute - but it works enough of the time to be one of the essential techniques.
Management fads are truly a bad sign. I don't think that there's anyone out there in the working world who doesn't realize this, on some level, but it's worth keeping in mind. When some higher-up at your company decides "You know, we'd make a huge leap in productivity if we just did everything totally differently than we've ever done them before - I read this great article!", then you really need to hunker down until the fit passes.
Well, some of the folks at GlaxoSmithKline down in Research Triangle are probably looking for somewhere to hide. Because according to this article, the company is (yes!) at the forefront of a movement that's (yes!) sweeping the nation: open office space. No assigned desks, no permanent locations, just everyone floating around in a cloud of happy productivity. Jim Edwards at Bnet is right when he calls this "slightly insane".
Um. . .haven't we been hearing about this wonderful innovation for years now? And haven't several companies tried it and abandoned it, because (strangely enough) their employees didn't like the idea of putting their possessions into lockers every morning, wandering (or scrambling) around for desk space, and being unable to leave the slightest sign of anything personal around their work area? Here are some tempting details:
All employees are assigned a storage unit where they can keep files, a keyboard, a power pack and a mouse. There will also be group storage spaces where files that need to be accessed by more than one person can be kept. Any files that are not accessed regularly will be stored off-site. GSK's document retention policy isn't changing; it just may end up being followed more closely.
Gosh, that does sound like what I've been yearning for all these years. Making the transition to this wonderful environment isn't easy, though:
The larger move will ultimately include an extensive education campaign to prepare employees for their new surroundings.
Employees will work in neighborhoods, each of which includes meeting rooms and quiet areas. They'll attend etiquette workshops, and each neighborhood will adopt a set of policies to deal with hypothetical situations that may come up.
The groups that are moving to the new layout are those whose managers embraced the change. (Admin Shelby) Bryant now sits at a desk directly across from her boss, David Bishop, GSK's director of site operations in RTP.
Bishop said as the move gets closer, more and more departments are expressing interest in unchaining themselves from their desk.
"I don't believe we will ever get to where everybody wants it," he said.
Maybe not! But that'll be their loss, won't it, not having to go through all that education, and attend those etiquette workshops, and then throw out all their stuff. Honestly, I think I'd rather chew on glass than attend a series of workplace etiquette seminars and get re-educated by someone who tells me that I'm not going to have a desk any more. And those meetings to set behavior policies, those will be delightfully excruciating, for sure. What on earth is the company thinking?
Well, they're thinking about how this will allow them to vacate several buildings, because housing the employees this way takes up less room. So once again, this conforms to a rule that has seldom let me down: any question that starts out with "I wonder how come they. . ." can be answered with the word "Money".
Here's a problem that I've seen at every company I've worked at, and there are good reasons to believe that it afflicts every company out there. That's because I think it's grounded in human nature: dog-and-pony-itis.
That's the phrase I use for what happens to meetings over time. Many readers will be familiar with the process: a company gradually accumulates regular meetings on its internal calendar - project team meetings, individual chemistry and biology meetings inside that, overall review meetings, resourcing, planning, interdisciplinary meetings. . .everyone who's anyone, in some companies, has to be calling a meeting of their very own.
Eventually, someone says "Enough!" and purges the schedule, replacing the tangle of overlapping meetings with A Brand New Meeting or two. These will actually discuss issues, for once, and people are encouraged to actually say what's really going on with their projects. For once. And who knows, maybe that's the case (for once) - but it doesn't last.
Because every time, in my experience, the Brand New Meeting itself starts to collect barnacles. Over time, it becomes less useful, and more of a show. The music starts up, the Pomeranian dogs start hopping around and barking, and the trained horses make their entrance from the wings. It becomes more expedient to just get up and tell people the broad strokes of a project, especially the broad strokes that are actually working, and leave the messy details out. And gradually, other meetings spring up to try to take up the slack, since nothing ever seems to get done at the Brand New. . .
The thing is, I don't know how to stop this from happening. It comes on like rust. I've lost count of the we've-got-to-get-rid-of-this-stupid-meeting initiatives I've seen over the years, and every time the cycles eventually repeats. So here's a question: has anyone broken out? And if you have, how? Suggestions welcomed in the comments. . .
One of the comments to this post brought my attention to this paper in JACS on a new fluorinating agent. I just finished writing a column on fluorinated drugs for Chemistry World, so the subject is on my mind.
I have to say, this looks like it could be a very useful reagent. I've never worked with any arylsulfur trifluorides, but that looks to change soon, since I'd guess that this stuff will shortly be commercialized. An air-stable, non-runaway reactive fluorinating reagent would hit the spot. It would be fine with me if I never open another bottle of DAST again, and my experiences with the likes of xenon difluoride haven't been wonderful, either. If anyone gets a chance to try this compound out, let us know if it's all it's billed to be!
Here's a lab equipment question that someone probably knows the answer to, but that someone isn't me. Anyone know where you can buy Corex glass? I'm looking for a tube of the stuff, about 3cm by 28cm, but the only thing I can find are centrifuge tubes. The stuff is (or at least was) made by Corning. It's an aluminosilicate and it's mechanically quite strong, so the centrifuge use makes sense, but no one seems to sell a plain tube of the stuff. Any ideas?
Here's a look (PDF) at the Novartis "Labs of the Future". This looks like another one of these "open lab" concepts, and it appears that Basel has really bought into the idea. The interview with the two biologists helping to head the project is. . .well, it's very Swiss, that's the best description. When asked "What indicators would you select to measure improvement?" after people move in, the answer is:
Bouwmeester: That depends on the monitoring period. I am assuming that one or two months after we move into the building, the employees will already be experiencing a new dynamic. If they report a positive difference, that will be a first measure of success. It is important that the people in the LOTF develop some
kind of ownership regarding their role in the building. It will be a more active role than usual. The LOTF is basically an open space where you can observe your peers across the hierarchies. This is a different type of social architecture compared to 10 years ago, or even today still. Everybody will be more under observation and observing more than before. The dynamism of the interaction between people will increase. The employees themselves will have to decide what is common practice on their floor. Of course the concept will need to be adapted over time; I would be surprised
if all concepts materialize exactly as anticipated.
I would be, too. In fact, I'd like to propose that last sentence be printed up on T-shirts, coffee mugs, and posters, but that's probably not going to happen. More on what this is actually going to look like:
Korthäuer: (There will be) big screens, placed where people typically pass by. There will be video cameras installed on each laptop, allowing easy and informal contacts. The information technology concept is an important part of it. We have also designed special furniture that serves the same goal. We want to get rid of functional cells such as coffee rooms, writing rooms, lab rooms etc. Our aim is to bring the walls down. But of course we still need differentiation. In the LOTF
there are still compartmentalized areas with particular qualities, constructed according to people’s needs and workflow requirements. . .
Korthäuer: On the information technology side, we are trying to implement a few applications which really support the concept. There will be videoconferences, ‘smart’ whiteboards that allow notes to be captured electronically. We focus on proven technologies. Gradually, we will bring new technologies into the building such as haptic interfaces. Removing the walls in a building can bring about big changes. There will be much less storage room. Therefore, a little robot operating in an elevator shaft will transport materials ordered by laptop up from the basement to the floors. . .
Bouwmeester: We have already implemented Virtual Reality Rooms with ultra-high-resolution video screens, so that the quality is as though you were in a real-life meeting; the effect is quite spectacular. As with any global project, it will all require a certain attitude, a set of skills that people will have to develop. Much energy and time will be needed to communicate efficiently between places as different as Basel, Cambridge and Shanghai. . .
Allow me to note a few difficulties. For one, those high-res video conference venues still have to deal with switching and transmission delays, especially across the distances that the Novartis guys are talking about. So if you try to have a spirited real-time discussion, you'll mostly be getting very clear, sharp, high-fidelity views of people interrupting each other and pausing awkwardly. (Update: see the comments section - some users are reporting more successful experience.) I have a more macro-scale worry about this sort of thing, too, having to do with my suspicion of plans that depend on people finally shaping up and acting the way that they're supposed to. As with any global project, y'know.
I think that I'll let Tom Wolfe have the last word here, since what he wrote in From Bauhaus to Our House is still applicable, over thirty years later:
I once saw the owners of such a place driven to the edge of sensory deprivation by the whiteness & lightness & leanness & cleanness & bareness & spareness of it all. They became desperate for an antidote, such as coziness & color. They tried to bury the obligatory white sofas under Thai-silk throw pillows of every rebellious, iridescent shade of magenta, pink, and tropical green imaginable. But the architect returned, as he always does, like the conscience of a Calvinist, and he lectured them and hectored them and chucked the shimmering little sweet things out.
Every great law firm in New York moves without a sputter of protest into a glass-box office building with concrete slab floors and seven-foot-ten-inch-high concrete slab ceilings and plasterboard walls and pygmy corridors. . .Without a peep they move in!—even though the glass box appalls them all. . .
I find the relation of the architect to the client in America today wonderfully eccentric, bordering on the perverse. . .after 1945 our plutocrats, bureaucrats, board chairmen, CEO's, commissioners, and college presidents undergo an inexplicable change. They become diffident and reticent. All at once they are willing to accept that glass of ice water in the face, that bracing slap across the mouth, that reprimand for the fat on one's bourgeois soul, known as modern architecture.
And why? They can't tell you. They look up at the barefaced buildings they have bought, those great hulking structures they hate so thoroughly, and they can't figure it out themselves. It makes their heads hurt.
From reader Jose, in the comments thread to the most recent post:
"Published I find it ironic that so many pharma sites who hired hotshot architects to design labspaces that foster as much personal interaction as possible, are now pumping the virtues of collaborations across 10 time zones."
Medicinal chemists spend an awful lot of time working with SAR, structure-activity relationship(s). That's how we think: hmm, what happens if I put a chloro there? If I make that ring one size larger? If I flip that stereocenter/add a nitrogen/tie back that chain? Ideally, you pick up on a trend that you can exploit to give you a better compound, but the problem is, no SAR trend lasts forever. Methyl's good, ethyl's fine, anything bigger falls off the cliff - that sort of thing.
Activity "cliffs" of this sort are the subject of a paper earlier this year in the Journal of Chemical Information and Modeling. (For some earlier approaches to this same type of question, see here, here, here, and especially here).This group (from Germany) looked over several public SAR databases and used a new algorithm to extract "matched molecular pairs", which are compounds that differ only at one point in their structure. And what they were looking for wasn't the orderly progressions; they were after the changes that tended to suddenly change the activity of a compound by at least 100-fold. Were there, they wondered, functional group shifts that have a greater or lesser chance of doing that, over a wide range of targets and compound classes?
It looks like there are, and they're the transformations that you might well imagine. Messing around with a carboxyl group, for example, seems rarely to be a neutral event. Carboxylates are so relentlessly polar and hydrogen-bonding that your SAR is probably going to love 'em or hate 'em. The next two liveliest groups were carbonyls (in general) and amines. Of less interest (but equally believable) is the transformation from methyl to bulky alkyl (or vice versa, which is the direction I'd recommend people try to go if at all possible - other things being equal, no one should grease up their compounds unless there's absolutely no choice).
Well, it needs no ghost come from the grave to tell us this, either. How about any surprises? Adding a secondary hydroxyl group was surprisingly silent, compared to what you might picture. And switching from secondary to tertiary amines (just with methyl groups) is a much less conservative switch than you might imagine, with several huge activity shifts across different target classes. Introduction of methyl ethers rarely affected things much one way or another, and that might account for the low tendency of dimethylamine-to-morpholine doing anything. Small halogens on aryl rings (fluorine, chlorine) had low potential to cause big shifts, with ortho-chloros showing no examples of that happening at all. Oddly (at least to me) was the fact that morpholine-to-alkylpiperazine showed almost no big changes, either.
But it has to be emphasized that these are (1) averages and (2) averages over a large (but not gigantic) data set. For example, one of the "no changes at all" transformations is a favorite med-chem isostere, thiophene for phenyl. And that's true - most of the time, that does nothing. But I've seen two examples in my career when that one actually caused a big change in activity, so it's rare, but not impossible. That's the thing that makes med-chem so enjoyable and so frustrating at the same time. It's full of things (like actually discovering a drug) that are rare, but not quite impossible.
I'm enjoying myself very much in the lab today, doing something I haven't done in 20 years: photochemistry. I did some during my post-doc (with Bernd Giese, which is also the last time I've done free radical chemistry, at least on purpose). Since then, though, it's one of those things that's never come up. We had a mercury lamp apparatus in my grad school group, which I saw used a few times - one of which resulted in one of those nose-wrinkling "What's that funny smell?" moments, when the person running it forgot to turn on the cooling water. Don't do that. Medium-pressure mercury lamps can get pretty toasty. (They'll also permanently tan your eyeballs if you're so foolish as to look at them, I should also note, so don't do that, either!)
Most synthetic chemists will have had a brief experience with the technique - it's very appealing to think of doing chemistry just by shining a light on the reaction. But there can be a lot of variables - the sort of lamp you use (and thus the wavelengths and energy flux), various filters, sensitizing additives, hardware setups. Many people find that they use it for one reaction at some point, to make a specific compound, and never quite find a use for it again. In my experience, every decent-sized chemistry department has a photochemical rig of some sort, and no one quite knows where all its parts are.
That's probably a shame. There are a lot of unusual and interesting reactions that can be done photochemically - if you like 3- and 4-membered rings, this is certainly a field you should look into. I can recommend this recent bookas a general review of the field, for anyone who's thinking about it. We'll see how much use I get out of my current setup, but for now, I'm happily blasting away with the ultraviolet. . .
Update: blasting away is right! My cooling water dribbled down and then cut out on me after I tried to turn it down a bit, and, well. . . now I'm cleaning melted goo off of the quartz. A razor blade is working pretty well, but that's no way to treat a working piece of equipment.
Earlier this year, I wrote here about using calorimetry in drug discovery. Years ago, people would have given you the raised eyebrow if you'd suggested that, but it's gradually becoming more popular, especially among people doing fragment-based drug discovery. After all, the binding energy that we depend on for our drug candidates is a thermodynamic property, and you can detect the heat being given off when the molecules bind well. Calorimetry also lets you break that binding energy down into its enthalpic (delta-H) and entropic (T delta-S) components, which is hard to do by other means.
And there's where the arguing starts. As I mentioned back in March, one idea that's been floating around is that better drug molecules tend to have more of an enthalpic contribution to their binding. Very roughly speaking, enthalpic interactions are often what med-chemists call "positive" ones like forming a new hydrogen bond or pi-stack, whereas entropic interactions are often just due to pushing water molecules off the protein with some greasy part of your molecule. (Note: there are several tricky double-back-around exceptions to both of those mental models. Thermodynamics is a resourceful field!) But in that way, it makes sense that more robust compounds with better properties might well be more enthalpically-driven in their binding.
But we do not live in a world bounded by what makes intuitive sense. Some people think that the examples given in the literature for this effect are the only decent examples that anyone has. At the fragment conference I attended the other week, though, a speaker from Astex (a company that's certainly run a lot of fragment optimization projects) said that they're basically not seeing it. In their hands, some lead series are enthalpy-driven as they get better, some are entropy-driven, and some switch gears as the SAR evolves. Another speaker said that they, on the other hand, do tend to go with the enthalpy-driven compounds, but I'm not sure if that's just because they don't have as much data as the Astex people do.
So as far as I'm concerned, the whole concept that I talked about in March is still in the "interesting but unproven" category. We're all looking for new ways to pick better starting compounds or optimize leads, but I'm still not sure if this is going to do the trick. . .
I had an email this morning asking me to settle a bet on lab technique. I'm not sure I know the answer myself, so I figured I'd throw the question out to the readership.
So here goes: in your vacuum cold trap, which I'll assume is cooled by dry ice (and not liquid nitrogen, for the most part), what solvent do you use: acetone or isopropanol? (If you use something else, feel free to add it to the list, but I think you'll be in a distinct minority). As for me, I used acetone back in grad school, but switched over to isopropanol years ago, because I didn't have to change it (or add to it) so often.
Mat Todd of the University of Sydney looks over the SciFoo conference that we both attended during the summer, and contrasts that to an ACS meeting. The comparison isn't kind, as you'd imagine:
. . .with a few very notable exceptions the talks I saw were a) presented in a dull Powerpoint-heavy series of slides with verbal commentary about what was on the slides where even the presenter was visibly bored with what they were saying and b) on published material that was c) way too predictable and incremental. So both the presentational style and the content were disappointing. So many talks at the ACS would have been more interesting if the speaker had simply given out paper copies of their latest paper and given us 10 minutes to read it in silence then 10 minutes to talk about it. Now of course specialism necessitates incrementalism in content, but it’s no good if the meeting becomes a chore to sit and listen to. Nor is it good if the talks come out of the Powerpoint Machine (the genius of the “Chicken Talk” is that you can kind of follow the talk structure without listening to the content – it sounds exactly like most academic talks right up to the last supplementary slide in response to the second question at the end). In maybe 80% of the talks I attended nobody asked questions, or nobody was allowed to, or people asked “pity questions” just to break the awkward silence, but which were in no way interesting in themselves.
"A chore" is exactly what I find too many presentations and conferences to be, unfortunately. If we limited presentations, as Mat suggests, to people who are excited about their results, we'd have a lot of short meetings in this field. . .
Here's something that you don't think about until you actually work in a department full of chemists: how do you keep track of who's got what, and where it is? Everyone has reagents on their bench, and hidden away under the fume hood, and they're ordering more (and using up the current bottles) all the time. And people are wandering from lab to lab, borrowing and pilfering, sometimes when the original owners are there, and sometimes not. So how do you know what you have?
I've seen a number of approaches to this chemical inventory problem. The essential thing is that every bottle of every reagent be trackable. That means some sort of bar-coding system, most likely. Those bar codes need to go on when the compounds come in the door, ideally, so there aren't a lot of invisible reagents floating around. I think the best way to do this is to have the shipping and receiving people involved - if you trust the chemists to bar-code things, many of them just won't quite get around to it.
The next big question is whether you're going to have a centralized chemical stockroom or not (I've worked under both systems). The stockroom probably makes it easier to keep track of things, in general, since otherwise the available reagents are distributed throughout the labs at all times (instead of the ten per cent or so that are actually in active use). And it helps to have some place to send all those bottles back to - when you clean up your bench, you know that there's one thing you can do immediately, which helps keep the chemicals homing back to the central location.
A stockroom, though, requires dedicated space and dedicated head count, and neither of those are always feasible. The spread-throughout-the-labs approach puts the work back on the chemists. Its biggest disadvantage is entropy: bottles move around, get silently consumed, or get just plain lost. (That happens with a stockroom system, too, but at a slower rate). After a while, your map of the chemical inventory is useless - and for popular reagents, "a while" might be about two weeks.
That brings up the moving-chemicals problem, and to be honest, I've never seen a good solution to that one. Ideally, any time a person borrows some reagent from its known location, they scan the bar code so the system knows that it's moved. In practice, you know, you're just using it for a couple of days. Or you're just running one reaction, and you're going to take it right back. It's just right down the hall; the folks down there know where it is. Right. A stockroom system keeps this from randomizing things as quickly, but no matter what, this sort of Brownian motion is going to scramble things eventually.
So there has to be a regular inventory taken, no matter whose system you're using. Whether that's someone from the stockroom coming through and scanning all the benches and cabinets, or whether you declare Inventory Day and make all the chemists do it themselves, it has to be done. Twice a year is not too often, in my experience.
If anyone has solutions to some of these problems that I haven't touched on, feel free to share them in the comments. But please, no "Just Make Everyone Act Responsibly For Once" recommendations. Let's assume that people are intrinsically looking for the easy ways out, and work from that - it's a worldview that has never disappointed me.
This year's Nobel for palladium-catalyzed coupling reactions highlighted how useful these have become. But what every practicing organic chemist knows is how complicated they can be, particularly when you first couple of favorite recipes don't work. I've long thought that almost any metal-catalyzed transformation can be optimized, if you're just willing to devote enough of your life to it. But you have to have a good reason to wade into the swamp, because there sure are a lot of variables that can be tweaked. Here's a good case in point, recently published in Organic Letters. A perfectly reasonable reaction (C-H arylation of a chloropyrazole, which had been demonstrated before) was run through the statistical wringer to track down the best conditions.
They looked at 6 solvents, 10 bases, 4 catalysts, 5 ligands, and 4 additives, which would give you 7200 combinations if you ran the whole shebang. A Design of Experiments approach cut the number of actual runs down significantly, and then (fortunately) some of the variables turned out to be pretty insensitive. So this one wasn't as bad as some of them get - the ligand didn't seem to have too much effect, for example, whereas in some other Pd couplings it's crucial. (The choice of base had a much bigger effect, in case you're wondering). Their best set of conditions seems to work reasonably well across a range of possible substrates.
DoE is worth a post of its own, and that'll be a timely thing for me. After brushing up against it for years, I may finally have a use for the technique soon. For those who don't know it, it's basically a way to figure out how to most efficiently sample "experiment space", by getting the most information out of each different run. And then you use principal components analysis (or something similar) to see what the most important changes were, and how they correlate to each other. It's asking, mathematically, what a synthetic chemist wants to know about a complicated reaction recipe: what changes are responsible for most of the variation in the results, and how can I track them down by running a reasonable number of experiments? In the drug industry, process chemists think about this sort of thing a lot more than discovery chemists do, but it's worth keeping an eye out for any time the approach could be helpful.
Update: here's a trip report on this conference over at Practical Fragments
I'm back from Philadelphia and the FBLD conference. I'm not going to put a trip report up on the blog - although I'm certainly writing one up for my colleagues at work - but a number of people at the meeting asked me what I might say about it here.
Well, I enjoyed it. I tend to like more focused conferences like this one, anyway, where most of the people doing the best work in a field can attend what's still a fairly small meeting. It probably helps that this isn't a very old series of meetings, too. Over time, some sort of scientific entropy sets in, and the topics covered can begin to smear out a bit. Some of the longer-running Gordon Conferences are (to me, anyway) a little blurry about what they're trying to cover.
That same tendency can affect individual talks. We medicinal chemists are particularly guilty of that, since our discipline spreads over a pretty wide area. At a meeting like this one, which was all about fragment-based techniques, people had to resist the temptation to keep going past the fragment-based parts of their talk. Once you get up toward 400 molecular weight, you're not talking about fragments any more - you're doing good old medicinal chemistry. Maybe it's structure-based at that point, maybe not, and maybe you're using some of the biophysical techniques that help out with moving fragment leads forward - but the fragment techniques are what got you to that point, not what's carrying you forward through the concerns about PK, formulations, polymorphs, and all the other later-stage worries of a drug program.
The speakers at this meeting generally did a good job avoiding this pitfall, but I have to admit that the few times I saw PK data come up on the screen, I stopped taking notes. I didn't stop listening, on the chance that there might be something interesting, but it certainly wasn't what I was there to focus on. One could imagine a whole meeting about solving PK problems in drug development - there probably is one, actually. But at that one, you'd have to make sure that the speakers didn't spend time telling you about the neat fragment-based techniques that led to their drug candidate.
As I said, though, there were a lot of interesting speakers at this one, and not a single talk was anything close to a complete waste of time. How many meetings can you say that about? Things ran smoothly, and with notably better food than some of the other conferences I've attended. Some meetings just pitch a bunch of Wissenschaftlerfutter out onto the tables, figuring that people will deal with it - and to be honest, they're usually right. We'll eat most anything in this field, although I've been told that physicists are even less discriminating, so at least we have that.
A group at GSK has published a paper in Angewandte Chemie on the kinds of reactions that medicinal chemists use, and why they use them. The conclusions will come as no surprise to anyone practicing in the field. The workhorse reactions were condensations (amides, etc.), palladium-catalyzed couplings, and alkylations. And if you look at the reactions used to generate arrays (small libraries) of compounds simultaneously, those reactions almost take over the list.
Why is that? Well, for one thing, because those reactions tend to work. You'll almost always get product out of them - no small thing. You really, really don't want to spend time tweaking a reaction just to make it produce something, not when the odds of any individual product working are still small. And you can also get a good amount of structural diversity off-the-shelf, by using the huge numbers of commercial amines, acids, aryl boronic acids, and so on. They're also fast reactions, for the most part: set 'em up one day, work 'em up the next, and on to the next analogs.
The authors list some criteria that new reactions would need to order to make the list: not fussy about conditions (temperature, time, order of addition, atmosphere, etc.), compatible with polar solvents, tolerant of a wide range of functional groups, easy to dispense, easy to clean up, and so on. They mention that there's been funding in the UK over the last few years (as there has been here) for discovery of new chemistries that would meet this standards, but (reading between the lines) it doesn't seem as if anything major has made it up the charts yet.
Their other take-home is that people who specialize in running arrays can usually do them more efficiently than those who set them up just once in a while. They suggest that it takes a slightly different sort of person to be good at this:
. . . Owing to their focus on and expertise with arrays, we have found that the array chemists can make, purify, analyze, and register array compounds more efficiently than the program medicinal chemists. They are frequently also more effective in delivering a higher percentage of products from the array in greater yield and purity.
The team has a unique skill set and mindset. We have found that an array chemist should be highly organized, show attention to detail, be manually dexterous, be comfortable with repeatedly delivering to deadlines, and have an ability to work with often introverted and occasionally obstreperous program chemists ! This combination of characteristics is uncommon amongst chemists.
As an obstreperous program chemist myself, I should resent that remark. But you know, they're probably right. . .
A comment to the most recent post on puns mentioned the famous JOC paper in verse from the 1970s, and prompted another comment that "If you have to report your results as a villanelle, I think we'll see fewer methyl, ethyl, butyl, futile papers. . ."
Well, it's not a whole paper, for sure. But here's the best that I can do in thirty minutes:
Put In Another Methyl Group: A Villanelle
I shouldn't have to put a methyl there
No matter what the modeling group might say
So it docks to perfection: I don't care.
The project head gave me an evil glare
When I spoke up at our review today.
I shouldn't have to put a methyl there.
"The glutamate will pick up that lone pair".
Who knows? That might be right; I couldn't say.
So it docks to perfection: I don't care.
How do these really bind? We don't know where.
It's not like we can get a good X-ray.
I shouldn't have to put a methyl there.
Quaternary chiral centers? I don't dare.
I'd need two months if I needed a day.
So it docks to perfection: I don't care.
But no one ever said research was fair.
I'm going to have to come up with a way.
I shouldn't have to put a methyl there.
So it docks to perfection: I don't care.
Update: yes, I'm going to give the molecular modelers their own poem. It's only fair!
I had an interesting email about a 2009 paper in Drug Discovery Today that has some bearing on the "how much compound to submit" question, as well as several other areas. It's from a team at AstraZeneca, and covers their application of "Lean Six Sigma" to the drug discovery process. I didn't see it at the time, but The title probably made me skip over it even if I had.
I'll admit my biases up front: outside of its possible uses in sheer widget-production-line settings, I've tended to regard Six Sigma and its variants as a buzzword-driven cult. From what I've been able to see of it, it generates a huge number of meetings and exhortations from management, along with a blizzard of posters, slogans, and other detritus. On the other hand, it gives everyone responsible a feeling that they've Really Accomplished Something, which is what most of these managerial overhauls seem to deliver before - or in place of - anything concrete. There, I feel better already.
On the other hand, I am presumably a scientist, so I should be willing to be persuaded by evidence. And if sensible recommendations emerge, I probably shouldn't be so steamed up about the process used to arrive at them. So, what are the changes that the AZ team says that they made?
Well, first off is a realization that too much time was being spent early on in resynthesis. The group ended up recommending that every lead-optimization compound be submitted in at least a 30 to 35 mg batch. From my experience, that's definitely on the high side; a lot of people don't seem to produce that much. But according to the AZ people, it really does save you time in the long run.
A more controversial shift was in the way that chemistry teams work. Reflecting on the relationship between overall speed and the amount of work in progress, they came up with this:
Traditionally, chemists have worked alongside each other, each working on multiple target compounds independently from the other members in the team. Unless managed very carefully by the team leader, this model results in a large, and relatively invisible, amount of work in progress across a team of chemists. In order to reduce the lead time for each target, it was decided to introduce more cooperative team working, combined with actively restricting the work in progress. The key driver to achieve and sustain these two goals was the introduction of a visual planning system that enables control of work in progress and also facil-
itates work sharing across the team. Such a visual planning system also allows the team to keep track of ideas, arrival of starting materials, ongoing synthesis and compounds being purified. It also makes problems more readily recognizable when they do occur.
We have reflected on why chemistry teams have always been organized in such an individual-based way. We believe that a major factor lies in the education and training of chemists at universities, in particular at the doctoral and postdoctoral level, which is always focused on delivery of separate pieces of work by the students. This habit has then been maintained in the pharmaceutical industry even though team working, with chemists supporting each other in the delivery of compounds, would be beneficial and reduce synthesis lead times.
OK, that by itself is enough to run a big discussion here, so I think I'll split off the rest of the AZ ideas into another post or two. So, what do you think? Is the "You do your compounds and I'll do mine" style hurting productivity in drug research? Is the switch to something else desirable, or even possible? And if it is, has AstraZeneca really accomplished it, or do they just say that they have? (Nothing personal intended there - it's just that I've seen a lot of "Now we do everything differently!" presentations over the years. . .) After all, this paper is over a year old now, and presumably covers things that happened well before that. Is this how things really work at AZ? Let the discussion commence!
Talking about the amounts of compound to submit as a medicinal chemist brings up another topic. In every med-chem department I've worked in, there have been periodic exhortations for the chemists to register their intermediates. But too few people do.
For those outside the field, what I'm referring to are the "stepping stone" compounds along the way to structures that you're actually targeting. We try not to have these pathways go on too long, but there are often compounds that lack a key methyl group, or don't have the right stuff on the nitrogen yet, and so on. From the way that the compounds in a project have been running, you can be pretty sure that these things aren't going to be of much use for your current biological target - but the point is that they could be useful for someone else.
I've always been surprised by how many compounds sit on the benches, or in drawers, and never quite make it into the compound repository. To be sure, there are plenty of intermediates that shouldn't go in there - anyone who compound-codes a red-hot acid chloride should be whacked over the head. But plenty of things that people think of as "just starting material" or "just an intermediate" have nothing wrong with them, and should be added. I don't even mind a Boc group on an amine - t-butyl's not anyone's favorite, but there are plenty of drugs out there with carbamates on them. Fmoc is where I'd draw the line, though, since I think there's too much of a possibility for the binding to be driving by that big ol' fluorenyl, which is the first thing you'd want to get rid of if the compound hits. I don't think I'd go for any silyl groups on the alcohols, but benzyls and the like are fine.
So do a good deed today if you're in the lab: clear out a few of those compounds you have sitting around and put numbers on 'em. In your heart, you know it's the right thing to do!
Here's a question for those of you out in the industry: how much compound do you make, when you make a new one? Sometimes this question is equivalent to asking "How little will they let you get away with?" Different organizations have different requirements, on paper and for real, as to what that amount is. Five mg? Ten?
I've worked with people who kept coding these little 1.5mg amounts on most of their compounds, but I only do that if I'm desperate. That's really only going to do the immediate project any good, and not much, at that, if you want to do anything beyond the first in vitro assays. You'd like to have something living in the screening files so it can perhaps do some good later on. I try to aim for 10 to 20mg of compound, myself, although I don't always make it. And you?
Here's a lab question for everyone. I have a bottle of Aldrich copper oxide nanopowder on my lab bench; I've been meaning to try it out for some Ullmann reactions. I note that Aldrich (and others) are now selling a variety of such nanopowders, mostly metals and insoluble metal compounds.
And that makes sense, because these are the things that tend to react at their surfaces, and you'd have to think that a real nanopowder would have a tremendous surface area. My question is: does this really work out? Has anyone noticed a difference between the nanopowder form of a particular reagent and its more traditional one? I can imagine there being one - but I can also imagine the particles clumping up under some conditions and giving you back the equivalent of the cheaper stuff, too. Any hands-on experience out there?
One of the folks over at Chemistry Blog has run into a shortage: he and his labmates have tried to order (-) sparteine from every supplier in the book, and there's none to be had. So if anyone has a big dusty bottle of it sitting around, you might drop these desperate chemists a line. But that got me thinking about the way things suddenly dry up like this.
The situation is different than for an industrial chemical shortage, like the acetonitrile crunch that we went through a while back (and which has long since eased up). It's quite unusual for a bulk chemical like that to go down; several factors hit all at once in that case, and it affected an awful lot of people who needed the solvent. But fine chemicals are much weirder. When you trace some of them back to their real sources, you sometimes find that there are really only a couple of people in the world at any given time making some of these things. Or, in many cases, you find that there's no one making it at all - someone made a bunch a few years ago for some reason, sold the excess to a supplier, and everyone else has been buying it from that same bottle ever since.
So when one of these small-scale itemsevaporates, the reason can be supply: no one makes it any more. Or it can be demand-driven: a single drug company's scale-up group can deplete the world's commercial supply of some strange little molecule when they suddenly switch to a 500-gram run. Everyone working in such a group knows to call all the suppliers when they have a prep calling for some weirdo starting material, and they'll often take the precaution of ordering whatever's out there to be had. (That serves as a cushion while they contract someone else to crank out a batch or figure out how to make it themselves). Naturally, you'd rather have your drug candidates depend only on things that can be ordered in tank car lots, but that's just not always possible.
So it could be that someone needed a lot of (-) sparteine for an asymmetric synthesis recently, and bought up the existing world stocks. But this one sounds like more of a supply problem. There would appear to be customers out there, who have been draining the existing stocks, but no one's been able to replenish them. TCI apparently stated that it's the starting material for (-) sparteine that has become unavailable, but that sounds a bit funny, since it would surprise me if the material on the market is synthetic. Sparteine is a naturally occurring alkaloid, found in several species of plant, and it's very hard to compete with isolation of the natural product in those cases.
Perhaps TCI means that the usual plant source is unavailable - that's happened before, too. A spike in Tamiflu demand a few years ago suddenly sent the price of star anise up to record levels, since the chiral starting material (shikimic acid) in the usual synthesis was most conveniently isolated from that source. But for sparteine, it looks as if the isolation comes from plants in the broom family, which are not exactly rare shrubs, so I'm not sure what's going on. Any ideas?
As a brief followup to my "Elements I Have Yet to Use" post, I note this new paper on cleavage of molecular nitrogen by a hafnium complex. And to get right down to organic synthesis, here's a paper from last year that used hafnium triflate as a Lewis acid.
OK, here goes: has anyone out there ever used hafnium for anything? Anything at all? I sure haven't. (N.b. - ordering some on purpose to raise your desktop monitor or prop the door open does not count).
In keeping with the problem discussed here ("sticky containers"), there's a report that a lot of common spectrometric DNA assays may have been affected by leaching of various absorbing contaminants from plastic labware. If the published work is shown relative to control tubes, things should be (roughly) OK, but if not, well. . .who knows? Especially if the experiments were done using the less expensive tubes, which seem to be more prone to emitting gunk.
We take containers for granted in most lab situations, but we really shouldn't. Everything - all the plastics, all the types of glass, all the metals - is capable of causing trouble under some conditions. And it tends to sneak up on us when it happens. (Of course, there are more, well, noticeable problems with plastics in the organic chemistry lab, but that's another story. Watch out for the flying cork rings!)
For the medicinal chemists in the audience, I wanted to strongly recommend a new paper from a group at Roche. It's a tour through the various sorts of interactions between proteins and ligands, with copious examples, and it's a very sensible look at the subject. It covers a number of topics that have been discussed here (and throughout the literature in recent years), and looks to be an excellent one-stop reference.
In fact, read the right way, it's a testament to how tricky medicinal chemistry is. Some of the topics are hydrogen bonds (and why they can be excellent keys to binding or, alternatively, of no use whatsoever), water molecules bound to proteins (and why disturbing them can account for large amounts of binding energy, or, alternatively, kill your compound's chances of ever binding at all), halogen bonds (which really do exist, although not everyone realizes that), interactions with aryl rings (some of which can be just as beneficial coming in 90 degrees to where you might imagine), and so on.
And this is just to get compounds to bind to their targets, which is the absolute first step on the road to a drug. Then you can start worrying about how to have your compounds not bind to things you don't want (many of which you probably don't even realize even are out there). And about how to get it to decent blood levels, for a decent amount of time, and into the right compartments of the body. And at that point, it's nearly time to see if it does any good for the disease you're trying to target!
The discussion of "privileged scaffolds" in drugs here the other day got me to thinking. A colleague of mine mentioned that there may well be structures that don't hit nearly as often as you'd think. The example that came to his mind was homopiperazine, and he might have a point; I've never had much luck with those myself. That's not much of a data set, though, so I wanted to throw the question out for discussion.
We'll have to be careful to account for Commercial Availability Bias (which at least for homopiperazines has decreased over the years) and Synthetic Tractability Bias. Some structures don't show up much because they just don't get made much. And we'll also have to be sure that we're talking about the same things: benzo-fused homopiperazines (and other fused seven-membered rings) hit like crazy, as opposed to the monocyclic ones, which seem to be lower down the scale, somehow.
It's not implausible that there should be underprivileged scaffolds. The variety of binding sites is large, but not infinite, and I'm sure that it follows a power-law distribution like so many other things. The usual tricks (donor-acceptor pairs spaced about so wide apart, pi-stacking sandwiches, salt bridges) surely account for much more than their random share of the total amount of binding stabilization out there in the biosphere. And some structures are going to match up with those motifs better than others.
So, any nominations? Have any of you had structural types that seem as if they should be good, but always underperform?
Update, March 19: I've added a few more suppliers to the list, and broken out a third category for the mixed reviews. And I note in the comments that someone claiming to be Kathy Yu from 3B Chemicals is threatening me with legal action. The IP address resolves only to AT&T Internet Services, but there does appear to be someone from that name who works at 3B. I hope, for her sake and that of the company, that this is someone impersonating her, because whoever is leaving these comments is doing 3B no favors.
And since I am reporting opinions, both my own and those of other contributors that I have no reason to doubt, and am doing so without malicious intent, I will cheerfully ignore all legal threats.
OK, here are the lists of good companies and not-so-good companies, based on my experience and those of readers. I've had some personal communications, too, which I've added to the data set. As more reports come in, this will be the post that's updated, so it can serve as a reference.
I should note up front that I'm not listing the Big Guys, since (while they can have their ups and downs), you generally know that they're going to send you something. What we're looking at are the companies that you might not have dealt with, but want to know if they're reliable. And that brings us to the:
Good Suppliers
ABCR: good prices and hit rate on orders. Very professional.
Activate: expensive, but what's there is there, and it's the right stuff.
Adesis: not cheap, but very reliable and willing to work with customers to deliver similar compounds.
Advanced Chem Tech: recommended for peptide/amino acid stuff.
AK Scientific: several good reports on availability and purity.
Alinda: have ordered one thing from them, which was fine.
Anaspec: good reports on reliability
Apollo: good stuff, but catalog needs to be a bit more in line with their real stock.
Array: very pricey, but it's all there.
Astatech: good experience reported
Bionet: interesting catalog, doesn't back-order you.
Chembridge: a big catalog, but it's all real. Occasional purity problem.
Chem/Impex: good hit rate on availability. Some questions on their chiral purities.
Combi-Blocks: good list of useful intermediates, delivers on them.
Enamine: similar to ChemBridge in many ways. Big catalog. Not the fastest out there.
Florida Center for Heterocyclics: occasional purity issues, but they do deliver.
Frontier: great source for boronic acids and the like.
Life Chemicals: have had good experiences with compound purity here.
Lu: good source for custom peptides.
Matrix: interesting catalog, which they will really ship to you.
Maybridge: on the border of being one of the big guys. Very reliable.
Midwest: good reports on reliability.
Netchem: custom synthesis, but (for once!) with good turnaround and purity.
Oakwood/Fluorochem: good prices and reliability.
Peptide Protein Research: good for custom peptides.
Pharmacore: good stock of intermediates.
Rieke: reliable, only game in town for many odd reagents.
Strem: well known for quality inorganics and organometallics.
Synquest: used to be PCR. Good customer service.
Synthonix: stuff is in stock, customer service is responsive.
TCI: has always delivered, and quickly.
Transworld: very reliable and responsive.
Tyger: have never had a problem with them.
Waterstone Chemicals: good experience on pricing and availability
Mixed Reviews
American Custom Chemicals (ACC): several tales of bad purity and customer service, but others have had nothing but good experiences with them.
3B Chemicals: "will lead you on for months". Several bad experiences reported. On the other hand, I've just heard directly from a colleague who's had good luck with them.
J&W Pharmlab: bad experience reported (delays and purity), but others OK.
Ontario: one good report, but others complain of availability and leads times.
SPECS: mixed reports, but overall positive.
The Not So Good:
Ambinter: seems to source a lot of stuff from mystery suppliers. Many delays.
Any supplier, sad to say, with "Hangzhou" or "Shanghai" in the name. Tend to have absolutely nothing on the shelf, and if there's even a listed price, it's science fiction.
Anichem: very bad experience here with unexplained delays.
Beta Pharma: bad experience reported.
ChemMaker: very negative report on customer service and responsiveness.
City Chemicals: several bad experiences reported
Combi-Phos: several reports of purity problems.
Rarechem: haven't come across anyone with a good report here.
UK Green: a bad experience reported.
Uorsy: nothing ever seems to be in stock.
Zelinsky: several bad experiences reported.
Is it just me, or is the fine chemicals supply business getting even more out of hand than usual? I was just talking with a colleague who'd sourced an interesting intermediate, at the (steep!) price of about $900 for a gram. She placed the order and. . .you guessed it, the supplier immediately back-ordered it, saying the price had changed. It took someone from Purchasing to drag the new quote out of them (they apparently wouldn't give it over the phone). Now (to no one's surprise, I'm sure) the material is over $3000/gram, and will have a lead time of weeks.
This sort of thing has gone one for a long time, of course. But my impression is that there's more of it than ever. When the Chinese and ex-Soviet suppliers began to appear some years ago, they were often a pretty cheap source of some unusual compounds. But that's changed.
My belief - and I'll be glad to hear from people who do more compound purchasing than I do - is that the Chinese outfits especially have decided in recent years that they have some real pricing power, and are pushing it to see how far they can get. Add that to the hand-waving don't-you-worry-now aspect of many of their product lists, and you have a recipe for irritation and wasted time. (Another colleague described some of these online catalogs as "things they wish they sold".)
A previous comment on to a post like this listed some suppliers that had been found to be reliable, and I'll reproduce that here, in no particular order: Maybridge, Enamine, Asinex, Key Organics, ChemBridge, Specs, ASDI Biosciences, InterBioScreen, Vitas M Labs, Life Chemicals, Labotest, and TimTec. Suppliers of weirdo outlier compounds that nonetheless tend to come through were Albany Molecular, Chem T&I, Florida Center for Heterocyclic Compounds, and Princeton Biomolecular. I've used many of those folks myself (and have had particular success with Life Chemicals and Specs, as far as availability and purity). Some of these companies are faster to ship than others. But the thing that stands out with all of them is that they have what they say that they have, and what's more, it costs what it says that it costs.
For intermediates, as opposed to final-compound-like structures, I'd say that I've had good dealings with Apollo, Synthonix, Matrix, Pharmacore, Adesis, Tyger, Fluorochem, Oakwood, and Astatech. There are, I'm sure, several other suppliers in this category, and I'd be glad to list more of them after seeing the comments.
But now let's reverse the polarity. What's a blog for if you can't say what you really think? Here, then, is a preliminary blacklist of suppliers. These people either have product listings that overlay poorly with reality, try to jerk you around on the price, take much longer to deliver than their initial estimates, or (lucky you) can do all of these at once. My personal recommendation is to be quite careful with ChemPacific, Uorsys, CTI, Zelinsky, and everyone with the words "Hangzhou" or "Shanghai" in the company name.
Please feel free to add others to the lists. I'll do a consolidated post reflecting everyone's experience - that way, we can give business to deserving companies you might not have worked with before, and we can perhaps shame some others into acting more reasonably.
The entries I've done on the "open-plan" Biochemistry building at Oxford (see also Jim Hu) generated a lot of comments from people who've worked in poorly designed science facilities. I've heard from Linda Wang, a reporter at C&E News, who's writing article on this very subject. She's looking for chemists who are willing to talk about both good and bad experiences working in various building designs, so if you fit that description, feel free to email her at l_wang-at-acs.org (email address de-spammified, just substitute the usual symbol) or give her a call at 202-872-4579.
A discussion at work the other day got me to thinking: what structures do you medicinal chemists out there just refuse to work on? Any? We all have our own prejudices - in fact, if you get enough chemists into one conference room, one or another of them will probably rule out just about any structure you propose. Try that sometime, and be sure to sneak a few marketed drugs in there to tick people off. Don't like organoazides? Michael acceptors? Nitroaromatics? Epoxides? Chloromethyl ketones? They're out there working in the real world and making real money.
Now, I'm not saying that you should concentrate on these things. The success rate for (say) chloromethyl ketones is surely lower than for a lot of other compound classes, and there's only so much time and money available. That's why I have personal rules like "No Naphthyls". If someone shows me a structure with a raw naphthalene hanging off it that works, well, good for them, and I guess I'd work on it on that basis. But I won't contribute any myself, because I think the odds are too low.
But I have even more deep-seated prejudices. There are some structures that I just don't think have a chance, even if it looks like they work at first. I'd rather kill them immediately than take the (grave) chance of wasting everyone's time. The first thing I can think of on such a list would be quinones and their ilk. There are just too many other bad things that they're capable of. Now that I've said this, I feel sure that someone is come up in the comments with an example of a quinone that's making five hundred million dollars a year or something. But I sure can't think of one myself, and I just don't see the point of trying to make a drug out of such a structure (unless their lively reactivity is part of some nasty mechanism all its own, in which case, good luck to you).
I've written both here and elsewhere about flow chemistry, the technique where you pump your reactions through a reaction tube of some sort rather than mixing them up in a flask. And I freely admit that I have a fondness for the idea, but it's definitely not the answer to every problem.
For one thing, I tend to like the idea of sending reactants over a bed of catalyst or solid-supported reagent (what I call Type II or Type III flow reactions in that 2008 link above). Type I reactions, in my scheme, are the ones where you just use a plain tube or channel, and all the reactants are present in solution. A big advantage of those, as far as I can tell, is to handle tricky intermediates that you wouldn't want to have large amounts of or to control potential runaway exothermic reactions. There's also the possibility of running the reaction stream through some solid-phase purifications and scavengers, the way Steve Ley and his group like to work, which is convenient since you're already pumping the stuff along anyway.
But the sorts of reactions that you often see in the flow-chemistry equipment brochures. . .well, that's something else again. More than one outfit has earnestly tried to sell me a machine based on how well it did a Fischer esterification. My problem wasn't that the reaction was discovered almost in Neanderthal times - it was that Thag run reaction in round bottom flask, work fine, not need flow reactor. I mean, really, this is a nonexistent problem and needs no solution.
So I read this new paper in Angewandte Chemie with interest. The authors are looking at some standard catalytic organic transformations and comparing them carefully between batch mode and a flow setup. They stipulate at the beginning that flow chemistry has the advantages mentioned above, but they're wondering about what it can do for more ordinary chemistry:
"In addition to these developments, general and rather sweeping claims have been made that microreactor systems accelerate organic reactions and that lower catalyst loadings and higher yields can routinely be achieved in these systems compared to those of reactions carried out in flasks. Despite these potential advantages, examples of successful implementation of microflow reaction technologies in either academic organic synthesis or industrial process research and manufacturing remain more isolated than these reports would suggest. However, the implication is that it is only a matter of time before microflow reactors will dominate laboratory studies aimed at both fundamental research and practical applications of complex organic reactions, with our current mode of operation in reaction flasks ultimately becoming a relic of the past. It seems therefore worthwhile to examine the assumptions behind this viewpoint to provide a critical analysis of “flask versus flow” as a means for effecting reactions."
What they find is that there's very little difference. A catalyzed aldol reaction that was studied under flow conditions by the Seeburger lab is shown to perform identically to a batch reaction, if you make sure to run them at the same temperature and with the same catalyst loading. The paper then looks at asymmetric addition of diethyl zinc to benzaldehyde, a model reaction that I often wish would disappear from human consciousness so it would afflict us no more. But here, too, under more challenging heat-transfer conditions, flow showed no differences from batch. The authors point out that this reaction is, in fact, run under industrial conditions, but not in a flow apparatus. Rather, it's done in batch mode, but though good old slow addition of reagent, which also gives you control over exotherms.
The authors specifically exempt all supported-reagent chemistry from their analysis, so that preserves what I like about flow systems. But for homogeneous reactions, the only time they can see an advantage for the flow reactors is when there's a potential for a dangerous rise in temperature. So now we'll see what some of the more flow-oriented people have to say in reply. . .
Not as much time to blog this morning (and it's been hard getting into the site, since there are a lot of people who apparently want to know how to order some dioxygen difluoride). For one thing, I'm clearing a bunch of reactions out, and I've been devoting thought to how to do that in the laziest possible manner.
Maybe I should clarify that. What I mean is, how do I work up all these reactions quickly, in such a way as to make clean compounds that are worth testing, but spend the least amount of effort doing so? There are, of course, all sorts of brute-force ways to bang these things through, some of which would involve me not leaving my lab for the next three days or so, but I have other demands on my time. It's worth thinking about the most efficient way to do it.
Since these things I'm making all have acidic groups hanging off them, the most appealing idea I have right now is to use a basic resin to clean them up - as most med-chem types know, you can generally stick acidic compounds onto such resin, wash a lot of the crud off and throw that away, then bump your desired compounds off with some sort of acidic wash. This sort of solid-phase cleanup became popular in the combichem era, and has persisted for situations like this.
That's probably how I'll go, as opposed to, say, individually loading every single one of the compounds onto the HPLC machine. That would make me rather unpopular with the other people who might want to use that instrument before March is upon us, for one thing, and it would be complete overkill as well. These compounds are all pretty clean looking - a wash-and-rinse protocol should turn them out in good shape, and there's no need to use Super Ultimate Purification on them. (And besides, I'm making them all in reasonable quantity, which would bog down the HPLC even more).
An even more brainless way to do this workup would be to run every single compound through an automated column (like a Biotage). At least the HPLC has a liquid handler on it - I could set the thing up with a few rows of samples to inject, and walk away with some degree of confidence that it would run them. But the Biotage-type machines are usually one-at-a-time things, for larger samples. One batch of five grams of stuff would be perfect - two or three dozen at 100 mgs each, not so.
And all this makes me think of someone who used to work down the hall from me (no more clues than that!) I noticed that he was always cranking away in the lab, every time I went past. I mean, this guy looked like one of those multi-armed Hindu god statues, with each hand holding a round-bottom flask or a TLC plate. Impressive! Until I realized, after dealing with him a while, that the reason he was zipping around in there like a hamster was because he was doing everything in the most brutal and time-wasting way possible. He seemed to pick his reactions and protocols according to how much hand labor they involved: the more, the better.
I took a vow never to be him, and today I plan to live up to that. Measure twice, cut once and all that.
Here's a quick question for those of you that order a lot of odd little compounds. A correspondent tells me that he's been ordering resupplies from some of the usual suspects in this area (ChemBridge, ChemDiv - you know the sorts of companies, if you're in the med-chem business). And a higher than usual percentage of compounds are coming back as "Unavailable". . .only to show up available, at a significantly higher price, from Aurora.
Now, I certainly don't know the business arrangements between all these companies. I know that some of the compounds themselves are clearly coming from the same original sources, often somewhere in Russia, and make their way into a number of catalogs at once. But is this Aurora business a coincidence. . .or a business model? Anyone seen this happen personally?
Yesterday's post touched on something that all experienced drug discovery people have been through: the compound that works - until a new batch is made. Then it doesn't work so well. What to do?
You have a fork in the road here: one route is labeled "Blame the Assay" and the other one is "Blame the Compound". Neither can be ruled out at first, but the second alternative is easier to check out, thanks to modern analytical chemistry. A clean (or at least identical) LC/MS, a good NMR, even (gasp!) elemental analysis - all these can reassure you that the compound itself hasn't changed.
But sometimes it has. In my experience, the biggest mistake is to not fully characterize the original batch, particularly if it's a purchased compound, or if it comes from the dusty recesses of the archive. You really, really want to do an analytical check on these things. Labels can be mistaken, purity can be overestimated, compounds can decompose. I've seen all of these derail things. I believe I've mentioned a putative phosphatase inhibitor I worked on once, presented to me as a fine lead right out of the screening files. We resynthesized a batch of it, which promptly made the assay collapse. Despite having been told that the original compound had checked out just fine, I sent some out for elemental analysis, and marked some of the lesser-used boxes on the form while I was at it. This showed that the archive compound was, in fact, about a 1:1 zinc complex, for reasons that were lost in the mists of time, and that this (as you can imagine) did have a bit of an effect on the primary enzyme assay.
And I've seen plenty of things that have fallen apart on storage, and several commercial compounds that were clean as could be, but whose identity had no relation to what was on their labels (or their invoices for payment, dang it all). Always check, and always do that first. But what if you have, and the second lot doesn't work, and it appears to match the first in every way?
Personally, I say run the assay again, with whatever controls you can think of. I think at that point the chances of something odd happening there are greater than the chemical alternative, which is the dreaded Infinitely Active Impurity. Several times over the years, people have tried to convince me that even though some compound may look 99% clean, that all the activity is actually down there in the trace contaminants, and that if we just find it, we'll have something that'll be so potent that it'll make our heads spin. A successful conclusion to one of these snipe hunts is theoretically possible. But I have never witnessed one.
I'm willing to credit the flip side argument, the Infinitely Nasty Impurity, a bit more. It's easier to imagine something that would vigorously mess up an assay, although even then you generally need more than a trace. An equimolar amount of zinc will do. But an incredibly active compound, one that does just what you want, but in quantities so small that you've missed seeing it? Unlikely. Look for it, sure, but don't expect to find anything - and have 'em re-run that assay while you're looking.
Update: I meant to mention this, but a comment brings it up as well. One thing that may not show up so easily is a difference in the physical form of the compound, depending on how it's produced. This will mainly show up if you're (for example) dosing a suspension of powdered drug substance in an animal. A solution assay should cancel these things out (in vitro or in vivo), but you need to make sure that everything's really in solution. . .
I am here to confess to a deep-seated prejudice, one that has been with me for many years now. I know that others feel differently, but I'm sticking to my rule: No Naphthyls.
OK, pile on me now for having a closed mind. I know that there are drugs that are more successful than anything I'll ever make that have a naphthalene in them. (At least that structure's a small one). It's just that I see a naphthyl as the worst sort of "potency through greasiness" move in drug design. They hurt your solubility, drive up your molecular weight, open you to metabolites that you may not care for, and all for what? A little activity in your in vitro assay.
I'm getting close to putting cyclohexyl on the same list, if you want to know the truth. Problem is, people make these things "just as SAR compounds". You know, they'll trowel this hunk of grease onto the side of the molecule, just to see what happens, and if it really looks good, well, they'll. . .find some way to make it better. Right. Tetrahydropyranyl instead, that'll do it. But my attitude is, why not just make the THP derivative in the first place, if that's where you're going to go?
SAR is long, and life is short. There isn't time to make everything. So I decided a long time ago that I'd try to only make structures that I could live with. That still admits a lot of weird stuff, don't get me wrong. I have functional groups on my go-to lists that make people roll their eyes. But I draw the line at flat slabs of lard. No naphthalenes.
It's been a busy day on the front lines of science around here; apologies for not getting anything up until now. Here's a topic that I was discussing with some colleagues not too long ago: how much do we need to know about each other's specialties, anyway? I'm assuming that the answer is "more than nothing", although if someone wants to make the zilch case, I'd be interested in hearing it done.
But once past that, what's the optimum? I (for example) have never done cell culture. Nor do I see myself ever needing to do it (and anyone who needs me to is clearly in a bad way). I know the broad outlines of the field, but almost none of the details, and I'm sure that even my broad outlines have some faint parts in them. So if I'm at some sort of meeting where the problem-of-the-day turns on cell culture issues, I can be of no help at all. Is this a problem? I understand that different cells take to culture conditions differently, have varying growth rates, need media changes and whatnot, can generally only be passaged a certain number of times, etc. In short, I know roughly what to expect from my cell culture colleagues, and what would be silly of me to demand. Is that about right?
After all, I don't expect them to know the ins and outs of medicinal chemistry, particularly the synthetic organic lab part of it. Things like methylene chloride being rather more weirdly polar as a solvent than you'd expect, or the fact that some amines will stick to solid magnesium sulfate drying agent (but not sodium sulfate), or how you can azeotrope out acetic acid with toluene, or how you want palladium tetrakis to be lemon yellow and not orange - these and dozens of others are the tricks of my lab trade, and they don't know mine in the same way that I don't know theirs.
But I do like it when my biology colleagues have the broad outlines - that molecules with chiral centers, other things being equal, are often harder to make than achiral structures, that sticking a lot of cycloalkyl grease on a molecule is asking for metabolic trouble (no matter what it does for the potency in the assays), what sorts of things tend to make a molecule more (or less) soluble, and so on. Those are the equivalent of me knowing that primary cell lines lose some of their functions in culture, the difference between transient transfection and a stable cell line, etc.
It seems to me that each discipline in our business could draw up a list of What Everyone Else In the Company Should Know about their area. So, to start off with, I'm throwing the comments section over to what biologists (and others) should know, at a minimum, about med-chem. Take it away!
About a year ago, I wrote here about the impressive-looking new biochemistry building at Oxford, and wondered if it would work out quite the way the architects intended. Now I see a report from a post-doc who actually works there:
My first thoughts setting foot into the new building were the following: How are we supposed to concentrate with our offices in the atrium? How are we going to manage to work at such tiny desks?
I have to say, these initial concerns were justified. We hear the lab phone of every single floor ringing through the atrium, including people's mobile phones (which also causes envy towards those who actually have reception). When people really need to concentrate on writing, reading or thinking while others are discussing their work or are simply chatting, the atmosphere can get pretty tense. And even if it was completely silent in the atrium, the small size of the desks already makes working difficult. . .I once discussed the lack of space with our head of department. He simply replied: when you have to write a paper, you work from home anyway... I'd say £47 million well spent!
Perhaps I should talk a bit about this phrase "raw data" that I and others have been throwing around. For people who don't do research for a living, it may be useful to see just what's meant by that term.
As an example, I'll use some work that I was doing a few years ago. I had an reaction that was being run under a variety of conditions (about a dozen different ways, actually), but in each case was expected to either do nothing or produce the same product molecule. (This was, as you can see, a screen to see which conditions did the best job at getting the reaction to work). I set this up in a series of vials, taking care to run everything at the same concentration and to start all the reactions off at as close to the same time as I could manage.
After a set period, the reaction vials were all analyzed by LC/MS, a common (and extremely useful) analytical technique. I'd already given the folks running that machine a sample of the known product I was looking for, and they'd worked up conditions that reproducibly detected it with high sensitivity. They ran all my samples through the machine, and each one gave a response at the other end.
And those numbers were my raw data - but it's useful to think about what they represented. The machine was set to monitor a particular combination of ions, which would be produced by my desired product. As the sample was pumped through a purification column, the material coming out the far end was continuously monitored for those specific ions, and when they showed up, the machine would count the response it detected and display this as a function of time: a flat line, then a curvy, pointed, peak which went up and then came back down as the material of interest emerged from the column and dwindled away again.
So the numbers the machine gave me were the area under the curve of that peak, and that means, technically, that we're one step away from raw numbers right there. After all, area-under-the-curve is something that's subject to the judgment of a program or a person - where, exactly, does this curve start, and where does it end? Modern analytical machines are quite good at judging this sort of thing, but it's always good to look over their little electronic shoulders to make sure that their calls look correct to you. If you want to be hard-core about it, the raw data would be the detector response for each individual reading, at whatever frame rate the machine was sampling at. That's even more raw than most people need - actually, while writing this, I had to think for a moment to picture the data at that level, because it's not something I'd usually see or worry about. For my purposes, I took the areas that were calculated, since the peak shapes looked good, and the machine's software was able to evaluate them consistently and didn't have to apply any sort of correction to them to meet its own quality standards.
So there's one set of numbers. But the person running the machine had taken the trouble (as they should have) to run a standard curve using my supplied reference compound. That is, they'd dissolved it up to a series of ever-more-dilute solutions, and run those through the machine beforehand. This, plotted as peak area versus the real concentration, gave pretty much a straight line (as it should), and the machine's software was set up to use this information to also calculate a concentration for every one of my product peaks. So the data set that I got had the standard plot, followed by the experiments themselves, with both the peak areas and the resulting calculated amounts. Since these were related by what was very nearly a straight line, I probably could have used either one. But it's important to realize the difference: by using the calculated concentrations, I could either be correcting for a defect in the machine (if its detector response really wasn't quite linear), or I could be introducing error (if the standard solutions hadn't been made up quite right) It's up to you, as a scientist, to decide which way to go. In my case, I worked up the data both ways, and found that the resulting differences were far too small to worry about. So far, so good.
But there's another layer: I had done these experiments in triplicate. There were actually thirty-six vials for the twelve different conditions, because I wanted to see how reproducible the experiments were. For my final plots, then, I used the averages of the three runs for each reaction, and plotted the error bars thus generated to show how tight or loose these values really were. That's what I meant about the area numbers versus the concentration numbers question not meaning much in this case. Not only did they agree very well, but the variations between them were far smaller than the variations between different runs of the same experiments, and thus could safely be put in the "don't worry about it" category while interpreting the data.
What I did notice while doing this, though, was something else that was significant. My mass spec colleague had done something else which was very good practice: including a standard injection every so often during the runs of experimental determinations. Looking these over, I found that this same exact sample, of known concentration, was coming out as having less and less product in it as the process went on. That's certainly not unheard of - it usually means that the detector was getting less sensitive as time went on due to some sort of gradually accumulating gunk from my samples. Those numbers really should have been the same - after all, they were from the same vial - so I plotted out a curve to see how they declined with time. I then produced another column of numbers where I used that as a correction factor to adjust the data I'd actually obtained. The first runs needed little or no correction, as you'd figure, but by the end of the run, there were some noticeable changes.
So now I had several iterations of data for the same experiment. There was the raw raw data set, which I never really saw, and which would have been quite a pile if printed out. This was stored on the mass spec machine itself, in its own data format. Then I had numbers that I could use, the calculated areas of all those peaks. After that I had the corresponding concentrations, corrected for by the standard concentration curve run before the samples where injected. Then I had the values that I'd corrected for the detector response over time. And finally, once all this was done, I had the averages of the three duplicate runs for each set of conditions.
When I saved my file of data for this experiment, I took care to label everything I'd done. (I was sometimes lazier about such things earlier in my career, but I've learned that you can save ten minutes now only to spend hours eventually trying to figure out what you've done). The spreadsheet included all those iterations, each in a labeled column ("Area" "Concentration" "Corrected for response"), and both the standard curves and my response-versus-injection-number plots were included.
So how did my experiments look? Pretty good, actually. The error bars were small enough to see differences in the various conditions, which is what I'd hoped for, and some of those conditions were definitely much better than others. In fact, I thought I saw a useful trend in which ones worked best, and (as it turned out), this trend was even clearer after applying the correction for the detector response. I was glad to have the data; I've had far, far worse.
When presenting these results to my colleagues, I showed them a bar chart of the averages for the twelve different conditions, with the associated error bars plotted, which was good enough for everyone in my audience. If someone had asked to see my raw data, I would have sent them the file I mentioned above, with a note about how the numbers had been worked up. It's important to remember that the raw data are the numbers that come right out of the machine - the answers the universe gave you when you asked it a series of questions. The averages and the corrections are all useful (in fact, they can be essential), but it's important to have the source from which they came, and it's essential to show how that source material has been refined.
I asked recently for suggestions on the best books on med-chem topics, and a lot of good ideas came in via the comments and e-mail. Going over the list, the most recommended seem to be the following:
A comment to yesterday's post made a point that seemed instantly familiar, but it's one that my own thoughts had never quite put together. All of us who do medicinal chemistry came out of academic labs; that's where you get the degrees you need to have to be hired. Many of us worked on the synthesis of complex molecules for those degrees, since that's traditionally been a preferred base for drug companies to hire from. (You get a lot of experience of different kinds of reactions that way, have to deal with setbacks and adversity, and have to learn to think for yourself. Plus, if you can put up with some of the people who do natural products synthesis, the thinking goes, you can put up with anything).
Here's the interesting part, though. People who do the glass-filament spiderweb-sculpture work that is total natural product synthesis will defend it on many grounds (some more defensible than others, in my view). They have, naturally enough, a bias in favor of that kind of work. But have those of us who've done that kind of chemistry and then moved on to industry ended up with the opposite bias? Have we reacted against the forced-march experience of some of our early training by resolving never to get stuck in such a situation again (which is reasonable), but at the same time resolved never to get stuck doing fancy synthesis again?
That one may not be so reasonable. And I don't mean that we avoid twenty-step syntheses for irrational reasons, because there are perfectly rational reasons for fleeing from such things in industrial work. But this bias might extend further. Take a workhorse reaction like palladium-catalyzed coupling - that's just what people tend to think of when they think of uninspiring industrial organic synthesis, two or three lumpy heteroaromatics stuck together with Suzuki couplings, yawn. One of my colleagues, though, recently mentioned that he saw too many people sticking with rather primitive conditions for such reactions and taking their 50% yields (and cleanup problems) as just the normal course of events. And he's got a point, I'd say. There really are better conditions to use as your default Pd coupling mixture than the ones from the mid-1990s. You don't have to always clean all the red-brown gunk out from your product after using (dppf) as your phosphine ligand, and good ol' tetrakis is not always the reagent of choice. But a lot of people just take the standard brew, throw their starting materials in there, and bang 'em together. Crank up the microwave some more if it doesn't work.
I can see how this happens. After all, the big point that people have to learn when they join a drug research effort is that chemistry is not an end in itself - it's a tool to make compounds for another end entirely. If you're just making analogs in the early stages of a new project, no one's going to care much if your yields are low, because the key thing is that you made the compounds. I've said myself (many times) that there are two yield in medicinal chemistry: enough, and not enough. Often, perhaps a little too often, five milligrams qualifies as "enough", which means that you can check off a box through some really brutal chemistry.
But at the same time, if you could make simple changes to your reaction conditions, or to the kinds of reactions you tend to run, you could potentially make more compounds (because you're not spending so much time cleaning them up), make them in higher yields (or make your limited amount of starting material stretch further), or make more interesting (and patentable) ones, too. I think that too many of us do tend to get stuck in synthetic ruts of various sorts.
Perhaps the main cause of this is the pressure of normal drug discovery work. But I do have to wonder if some of the problem is a bit of aversion to the latest, hottest reagent or technique coming out of the academic labs. To be sure, a lot of that stuff isn't so useful out here in what it pleases us to call the real world. But there are a lot of things we could stand to learn, as well. Palladium couplings used to be considered kind of out-there, too, you know. . .
I'm home today (sick children, etc.), so I'm blogging from next to my daughter's guinea pig cage rather across the hall from my lab. But I have a lab-based question to throw out: what would you say is the chemistry technique or reagent with the worst publication-to-real use ratio?
I have a couple of nominees to get things rolling. For reagent, I would like to advance the montmorillonite clay stuff. I cannot count how many papers I have seen on its use as a Lewis acid, catalyst, and all-around good thing to have, but I have never used it myself, never spoken with anyone who has, and never (to my recollection) heard it suggested as a possible thing to try when someone encountered a synthetic problem. For all I know it's a fine reagent, but its footprint does not seem to be very large. I actually have used benzotriazole, but I've never seen an actual container of montmorillonite K-10.
For general technique, I'm tempted to nominate ionic liquids. Man, are there ever a lot of publications on those things, but again, I've never actually encountered them in actual practice. I have heard second-hand of people trying them, so I guess that counts for something, but it still seems to be disproportionate compared to the avalanche of literature citations for the things. The craze seems to have peaked, but still not a week goes by that I don't see a paper.
Nominations? As with the book recommendation post, I'll assemble things into master lists.
I get regular requests to recommend books on various aspects of medicinal chemistry and drug development. And while I have a few things on my list, I'm sure that I'm missing many more. So I wanted to throw this out to the readership: what do you think are the best places to turn? This way I can be more sure of pointing people in the right directions.
I'm interested in hearing about things in several categories - best introductions and overviews of the field (for people just starting out), as well as the best one-stop references for specific aspects of drug discovery (PK, toxicology, formulations, prodrugs, animal models, patent issues, etc.)
Feel free to add your suggestions in the comments, or e-mail them to me. I'll assemble the highest-recommended volumes into a master list and post that. Just in time for the holidays, y'know. . .
When we screen zillions of compounds from our files against a new drug target, what can we expect? How many hits will we get, and what percentage of those are actually worth looking at in more detail?
These are long-running questions, but over the last twenty years some lessons have been learned. A new paper in J. Med. Chem. emphasizes one of the biggest ones: if at all possible, run your assays with some sort of detergent in them.
Why would you do a thing like that? Compound aggregation. The last few years have seen a rapidly growing appreciation of this problem. Many small molecules will, under some conditions, clump together in solution and make a new species that has little or nothing to do with their individual members. These new aggregates can bind to protein surfaces, mess up fluorescent readouts, cause the target protein to stick to their surfaces instead, and cause all kinds of trouble. Adding detergent to the assay system cuts this down a great deal, and any compound that's a hit without detergent but loses activity with it should be viewed with strong suspicion.
The authors of this paper (from the NIH's Chemical Genomics Center and Brian Shoichet's lab at UCSF) were screening against the cysteine protease cruzain, a target for Chagas disease. They ran their whole library of compounds through under both detergent-free and detergent conditions and compared the results. In an earlier screening effort of this sort against beta-lactamase, nearly 95% of the hits (many of them rather weak) turned out to be aggregator compounds. This campaign showed similar numbers.
There were 15 times as many apparent hits in the detergent-free assay, for one thing. Some of these were apparently activating the enzyme, which is always a bit of an odd thing to explain, since inhibiting enzyme activity is a lot more likely. These activators almost completely disappeared under the detergent conditions, though. And even looking just at the inhibitors, 90% of the hit set in the detergent-free assay went away when detergent was added. (I should note that control cruzain inhibitors performed fine under both sets of assays, so it's not like the detergent itself was messing with the enzyme to any significant degree).
They point out another benefit to the detergent assay - it seems to improve the data by keeping the enzyme from sticking to the walls of the plastic tubes. That's a real problem which can kick your data around all over the place - I've encountered it myself, and heard a few horror stories over the years. But it's not something that's well appreciated outside of the people who set up assays for a living (and not always even among some of them).
So, let's get rid of those nasty aggegators, right? Not so fast. It turns out that some of the compounds that showed this problem during the earlier beta-lactamase work didn't cause a problem here, and vice versa. Even using different assays designed to detect aggregation alone gave varying results among sets of compounds. It appears that aggregation is quite sensitive to the specific assay conditions you're using, so trying to assemble a blacklist of aggregators is probably not going to work. You have to check things every time.
One other interesting point from this paper (and the previous one): curators of large screening collections spend a lot of time weeding out reactive compounds. They don't want things that will come in and react nonspecifically with labile groups on the target proteins, and that seems like a reasonable thing to do. But in these screens, the compounds with "hot" functional groups didn't have a particularly high hit rate. You'd expect a cysteine protease to be especially sensitive to this sort of thing, with that reactive thiol right in the active site, but not so. This ties in with the work from Benjamin Cravatt's group at Scripps, suggesting that even fairly reactive groups have a lot of constraints on them - they have to line up just right to form a covalent bond, and that just doesn't happen that often.
So perhaps we've all been worrying too much about reactive compounds, and not enough about the innocent-looking ones that clump up while we're not looking. Detergent is your friend!
I mentioned the H-Cube hydrogenation machine here a couple of years ago as an early example of a commercial flow chemistry machine. As some readers may have guessed, my recent post on hydrogenations was partly inspired by a recent run of activity on this instrument, which came in quite handy.
Until the last couple of days, that is. Now there's a problem, and I'd be glad to hear from any H-Cube users who might know how to solve it. (If you haven't used one, you can probably bail out right now!) What's going on is: when I try to run a hydrogenation in "Full H2" mode, everything works fine until the H2 valve closes. The pump's fine, the flow through the instrument is fine. . .until the status switches to "Running". At that point the flow stops momentarily, then a gout of solvent runs from the outlet all at once, and then. . .nothing. Well, nothing except hydrogen gas - if I dip the outlet tube below the surface of some solvent, I can see that it's still producing that. But there's no flow. Lifting the solvent inlet from the reservoir, I can see that nothing's being taken up - an air bubble forms at the inlet, and just moves up and down.
So there's something going on when the system starts letting hydrogen into the flow, but I'm not sure what that might be. I can always call in the $250/hr folks, but I thought that throwing my problems out onto the blog was at least worth a try. Just to take care of some obvious fixes, so far I've cleaned the metal frit, replaced the Teflon membrane, sonicated the check valve, and tried changing catalyst cartridges. Anyone got any clues after that?
Here are a few more of those questions that medicinal chemists have to deal with from time to time. Most of these have no definitive answers (which is why they keep coming up!)
1. You're making a compound that looks to be important in the project - maybe even the clinical candidate, if things go right. But there's a step in the synthesis which - while it does work - is clearly not something that's going to scale up too well. You need more compound right now, and you can push things through. But you're eventually going to have to ditch that step (unless this compound gets overtaken by another one), so. . .when's the right time to worry about that?
2. Your compound series is in a pretty crowded patent landscape. In fact, another application has just published that really looks to be breathing down your neck. Of course, that means the work in it was done a year and a half ago (or more). Can you assume that Company X has followed the same course that you have, and has already investigated the series you're working on? Should you drop them, or go in in the chances that six months from now another application will drop that covers you like a tarp?
3. You're finally writing up one of your old projects for publication. But it's been a while, and the details of what happened are not as sharp as they were when thing were going on. What's more, on looking the work over, you realize that there are some obvious gaps in it, stuff that didn't look that way at the time, but sure does so now. You can write things up to make it look more coherent, but only by rearranging the way it really happened. Where do you draw the line?
4. Your lead compound is ready to go into toxicology testing, the last big step before declaring victory and naming it as the development candidate. Trouble is, there's something funny about it in rats. They just don't get the blood levels that mice and dogs do, and your tox people would really, really rather run the tox study in rats (since that's the standard, and what they have the most comparison data for). Update: I mistakenly switched rodents mentally this morning on the train, now they're switched back to what they should be). You can get the blood levels up to where they need to be - but only by using a dosing vehicle that might have problems of its own, and that the toxicologists haven't had much experience with either. What to do?
I've been occupied all morning with voodoo. Well, the technical name for it is catalytic hydrogenation, but let's call it for what it is: witchcraft. It's a widely used reaction in organic chemistry, and you can use it to reduce all kinds of different functional groups on your molecules. But once you get off the well-traveled roads, it's all jungle drums at midnight.
One reason we chemists like this reaction so much is that it's simple. You add some dark insoluble powder to your compound - which is some metal like palladium, platinum, nickel or the like, adsorbed onto carbon black or another solid. Then you add solvent and put the whole thing under an atmosphere of hydrogen gas. That soaks into the metal particles, your compound sits on them and gets magically reduced, and after a while you filter everything off and there's your clean, transformed product.
Most of the time. You'll note that I've skipped over a lot of variables there. For one thing, there's the choice of metal catalysts. Pt and Pd get the most use, but they come on a variety of solid supports. Carbon, alumina, barium sulfate, calcium carbonate. . .they all act differently. And don't stop with those guys: nickel's not to be ignored, then rhodium's available, and even ruthenium if you want to crank up the pressure. The pressure of all that hydrogen, there's another variable. Just a balloon on top, atmospheric pressure? Or put in a thick glass bottle on a shaker and turn it up to 50 pounds per square inch? Higher, in a metal apparatus? And what temperature did you have in mind? Ambient, or would you like to heat things up? Remember, as the pressure goes up, so does the temperature you can run the solvents up to.
Ah yes, the solvents. A lot of the time you see this work done in methanol or ethanol, but the reactions will often go quite differently in ethyl acetate or even something less polar. I've even seen some done in dichloromethane, although that somehow just seems wrong. Acids often have a profound effect on things, particularly if there's a basic amine in your compound.
And I haven't mentioned poisoned catalysts yet, have I? A bit of lead, or the addition of (non-protonated) amines or sulfur-containing compounds can dial down the reactivity of a lot of these metals - often down to zero, but sometimes to a useful level that you can't reach any other way. And then there's transfer hydrogenation, where you don't use the gas itself, but let some other compound give up hydrogen inside the reaction and transfer it over to your substrate. Paraformaldehyde, formic acid, phosphites, cyclohexene - all of those will work, and they can all work differently.
So. . .how many variations are we up to? Do you want to use 5% palladium on carbon in methanol, room temperature at 50 psi? Or platinum oxide in acetic acid at 50 degrees? Rhodium on alumina, ethanol, 100 psi at 100 C? Or wet 10% platinum catalyst with formic acid? That should get you started on this simple, well-known reaction. I've run 22 of them in the last two days, with the assistance of the H-Cube reactor, and I have to say: I'm about hydrogenated out.
I wrote about this topic a few years ago, and thought I'd update it. Many chemists find themselves looking at a periodic table and wondering "How many of these things have I personally handled?" My list is up to nearly 45 elements (there are a couple that I've got to think about, one-off catalyst reactions from twenty-two years ago and the like). And there are at least 29 that I hope to never use at all, since they're radioactive and I'm generally not in the mood for that. So what does that leave me?
Well, I've never used beryllium, although it's not that I'm tapping my foot waiting for any. It's pretty toxic stuff, for the most part, and there are hardly any organic chemistry reactions that get near it. That means that I can't even think what I might use it for, and I could easily go my whole career without seeing any.
The next lowest molecule weight element I haven't messed with (excluding unreactive neon, which you at least get to see in its excited state) is probably scandium. That whole first column of transition metals is pretty useless for organic chemists, to be honest (Yttrium? Lanthanum?), and I've never seen any reactions that leapt out at me as things I had to try. No, if the answer is scandium, it must have been a pretty odd question.
Next up, I haven't used either of the G twins, gallium and germanium. They're not too well studied compared to their family members above and below: aluminum and even indium are more widely used than gallium, and silicon and tin show up in organic labs a million times more often than germanium. But with those relatives, you'd have to think that there's something interesting that can be done with these, so it depends on whether anyone finds out what that might be during the rest of my chemistry career.
And right next to these is arsenic, which I've also managed to avoid. It's famously poisonous, although it's really not worse than a lot of other things that get used much more often. But again, there's not a lot of compelling chemistry to be done with the stuff, not that I know of, anyway, and there are always those unfortunate nomenclature problems to be dealt with, especially if you have a British accent.
Krypton I've never had a use for, and I'd have to rate the chances as very low indeed. In the next row, I've handled strontium chloride, but only to make red-colored flames for a school demonstration show. I have yet to touch yttrium, as mentioned above, and I've managed to miss zirconium so far as well. There are actually a number of organometallic reactions that use that one, so it's at least a real possibility. Niobium I have yet to encounter, and at the rate it's used, I probably never will. Cadmium's another toxic beast - there are some old reactions that use organocadmiums, but I can't think when I saw a modern reference that used any of them, and I don't see this one in my future, either. Antimony I might use if I never need some horrible superacid. Tellurium, well. . .there would have to be a pretty good reason, given its reeking, nose-wrinkling sulfur and selenium relatives, but someone might yet come up with one. Can't rule that one out, unfortunately.
Now we're getting into the heavy metals, and a lot of gaps start to appear. Has anyone in an organic chemistry lab ever used hafnium or tantalum? Didn't think so. The best candidate for "something I could use, but haven't" in this bunch is osmium. The tetroxide is a very useful reagent that I just haven't had the need for. It wouldn't surprise me if that's the next addition to my list. I've no desire whatsoever to use thallium. It's part of a short run of nasties that you hit right after the jewelry metals - you have your platinum, then gold, and you think you're in the high-rent district, and suddenly it's mercury, thallium, and lead right in a row. Reminds me of the way towns were stuck next to each other in New Jersey.
And as far as the lanthanides, well, I've used cerium as a TLC stain, and once I used samarium iodide - which, true to its reputation, didn't work. None of the others have I touched, and unless I need some funky NMR shift reagent, which fewer and fewer people do these days, I don't see it happening. There are a lot of funny rare earths down there, but little reason for an organic chemist to go digging around among them.
Weirdest element I actually have handled? Xenon would have to be the winner - I've used the difluoride, and yes, that was the recourse of a desperate chemist. But it did work to turn a silyl enol ether into an alpha-fluoro ketone, so I can't say anything bad about it, other than its rather penetrating smell, which I probably should have taken more care not to experience. . .
Most chemistry departments in the drug industry have some academic consultants who come in every so often. The idea is that they'll have some useful suggestions about synthetic problems (there aren't so many academic consultants who are useful on drug discovery questions as opposed to pure chemistry ones). At the companies where I've worked, the consultants will spend the day in a conference room, while project teams troop in and out with presentations.
How useful this process is varies, to say the least. The first variable is the consultant, because some people are just better at that sort of thing. Ideally, you want someone who has a lot of ideas, has them relatively quickly, and enjoys putting them out for people to comment on them. Not everyone fits that description. While those can all be useful qualities, there are plenty of world-class scientists whose working style doesn't fit those requirements, and these people tend to be less valuable for drop-in sessions.
Another variable is the sorts of problems the drug discovery teams are dealing with. We try, in the industry, to reduce our chemistry to the simplest possible routes. Time is money (and money is money, too), and we always need methods that will reliably crank out plenty of different analogs without a lot of work. When that works, it often doesn't lead to especially exciting chemistry - in fact, the Venn diagram would show that "smoothly running project" and "exciting chemistry" don't overlap much. That means that the projects where things are going fine don't have much to talk about when the consultants appear, and those sessions sometimes end up spending more time on peripheral problems.
Much of the time, too, the biggest problems aren't chemical ones. If you're having trouble with metabolism, tox, or absorption, there aren't going to be many consultants who can help you out. Most of the ones who can are ex-industry people. (And with problems like these, sometimes no one can help you out at all). But asking someone about oral bioavailability when their research is all about interesting new synthetic organic methods is a waste of time - yours and theirs.
I've had some useful and interesting consulting sessions over the years, but some really disastrous ones, too. Many of the latter feature those "Well, now what do we talk about?" moments, which seem to be a cue for Satan to emerge and fill out the hour. So plan ahead. Make sure that you've got plenty to talk about. Actually, you'd better have more than you think you'll need, because some of your topics may either get a fast answer, or an equally fast shrug of the shoulders. . .
Hmmm. As a colleague just pointed out to me, I've spent some time here defending "me-too" drugs. And just this morning (see the previous post) I take off after what can only be described as "me-too reactions", saying that I don't see the use for so many of them.
Well! The only defense I can offer (until I think of a better one) is that there is no drug category so populated as the aldoxime-to-nitrile conversion is in synthetic chemistry (or acetal formation/deprotection, desilylation, or the other categories I spoke of in that other post). I suppose I might have a tougher time standing up for me-too drugs if there were (say) twenty-nine statins on the market. But still. . ."I'd better put up a post on that", I said. "Better you than someone with a funny pseudonym in your comments section", came the reply.
Here's a question you don't hear discussed very often: are there some synthetic organic chemistry reactions that don't need any more work? I'm moved to ask this because I just came across yet another way that someone has reported to dehydrate an oxime to a nitrile. (No, I won't link to it. You don't need it. No one needs it).
If asked to count the number of times I have seen new reagents that dehydrate oximes to nitriles, I would be at a total loss to even try to guess. But I've seen it over and over and over. Is it possible that we now have enough ways to do this? And that anyone who is contemplating adding another one to the list should instead go do something else?
I'll vote for that. And there are several other transformations that could go on the same list. That doesn't mean that I think that our existing methods for these are all perfect, or that they couldn't be improved. I mean, even for forming amides, I would like an inexpensive reagent that never fails, even with crappy unreactive hindered coupling partners, works at room temperature in about five minutes, and has a ridiculously simple workup. We don't quite have that, do we? But no one's publishing on coupling reagents like that, because they're rather hard to realize. What we get are a bunch of things that are about as useful as what we have already.
And I agree that it's worth having multiple methods to accomplish the same reaction. I've been saved several times by being able to move down the list and find something that works. But how long should the list be? Eight reagents? Ten? Twenty? At what point should something like this cease to become an acceptable field for human effort?
My first nomination, then, for the Retirement Home for Organic Transformations is aldoxime to nitrile. I am willing to face the rest of my chemistry career with only the monstrously long list of reagent systems we have today for that reaction. Further nominations can be made in the comments - I'll assemble a list for another post.
Last Saturday night I stayed out until 3:30 AM, then slept in the back of our van. Now, that may sound like a pretty good evening for some of you, but it might seem a little odd for a guy like me. There's a good reason, though - I was out at the Connecticut Star Party, a meeting of amateur astronomers out in the boonies of eastern CT. Fall is a good season for those get-togethers - there are a lot of interesting things in the sky, the weather tends to clear out as cold fronts come through (but it's still temperate, overall), and it gets dark at a reasonable hour. Conditions last weekend were about as good as they can get, actually - I won't go into what I observed, unless it turns out that there are a lot of readers to whom phrases like "Minkowski's Footprint" and "G-numbered globulars around M31" mean something.
There were good views of Jupiter, though, and that always reminds me of the lab. I didn't spend much time looking at the planet (it tends to ruin your night vision for a while!), but the colors of the cloud belts are striking: yellow, brown, orange, tan, and (of course) the Great Red Spot, which is sort of a light brick color these days. (That's about the right color there in the photo, although that's a lot higher-resolution than you can see with the naked eye, taken as it was from the Cassini spacecraft on its way to Saturn. The black dot is the shadow of one of the moons, giving anyone in Jupiter's cloud deck a total solar eclipse).
What it reminds me of are the reactions on my bench (and some of those older stored samples), which are turning the same colors. And they're doing that for the same reasons. Jupiter's a gigantic stew of organic chemicals, which are being run through all kinds of temperatures and pressures (including plenty of conditions that are too bizarre to reproduce - so far - on Earth), being irradiated by the Sun and constantly zapped by huge lightning storms. The side reactions in my lab tend to make yellow, orange, red, and brown stuff, and Jupiter is nothing but side reactions.
So what is all that stuff? It's rather hard to characterize it, naturally, but I've always assumed that they're some sort of high-molecular-weight condensation products. (There's been some work done on trying to figure out what the astronomical versions of it, called tholins, must be). There must be a fair number of double bonds and a lot of conjugation in there, to get all those chromophores which push the transmitted light down to the yellow-orange part of the spectrum. All the higher-energy wavelengths of light, the purple/blue/green stuff, are being soaked up. No organic compound in my experience has ever decomposed to anything colored blue. They start by going yellow and then head down through orange and red, towards deep brown and thence to black.
So when I purify these things, and all the colorful stuff sticks to the top of the chromatography column and makes bands of who-knows-what up there, I often glance up at the stuff I'm throwing away, and think "Jupiter". And that's probably accurate.
Let us now praise sulfur. Well, some kinds of sulfur, anyway. The +2 oxidation state (earlier typo fixed, aargh) is a bit hard to handle, what with all those angry-skunk, burnt-tire overtones. But move up the ladder to +4, and you've got some possibilities.
Those do not include the hideous thioacetone, but bring some oxygen into the picture and you get a sulfoxide. And these guys I like, probably because one of the best compounds I've made in my career had one as a prominent feature.
Not every medicinal chemist shares my enthusiasm, that's for sure. Sulfoxides have a reputation for being potentially metabolically unstable - and they can go either way, being oxidized up to sulfones or reduced back to the parent sulfide. (I believe the former is more common, and is the clearance mechanism for DMSO, among other members of the tribe). But there are some out there in the market, chief among them esomeprazole (Nexium). Then there's armodafinil (Nuvigil), Cephalon's follow-up to Provigil, like Nexium another single-enantiomer-of-a-racemate drug.
But sulfoxides aren't just for extending your patent life and raking in the money. They can make a big difference in activity. The group has a strange character to it, because that oxygen atom is about as close to a naked O-minus as you're going to find. And the tetrahedral geometry of the sulfur means that this electronegative group is held is a very specific orientation relative to the other parts of your molecule. Like a nitrile, a sulfoxide is sui generis: there's nothing else that will do what it does.
And they're chiral. That can either be a bug or a feature, depending on your project and on your view of the world. If your target protein recognizes that chirality, it's probably really going to recognize it, because of that strong character. But that chirality is yet another reason that people avoid the sulfoxide, because that means chiral synthesis, which is a pain. All sorts of methods have shown up - chiral oxidation of sulfides, displacement with inversion at the sulfoxide sulfur - but there's no good general solution. The existence of the commercial drugs shows that this problem can be overcome, but there's no use denying that it's a problem.
All these problems can, at times, blend together. I was told some years ago about a Merck clinical compound that had a chiral sulfoxide. When they checked for metabolites, they found what looked like unchanged drug substance coming back out. A closer look, though showed that this was actually the enantiomer of the starting drug! What happened, as I heard it, was that the sulfoxide was first getting reduced, then oxidized back up to the opposite sulfoxide, when then passed out unchanged. Eating your starting material and collecting your own urine has yet to catch on as a sulfoxide inversion method, though. . .
For those who don't work in the industry, and wonder what goes on behind the closed doors of the research buildings, allow me to give you a fly-on-the-wall view of a typical meeting of a drug discovery project team. There are no huge revelations here, and I'm not going to try to reproduce 45 minutes worth of talk, but I think that my industrial readers will find this to be a pretty accurate depiction:
(Camera view of the inside of a small conference room, with six or eight people seated around a table)
CHEMIST A: OK, is this everyone that's going to show up? We have to stop this thing of starting all the meetings fifteen minutes late. (Slide goes up on screen from laptop). All right, here are the two scaffolds, and here's where we were last time with them. You guys should know that at the last Senior Review Meeting everyone kept asking when we were going to narrow down on just one of these, and I kept having to tell them that we're not ready to do it yet. But they're getting tired of hearing that.
CHEMIST B: Not as tired as we are of them asking the question. But I guess you probably didn't say that? OK, I'll do Scaffold 1; my lab's been working on that one the whole time. (New slide goes up). As usual, these things are potent out the wazoo, but we can't shake that Other Enzyme activity, and none of these compounds have the plasma stability that we want.
Last time we said that we were going to hang a bunch of stuff off the 4-position to try to fix that metabolic problem, but we only got a few of the things made. Every time you try to put anything useful out there, you get this side product, and most of the time you can't separate the stuff, you can just see it in the NMR and maybe on the LC/MS trace.
But we've made these four analogs - the potency isn't getting any better, but it isn't getting any worse, and we've put 'em in for PK. If they work, though, we're going to have to find another way to do this stuff.
CHEMIST C: Why don't you try to put in those groups via (obscure name reaction)?
CHEMIST B: Because (obscure name reaction) doesn't flippin' work on this system - we tried that, too, and all we get back is starting material. At least the route we've got gives us something. Sometimes. Sort of.
CHEMIST A: What are we going to do if those come back from PK with the same short half-life?
CHEMIST B: Well. . .work on something else, I guess, because if the problem is out here in the 4-position, you'd think that these changes would fix it. Unless we suddenly made some other part of the molecule more likely to be metabolized by messing with this end of it. But you can assume stuff like that all day, and it doesn't get you anywhere. Keep thinking like that, and you'll never make anything.
CHEMIST A: OK, we'll wait for the numbers. My group's been doing the second scaffold, so I'll take that one. (New slide goes up). These have always been the most selective compounds we've got against That Other Enzyme, and they have pretty good PK numbers, but we keep trying to get more potency. We made this series of amines, trying to pick up a hydrogen bond out there in the far binding pocket, but. . .well, most of them don't seem to work. They're really soluble, though. Every time we make something that's really soluble, it doesn't bind.
BIOLOGIST A: Yeah, those things were nice. Should have known.
CHEMIST A: The outlier is that third one, the piperazine. That looks like it might be picking up something, so we're going to make another series off of that one. What we really need is the piperazine with this funky group on it, and you're supposed to be able to buy it from insert name of fly-by-night supplier, but I don't want to depend on those guys.
CHEMIST D: So how long are we going to keep beating on these things? Have you guys ever made anything that's below, like, fifty or a hundred nanomolar?
CHEMIST A: Well, that thiophene compound was the best, and that's what got us excited, you know, but none of the other aryls seemed to work as well. So we've still got the three-position to try out there, and I think we've got some intermediates that we can use to get some analogs. I don't want to pull the plug until we've made those. And we need to make that piperazine series that's up there.
CHEMIST D: But last time you didn't want to pull the plug until you'd made these compounds. Does the plug ever actually come out, or not?
CHEMIST A: Well, not yet, partly because, hey, when you get down to it, this is probably the best series we have to work on. Nothing else gives us those plasma levels.
CHEMIST B: But there's only so much that blood levels can do for you if the potency isn't there. Would you put a hundred nanomolar compound into the animal model?
BIOLOGIST B: I hope not, because as you guys know, that model is a pain in the neck to run, and we'd rather not spend three or four weeks on it unless you've got something that you think is going to actually work.
CHEMIST C: What if you try to mimic that right-hand part of the first scaffold with some sort of cyclic amine goes to screen and waves hands like over here? Piperidine, pyrrolidine - would that hit the same part? It looks like there's space in the X-ray structure to get over there.
CHEMIST A: You want to try it?
CHEMIST C: Well. . .OK. I'll take a look, see if we can get something like that. You guys have any of the ester left, or did you burn it all up already?
(camera pulls back out of the conference room)
. . .and that's how it goes. In fact, that's almost exactly how it goes, most of the time. That's science as it's being done.
Things are pretty quiet around the industry these days, so my blogging thoughts have been turning to Big General Problems. And here's one that I know that people are working on, but which I think we as chemists are going to have to understand much better: localization.
"Say what?" is the usual response to that, but hear me out. What I mean is the trick that living cells use for their feats of multistep synthesis. Enzymes aren't generally just floating around hoping to bump into things - well, some of them are, but a lot of them are tied to specific regions. They're either membrane-bound, or they're expressed in structures where they don't get a lot of chances to diffuse out into the mix. The interior of a cell, on the whole, is a pretty intensely structured place (as it would have to be).
And that allows specific reactions to take place away from other things that might interfere, which is something that we have a hard time doing in the lab. If you have a five-step synthesis, it's a pretty safe bet that you don't dump the reagents for all five steps into the pot at the same time and hope for the best. No, we generally have to fish out the product and take it on separately. It's often a real achievement (especially on larger scale) to be able to "telescope" two steps into one flask and skip any sort of product isolation between them. Doing it with more than one step is even more rare (and more useful when you can bring it off).
There's been a lot of work on one-pot cascade or domino reaction systems, and that's a step toward what we need. But most of these cases are reaction-driven: people find chemistries that can be run in this fashion, and then try to exploit them to make whatever can be made. Nothing wrong with that, but it would be nice to have product-driven approaches, where you'd look at a particular structure and figure out which multicomponent reaction scheme would work best for it. Generally speaking, we just don't have enough worked-out systems to be able to do that.
And that's where I think that some new technologies could help, specifically flow chemistry and/or microfluidics. Instead of figuring out reactions that can exist while all stirring around together in one pot, this approach takes it as a given that many transformations probably just can't be done that way. And if you can't have one big reactor with multiple things in it, then why not make multiple reactors, each with a different thing in it? Flow systems can, in theory, send compounds through a series of isolated reactions, moving the material physically through various zones and reagents. Not every reaction is perfect of course, but you can often use scavenger reagents along the way to strip out potential interfering impurities before the next step.
I like the idea, but there are a lot of things to be done to make it work. Probably the most advanced organic synthesis that's being done is this style is in Steve Ley's lab at Cambridge. I always enjoy reading their flow papers, which make clear that there's some significant optimization that needs to be done before you can throw the switch and stand back. Some other multistep flow work can be found here and here, and the same comment applies: there's a lot of preparation involved.
My hope is that these kinds of things will eventually move toward more of a plug-and-play system, where you put in the various cartridges and choose a protocol from the list of best-general-fits for your planned reactions. We're quite a ways from that, but I don't see why it wouldn't be possible.
What does it take for a new technology to catch on in the labs? There's an endless stream of candidates (I hope it's endless, anyway), from small gizmos that you can keep in your drawer to multi-hundred-thousand-dollar machines that need their own air handling systems. But all of them start out in the "is this thing any good?" zone, and not all of them emerge, no matter how much they might cost.
That's the first criterion: does the new equipment do anything useful? You'd think that this would have been worked out by, say, the team that developed the product in the first place, but hope does spring eternal. Companies do sometimes get some funny ideas about what their intended markets are clamoring for.
The second test is whether it does its thing in a way that doesn't mess up what you're already doing. "Useful but annoying" is an all-too-well populated category, and if the balance tips too far toward the latter, people will gradually find reasons to stop using the equipment. With some equipment, you start to feel as if you're paying twenty dollars for $20.03 in pennies, putting the whole process into the "not worth the trouble" bin.
Automation is often a factor here. Poorly engineering automation will drive people away like a skunk, of course. Lack of automation won't drive them away, but it won't give them an incentive to come back, either. But do it right, and you lower the perceived cost of using the equipment. Microwave reactors for chemical reactions are a good example of this. The first buckaroos who did these things used kitchen microwave ovens and homebrew reaction vessels. Then there was a generation of reaction carousels that fit into the oven compartment, but that fell into the "annoying" category. The more recent crops of dedicated machines, though, have caught on. They don't look like microwave ovens at all (for example), since the reaction chamber is much smaller (built, in fact, to fit the reaction vials). And they run from a software interface, allowing you to put your tube in the rack, set up your conditions, and walk away.
That phrase "and walk away" is the key idea behind good lab automation. You shouldn't have to stand in front of a machine to make sure that it's going to do what it's supposed to. You can walk away from NMRs, from LC/MS machines, from fraction collectors and many other devices. But if you can't, because the machine hasn't evolved to the point where automation is possible - or worse, if it has automation you can't trust - then the benefit of using the thing had better be substantial.
Lab-scale flow reactors are a good example of equipment that hasn't quite reached the walk-away stage yet (although I have hopes). I know that there are several machines out there that have some ability to do multiple unattended runs, but I'd be interested to know how many users actually manage to leave the things alone while they're doing them. I'm a fan of flow chemistry, but until the machines are more like the microwave reactors, their user base will be confined more to hairy, wild-eyed types like me. The companies in the business seem to realize, though, that my phenotype will not allow them to earn an honest living, and are taking steps.
Earlier this month I posted about rolofylline, which I noted has a rather unusual noradamantane attached to it. Now check out this ORL-1 compound from Banyu, complete with the not-so-widely-heard-of bicycloheptane-spirocyclopropane group.
This was not arrived at lightly, as you'd imagine. There's a table in the Supporting information for the paper, but I'll quote from the body of the main manuscript:
Various kinds of cycloalkanes, substituted or nonsubstituted cyclopropyl rings to medium sized rings (such as cyclopentylmethyl, cyclohexylmethyl, cycloheptylmethyl, cyclooctylmethyl, cyclononylmethyl, cyclodecylmethyl), spiroalkane (such as spiro[2.5]octanemethyl, spiro[3.5]nonanemethyl, spiro[4.5]decanemethyl, spiro[2.4]heptanemethyl, spiro[3.4]octanemethyl, spiro[4.4]nonanemethyl), bicycloheptane (such as methylbicyclo[2.2.1]heptylmethyl, dimethylbicyclo[2.2.1]heptylmethyl, spirocyclopropanebicycloheptanemethyl), and branched alkanes (such as 3,3-dimethylbutane, 3,3-dithylbutane, 1-methylcyclobutaneethyl, 1-methylcyclopentaneethyl, 1-methylcyclohexaneethyl) were tested.
No, that couldn't have been a lot of fun. Anyone else out there found themselves having to optimize grease recently?
Yesterday's post on so-called "ugly" molecules seems to have touched a few nerves. Perhaps I should explain my terms, since ugliness is surely in the eye of the beholder. I'm not talking about particular functional groups as much as I'm talking about the whole package.
First off, a molecule that does what it's supposed to do in vivo is (by my definition) not truly ugly. The whole point of our job as medicinal chemists is to make active compounds - preferably with only the activity that we want - and if that's been accomplished there can be no arguing. Of course, "accomplished" has different meanings at different stages of development. Very roughly, the mileposts (for those of us in discovery research) are:
1. Hitting the target in vitro.
2. Showing selectivity in vitro.
3. Showing blood levels in vivo.
4. Showing activity in vivo.
5. No tox liabilities in vivo.
And these all have their gradations. My point is that if you've made it through these, at least to a reasonable extent, your molecule has already distinguished itself from the herd. The problem is that a lot of structures will fly through the first couple of levels (the in vitro ones), but have properties that will make it much harder for them to get the rest of the way. High molecular weight, notable lack of polarity (high logP), and notable lack of solubility are three of the most important warning signs, and those are what (to me) make an ugly molecule, not some particular functional group.
My belief is that, other things being equal, you should guard against making things that have trouble in these areas. You may well find yourself being forced (by the trends of your project) into one or more of them; that happens all the time, unfortunately. But you shouldn't go there if you don't have to. It's also true that there are molecules that have made it all the way through, that are out there on the market and still have these liabilities. But that shouldn't be taken as a sign that you should go the same route.
Ars longa, vita brevis. There's only so much time and so much money for a given project, and your time is best spent working in the space that has the best chance of delivering a drug. A 650 molecular weight compound with five trifluoromethyl groups is not inhabiting that space. It's not impossible that such a compound will make it, but I think we can all agree that its chances are lower compared to something smaller and less greasy. If the only thing you can get to work is a whopper like that, well, good luck to all concerned. But we have to depend on luck too much already in this business, and there's no reason to bring in more.
I've been involved in another outbreak of the perennial debate about what kinds of compounds medicinal chemists should be making. I can summarize the way this usually goes:
Chemist A: "Look at all these ugly molecules! Why can't we institute some sort of "No-Suzuki-Coupling" rule for two days out of every week or something? Failing that, why doesn't everyone at least try to make things that look better from the start?"
Chemist B: "Nice thought - but the most potent molecules tend to be on the uglier end of the spectrum. And once you've made a single-digit nanomolar compound for the first time in a new project, it's impossible to walk away from it. It's almost like you get to choose: good physical properties, or good activity and selectivity."
Chemist A: "Don't look in these places if you don't want to find what's there. I'm tired of people making big insoluble monster compounds "just for SAR purposes". Because then some of them hit, and you're stuck with 'em."
Chemist B: "But I can't go give a project update and say that we found the most potent compound ever, but we're not going to follow up on it. And then spend the rest of the time telling everyone that we made a whole bunch of compounds with great properties, but hey, they have no activity. That's not going to do me (or anyone else) any good, right?"
Chemist A: "That's why you don't make the uglies in the first place, so you don't get put in that bind. Of course, what everyone says to do is to take that potent ugly compound and make it better, now that you've found it. Problem is, most of the time the things you do to make it "better" start to kill the activity. We'd be better off with fewer hot compounds, as long as the ones we had were decent."
And so it goes. This same debate has gone on in my other workplaces, too, and I believe that it's a general one across the industrial labs. Who's winning the argument at your shop?
I had some requests to answer my own "Random Questions" from the other day, so here goes:
1. Does it bother you, or by contrast make you a bit proud, when you tell someone that you're a chemist and (as happens in about seven out of ten cases) they say "Oh, that was my hardest/least favorite/most boring subject when I was in school"?
Well, whether it bothers me or not, this happens all the time. Like pretty much every chemist in the world, I get to hear all about how people couldn't stand my subject in school. I take the point that mathematicians have it even worse, but it's not like we miss many of them with chemistry, either.
When people ask me what I do, I tell them "drug discovery", and I mention the diseases that I'm working on. That never fails to get some interest, and only then I spring on my listener the (often unexpected) info that this involves chemistry. Coming at it from that angle almost always leads to a good conversation, while coming at it from the "I'm a chemist" angle often leads to "Hey, look at the time!" effects. It's worth doing it in the right order, though - I like the effect when of showing that this boring/hard/useless subject actually leads to what most people find is a really interesting job.
2. How many thousands (10s, 100s of thousands) of dollars of unused equipment is sitting in dusty, unused storerooms at your company, because someone ordered it years ago and either (1) never got it to work, (2) was the only person ever to get it to work, or (3) found that it worked, but what it did wasn't worth doing that way?
Disused equipment? What is this disused equipment you speak of? Never have I seen such a thing. Why, those elaborate combichem machines in the sub-basement, they're just down there because they're so valuable. That rotating split-and-mix thingamabob and the multichannel parallel doohickey, we guard those closely.
Hah! Actually, I remember a couple of labs where this stuff wasn't in the basement at all. No, it was out there in the hoods, taking up space and slowly gathering dust, a standing reproach to everyone who walked past it. It would have been better off out of sight, but no one quite had the heart. And besides, it would sometimes get turned on for visiting groups - there was that.
3. Have you ever set up a reaction and thought "Boy, I sure hope that this doesn't work"?
I suppose that this is somewhat shameful, but yes, I have set up reactions hoping that they would fail. Usually it's been when I've had to use a particularly distasteful reagent (sodium ethanethiolate, for example), and I don't want to end up using it on a larger scale. I remember a fellow grad student presenting his work while we were trying to get our PhDs, and he detailed a deoxygenation step which only worked when his intermediate was made using a hefty excess of thiophosgene. "As fate would have it", said his long-suffering labmate from the back of the room.
And I've had less honorable instances, dating back to grad school or early in my industry career, when I was more or less forced to run a reaction a particular way even though I felt there was no chance for it to work. So yeah, in those cases I did look forward to saying "Yes, I tried your idea. And no, it didn't work any better than mine."
4. For the drug discovery people out there, what per cent of compounds you've made over the years would you guess dissolve in plain water to any real extent? Is that figure going up, or down?
The figure is hard to estimate, but it sure isn't high. Things that dissolve in straight water are hard to work with, y'know - they tend not to extract so well into ethyl acetate or dichloromethane, and they don't run so great on silica when you try to clean them up. That's worth another blog post in itself - the way that our standard chemistry techniques tend to push us away from a lot of polar molecules that might be just what we need for med-chem.
5. What, off the top of your head, would you say in retrospect is the most time-wasting chemistry you've ever ended up doing?
Tough competition. I'm tempted to say vacuum pyrolysis of corn starch to make levoglucosan, but I needed that for my dissertation, so it can't be called useless.
The real winner, in retrospect, has to be a series of reactions I did in my first couple of months in my grad school group, when I was still taking classes and working in the lab part time. I was presented with a route to a tetrahydropyran compound that we needed, a four-or-five stepper that involved an aluminum alkyne opening an epoxide, a Lindlar hydrogenation, a ring closure. . .I can still draw the damn thing on the board, now that I think about it, and it's been twenty-five years ago this spring. Being a first-year grad student, I hopped to it - and hopped right into the mud, since the route bogged down (and how) at the ring closure stage). I kept at it for a while, and then one evening I decided to look up my target compound in Chemical Abstracts.
That wasn't so easy in those stone ax and bearskin days - command-line access to CAS via a rockin' 1200 baud modem and a terminal was still a few months away. I paged through the five-year indices, and found. . .my compound. In a Tetrahedron paper. Two steps, from stuff you could buy from Aldrich, and you form the ring in the first step through a Prins reaction. I was shocked. Surely this couldn't be a known compound. Surely someone must have looked the structure up before coming up with that route I'd been given.
Surely not. And thus did my lab education begin. So you know, when I think about it, even though those first couple of months were a waste of chemicals and effort, perhaps they weren't as much a waste of time as I thought. . .
I didn't note it here when it came out last year, but I wanted to recommend this paper to all the readers who are medicinal chemists. It's an effort by M. Paul Gleeson of GSK to generalize some rules from huge piles of oral dosing data in the company's files. It's all boiled down to a set of charts, for different classes of compounds (neutral, acidic, basic, and zwitterionic), and you can see the effects of changing molecular weight and/or polarity on things like bioavailibility, potential for hERG problems, clearance, etc.
There are no major surprises in the charts. But it's very useful to have all these "rules of thumb" in one spot, and to have them backed up by plenty of data. For experienced medicinal chemists, it's a distillation of everything that we should have been learning. And for those starting out, it's a way to get a fast understanding of what matters when you're making new structures. Check it out!
I've been evaluating an interesting and useful piece of equipment the last few days, and getting a lot of things done with it. At about 8:20 this morning, though, I marched into the lab and proceeded to clog, mis-plumb, and generally absurdify the thing, and I've spent the rest of the day trying to get back to the way things were at 8:15. You know, before I laid my magic hands on the apparatus and gave it the healing touch. Honestly, I couldn't have done a more complete job if I'd been wearing a rainbow wig and honking a horn.
At the moment, all seems to be working, but I've labored under that illusion several times today. If this doesn't do the trick, I'm going to bring in a troupe of Pomeranians and train them to jump over the thing. Sheesh.
Here's something that I'll bet every bench chemist has experienced: thinking that you've quenched some nasty reagent (it has to be gone by now!) only to find that it's very much still with you. These guests that won't leave can be smelly, corrosive, or downright dangerous when they finally yawn, stretch, and decide that it's time to move off the couch.
Alkylaluminum species, in my experience, take their time for longer than you'd think possible, and then depart in a tearing hurry. I used to use several diethylaluminum-X things (cyanide, alkynes, and so on), and was taken by surprise early on by their lackadaisical response to methanol or water at the end of the reaction. "Surely there's some excess aluminum junk in there", I remember thinking the first time this happened, "but there's nothing happening. Maybe I should just squirt in some more." That last phrase has been the prelude to many exciting chemistry moments, and so it was here. Not long after I acted on that impulse, the reagent caught on the fact that it had lots of methanol surrounding it ("Hey, I react with this stuff, don't I?"), and another geyser was born.
Perhaps the king of the "I thought it was hydrolyzed" bunch is phosphorus oxychloride. That stuff takes forever to get around to reacting with water, although on the face of it, you'd imagine it fizzing and sputtering as soon as it got within range. But no, many chemists who've used this reagent have returned to their fume hoods to find the contents of their sep funnels or waste jars gradually coming back from the dead. Milkshake can tell you all about it at Org Prep Daily, and so can many others: never take this one for granted.
There are a lot of ways to think about the chemical reagents that we have stirring around in our flasks. But one of the basic ones, and one of the most useful, divides them into classes according to whether they’re in solution or not.
When things are in solution, they may act funny, but at least everything’s starting out on the same footing. If all the components are dissolved (and if everything’s stirring the way it should), then they all have the chance to find each other and do their respective things. But if some reagent is still a solid in there (powder, chips, what have you), that takes you into the nonintuitive world of surface chemistry.
This actually happens quite a bit. Plenty of standard organic reactions involve insoluble things where the chemistry takes place on the surface. There’s formation of a Grignard reagent from magnesium turnings, deprotonation with powdered sodium hydride, hydrogenation over palladium-on-charcoal – these are all classics. And I'm not even mentioning the surface-driven industrial scale catalyst systems today, which is unfair of me, since the economies of the entire industrialized world depend on them. But in all cases, the real details at the molecular level of these reactions are not easy to work out.
People are still arguing, for example, over just how catalytic hydrogenation works on the metal surface, although the general details of the mechanism are known. That one’s complicated by not just being the plain metal, but a weirdo solution of hydrogen in the metal lattice. There’s no dispute, though, that the reaction is taking place on the surface of the metal, and that the higher the surface area the better off you are.
That’s one big variable right there: surface area. Finely divided substances are very different players in these systems, and many chemists find (early in their lab careers) that they’ve unwittingly bought front-row seats for a demonstration of just how different they can be. Finely divided powders have a lot of surface area in them, and if that’s a rate-limiting factor, you can find yourself with something that’s easily a hundred times more reactive just by picking up a different bottle of what appears (at first glance) to be the same substance. I once saw someone substitute lithium powder for lithium sand in a prep without thinking about this issue, and not so much later, I got to see the same guy clean the inside of his fume hood out with a scrub brush.
But there’s more than just surface area affecting some of these reactions. Grignard formation, for example, seems to take place (at least initially) in fresh breaks or cracks on the magnesium surface. That exposes metal that hasn’t had a chance to become coated with anything (like a layer of magnesium hydroxide), and (zooming in) it also may reveal individual reactive magnesium atoms, left out on the edge and insufficiently surrounded by their teammates. Once these react and fly off into solution, the ones around them become exposed, and so on, and the oxidized layers become undermined and flake off. The standard Grignard-initiation tricks are all designed to speed this process along. A drop of iodine will react quickly with any magnesium points or edges, exposing still more fresh rough surface, as will reaching down under the solution and breaking the turnings with a spatula (or, alternately, grinding them with a heavy stir bar).
These days, what’s really complicating things is the ability to generate (and characterize) nano-sized particles. At some point, these things can stop behaving like tiny bits of the bulk substance (which can be enough of a difference in itself, as mentioned above), and start acting like completely new beasts. And the really nano-sized stuff has a better chance of actually being in solution – but that brings on various headache-inducing thoughts about what “being in solution” means on this scale. If you have clumps of (say) palladium a few dozen atoms wide, which manage to be solvated enough to float around, is that a heterogeneous reaction or a homogeneous one? At that size, is that a "surface", or not (and is the reaction really taking place on it?) What if the nanoparticles are immobilized on a solid support - do they stay and react there, or is the reaction driven by the few that escape? (That effect has been noted in the Heck reaction, among others).
We need to understand these things better than we do - there are surely a lot of very useful things that could be done if we had better control over catalysis and surface chemistry. It's going to keep a lot of people occupied for a very long time.
We use a lot of automated equipment in the drug discovery business. There’s an awful lot of grunt work involved, and in many cases a robot arm is better suited to the task – transferring solutions, especially repetitive transfers of large numbers of samples, is the classic example. High-throughput screening would just not be possible if you had to do it all by hand; my fingers hurt just imagining all the pipetting that would involve.
But I wouldn’t say that the process of medicinal chemistry is at all automated. That’s very much human-driven, and a lot of the compounds on most med-chem projects are made by hand, one at a time. Sure, there are parallel synthesis techniques, plates and resins and multichannel liquid handlers that will let you set up a whole array of reactions at once. But you do that, typically, only after you’ve found a hot compound, and that’s often done the old-fashioned way. (And, of course, there are a lot of reactions that just don’t lend themselves to efficient parallel synthesis).
But I remember the first time I saw an automated synthetic apparatus, back at an ACS meeting in the mid-1980s. There was a video in the presentation (a real rarity back then), and it showed this Zymark arm being run to set up an array of reactions, assay each of them after an overnight run, and report on the one that performed the best. “Holy cow”, I thought, “someone’s invented the mechanical grad student”. Being a grad student at the time, I wasn’t so sure what I thought about that.
This all comes to mind after reading a report over at Wired about a robotic system that has been claimed to have made a discovery without much human input at all. “Adam”, built at Aberystwyth University in Wales, seems to have been set up to look for similarities in yeast genes whose function hadn’t yet been assigned, and then (using a database of possible techniques) set up experiments to test the hypotheses thus generated. The system was also equipped to be able to follow up on its results, and eventually uncovered a new three-gene pathway, which findings were confirmed by hand.
And Ross King, leading the project at Aberystwyth, is apparently extending the idea to drug discovery. Using a system that (inevitably) will be called “Eve”, he plans to:
. . .autonomously design and screen drugs against malaria and schistosomiasis.
"Most drug discovery is already automated," says King, "but there's no intelligence — just brute force." King says Eve will use artificial intelligence to select which compounds to run, rather than just following a list.
Well, I won't take the intelligence comment personally; I know what the guy is trying to say. I’ll be very interested to see how this is going to be implemented, and how it will work out. (I'll get an e-mail off to Prof. King asking for some details). My first thought was that Eve will be slightly ahead of a couple of the less competent people I’ve seen over the course of my career. And if I can say that with a straight face (and now that I think about it, I believe that I can), then there may well be a place for this sort of thing. I’ve long held that jobs which can be done by machines really should be done by machines.
But how is this going to work? The first way I can see running a computational algorithm to design drugs would be some sort of QSAR, and we were just talking about that here the other day – most unfavorably. I can imagine, though, coding in a lot of received wisdom of drug discovery into an expert system – Topliss tree for aryl substituents, switch thiophene for phenyl, move nitrogens around the rings, add a para-fluoro, check both enantiomers, put in a morpholine for solubility, mess with the basicity of your amine nitrogens, no napthyls if you can help it, watch your logD - my med-chem readers will know just the sorts of things I mean.
Now, automating that, along with feedback from the primary and secondary assays, solubility, PK, metabolite ID and so on. . .mix it in with literature-searching capability for similar compounds, some sort of reaction feasibility scoring function, ability to order reagents from the stockroom, analyze the LC/MS and NMR traces versus predictions, weight the next round of analogs according to what the major unmet project goals are. . .well, we're getting to the mechanical medicinal chemist, sure enough. Now, not all of these things are doable right now. In fact. some of them are rather a long way off. But some of them could be done now, and the others, well, they're certainly not impossible.
I'm not planning on being replace any time soon. But the folks cranking out the parallel libraries, the methyl-ethyl-butyl-futile stuff, they might need to look over their shoulders a bit sooner. That's outsourcing if you like - from the US to China and India, and from there to the robots. . .
The comment that showed up recently about unearthing an "original Cable and Wireless dephilostagenator" in a lab reminded me of the huge lab moving job I was in on some years ago. We were packing up the entire company's research site and moving it to another spot in New Jersey (Bloomfield to Kenilworth), and this was supposedly the biggest moving job in the US that year. I do know that the Garden State Parkway was used for the parade of 18-wheel trucks at like 3 AM several times, by special arrangement with the state. (You normally can't take trucks on the thing; that's for the Jersey Turnpike, which doesn't go anywhere real close to Kenilworth).
At any rate, as we started clearing things out, there were several layers of equipment. First were the things that we'd either ordered or had used fairly recently - fine. Behind that, or in the less traveled cabinets, were things that we recognized, but (in many cases) didn't even know that we had. Finally, we began to unearth things that we hardly even knew the names of. I remember finding a dropping mercury electrode apparatus down our way; it's still the only one I've ever seen. It had that solid, black-enameled 1952 look to it, with the name of the company written in silver script lettering on the side, "Dyno-Electromat" or something of the sort. It reminded me somehow of those solid old electromechanical adding machines.
That one was only going to find a home in a museum or in a hazardous waste collection dumpster, and you can guess which alternative won out. But when a site shuts down or moves, there are generally large piles of perfectly usable equipment left sitting around, and it finds its way out into the market one way or another. Courtesy of another commentator, here are some folks from Yale digging through stuff that I might have leaned up against at some point. . .
Since I was talking the other day about getting published procedures to work (or not!), I thought I should mention that most chemists have, at one time or another, had reactions of their own that not even they can get to work right every time. Most chemical reactions are reasonably robust, within limits (see here for a proposal to establish some!) But every so often, you come across one that has a narrow tolerance, sometimes for things that you can’t even put your finger on.
I’ve seen this happen particularly in low-temperature carbanion reactions. Some of these anions don’t particularly want to form in the first place, and they can be quite sensitive to concentration, the presence of different amounts of salts and counterions, variations in temperature, and so on. The rates and efficiencies of cooling and stirring can affect some of these factors, as can the age and handling of the reagents, and the rates at which they’re added into the reaction mixture. If you’ve got a system that just barely works, a lot of things can push it over the edge.
My personal experience with this first came in grad school, when I had a cyanocuprate reagent opening an epoxide. As I mentioned on the blog a few years ago, I tried that system out, after several other reagents had given not-so-great yields, and it worked really well. So I tried it again – same results! I scaled it up (at the time, “scaled it up” meant running it on about a gram), and it worked again. Problem solved! Little did I realize that the reaction would never work again. It failed the next time, and the next, and the next. I tried everything I could think of. I made everything cleaner, I made everything fresh: no product. I made everything sloppy, with no particular care, the way I’d done it in the beginning. No product. Nothing ever worked. I never did sort out what was going wrong; it was easier, in the end, to find another reaction.
Scaling up such a reaction is especially difficult – even relatively laid-back reactions have to be looked at closely when moved up to larger scales, much less a jumpy, skittish one that gets the vapours and passes out at the first sign of trouble. It’s the job of the process chemists to avoid such narrow-window chemistry whenever possible. The idea process reaction is one that provides the same yield, with the same purity profile, under a wide range of conditions: foolproof, in other words. Naturally, nothing is really foolproof (fools are too tricky), but you do what you can.
All the comments on the Lundbeck / Dr. Reddy's imbroglio got me to thinking: how good are patent procedures, anyway? I said in that earlier post that I didn't think that they were that much different from procedures in the open literature, but I'd like to throw the issue open for comment.
You might think that patent procedures would be better, actually. There are potential legal implications to bad patent writeups that don't apply to lousy procedures published in a journal. You're supposed to teach how to make the new chemical matter (or how to do the new process) that you're claiming, and if your patent's details really are insufficient to fulfill that requirement, you have a problem. Patents have been invalidated over such disputes. If you thought your invention worth the trouble of patenting, you'd presumably be motivated to provide sufficient detail to make sure the patent is granted, and that it holds up if challenged.
That said, not all that many patents get seriously challenged over such issues. It takes lot of time and a lot of money, and the number of cases where it's worth the trouble are limited. And a patent has to be pretty lousy (or pretty deceptive) to truly fail to teach what its procedures outline. I guess what I'm asking about is the wide middle ground - the various procedures that aren't necessarily make-or-break for the validity of the patent, but are in there as parts of synthetic schemes. What's your success rate following these? And is it better or worse than your success rate trying to reproduce things out of, say, The Journal of Organic Chemistry?
Time for some lab talk. There are usually a number of different ways to attack a given problem in organic chemistry. You go with what you know, or what looks most likely to work, or what you actually have the equipment (or funds) to realize. This range of choices goes all the way down to what you’d think would be pretty trivial questions, such as: how do I heat up my reaction?
The standard way to do this is to take the usual flask you’d run the thing in at room temperature and dunk it into something hot. That can be an oil bath with a heating coil in it (good temperature control, but messy), a solid heating mantle of ceramic or metal (clean, but doesn’t change temperature so readily), a woven glass heating mantle, a sand bath on a hot plate, what have you.
Then you can go a bit higher-tech, and heat up your reaction with microwaves. I talked about this here a few years ago (and I note that somehow that stretch of blog time has never been archived on this site; I'll have to work that in some time). The early days of the technology featured (first) kitchen models hauled directly into the lab, then carousel devices built to go inside their cooking spaces. But over the years things have settled down to custom-built chemistry microwave setups, walk-up instruments that let you drop a sample tube in, set the temperature and time that you want, and walk away to pick things up later. Microwave heating has become a preferred way to run a lot of palladium-catalyzed reactions.
Does the microwave do anything special other than heat things up, though? That’s been an arguing point for several years. Various “microwave effects” have been proposed, with mechanisms ranging from the unlikely to the pretty believable. In that last category is the thought that when you’re using powdered metal catalysts, that since these absorb microwaves strongly they give some sort of local micro-heating effect that drives the reactions forward.
Could be – but apparently isn’t. A recent paper from Oliver Kappe's lab in Graz, Austria looks at Heck reactions done that way. Kappe is a recognized pioneer and expert in microwave synthesis (see his latest book, linked below), and if you're interested in the field he's well worth reading. In this case, careful experimentation established that the microwave reactions work well because of their heating profile: they get up to temperature very quickly, which seems to be beneficial. But they found no evidence of a specific microwave effect when they ran the reactions under similar heat gradients but with different energy sources.
They also tried this reaction via yet another heating technique, flow chemistry, which I last spoke about here. That turned out to be pretty interesting, too. They were pumping their two starting materials hot over a cartridge of supported palladium-on-carbon catalyst, but found a couple of odd effects. For one thing, the first flow runs tended to give a lot of side reactions, which was surprising considering how clean the conventional runs were. Looking over the system carefully, the team found that the two reactants were separating from each other as they went down the catalyst tube. They couldn’t couple as efficiently because they were pulling away from each other – the alkene coupling partner came out first, while the aryl halide dragged behind, presumably slowed down by interactions with the powdered carbon support.
The other unexpected effect was that even after partially fixing that problem, after a dozen runs or so the reactions weren't working so well. Then the earlier fractions collected and left to sit turned out to be depositing shiny mirrors of palladium metal on the insides of the glass tubes, and all became clear. The Heck reaction was leaching the palladium metal off the solid support! This had been a mechanistic proposal before, but the flow apparatus provides some real evidence to back it up. When you do this in batch mode, via microwave or whatever, the palladium species get a chance to re-absorb onto the carbon as the reaction cools down, and you're none the wiser, but the flow system just washes 'em on through.
What finally did the trick was to add very small amount of the palladium to the starting system, pump that through a hot tube reactor, and use another scavenger column to clean out the metal. You can get away with that in a Heck reaction, since they can run using ridiculously low catalyst loads. I have to say, I hadn't thought so much about this possibility; that's somewhere in between my Type I and Type II flow reactions in my own scheme.
Have you ever worked for a company with its own corporate anthem? It would probably have to be a fairly large outfit; I don’t think a smaller shop would be able to afford such a thing, even if they somehow decided that they needed one. (Here’s some advice: if your small or medium-sized company rolls out its own song, strongly consider hitting the exits if you can. That’s the sort of mindless expenditure that only a behemoth can get away with).
I’ve encountered one of these, in one of my former positions. We were having some big site-wide meeting, and one of the honchos introduced the video clip. There are whole agencies who do these things – they write the songs, hire people to sing them, produce the video, and so on, and the product of one of these bizarre production companies was what we got to see.
And what a sight it was. A perfectly calibrated multiethnic assembly began to belt out our new company song with verve and enthusiasm. There were plenty of solo shots and different camera angles. It was all about dreams and teams, visions and decisions, exceeding and succeeding. The singers grinned, looked confidently up into the future, and joined hands as they got to the chorus. I watched all this with mounting dismay and horror, wanting to clap my hands over my ears, both to shut out the music and to keep my soul from trying to flee my body via my Eustachian tubes.
I don’t think that this was the reaction the song was meant to elicit, but I didn’t seem to be alone. As I left the auditorium with some of my fellow chemists, we speculated on whose idea this anthem might have been, how much it had cost, on whether the firm that produced it was from North Korea or not, and wondered how the experience of listening to it might have affected our lifespan and fertility. One of my group said that there surely must have been better songs available, and suggested that he personally would have been much more motivated by AC/DC’s “Highway to Hell”.
I had to agree; that would have done it for me, too. I started imagining a re-take of the video: the same blue backdrop, one of our executives striding out, giving the camera a manly smile, and saying: “Yes, here at _____ Pharmaceuticals, we truly are on a Highway To Hell. Won’t you join us?” The same happy singers would come streaming out from both sides, swinging into the chorus. . .oh, that would have been much better. And overall, rather more accurate than all that “driven by our vision” stuff, too, now that I think about it.
Medicinal chemists spend a lot of their time exploring and trying to make sense of structure-activity relationships (SARs). We vary our molecules in all kinds of ways, have the biologists run them through the assays, and then sit down to make sense of the results.
And then, like as not, we get up again after a few minutes, shaking our heads. Has anyone out there ever worked on a project where the entire SAR made sense? I’ve always considered it a triumph if even a reasonable majority of the compounds fit into an interpretable pattern. SAR development is a perfect example of things not quite working out the way that they do in textbooks.
The most common surprise when you get your results back, if that phrase “common surprise” makes any sense, is to find that you’ve pushed some trend a bit too far. Methyl was pretty good, ethyl was better, but anything larger drops dead. I don’t count that sort of thing – those are boundary conditions, for the most part, and one of the things you do in a med-chem program is establish the limits under which you can work. But there are still a number of cases where what you thought was a wall turns out to have a secret passage or two hidden in it. You can’t put any para-substituents on that ring, sure. . .unless you have a basic amine over on the other end of the molecule, and then you suddenly can.
I’d say that a lot of these get missed, because after a project’s been running a while, various SAR dogmas get propagated. There are features of the structure space that “everybody knows”, and that few people want to spend their time violating. But it’s worth devoting a small (but real) amount of effort to going back and checking some of these after the lead molecule has evolved a bit, since you can get surprised.
Some projects I’ve worked on have so many conditional clauses of this sort built into their SAR that you wonder whether there are any boundaries at all. This works, unless you have this, but if you have that over there it can be OK, although there is that other compound which didn’t. . .making sense of this stuff can just be impossible. The opposite situation, the fabled Perfectly Additive SAR, is something I’ve never encountered in person, although I’ve heard tales after the fact. That’s the closest we come to the textbooks, where you can mix and match groups and substituents any way you like, predicting as you go from the previous trends just how they’ll come out. I have to think that any time you can do this, that it has to be taking place in a fairly narrow structure space – surely we can always break any trend like this with a little imagination.
Another well-known bit of craziness is the Only Thing That Works There. You’ll have whole series of compounds that have to have a a methyl group at some position, or they’re all dead. Nothing smaller, nothing larger, nothing with a different electronic flavor: it’s methyl or death. (Or fluoro, or a thiazole, or what have you – I’ve probably seen this with methyl more than with other groups, but it can happen all over the place). A sharp SAR is certainly nothing to fear; it’s probably telling you that you really are making good close contacts with the protein target somewhere. But it can be unnerving, and sometimes there’s not a lot of room left on the ledge when you have more than one constraint like this.
Why does all this go on? Multiple binding modes, you have to think. Proteins are flexible beasts, and they've got lots of ways to react to ligands. And it's important never to forget that we can't predict their responses, at least not yet and not very well. And of course, in all this discussion, we've just been considering one target protein. When you think about the other things your molecule might be hitting in cells or in a whole animal, and that the SAR relationships for those off-target things are just as fluid and complicated as for your target, well. . .you can see why medicinal chemistry is not going away anytime soon. Or shouldn't, anyway.
This piece over at Science magazine's "The Gonzo Scientist", brought back some memories. John Bohannon, in the midst of an investigation of truffles, tried an experiment on some party guests: rank a series of five patés according to taste. There were three authentic ones, two fake ones (liverwurst and whipped Spam), and. . .dog food.
He did tell people that dog food was one of the choices. Interestingly, although it ranked last in the taste test, people were no better than chance at identifying it as such. Perhaps they expected it to taste better than it did? But the reason this made me smile was thinking about the usual behavior of scientists and engineers down by the coffee machine. You know what I'm talking about - put anything down there, and people will eat it. It's a standard way of clearing out dessert-like things from home that you don't want around the place; take it to work and it'll disappear.
Well, I saw that put to the test once at a former company of mine. One of the freer spirits down the hall put out a bowl of chocolate-flavored hamster treats and sat back to watch the results. Unlike the dog-food experiment, he did not inform his subjects - but in his defense, he told me that he'd tried one himself, and that although they were somewhat gritty, he'd had worse.
Results? The hamster treats disappeared, of course. I'm just glad he didn't press on with this line of research - and as for me, I made sure never to eat anything left by the coffee machine at that end of the hallway. . .
When I joined the Wonder Drug Factory in late 1997, you still had to buy chemicals by writing down the name and catalog number on a form (and press hard; it was one of those multicolor triplicates). I thought that was pretty primitive then, since at my previous company we’d already gone to electronic ordering (clunky, especially in retrospect, but a lot better than anything involving blue, white, and yellow forms). But to find out where to buy the chemicals you wanted – now that was a challenge.
ChemSources was the usual solution. That was (is, I guess) a large volume containing compounds indexed by name and formula, with the suppliers listed for each. There was a red one for domestic suppliers, and a similar-sized blue book for international ones. And although it came out regularly, it was perforce always out of date. How could it not be? The suppliers changed their catalogs constantly. For that matter, the list of suppliers changed constantly. It wasn’t unusual to look up a compound, find its only commercial source was some little outfit you’d never heard of, and find on tracking them down that they’d gone out of business the previous year.
No one does it that way any more, of course, and good riddance. ChemSources appears to still be in business, and you can even get their bound volumes for your shelves. But why would you do such a thing? Even they offer online searching - well, for a subscription fee. But why would you do that? There are free sources for basically the same information. If you just want data on some compound and where it might turn up, ChemSpider is a good place to look. And if you want supplier information, eMolecules looks like the place to go. Their model is "basic search for free", and if you want pricing, export of data, or integration with your in-house databases, you can sign up for their "plus" service and pay fees.
And that's pretty reasonable, because I get a lot of use out of the free service, myself. I can see prices in my company's in-house ordering software. But I'm not one of the most price-conscious chemical consumers out there, since I'm mostly ordering small quantities of a lot of different things. As long as someone isn't egregiously ripping me off, I'm fine (and that's what our Purchasing department is there to check on, anyway, and don't they just love me over there). One of the things that I enjoy about eMolecules, though, is that they help me figure out what a lot of these little bar-coded vials are. There are a lot of suppliers that will send you ten milligrams of stuff with no real label on the vial, just an eight-digit number or the like, which isn't much help. If you don't label them right then - which often involves loading a CD that they shipped with the vials - you can be puzzled in a few weeks or months when you need the stuff again.
But the eMolecules folks have all these people in their files - Life, ChemDiv, Asinex, Specs, ChemBridge, and the other members of the catalog-number-only club. The search isn't perfect (for one thing, they're missing a fair bit of the corresponding CAS numbers to search by), but it's a lot better than anything else I've come across for free.
I did something in the lab the other day that I hadn’t done in several years: run some preparative TLC plates. I had some small reactions that needed to be cleaned up, and the HPLC systems were all in use, so I thought “Why not?” (I wrote here about the decline of analytical TLC in general in some labs, and I think it's fair to say that the larger-scale prep version has seen an even steeper drop in use over the years).
Prep TLC, for those of you not in the business, is a pretty simple technique. You take a square glass plate that’s been coated with a dry layer of ground silica, a white slurry that for this application is about the grittiness of flour or ground sugar. You then take your mixture of gunk, dissolve it up in a small volume of solvent, and deposit it in a line across the bottom of the plate, an inch or so up from one side and parallel to it. Then you take a large glass container and add some solvent to the bottom of it, and put your plate in so that the streaked line of material is near the bottom. Here's one running.
The solvent soaks into the layer of silica, and after it gets up an inch or so it hits your line of stuff. As it continues to move up, soaking further and further up the glass plate, the different components of the mixture will be carried along at different rates. The compounds that stick to silica gel (for one reason or another) will lag behind, while the ones that don’t will move out into the lead. After an hour or so, the solvent line will be up near the top of the plate, and your mixture will now be spread out across it into a series of bands. (The TLC page at Wikipedia has some useful images of this). Up at the top, running with the solvent, will the the nonpolar stuff that didn’t have anything to slow it down. Right down near the bottom, not far up from your original streak, will be the most polar stuff, especially any basic amines – silica gel is mildly acidic, so the amines will stick to it very tightly indeed. And in between will be the other components, divided out according to how they balanced out the pull of the silica gel support with the attraction of the solvent moving them along. Sometimes you can see them as colored bands on the silica plate, but more often you shine a UV light on the whole plate to see them. The silica we use has an ingredient that makes it fluoresce green under ultraviolet, and our compounds usually show up as dark blue or purple bands against the green. It’s a color combination known to every working synthetic organic chemist.
You can see that picking different solvents for this process can change things a great deal. A weak solvent (like hexane) will allow almost everything to stick to the silica. (A compound has to be mighty greasy to be swept along by just hexane; I doubt if there’s a drug in the business that you’d be able to clean up that way). A standard mix is some proportion of ethyl acetate mixed with hexane. You can go up to straight ethyl acetate, or even further by mixing in methanol or the like. And if you’re desperate, you can go to most any solvent mixture you like – three-solvent brews, toluene, acetonitrile, acetone, whatever works.
So how do you get the things off? By the lowest-tech method you can imagine. You mark the position of the band (or bands) you want, and then take a metal spatula and scrape the silica there off the plate. You them dump that into a flask and stir it with a strong solvent, then filter off the silica and wash it some more to rinse your compound out.
This used to be much more of an everyday technique, but automated column chromatography (same principle, pumped through a tube) has taken over. But prep TLC still has its appeal. Done with skill, it can provide very clean compounds, with quite good recovery. In fact, its low cost and power have made it a favorite technique at places like WuXi, the outsourcing powerhouse in China. I've had several first-hand descriptions of their prep TLC room, with rows of plates being run, marked, and scraped in assembly-line fashion. It's the sort of thing you'd only do in a cheap-labor market, because of the unavoidable hand work involved, but it is effective.
I don't know where WuXi gets its plates, but if you make your own, it's an even cheaper technique (discounting labor costs, naturally). You take up the silica gel powder in water, make a thick, well-mixed slurry out of it, and spread it across a square of glass, shaking and tapping it to get the air bubbles out. Back when I was doing summer undergraduate work, I poured a number of these things, although it's certainly nothing I've had experience with since the first Reagan administration. For all I know, that's how WuXi does it now. Perhaps they've found a low-cost supplier of their own, but the idea of a cheap supplier for a Chinese outsourcing company is an interesting one all by itself. . .
I see that there’s a new biochemistry building at Oxford, written up here in Nature. It was designed by a London architectural firm, Hawkins\Brown (love that backslash, guys, so very modern of you), and according to the article, the design:
”. . .ensures that the 300 researchers working there communicate as much as possible. The traditional layout is reversed: here, labs are on the outside, divided by clear glass walls from the write-up areas, which are open to a vast, five-storey atrium. Everyone is visible. Open staircases clad in warm wood fly across the atrium at odd angles, and each floor hosts a cluster of inviting squashy leather chairs and coffee tables, giving the impression of an upmarket hotel.”
You can judge for yourself here. But as I was reading that, I kept wondering, where have I heard descriptions like this before? Oh yeah, the last time I moved into a new building. Actually, every single time I’ve moved into one, come to think of it. I was part of a gigantic corporate move in 1992 into what was billed as a “high-interaction facility”, which was nothing of the sort. And then at the Wonder Drug Factory, one of the new lab buildings had the whole research area behind a large glass wall; it was the first thing you saw when you came into the place. Unfortunately, since it was full of snazzy equipment, it became part of the standard tour for visitors (the combichem labs were largely abandoned by then), and the people working there sometimes felt like zoo animals. And my current building has the labs all around the outside walls, and a huge atrium in the middle of the building (to what purpose, no one is sure; it’s completely empty).
Most of the Nature article, though, is taken up with the artworks that were commissioned for the new building. I can’t pronounce on these without seeing them all, although the hanging birds display reminds me of a display I saw hanging in a shopping mall in St. Louis in the late 1980s. I do get a bit worried when I hear some artwork described as “rais(ing) questions about how we organize and view the world around us”, since that’s the worst kind of boilerplate artspeak. (Find a large abstract installation you can’t use it on). Another statement about how “if you have a greater degree of visual literacy, you reflect more on both the way you represent things, and also the way that may limit the way you think about them”, falls into the same vaguely depressing category.
“Time will tell if money spent on art gives a significant return in scientific discovery”, is how the article ends up. But how will we know? Set up a control building with no artwork at all, or one furnished only with the Pre-Raphelites? (Full disclosure – I’d rather work in that last one). My guess is that the people who work there everyday will gradually stop seeing the artworks at all; their biggest effect will be on visitors, for what that’s worth.
And as for laboratory building design in general, my suspicion is that there aren’t that many useful general design schemes. Once you’ve fallen into one of those slots, what will matter most for productivity will be the boring details about the size of the benches and hoods, the ease of using shelves and cabinets, the number and location of electrical outlets and sinks, and so on. As for interaction between the scientists, I agree that it does a lot of good: but how to force it? There seems to me to be a tradeoff between convenience and interaction – the most interactive buildings I’ve worked in were the ones that forced me, though a limited number of doors and stairs, to walk down long corridors past a lot of open (and rather cramped) offices and labs. Spread things out, put in a lot of access points, and people just won’t see each other as much.
So here’s the question: I’m sure that many of them can hurt it, but has anyone worked in a building that seemed to help discovery? Examples welcome, and feel free to link to pictures.
A run of bad accident news today, and all of the same kind. The Chemistry Blog has the story of a fatality in the labs at UCLA. The short and painful details are: inexperienced student, t-butyllithium, flammable clothing, and panic (as in not running toward the safety shower).
This is very sad to hear about, and as with so many lab accidents, one of the saddest parts is how easily it could have been prevented. t-BuLi is, of course, a well-known fire starter, and the student did know about that problem. But one of the keys to working with dangerous substances is to think through what you’ll do if something goes wrong. For a pyrophoric compound, that means knowing where the nearest fire extinguisher and safety shower are. It’s very easy to panic when something goes wrong, but if you’ve rehearsed what to do beforehand, you have a much better chance of doing the right thing in tough circumstances.
I pass this along to the students who read this site, and I’m sure the other experienced lab workers here will agree: always think “OK, what’s the worst thing that can go wrong with this reaction?”, and think about what you’ll do if that happens. Fire? Explosion? Sudden leak of nasty toxic stuff? Think it over. Anyone working in a laboratory should always know where the nearest fire extinguisher is. That is, the nearest appropriate one – if you’ve got a separate Class D model for metal fires, or even just a sand bucket, then when you need it you’re really going to need it. And everyone should know where the nearest safety shower is, because no one ever just sort of needs to use one of those. I’ve had to run and pull one once in my career, and let me tell you, it was a damned good thing that I knew where to go when the chips were down.
The other news I have was communicated to me privately, so I won’t go into details other than to say that it appears to be another fatality, this time involving inhalation exposure to trimethylsilyl diazomethane. The problem with these sorts of reagents is that you might think that they’d cause breathing trouble immediately, but you’d be wrong. Diazomethane, phosgene, methyl bromide and others can actually take hours to kill a person, and for a good part of that time, the only symptoms might be a slight cough. But serious lung damage can be coming on slowly during that period, and by the time it’s clear that there’s a problem it’s usually too late to do very much about it. Unfortunately, in some cases, it’s too late right from the start, but that takes quite a bit of exposure, and indicates a serious mistake somewhere along the line.
Anyone who works with such volatile and damaging reagents needs to be completely aware of what they’re doing, and to only handle them under good ventilation. I’ve used such things many, many times in my career, without incident, and so have most working organic chemists. But we should never lose respect for what we’re holding in our hands.
I’m not trying to scare beginning chemists out of doing lab work. It has it hazards, but so does driving to work in the morning or cutting up food for dinner. (When I was in graduate school, my mother once expressed her worries about my lab work, but I told her that the most dangerous thing I did was to drive 650 miles back home on holidays). But every well-appointed chemistry lab is full of death in screw-capped bottles, and that bears thinking about. Random, unforeseeable accidents are, fortunately, very rare. But that means that the others didn’t have to happen, and that’s painful to contemplate.
OK, solid chemistry around here today. It looks as if I'll be running a ring-closing metathesis reaction soon. Nobel Prize in 2005, all over the chemistry journals for years. . .and I've never had occasion to use one until now. And when I think about it, there are quite a few other reactions in that category for me.
For example, I'm not at all sure that I've ever done a directed ortho-metalation. I've come close a few times, and I couldn't absolutely swear that I've never done the reaction, but none come to mind. No Fischer indole synthesis for me, because I've always been able to buy the indoles I need, and the same thing applies (fortunately) to the widely disliked Skraup cyclization for quinolines. I've never done an Eschweiler-Clarke reaction, although there have been several times I've needed to form methyl amines, and it probably would have been a good idea.
I've never done any of those multicomponent condensation reactions (Ugi, Passerini, etc.), partly because I've never done much combinatorial chemistry. And I've never done a Rosenmund reduction, but jeez, in this day and age, who has? No Julia olefination, no Fries rearranagements, no Kolbe-Schmitts (or Kolbe anything, come to think of it). And no Paterno-Buchi reactions, because I haven't really done any photochemistry for about twenty years.
I suppose the biggest gaps in my record are the RCM (soon to be filled) and the directed metalation (unless I can think of one that I've blocked from my memory somehow). Most of the other big ones I've done at one point or another, albeit perhaps only once or twice: the last time I ran a Wurtz coupling, on purpose, anyway, was twice in the summer of 1983. I have run a lot of the obscurities as well, of course (Shapiro elimination, anyone?) And I have a lot of fondness for some lesser-known reactions, such as the Prins cyclization, which got me out of a tight spot in my first year of grad school (I've been grateful ever since).
So here's my chem-geek question for the laboratory-bound part of my readership: what famous reactions have you never done? Have you been avoiding it for some reason, or have you just never needed the thing (and wondered why it's so darn famous)? Confess below!
A colleague came by a while ago and said "You know, the comments to that last post of yours are in danger of turning into Monty Python's Four Yorkshiremen sketch". At the moment, things are running about 50/50 between the "lack of equipment teaches you skills" and "lack of equipment wastes your time" camps. . .
The late Peter Medawar once wrote about resources and funding in research, and pointed out something that he thought did a lot more harm than good: various romantic anecdotes of people making do with ancient equipment, of great discoveries made with castoffs and antiques. While he didn’t deny that these were possible, and admitted that you had to do the best with what you had, he held that (1) this sort of thing was getting harder every year as science advanced, and (2) while it was possible to do good work under these conditions, it surely wasn’t desirable.
His most interesting point was that lack of equipment ends up affecting the way that you think about your research. It’s not like people with insufficient resources sit around all day thinking of experiments that they can’t run and can’t analyze. If you know, in the back of your mind and in your heart, that there’s no way to do certain experiments, then you won’t even think about them. Your brain learns to censor out such things. This limits your ability to work out the consequences of your hypotheses, and could cause you to miss something important.
Imagine, say, that you’re working on some idea that requires you to find very small amounts of different compounds in a final mixture. A good LC/MS machine would seem to be the solution for that, but what if you don’t have access to one? You can spend a lot of time thinking about a workaround, which is mental effort that could (ideally) be better applied elsewhere. And if you had the LC/MS at your disposal, you might be led to start thinking about the fragmentation behavior of your compounds or the like, which could lead you to some new ideas or insights – ones that you wouldn’t have if you’d had to immediately cross off the whole area.
If you’re in a resource-limited situation, then, you’ll probably try to carefully pick out problems that can actually be well addressed with what you have. That’s a good strategy, but it’s not always a possible one. Huge areas of research can be marked off-limits by the lack of key pieces of equipment, and by the time you’ve worked out what’s possible, there may not be anything interesting or important left inside your fence. Medawar’s point was that being stuck inside such a perimeter would not only hurt the way that you did your work, but could eventually do damage to the way that you thought.
It occurs to me that this is similar to George Orwell's claim in "Politics and the English Language" that long exposure to cheap, misleading political rhetoric could damage a person's ability to think clearly. "But if thought corrupts language, language can also corrupt thought". There may be other connections between Orwell's points and scientific thinking. . .definitely a subject for a future post.
In fairness, I should mention that the flip side of this situation isn’t necessarily the best situation, either. Having everything you need at your disposal can make some researchers very productive – and can make others lazy. Everyone has stories of beautifully appointed labs that never seem to turn out anything interesting. There’s danger in that direction, too, but it’s of a different kind. . .
Organic chemisty can be a real high-wire act. If you’re taking a compound along over a multistep sequence, everything has to work, at least to some extent: a twelve-step route to a compound whose last step can’t be made to work isn’t a route to the compound at all. To get the overall yield you multiply all the individual ones, and a zero will naturally take care of everything that came before it.
Even very respectable yields will creep up on you if you have the misfortune to be doing a long enough synthesis. It’s just math – if you have an average 90% yield, which shouldn’t usually be cause for distress, that means that you’re only going to get about 35% of what you theoretically could have after ten steps (0.9 to the tenth). An average 95% yield will run that up to 60% over the same sequence, and there you have one of the biggest reason for the importance of process chemistry groups. Their whole reason to live is to change those numbers, to make sure that they stay that way every time, and without having to do anything crazier than necessary along the way.
When you’re involved in something like this and you know you’re going to be approaching a tricky step, the natural temptation is to try it out on something else first. Model systems, though, can be the road to heartbreak. In the end, there are no perfect models, of anything. If you’re lucky, the conditions you’ve worked out by using your more-easily-available model compound will translate to your precious one. But as was explained to me years ago in grad school, the problem is that if you run your model and it works, you go on to the real system. And if you run your model and it doesn’t work, well. . .you might just go on to the real system anyway, because you’re not sure if your model is a fair one or not. So what’s the point?
This gets to be a real problem in some labs. While ten steps is medium to long for a commercial drug synthesis, it’s just the warmup for a lot of academic ones. Making natural products by total synthesis can take you on up into the twenty- and thirty-step levels, and some go beyond that, most horribly for everyone concerned. In such cases, you’d much rather have several segments of the big honking molecule built separately and then hooked together, rather than run everything in a row.
But what if you spend all that time on the segments, but you can’t put the things together? The most famous example of that I know happened in Nicolaou’s synthesis of Brevetoxin B. The initial disconnection of this terrible molecule into two nearly-as-awful pieces turned out to have been a mistake. Despite repeated attempts, no way could be found to couple the two laboriously prepared pieces to make the whole molecule, and untold man-hours of grad-student and post-doc slave labor had to be ditched for a new approach. If you want to see the approach that worked, here’s a PDF of a talk about it.
But if you go linear, you’re taking the same risk, and the math will absolutely eat you alive. A 90% average yield will ensure that you throw away 95% of your material if you keep going for 28 steps. And keeping a 90% average over twenty-eight steps is just not possible with real-world chemistry, either – and yes, I’ve seen those papers where they do, but I don’t believe them. Do you? Make it 25 steps of average 90%, and three 60% losers, and now you’re down between one and two percent of your material left. Which is no way to live.
I note that the above summary of the Brevetoxin synthesis counts 123 synthetic steps. It calculates an average yield of 91%. A 2004 synthesis from Japan comes to 90 steps with an average yield of 93%.
About a year ago I wrote a post on flow chemistry. That, broadly speaking, is the practice of doing reactions by pumping them through some sort of reaction zone, instead of putting everything into a flask and letting it rip.
There are refinements. In batch mode, you can of course add reagents in sequence, or trickle them in by slow addition. And there are several variations to flow chemistry - in my mind, I have three categories. Type I flow reaction, in my numbering, are the ones that don't depend on any reagents in the tubes themselves. Everything you need is in solution, and you're just using temperature and/or pressure to make them do what you want. Nucleophilic displacements and cycloadditions are in this category: mix up your starting materials, pump 'em down the hollow tube, and get your product out the other end. Ideally.
Type II flow reactions, then, are the ones that need some sort of solid-supported catalyst. Palladium couplings (or other metal-catalyzed processes) are a perfect example of this, as is the H-Cube hydrogenator. Now you have some solid matrix inside your tubing, and you're pumping material over that. Heat and pressure are still very much a part of things, but the catalyst is, too - and the advantage here is that it doesn't end up in your reaction mixture. Starting materials should go in, and product should come out, and you should be able to use the catalyst again. Ideally.
And Type III flow reactions, in my scheme, are the ones that need full equivalents (or more) of solid supported reagent. I think that the companies getting into flow apparatus should keep these in mind. That's because you're going to use these things up, eventually, and the companies involved will be able to sell you more. ("Give 'em the razor and sell 'em the blades", as King Gillette said). All sorts of chemistry might fall into this category - reductive aminations are the first thing that come to mind from a med-chem perspective. All sorts of reactions with nasty workups are candidates for this sort of approach.
But there's a catch, the dirty secret of flow chemistry from my experience so far: you know how we medicinal chemists sometimes have trouble making soluble compounds? Well, brace yourselves when you go with the flow reactors, because you're going to be clogging things up left and right. Any flow apparatus that does not take this into account should be regarded with suspicion: "easy to clean out" is a very desirable quality. Things have to be run more dilute than you think they do, and in stronger solvents. That can mean trouble on the back end, with more (and more difficult) solvents to get rid of in the isolation.
If anyone out there is also involved in the flow world and can talk about it, I'd be glad to hear some experiences. For bench-scale medicinal chemistry, the field is still in its early days, and there are lot of things that haven't been tried yet.
I’ve noticed over the years that my patience in seminars and talks has been eroding. This started in graduate school – I certainly sat through my share of lousy talks back then, but I was starting to skip out on the occasional one, after a certain level of grimness was reached.
For example, I remember walking down the hall with a new post-doc, when the building’s speakers came to life. “May I have your attention, please. . . “ We stopped to listen. “There will be a seminar in the main auditorium in ten minutes, entitled “Raman Spectroscopy of Synthetic Asphalt Roofing Materials” (I swear that this is a real title, or something very close; it was appalling). The new guy asked, in a slightly worried tone “Do you guys in the group usually go to these things?”
And at that point, one of my fellow group members came lurching out into the hallway, pantomiming elaborate choking gestures as he pointed desperately at the speaker up on the wall, slumping against the wall as the horror of the seminar’s title overcame him completely. We watched him slide to the floor, still gesturing at the intercom, and I said calmly: “No, we skip a few of them now and then”.
Well, over the years I’ve continued to skip a few of them now and then, and my threshold has been steadily creeping up. I realize that many of the topics that keep me glued to my seat are, by any objective standard, rather dry. Give a detailed talk about enantioselective hydrogenation, the thermodynamics of multivalent binding, or even the latest thinking about the patent office’s requirements for obviousness rejections, and I’ll be right there, practically munching popcorn. To me, those things are interesting. But plenty of things aren’t.
It’s to the point now where there are single phrases that give me that “late for the door” feeling. After that hits, it’s a major effort for me to stay in my seat. So, speakers, if you see me out in the audience and think that the ambience would be improved without me, it isn’t hard. Just spend a few minutes going on about “cross-functional goal setting” or the wonders of ISO nine-thousand-whatever. I’ll spray gravel on my way out. One day I’ll probably end up dangling from a bunch of knotted tablecloths, having rappelled down the side of my building from an upper-floor conference room. “Vision statement”, I’ll gasp to the passers-by as I drop to the sidewalk in relief. “They invited me to work on a new vision statement. . .”
It’s worth examining your own scientific prejudices and biases from time to time, to see if they’re still valid. Of course, that begins with the difficult task of figuring out what they are – it’s hard to think of these things when you need them. So I try to make note of my presuppositions when I find myself acting by them, flagging them for later review.
One of these that’s come up recently is the bias that I (and many other medicinal chemists) have against symmetric compounds. (By that I mean palindromic compounds with a mirror-plane right down the middle of their structures). We tend not to make such compounds; we downgrade screening hits with that look to them, and if we start to work on one the first things we do is to desymmetrize it and see if it gets any better. Why?
I think that one reason must be that there aren’t many truly symmetric binding sites out there. Proteins, while they can have large-scale symmetric structures, are usually pretty twisty and heterogeneous on the scale of a drug-sized molecule. Even in the cases of real protein symmetry (a dimer of two identical subunits, say), your compound would have to be fitting into some very select spaces to be feeling that symmetrical environment perfectly.
So a symmetric drug molecule feels wrong, somehow unoptimized. But there’s no reason that its two seemingly identical ends have to be doing the same thing on each side. They could easily be binding to completely different residues, or in different ways – it’s worth remembering that the symmetric structure we draw on the board may not have much in common with the molecule’s real 3-D conformation: a few zigs and zags in the rotatable bonds, and things aren’t as balanced as they looked.
Perhaps we shouldn’t be so hard on these structures. I’ve crossed several of them off my lists over the years, but I think from now on I’ll give them more of a chance. Anyone with me?
We try to be delicate when we synthesize our molecules – really, we do. Delicate reactions often have better yields and fewer side products. Exotic catalysts in perfectly tuned metal coupling reactions – these things are wonderful when they work, because you go from pure starting material to darn near pure product.
But life in the drug labs is not always thus. We also have to turn back the clock, and break out reactions that our grandfathers would have recognized – dark, fuming things that will eat a hole in your lab coat. Nitration is one of these – good old nitric acid is still very much around for that reaction, often in vile mixtures with sulfuric and the like. It’s cheap, and it often works, so you can’t get away from it. And if 1:1 nitric/sulfuric won’t perforate your clothing, you must have put on armor instead of Armani.
Chlorosulfonic acid is another such reagent. It’s nasty by anyone’s standards, but it’ll stick a chlorosulfonyl group onto an activated aromatic ring in one step, which is nothing to take lightly. You don’t want to pour that into water to work it up, not unless you want to see it splatter all over your hood. Nope, you’ll need a trip to the ice machine – slow drizzling over crushed ice is the traditional workup, for good reason.
That’s a good acid for another brute-force reaction that we still have with us: the Friedel-Crafts. Fancier ways exist to acylate an aromatic ring – those metal-catalyzed ones, for example, often in the presence of carbon monoxide. But who wants to use CO if you don’t have to? And you need a leaving group where the acyl group is going to go. The Friedel-Crafts will just come in and jam one in on an unsubstituted carbon, if the electronics of the ring are right. And all you need to do is treat your molecule with some hammer-of-the-gods reagent like chlorosulfonic acid, polyphosphoric acid (which looks and acts like honey from Hell), or straight aluminum chloride powder. That last one is at least a solid, albeit a corrosive one, but you pay the toll during the workup. That’s when it hydrolyzes to piles of white aluminum oxide junk, often turning your reaction into a thick mess.
So no, it’s not all twenty-first chemistry, all the time. World War I-era chemistry is still very much with us at times. Actually, I sort of like it that way. When I have to break out the polyphosphoric acid, the powdered iron, or the elemental bromine, I feel as if I’m keeping faith with my predecessors. They wouldn’t know what to make of the LC/mass spec machine, but they’d grin when they saw me trying to work up my aluminum chloride reactions.
For the most part, the biologists on a drug discovery project expect us in the med-chem labs to be able to make pretty much anything we need to make. Actually, I don’t have to go that far – the other chemists more or less expect that, too. Chemistry’s a big field, with a lot of reactions and techniques, and if you want some particular structure badly enough, there are usually ways to get to it.
But not always, and not always by routes that you’re willing to put up with. That’s especially true early in a project when you need some robust chemistry to turn out a lot of diverse analogs quickly, so you can have some idea of which parts of the molecule are most important. Synthetic trouble at this stage is frustrating for everyone involved.
I was on a project a few years ago that ran into this exact problem. Compounding the pain was the way the lead compound looked when it was up on a screen during a meeting: small, perfectly reasonable, easy to deal with. Hah! It was a werewolf, that thing. None of the ideas that we had ever worked out the first time, and many of them never worked out the last time, either. Meeting after meeting would take the same format when there were outside managers or other chemists present: “But why don’t you just. . .” “We did. It doesn’t work.” “But then you should try. . .” “We know. We tried that. It doesn’t work.” “Well, OK, but then you could always come around and. . .” “We could. If it worked. But it doesn’t.”
New chemists would be added on to the program to try to get things moving, and they’d always come in rolling up their sleeves, muttering “Do I have to do everything myself around here. . .” How do I know? Because I was one of them. Within a month or two, though, I was in the same shape as everyone else on the project, looking at a bunch of NMRs and mass spec traces and trying to figure out what went wrong. Meanwhile, helpful folks would wander past the whiteboard and ask me how come we hadn’t tried the reaction that had just failed for the eleventh time. Eventually we learned to offer the more persistent questioners a supply of our starting material so they could solve the problem themselves and be heroic, but nothing ever came of that.
The project managed to stagger to a clinical candidate, but ran into mechanistic problems in the more advanced animal models. (That was really the hot fudge topping on the whole sundae – this was one of those therapeutic areas whose definitive animal models were too complex and costly to run until you were absolutely sure you had The Compound). I haven’t run into one quite like this since, and with any luck, I never will.
I did carbohydrate chemistry for my PhD - well, I used carbohydrates as starting materials to make other molecules, but I did my share of pure carbohydrate stuff along the way. And although that was over twenty years ago, the stuff I did is still considered by most people to be a sort of esoteric thing, an odd specialty that not many people have experience with. Time has clearly not mainstreamed sugar chemistry.
It's not like people don't use the things, often for just the reasons that I used to (as versatile chiral starting materials). But the reputation of the compounds lingers. I think it's because of all the odd little reactions that sugars do. There's a certain amount of knowledge that has to be learned - all that stuff with the anomeric center, for starters, and all the name reactions that only occur in sugars, like the Ferrier rearrangement.
Then there are the protecting groups. With all those hydroxys hanging around, a lot of them are going to have to be tied up for extended periods while your work gets done. But every hydroxy group on a sugar ring has a slightly different personality - they acylate and deacylate in a particular order, for one thing, which varies from one sugar system to another. And there are the acetals and ketals to tie up two hydroxyls at once - very useful, but there are a lot of different combinations that can form under different conditions and with different carbonyl reactants.
The closest analog to the field that I can think of is steroid chemistry. In its day, that was a hugely popular and important field, with all sorts of ins and outs - tricky transformations that you learned from the old hands. But these days, hardly anyone cares - pure steroid chemistry is a backwater, and many of the esoteric reactions are largely forgotten. Sugar chemistry has escaped that fate - it's still relevant - but hasn't escaped the atmosphere of an eccentric club.
My own sugar knowledge, while still sound, is not exactly up to date. I know that the field has moved on over the years, but I've had only sporadic need to keep up, since carbohydrates don't appear in many drug structures. I've been able to work in some of them once in a while, but I've never worked on a project where my sugar experience has been front and center.
You need access to vacuum if you’re going to work at the bench in chemistry. In fact, you need more than one kind. Reasonably hard vacuum (well, by our standards, which is laughable by the standards of the physicists) is down in the single Torr or below – that is, less than about 1% of normal air pressure. We use that for pulling out residues of water or organic solvents from our compounds. You can’t usually see it happening from the solid ones, but the syrupy liquids will foam up or blow a long series of thick bubbles when the vacuum is applied. The foam can be an irritating problem at times; some things will fill your flask with sticky bubbles and go right on up into the vacuum line if you’re not watching them.
The lesser vacuum lines are used for bulk evaporation of solvent (on your rotavap) and for filtering things off. We do an awful lot of both of those, too, and a full vacuum-pump pull is too vigorous for them in most cases. Evaporating down reactions is a constant task in an organic chemistry lab; I’d rather not think about how much of it I’ve done over the years. As for filtration, there are many cases where a solid product can be filtered out of the bulk liquid (which is good) or where some undesired solid by-product has to be filtered out before you can go on (not as good).
The low-tech way to get the sort of pull-it-though vacuum you need for these things is a water aspirator. You don’t see these as much any more, and you don’t see them at all in industry, since they necessarily pull solvent vapors into the water stream. But they work. An aspirator is basically a narrowing tube that hooks up to a hard-spraying water tap and has a sidearm fitting. The accelerating blast of water pulls the air in the tube along with it as it goes, creating a useful vacuum. If you wanted to make one rather more environmentally friendly, you’d keep a well-stocked dry ice condenser in line with it to trap out the solvent vapors before they go down the drain (which is what your rota-vap should have on it, anyway), but even with that, you’re always going to be turning the water flow into a waste stream. As I say, you don’t see them as much these days.
But we used them back when I was in grad school, that’s for sure, mostly for the rotavaps. If you wanted to keep things from splashing around back in your hood, you attached some rubber tubing to the other end of the thing and ran it further down the drain a bit.
Well, one day, one of the guys in the lab next door to me was shocked to see water blasting around in his hood. It was a real fountain, just geysering out full blast from what must have been a cracked water line or something in the back. He ran over and immediately shut off every tap – but to no avail. Roaring, showering water everywhere. Getting a look at the source, he realized, to his consternation, that the water was coming up out of the drain in the back of his hood. I remember standing there with him, staring at this in disbelief. It looked like a special effect. How on earth could you get water blasting up out of a drain pipe?
Suddenly it hit me. I ran around to the other side of the lab, where a new Japanese post-doc had taken up residence. “Masa”, I asked him, “Did you just put that rota-vap in your hood today?” “Yes, yes, just started it today”. There was a water aspirator flooshing away back in the back of his hood. “Did you put some rubber tubing on that thing?” “Tubing? Oh, yes” “How much?!” “Whoaaa. . .” He spread his arms to indicate the mighty extent of the rubber tubing he’d added.
Mighty, indeed. He’d run the stuff down his drain, through a horizontal pipe and right through a T joint, and back up out of the drain of the other guy’s hood, which backed on to his. So when he turned his water on full throttle, he immediately started irrigating his labmate’s space. We finally go thing turned off, and trimmed back the rubber tubing to a more reasonable length (like, not seven feet), and order was restored. For a while.
Note: if you want to see How Not To Do It to a really expensive vacuum rig, try here.
I wanted to recommend this post by Milkshake over at Org Prep Daily (and not just because he liked the recent column I wrote for Chemistry World). I was writing about the limited number of reactions that some med-chem labs get locked into, and the effect of this both on the compounds that get made, and on the motivation of the chemists. Milkshake has a good set of recommendations on how to avoid the boredom trap, and I recommend checking them out. He ends with the following:
You should care about the chemistry methodology and do things not just to crank out the final compounds to fill up the testing queue. Your boss (has) perhaps lost all his chemistry interest already and maybe he is unnerved about the project progress and pushes people hard - but while you try not to get fired you don’t necessarily want to think like your boss (and end up wretched). If you continue to look at your research project with curiosity and do things also for the sake of your chemistry interest you are likely to be more original because thinking about the methodology will suggest new directions in your medchem project. You may get accused of playing with chemistry and going off-tangent but you will likely remain more content and productive. . .
And this is all true. Most projects need some oddball compounds thrown into them, to keep things interesting (and honest), and it’s the people who are keeping up with the literature who will probably make them. I went through a period some years ago when I didn’t stay current with the journals very well, and if I’d let that continue to slide, it would have had a bad effect. (RSS was one of the things that saved me!)
But there’s another very good reason to stay sharp and run the unusual reactions, though: the boring reactions are increasingly going to be shipped to someone else, someone who probably works in a very different time zone. Yep, this is my “give ‘em something they can’t get in Shanghai” talk again. The outsourcing shops are there to pound out molecules as quickly as possible, and they’re going to use well-established chemistry as much as they can. Now, that’s the same pressure that operates in most med-chem projects, but I strongly recommend differentiating yourself if possible.
Be the person who runs the new stuff, who reads the literature and adopts things quickly, and who makes compounds that aren’t like all the stuff that’s already in the screening deck. You don’t have to go completely crazy, you know. There are plenty of good, reasonable structures that no one else is making at your company – have no doubt – and if you’re the person who makes them and who introduces new chemistry into the department, you have something with which to justify your salary (or a higher one!) On the other hand, if you’re the person who cranks out the sulfonamide libraries, well. . .they can get that cheaper somewhere else, you know.
1. “Hey, who dropped that condenser out on the floor in front of my hood? That looks just like the one I had on my reaction flask. . .”
2. “How come the toxicology people haven’t called me about our lead compound yet? Two-week tox finished a while ago, and usually they’re a lot faster than this. . . “
3. “Is there any active aluminum compound left in this reaction or what? I keep dripping methanol into it to quench it, and nothing’s going on at all so far. . .”
4. “Who’s going to scale up our candidate compound, anyway? We need 300 grams of the stuff, and the scale-up group is booked solid. . .”
5. “So, is this the high-pressure hydrogen line or the low-pressure one that I’m opening?”
6. “I wonder what the error bars are on that behavioral assay. . .”
One of the things that no one realizes about research (until they’ve done some) is how much time can be spent going back over things. Right now I’m fighting some experiments that should be working, have worked in the past, but have (for some reason) decided not to work at the moment. Irritating stuff. There’s a reason buried in there somewhere, and when I find it things will be that much more robust in the future, but I’d hoped that they were that solid already.
And across the hall, a check is going on of some screening hits. When you get a pile of fresh high-throughput screening data, including some fine-looking starting compounds for a new project, what do you do with it? Well, if you have some experience, the first thing you do is order up fresh samples of all the things you could possibly be interested in, and check every single one of them to make sure that they actually are what they say on the label. Don’t start any major efforts until this is finished.
In fact, you should order up solid samples from the archives along with some of the DMSO stock solution that they used in the screening assay. They might not be the same, not any more. False negatives and false positives are waiting in your data set, depend on it: compounds that should have hit, but didn’t because they decomposed in solution, and compounds that (sad to say) did hit only because they decomposed in solution. You’ll probably never know about the first group, and you can waste large amounts of time on the second unless you check them now.
Getting a project going, then, can seem like trying to get a dozen nine-year-olds into a van for a long trip. Someone’s always popping out again, having forgotten something, which reminds someone else, and your scheduled departure time arrives with everyone running in circles around the driveway.
But nine-year olds can eventually be corralled, as can the variables in most scientific projects. But not always. Where you don’t want to be is the situation people had with the early vacuum-tube computers. Vacuum tubes have not-insignificant failure rates. So if you have, say, twenty thousand of the little gizmos in your ENIAC or whatever, doing the math on mean-time-between-failures shows you that the thing can run for maybe forty-five minutes before blowing a tube (unless you take heroic measures). And the more vacuum tubes you have, the worse the problem gets: make your computer big enough, and it’ll blow right after you throw the switch, every time.
So that’s the other thing you have to watch when troubleshooting: try to make sure that your problems aren’t built into the very structure of what you’re trying to do. In med-chem projects, look out for statements like “we have all the activity we need, now we just need to get past the blood-brain barrier”. Sometimes there’s a way out of those tight spots, but too often the properties that (for example) could get your compound into the brain are just flat incompatible with the ones that gave you that activity in your assay. You’d have been better off approaching that combination the other way around, and better off realizing that months ago. . .
Here’s an appropriate topic for a Friday, although at first many of you may think I’ve lost my mind. What would happen if you combed the full text of the experimental sections of the chemistry journals, looking for how long people ran their reactions?
I’m pretty sure that I know what you’d see: there would be a lot of scatter in the short time periods, with some peaks at the various half-hour and hour marks just for convenience. But as you went out into the multiple-hour procedures, I feel sure that you’d see pronounced spikes in the data at around sixteen to twenty hours and again at around 72 hours.
Some readers have doubtless started nodding their heads, having done the math. Those times correspond to "overnight" and "over the weekend", and I'm willing to bet that they're over-represented (and how) in the data set. I'll go on to predict scarce examples in, say, the 14-hour or 38-hour ranges - there's not much way to run a reaction for those intervals and not be in the lab too early in the morning or too late at night.
A second-order prediction is that when such reactions are found, that their origins will skew heavily toward academia rather than industry. And I'm also willing to bet that patent procedures will tend to follow the working-day timelines more than the general literature, for the same reasons. My last higher-order prediction is that the reaction times would not, in fact, obey Benford's Law, as many other data sets of this kind do.
As far as I know, no one's ever done this sort of analysis, but I suppose it would be possible, especially for someone at Chemical Abstracts or at one of the scientific publishers. If someone wants to try it, please let me know what comes out. And if the results follow my predictions, please feel free to refer to the title of this post or something similar. I won't object.
Here’s a question that came up in a discussion at work the other day: when a new head of research comes in, how long should you give them before judging how they’re doing?
That’s a tough one to answer, I think, because there are a lot of variables. First is the size of the outfit, coupled with the scope of the position. A really big organization is a very, very hard thing to change, no matter how powerful the new person might be. I’m not at all sure how possible it is to change a company’s culture, but I’m pretty sure that it requires major shock therapy to do it. (If any of you have read C. N. Parkinson on what he calls “injelititis”, you’ll know the sort of thing I have in mind).
And different levels of authority affect processes with different timelines. A head of chemistry will be able to show results in less time than a head of research, who will need less time than a head of total R&D, because that person has to wait for the clinical results. As I’ve mentioned before, that seems to me to be one of the biggest challenges in this industry – the way that big changes can take years to work their way through to the results stage. It’s hard to steer intelligently if the front tires respond ten miles after you cut the wheel over hard.
You also have to ask what the new person is being asked to do. Steer the course on something that already seems to be working? Or shake the place up and make things happen (for once)? Expand the workforce, contract it, spend money or save it, stick with the existing therapeutic areas or branch into new ones? The job descriptions on these things are pretty wide-ranging, so the evaluations have to be, too. Without a clear idea of what the new boss is trying to do, it’s impossible to say how well it’s being done. You could wind up giving bozos credit for something that had nothing to do with them, or blame excellent managers for things that were completely out of their abilities to control. (I know, I know, that kind of thing happens all the time, but you don’t have to add to it if you can help it).
So, how long for an evaluation, then? One to three years for head of chemistry, five or six for head of research, up to ten for head of R&D (if they last that long?) I'd be interested in hearing other estimates. . .
A colleague and I were talking the other day about some of the molecules that turn up when you dig through a company's internal database. This was a favorite sport of mine during slow afternoons at the Wonder Drug Factory - I would put in a query for bizarre or unlikely chemical groups and see what fell out. I was rarely disappointed - eventually I assembled a folder of the most hideous examples, which never failed to astound.
The compound collection at my current employer isn't nearly so weird, fortunately. But every drug company has large lists of compounds that aren't so attractive as leads, because they were made in the last stages of previous projects. This is a well-known problem, often referred to as a gap between "drug-likeness" and "lead-likeness". For the most part, the compounds you start a project with don't get smaller - they get bigger, as people hang more things off of them to get more potency, selectivity, or what have you. So you're better off starting as small as you feasibly can, giving you room for this to occur without taking you off into the territory of too-huge-to-ever-work. (That's one of the fundamental ideas behind the vogue for fragment-based drug discovery, for example).
"Too-huge-to-work" is a real category, as my industry readers will gladly verify. I think that the "Rule of Five" cutoffs have been sometimes applied a little too mindlessly, but there's no getting around the fact that if your latest molecule weighs 750 and has thirteen nitrogen atoms in it, the odds of it being a drug are rather slim. As my colleague put it, when you make something like that and send it in for testing, what you're saying is "I know that almost every molecule that looks like this fails. But I'm different. I feel lucky". And that's no way to run a research program. Given finite time and finite money, you're better off prospecting in chemical areas with better chances.
So what to do? We kicked around the idea of setting up some filters in the compound registration system itself - if someone tries to send in some horrible battle cruiser of a molecule, the system would make a puking noise or something and refuse to register the compound at all. There would have to be be some sort of override (perhaps for a higher-level manager to authorize) for those times when you actually have evidence that the ugly molecule works, but maybe the "You Lose: Make Something Else" screen would focus attention on the properties of what's being made. Of course, if anyone ever implemented this, the arguing would begin about where to draw the line (maybe there'd be a yellow "warning zone" in between), but I think that everyone agrees that at some point a line should be drawn.
So, for my readers around the industry - do you have such a cutoff? Can you register any crazy compound that crosses your bench, or does your company's software fight back? If so, what's the feedback - beep, e-mail warning, electric shock? Inquiring minds want to know.
I was talking with a colleague recently about the different cultures that have grown up in different drug companies where lab associates are concerned. For those outside the industry, those are non-PhD-holding scientists, who (for the most part) do not move into managerial positions. There's room for a whole separate blog post on the people who (for one reason or another) never got the PhD degree but are the equal or superior of anyone who has, but for now I'm talking about the rest of the associate population.
As people get more experienced, they become more valuable, or at least they should. An experienced chemistry lab associate is one of the most readily employable people in the industry, under normal conditions. A company may or may not feel a need for another twenty-year middle manager type, but there's always a need for hands at the bench to make compounds, and good associates are the people who make the most. And with some time in the industry, they have a far better understanding of the real world of drug discovery than any PhD coming in fresh out of their post-doc.
Or at least they should. There are, though, some companies that treat their associates more like draft animals, putting them in the position I held in the summer of 1979 when I worked for in a greeting card factory before going to college. I was a "materials transport handler", which meant "See that big pile of stuff here? Haul it over there." It's the only time I've done manual labor for money for more than an afternoon, when I think about it. But I'm told that there are shops in this industry that tell their associates exactly what to do at every turn, up to the point (so I hear) of having them take spectral data and turn it over to their supervisors rather than interpret it themselves.
That's something you associate with the old-style German and Swiss labs, where there's a clear heirarchic division between the PhD holders in their offices and the "laboranten" out in front of the hood. Even there, I don't think this is quite as rigid as it used to be, so the thought of this here in the US is quite odd. But it does seem to go on, so I'm asking the readership: what's the status of the usual lab associate where you work?
We recently encountered a problem that’s (unfortunately) a rather common one. An enzyme assay turned up an interesting hit compound, with some characteristics that we were hoping to see for leads against our target. A re-test showed that yes, the activity appeared to be real, which was interesting, since this hit was a welcome surprise from a class of compounds that we weren’t expecting much from.
It was a comparatively old compound in the files, and all we could find out was that it had been purchased rather than made in house. Looking around, it seemed that there were very few literature references to things of this type, and only one commercial source: the Sigma-Aldrich Library of Rare chemicals, known as SALOR. That, though, was a potential warning flag.
Those compounds come from an effort started by Aldrich’s Alfred Bader many years ago, who started trolling around various academic labs looking for unusual compounds that no one wanted to keep around any more. Over time the company has accumulated a horde of oddities that are often found nowhere else, but there are several catches. For one, these things are usually available only in small quantities, tens of milligrams for the most part. That’s plenty for the screening files, but you’re not going to make a bunch of analogs starting from what comes out of a SALOR vial. Another catch is that the compounds are sold, very explicitly, as is: the university sources tell Aldrich what’s on the label, so that’s what they sell you and caveat emptor all the way, dude.
So often as not, you get what we got, a nice-looking white powder which, on closer analysis, turned out to only have a vague relationship to the structure on its label. We knew that we were in trouble as soon as the first NMR came out: way too much stuff in one region, nowhere near enough in some others. Mass spec confirmed that this thing weighed more than twice as much as what it was supposed to. We’ve since pretty much nailed down what the stuff really is, and our interest in it has decreased as each of the veils has been removed from the real structure.
We’re correcting the data in our own screening files, of course. And yes, we’re going to tell the folks at Aldrich to change their label, too, assuming they have any of this stuff left. At least the next person will know what they’re getting. For once. But there are more of these things waiting out there – in every large compound collection, in every catalog, in every collection of data are mistakes. Watch for them.
Well, while the mail continues to come in about my post yesterday, I’m going to pull back from the global perspective and zoom back into the glassware drawers of my lab bench today. A while back I wrote about the different sizes of ground glass joints that organic chemists typically use. People from outside the field are sometimes struck by the fact that we don’t have to do as much glassblowing and the like as they might have thought. Decades ago there was a lot more, but for a long time now we’ve been able to build up all sorts of apparatus (apparati?) by connecting standardized glass fittings together.
This has all sorts of advantages, letting us assemble odd custom configurations pretty easily, and change them without too much work. The downside is that the ground glass joints aren’t by themselves vacuum tight – not by the standards of inorganic chemists, for sure – and need to be anointed with thick, nasty vacuum grease before they can be trusted to that level. And if you don’t grease them for normal work, which we tend not to because the grease gets into your compounds, then the joints tend to freeze if left too long or too tight.
There are all sorts of voodoo tricks for unsticking them. I pride myself on being able to do it, but (objectively) I don’t think my success rate is all that greater than the norm. For the record, my technique is to put a few drops of silicon bath oil up around the edge of the stuck connection and let it soak in for a few hours. Then I rapidly heat the outside joint, grab it with a towel, and do the usual pulling and tapping while hoping for the best. There are better ways, but they're typically found only in a glassblowing shop.
When I last wrote about this fascinating subject (hey, chemists like their glassware), I mentioned that I’d gotten in the habit of using 29/42 size joints. (That’s a measure of size: the first number is a diameter, and the second is the length or taper). That’s a larger one than is common in American labs; you see it more in Germany, among other places. I’m so used to it now that the standard 24/40 glass joints you see all over the place look narrow and shrunken to me – will I really be able to get my product out of that?
The standard small size these days is 14/20 – that’s the size of all our 5, 10, and 25 milliter flasks. (You can get 100 mL flasks (or larger) with that size joint, too, but they start to look disproportionate and weird, and there’s no real reason for large flasks to have such a small neck). In between that and good ol’ 24/40, though is the 19/22 size, which I really should look at again. It would be the wide-mouth counterpart to 14/20, in the same way that 29/42 is to 24/40. I’d probably like it.
But I’ve hardly seen a flask of that size since I was an undergraduate, and that whole range of glassware immediately recalls sophomore organic chemistry labs. I wondered why that was, but now I have the story thanks to reader Norm Neill of glassmaker NDS Technologies, who saw its birth at Kontes:
"The 19/22 Glassware kit was developed jointly by Eric Nyberg from Kontes Glass and Dr Howard Martin from Lake Forest College in the late 1950's. . .they wanted to scale down the size of the glassware from the traditional 24/40 glassware to something smaller so it could be issued as a complete kit to a student and locked in his lab drawer. . .The next size down from 24/40 is 19/38 but the joint length was too long to allow us to scale down the kit (and) fit into a standard lab bench drawer. The 19/22 medium length joint was the best trade off at the time. . .The packaging of the kit was so popular that during the early 1960's production had to be allocated. The overwhelming success of the 19/22 glassware started the development of an extensive line of 14/20 glassware under the Bantamware® brand."
It's my impression that the 14/20 glassware has been taking over the student market in recent years as well, what with the move to smaller and smaller amounts of solvents and reagents. That makes me wonder if 19/22 glass has a future, which means that I'll probably find some lunatic reason to switch my small-scale stuff to it really soon, giving me the most oddball glass collection in the place. . .
For those who were wondering, my copper reactions the other day worked out just fine. They started out a beautiful blue (copper iodide and an amino acid in straight DMSO – if that’s not blue it’s maybe going to be green, and if it’s not either one you’ve done something wrong). Of course, the color doesn’t stay. The copper ends up as part of a purple-brown sludge that has to be filtered out of the mix, which is the main downside of those Ullman reactions, no matter how people try to scrub them up for polite company.
And DMSO is the other downside, because you have to wash that stuff out with a lot of water. That’s one of the lab solvents that everyone has heard of, even if they slept through high school chemistry. But it’s not one that we use for reactions very much, because it’s something of a pain. It dissolves most everything, which is a good quality, but along with that one comes the ability to contaminate most everything. If your product is pretty greasy and nonpolar, you can partition the reaction between water and some more organic solvent (ether’s what I used this time), and wash it around a lot. But if your product is really polar, you could be in for a long afternoon.
That mighty solvation is something you need to look out for if you spill the stuff on yourself, of course. DMSO is famous for skin penetration (no, I have no idea if it does anything for arthritis). And while many of my compounds are not very physiologically active, I’d rather not dose myself with them to check those numbers. At the extreme end of the scale, a solution of cyanide in DMSO is potentially very dangerous stuff indeed. I’ve done cyanide reactions like that, many times, but always while paying attention to the task at hand.
Where DMSO really gets used is in the compound repository. That dissolves-everything property is handy when you have a few hundred thousand compounds to handle. The standard method for some years has been to keep compounds in the freezer in some defined concentration in DMSO – the solvent freezes easily, down around where water does (Not so! Actually, I've seen in freeze in a chilly lab a couple of times, now that I'm reminded of that in the comments to this post. Pure DMSO solidifies around 17 to 19 C, which is about 64 F C - a bit lower with those screening compounds dissolved in it, though).
But there are problems. For one thing, DMSO isn’t inert. That’s another reason it doesn’t get as much use as a lab solvent; there are many reaction conditions during which it wouldn’t be able to resist joining the party. You can oxidize things by leaving them in DMSO open to air, which isn’t what you want to do to the compound screening collection, so the folks there do as much handling under nitrogen as they can. Compounds sitting carelessly in DMSO tend to turn yellow, which is on the way to red, which is on the way to brown, and there are no pure brown wonder drugs.
Another difficulty is that love for water. Open DMSO containers will pull water in right out of the air, and a few careless freeze/thaw cycles with a screening plate will not only blow your carefully worked out concentrations, it may well also start crashing your compounds out of solution. The less polar ones will start decided that pure DMSO is one thing, but 50/50 DMSO/water is quite another. So not only do you want to work under nitrogen, if you can, but dry nitrogen, and you want to make sure that those plates are sealed up well while they’re in the freezer. (As an alternative, you can go ahead and put water in from the start, taking the consequences). All of these concerns begin to wear down the advantages of DMSO as a universal solvent, but not quite enough to keep people from using it.
And what about the compounds that don’t dissolve in the stuff? Well, it’s a pretty safe bet that a small molecule that can’t go into DMSO is going to have a mighty hard time becoming a drug, and it’s a very unattractive lead to start from, too. That’s the sort of molecule that would tend to just go right through the digestive tract without even noticing that there are things trying to get it into solution. And as for something given i.v., well, if you can’t get it to go into straight DMSO, what are the chances you’re going to get it into some kind of saline injection solution? Or the chances that it won’t crash out in the vein for an instant embolism? No, the zone of non-DMSO-soluble small organics is not a good place to hunt. We’ll leave proteins out of it, but if anyone knows of a small molecule drug that can’t go into DMSO, I’d like to hear about it. Taxol, maybe?
I was running a copper-catalyzed coupling reaction the other day when my summer intern asked me how it worked. I showed her the mechanism that the authors of the paper had proposed, but pointed out that it was mostly hand-waving. The general features are probably more or less right: the copper iodide presumably does form some kind of soluble complex with the amino acid that’s needed in the reaction mix, and that may well form some sort of complex with the aryl halide, which opens up the ring to nucleophilic substitution, etc. If this were an exam, I’d give full points for that one.
But a lot of these couplings are, as I pointed out to her, very hazily worked out. The Ullman reaction, in various forms, has been with us for many decades, and there are more variations on it than you can count. If it always worked reasonably well, or if people had any strong ideas about how it did so, the literature on it wouldn’t be in the shaggy shape it is. Copper chemistry in particular has been (simultaneously) a very useful area for people to discover new reactions, and a horrible trackless swamp for people trying to explain how they work.
All you have to do is look at the vicious exchanges between Bruce Lipschutz and Steve Bertz during the 1990s about whether such as thing as a “higher-order cuprate” exists. I have absolutely no intention of reconstructing this argument; I would have to be paid at a spectacular hourly rate to even attempt it. It's enough to say that the arguments raged, in an increasingly personal manner, about what state the copper metal was in, what ligands coordinated to it, and what the active form of these reagents might be (as opposed to what the bulk of the mixture was at any given time). It culminated in what must be one of the most direct titles for a scientific paper I've ever seen: It's on lithium! An answer to the recent communication which asked the question: 'if the cyano ligand is not on copper, then where is it?'. That's in Chemical Communications7, 815 (1996), if you're interested (here's the PDF for subscribers). Bertz continued to shell Lipshutz's position past the time when any fire was being returned, as far as I can tell, and continues to work in the area. Lipshutz, for his part, hasn't published on the higher-order cuprates in some time (being no doubt heartily sick of the whole topic), but has kept up a steady stream of work on new reactions involving copper, nickel, and other metals.
So if well-qualified researchers, brimming with grad students, postdocs, and grant money, can argue for years about copper mechanisms, I'm going to stay out of it. As time goes on, I'm increasingly indifferent to reaction mechanisms, anyway. I want to get product out the other end of the reaction. And while there are times when knowing the mechanism can help reach that goal, those times do not occur as frequently as you might hope.
I have a summer intern this year, and she has (so far) not caused anything to burst into flames. That’s the first thing you ask of a summer student, and the fact that she’s gotten several reactions to work is just a welcome extra. A summer with no laboratory bonfires will be a successful summer, as far as I’m concerned.
That’s because I’ve experienced the alternative, as I’ve detailed here before. If most of the lab fire stories you hear start out with the phrase “We had this solvent still. . .”, the rest of them all seem to begin with “We had this summer undergrad student. . .” (You can imagine the flame-filled end to any story that starts out with a summer student distilling some solvent – that Venn diagram leaves you with no way out at all).
No, after watching an undergrad next door to me kick a four-liter jug of pyridine all over the floor, causing a shimmering wave of unspeakable pyridine vapors to almost knock me off my feet. . .and after watching another one walk away for two hours after setting up a reduced-pressure DMSO still, which inadvertently turned into a high-pressure apparatus and blew DMSO and calcium hydride all over the inside of a hood. . .and after watching them charcoal reactions by plugging heating apparatus straight into the wall outlet instead of into the Variac. . .and, well, you get the idea.
I should add that I was no great shakes as a summer undergrad myself. I did a summer after my sophomore year with Tom Goodwin, but didn't get a great deal accomplished (through no fault of his!) Then after my junior year, I worked with Dale Boger, back when he was at the University of Kansas, but I mostly (and rather slowly) found a list of conditions that don't work for inverse electron demand Diels-Alder reactions. But although I spilled some generous amounts of solvent, I didn't set anything on fire.
No, we're going to have a calmer and more productive summer around here. I have my student working on a problem I've had a longstanding interest in, one that needs some variables chased down and figured out. With any luck, enough data will be generated to make for an interesting publication late in the year, and everyone will come out ahead.
One of the reasons I starting this blog was that many people I met were interested in my job. Very few of them had ever talked to someone who discovered new medicines for a living, and a surprising number of them (well, surprising to me) had no idea of where medicines came from in the first place.
Talking to such folks (interested, but with no particular training in science) gave me some good practice in explaining the work. It helps that the kind of work I do is actually fairly easy to explain. There are a lot of details – as with any branch of science, the closer you look, the more you see – but I haven’t run across any key concepts that can’t be communicated in plain language. (It also helps that medicinal chemistry, as it’s actually practiced, uses an embarrassingly small amount of actual mathematics).
The toughest things to deal with are the parts of the field that actually touch on physics and math. My vote for the hardest everyday phenomenon to explain at anything past a superficial level is magnetism. So that means that explaining how an NMR machine works is not trivial. At least, explaining it in a way that a listener has a chance of understanding you isn’t – a while ago, I took up the challenge to try to explain it here in lay terms, and I haven’t done it yet, for good reason.
Explaining statistical significance is doable, but going much past that (principal components, the difference between Bayesian and frequentist approaches) takes some real care. And, of course, when you open the hood on chemical reactivity, the mechanisms of bond-forming and bond-breaking, you quickly find yourself in physics up to your armpits. It’s easier to stipulate, openly or by assumption, that there are such thing as chemical bonds, and that some of them are stronger than others. You don’t want to start answering a question about why one group falls off your drug molecule easier than another one does, only to find yourself fifteen minutes later trying to explain the Pauli exclusion principle. Counterproductive.
But the basics of medicinal chemistry can be sketched out pretty quickly, which makes some of the more curious listeners wonder, after a while, why we aren’t better at it. The best example I can give them is to advance a quick, hand-waving explanation of, for example, how compounds get into cells. Then I point out that that explanation is unnervingly close to the best understanding we have of how compounds get into cells. The same holds for a number of other important processes, way too many of them.
And that's why drug discovery is simultaneously frustrating and fascinating. We know huge numbers of things, great masses of detail that can take years to piece together. And it's not enough. Some of the most important puzzle pieces are still weirdly ill-defined, and there are probably others whose existence we haven't even realized yet. I'd be willing to bet that if you scanned the whole history of pharmaceutical discovery, you'd find people at every point thinking "You know, in any thirty years they should have all this figured out". But the years go by, and they - we - don't. Give it another thirty years, you think?
We order chemicals from all sorts of suppliers – big, reputable outfits like Sigma-Aldrich-Fluka all the way down to places that none of us even have heard of before. In those latter cases, the primary question is always whether or not the reagent will actually show up, and the secondary one is how long it’ll take. There are some of those small suppliers who pad their catalog with things that aren’t exactly available, not yet – but hey, they will be if someone orders them. They’ll just tell you it’s back-ordered, and tell someone in the lab to get cracking.
And when you get your compound in, they arrive in various forms. Glass or plastic bottles are the norm, naturally, with the occasional irritating (but presumably necessary) sealed-glass ampoule. But after some time in the lab, you can tell some of the suppliers from across the room. For example, the Japanese company TCI sends a lot of its compounds in normal-looking glass bottles, but these are first put inside capped plastic containers, like larger translucent versions of the ones that 35mm film probably still comes in. And once you taken them out, their glass bottles have these odd plastic labels on them which come up around the screw cap and are perforated around the cap’s border. On the labels, they also have that same thin, fussy, serif font that the Japanese have been using for Roman-style letters for decades (since the war?) and is only in recent years disappearing from their world.
Maybridge, British vendor of all kinds of odd stuff, often sends its compounds in these weird little squat brown-glass bottles with small black caps on them. They must have the world supply of that particular bottle shape tied up, since I’ve never seen one anywhere else. It most resembles the small bottles that solutions for injection are packaged in. So many of the company’s catalog items are in such bottles (or even smaller ones) that it seems wrong somehow when you come across a huge (huge for Maybridge) hundred-gram bottle with their label on it.
Most of the suppliers have neutral-sounding names like those above. They could be chemical companies, vendors of kitchen cabinets, real estate trusts, who knows: Maybridge, Oakwood, Lancaster (now gone, and their blue labels with them). And some of them are unmistakably in the chemical supply business, but rather blandly named (Pharmacore, for example, or Chembridge). Some names are, perhaps, mistakes: the namers of Asinex, for example, seem to have been unaware that the closest Engish word is “asinine”, which means that they have to hope for people to pronounce that “s” as if it were a “z”. (I should mention that both Asinex and Chembridge indulge in one widely hated practice: putting no useful information on their tiny vials other than a catalog number or bar code – Bionet (Key) is a similar offender).
In this dull company, I’m always glad to see the weirdos. I miss the now-purchased-away British supplier called Avocado – green labels, naturally – and always wondered who named them and why. Tyger Scientific makes me wonder if there’s an English major in somewhere at their founding, fond of William Blake. And there’s one company that came into the industry under the glorious name of, I am not making this up, “Butt Park”, and many are the chemists they’ve made stand puzzled in front of the supply cabinet. (I'd provide a link, but I can't find a direct one, and Googling it can be a real minefield).
I refuse to consider that name a mistake. That's a feature, not a bug, and I wish that there were more competition in the category. I would proudly and purposely send business to, say, Batshit Chemical Supply, Inc., even if they back-ordered me every single time.
Not many chemists come into the drug industry knowing very much about biology. I certainly didn’t, not on the level that was needed. It’s not surprising, but it’s also not as much of a handicap as you’d think, at least not at first.
That’s because the first job of a new hire in the med-chem department is to crank out compounds, and that goes for both the PhD and Master’s levels. (Those roles diverge as time goes on, though). But with a few obvious rules in hand (no hot reactive functional groups, no huge greasy monster molecules, etc.), a person can contribute reasonable-looking compounds pretty quickly. No biological knowledge needed.
But if you’re going to be more valuable than a new hire (and as time goes on, you’d better be), then you have to start picking up some more of the broader science of drug discovery. That turns out to involve a lot more than chemistry, which is one of the things that chemists have to get adjusted to. If you’re going to move up to the point of being considered to lead a new project, you’re going to have to show that you can converse with the folks who know protein expression, assay development, molecular biology, PK, toxicology, and so on. You’re not going to be expected to come in and solve their problems (although if you do manage to solve one once in a while, it’ll do both you and them some good). But you are expected to understand what they’re talking about.
So that’s a piece of advice I can give to new chemistry hires in this business: get ready to learn everyone else’s business, too. Listen up when the people from the other departments talk about what they’re up to, and especially when they complain about their problems. Try to understand why they’re complaining, and ask them (especially one on one) about what they usually try when this sort of thing happens. The occasional paranoid might think at first that you’re compiling info in order to mess with them later, but you shouldn’t be the sort of person around whom that suspicion credibly lingers. In general, if the people in those other groups are any good at all, they’ll be glad to tell you what’s going on, and you’ll pick up a lot of practical knowledge.
The consequences of not doing this sort of thing become more severe as time goes on. At one of my former companies, we once brought in a job candidate from BNP (Big Name Pharmaceuticals). He’d been around seven or eight years, enough time to be considered fairly experienced. But people at that level vary a lot, and he was (as it turned out) on the low end. When we’d ask him about, for example, any formulation problems he’d had to deal with on his project compounds, he told us that well, he didn’t usually go to those meetings, his boss did. And when we asked him about how he got along with the PK group – well, they were over in another building, and he hardly ever saw them. And so on, and so on.
He was well along to being crippled by the way things were done at BNP. Actually, it may have been more the way he was doing things. From talking with other people from that shop over the years, it’s clear that it didn’t have to be that way – if you made the effort, you could go to those meetings, and if you took the time, you could go over to those other buildings and show your face. But you didn’t have to, and this guy (since he didn’t have to) didn’t bother to. And by keeping to his burrow, he hadn’t learned nearly as much as he could have. We didn’t make him an offer. So talk to people, talk to people outside your field. If you’re any good at all, they’ll learn something from you, too.
Hexane (or its cheaper, less well-defined cousin, petroleum ether) is the proto-solvent. Light and thin, it’s the weakest at actually dissolving anything, so it’s the background to most stronger mixtures in a purification. Like most other solvents, though, it’ll strip the oils right off your skin, leaving you spiderwebbed with white lines across your fingers and in need of some lotion. Its smell isn’t pleasant, but it doesn’t really stink, either. A nonchemist would easily place it in the oil / kerosene / gasoline end of things, which is exactly where it belongs.
When I first encountered ethyl acetate back in college, little did I realize that I was picking up the scent of the rest of my life. I've been in the lab ever since, and so has it. Pleasant, unspecifically fruity, vaguely bubblegum-like, the smell of that solvent is a daily companion to almost every synthetic organic chemist in the world. Mixed with hexane in different proportions according to your needs, it runs the majority of chromatographies in the world. Squirt bottles of it sit around on benches. By now, it’s an old, old friend, and the smell of it says that I’m actually getting something done.
Ether (the real ether, diethyl ether) seems like it’s close to not being there at all. No long for this world, it’s supremely light, and evaporates so quickly, that it just barely holds on to the liquid state. It has a slightly dangerous overtone to it, since it can ignite so easily and forms explosive peroxides if it’s left sitting around. The somewhat smothering smell can’t quite be described, but is instantly recognizable. Its oxygen atom gives it more dissolving power, so ether/hexane mixtures are good for delicate separations, although often impractical on a larger scale. To me, ether is sort of a lighter, stronger hexane, in the relationship that titanium has to steel.
Methanol, on the other hand, has no smell – no smell whatsoever to me, at any rate, despite what that Wikipedia link says, although I think I can tell it from air. Pure lab ethanol smells great, but methanol is a blank to me. It’s the most watery of the common solvents – it’s lighter, but that OH group gives it some surface tension, which (along with its bizarre weight) is one of water’s defining characteristics. You notice the difference, compared to thin, slippery hexane or ether – methanol is a solvent with some body to it. It’s powerful stuff in chromatography, too – one per cent added to a weaker solvent will totally change things.
Do you call it dichloromethane or methylene chloride? The latter probably gets more use, and rolls off the tongue a bit more easily. This stuff is like the demon form of hexane – it has no oxygen atoms like ether or ethyl acetate, but is a pretty strong solvent, in what always seems a mysterious way. With another immediately recognizable but hard to describe smell, its odor is the prototype of “chlorinated”. But the thing that stands out the most is its weight. This is the only common solvent that’s heavier than water, and you can build up your arms doing curls with jugs of the stuff. We don’t use its even denser cousin chloroform all that much; it would be even better bodybuilding material.
Acetone is one of the solvents familiar both in and out of the lab: nail polish remover, without the added scents. You hardly ever run an actual reaction in the stuff, though, and when you do it feels a bit odd. That’s because acetone has become the default flask-rinsing solvent of the chemistry world. I’m not sure when that was settled, but it was decades ago: a perfectly respectable solvent, stuck in the role of janitor to all that brown, red, and yellow stuff stuck to the inside of a million round-bottom flasks.
Time for another quick quiz on whether you have what it takes to be a big-time medicinal chemist. Prepare for some not-so-welcome old friends to visit you yet again:
1. Your two main assays refuse to act as if they’re part of the same project. Most of your potent compounds in the first enzyme assay don’t do much against the cells, and the best cellular compounds are no great shakes in the enzyme assay. There’s a narrow zone of overlap, but it doesn’t look big or robust enough to base the whole project on. Do you pursue the cellular activity, on the theory that that’s the effect you’re looking for, or pursue the enzyme activity (on the grounds that it’s the right target, and you just have to get the things into the cells), or consider revamping the assays completely, or what?
2. In the next case, your disconnect doesn’t occur until you get to metabolism and PK. When you run your compound across liver enzymes, they grind it into dust. But you did that after you dosed the animals, you buckaroo, and not only did the compound seem to work OK, but its blood levels weren’t bad, either. So how come it looks as if it should be disappearing? The most destructive of the enzymes, by the way, was the human one. Are you worried about that, or not?
3. The project you’re on has a compound profile as a goal – so much potency, at least so much selectivity, and the like. As time goes on, there’s one selectivity assay in particular that you just can’t seem to shake. The only time you see a decent separation between your activity and the one you don’t want is in a compound series that you don’t like – they’re big and greasy, and although they look very active in the enzyme assay, they never perform as well as they should in the animals. But it’s starting to seem as if you have a choice: good properties or selectivity, but not both at the same time. What to do?
4. OK, let’s back up some. You’re working on a project that hasn’t really made it to the medicinal chemistry stage. The screening folks have run the target, and forwarded you their data. Nothing shows up really potent, but there are some 500-nanomolar things scattered around. And “scattered” is the word, all right. You probably have two dozen near-singletons in that range – nothing seems to show much of a robust effect across a given class of compounds. But this is a target that everyone wants to start a program on - it's hot, it's happening. How do you proceed?
I ran a reaction the other day which gave me two very similar products. That's not so uncommon, but this one really shouldn't have been able to do that. (For the chemists in the audience, these two so similar, in fact, that the usual LC/MS conditions only showed one peak. NMR tells you different, though, and a painstaking multiple-elution TLC in some nonstandard solvent mixtures resolves the two spots).
I thought about the problem a bit, and decided that the first thing to do was to check my starting material. And there they were: two very similar starting materials, together in the same jar. Mind you, there's only one structure on the label. No wonder the stuff was so sticky. I'd received the Special Extended Edition without knowing it - odds are, the supply company sent it to me without knowing it, either, although that'll change when they get my e-mail. One of the components, anyway, seems to be the right stuff, so I suppose it could be worse.
This happens more often than it should, often enough that every working chemist has a similar story or two. And it doesn't correlate that well with the size or renown of the company you're ordering from, since everyone sources material from all over the place. Little mom-and-pop operations have sent me plenty of fluffy, flawless stuff, while Aldrich has on occasion mailed me goo. (On another occasion they mailed me a perfectly empty sealed ampoule with a label on it, but since the label didn't read "Air", I thought I had reason to complain). That doesn't mean that reputations don't vary. Even though they're now part of the same company as Aldrich and Sigma, those Swiss fanatics at Fluka do this sort of thing to you comparatively less often than their cohorts.
Not all the unopened slime you encounter is necessarily the fault of the company that shipped it. Some things just aren't stable, or at least aren't so stable in the back of an unventilated truck or sitting out in the sun on a loading dock. And the longer it is after an order's been received, the more the problem is likely to be with the receiver. A look at the condition of the vials in a drug company's compound repository will convince anyone that the kinds of molecules we like may not have indefinite shelf lives.
In this case, it's going to be easier to clean up the starting material and run the reaction again than it would be to clean up my dueling products. Easier yet would be to get a bottle of the right stuff from the supplier, but this one isn't exactly a high-volume compound, and I suspect that it's all the same nasty batch on their shelves. Worth a try, though. And thus does science stagger on.
Lab fires don’t happen as often as you might think, at least to hear the way organic chemists talk. We all have alarming stories of alarming reactions (often set up by some rather alarming labmates), but these things are harvested over a fairly broad range of experience. It’s a familiar enough topic that I can remember someone sitting down at lunch while we were swapping lab stories and saying “Oh, this conversation. . .”
But happen they do, and it’s always worth taking a couple of minutes to think about what you do in such a situation. That depends on the fire, of course. For starters, a small one burning out of the neck of a flask can be put out quickly just by slapping a beaker over the top of it. Never neglect that possibility, because it’s fast, effective, and (truth be told) if no one saw you do it, no one necessarily has to know that your (minor!) fire even happened.
Larger ones aren’t going to be so easy, but there are some potential ways out of those, too. My wife had a labmate in her molecular biology department who was always setting off blazes with the ethanol she used to wipe things down with. (This person neglected to turn off the Fisher burner used for sterilizing wire loops, etc., before she started sloshing the alcohol around). A fire like that will just burn itself out if you close the hood sash and let it rip for a few seconds, as long as you’re sure that there’s no fuel source (like the wash bottle of ethanol you might have chucked in there in a moment of panic, for example).
Most chemistry hoods, though, have all too many sources of fuel in them, so you probably won’t be able to put out a blaze through benign neglect. If it comes to a fire extinguisher, make sure you already know where the nearest one is, for starters. You'd be surprised how hard it is to find one of the darn things when you really need it. And once you've found it, make sure that you know which kind you’re using. The carbon dioxide ones don’t make the horrible mess that the dry-chem ones do, which is one thing in their favor, although I think in general they’re a bit less effective. You can tell the difference immediately – the carbon dioxide ones have the big nozzle on them, while dry-chem is a short, plain hose. My lab is outfitted with the latter, which makes me wish more than ever that we never have to use them.
And if you happen to have halon extinguishers (are those still around?), make a note of that, because the technique you may have learned for using the other ones won’t work. Instead of coming in and aiming at the base of the fire, with halon you have to stand further back and let the stuff shower down on it. A colleague of mine once blew the contents of a flaming oil bath all over the lab because he hadn’t been trained in that distinction.
The safety people always tell you that if you’ve used up one extinguisher and the fire still isn’t out, to head for the door rather than reach for a second one. That’s probably good advice (although I’ve seen it disregarded), and I’d advise you to take it. Actually, I’d advise you never to have that decision to make at all, but that’s not always up to you. You may be doing nothing but adding sodium sulfate to a bunch of dichloromethane today, but who knows? The guys next door might be gearing up for Trimethylaluminum Fiesta Days. You never can tell.
I’m going to be working up an Arbuzov reaction this morning, which is an odd thing for me to say. That’s because to the best of my recollection, which is pretty good, I’ve only run any of those during one period in my lab career. That was back in grad school, along about 1985, I’d say. I hope this one proves more useful than that one did – I was trying to make some dimethyl diazomethylphosphonate, and the prep was a relentless barrage of No Fun. (The first part of the sequence was identical to this).
I keep a list in my head of songs that I’ve only heard one time (no, I don’t appear to be normal, thanks for asking), and perhaps it’s time for me to assemble a list of reactions that I’ve only run once. That’s a tougher one, because if a reaction fails, you may well run the thing again. Still, I’ve only done one hydrogenation at 2000 psi with rhodium on alumina (July 3, 1984, and it looked like used lawnmower oil afterwards, I should add), and I’ve only used samarium iodide one time (and it didn’t work). But for a longer list I might have to settle for some things that I ran for a brief period and never have since.
The Claisen rearrangement would fall into that class, for sure. A feature of my early grad school work, I’ve never had the need to run one since. I can't think of the reaction without smelling ethyl vinyl ether in my memory, which is not a feature, in case you're wondering. I did a lot of carbohydrate reactions back then that I haven’t had the need to return to, either – Ferrier rearrangements being just one of them. And, like many other chemists, I had a brief photochemistry period, in my case during my post-doc, and have never run one of those again, either. Others that enjoyed their day in the sun and have never been seen again in my hood are the Prins reaction, nitrone cycloaddition (not since I was an undergrad in 1983), Lindlar hydrogenation, and the Henry reaction.
The thing is, any of these could make a comeback at any time. They’re still all perfectly reasonable reactions, and depending on what comes out of the next high-throughput screen or literature search, I might be setting one up next week. You never know. But there are some reactions that I think I’ve said goodbye to forever. In some cases, that’s because better alternatives are now available - I mentioned here that I haven’t used PCC for oxidations in years, and I think that one’s been pretty much superseded.
Others are history because I either very much doubt I’ll have the need for them, or because I just flat out Don’ Wanna. For example, I made Dess-Martin periodinane three times on a hundred-gram scale, during a period in the early 1990s when it wasn’t commercially available, and I plan, with any luck, never to do that again. The prep has been improved since those days, but that explosive intermediate was never something I enjoyed seeing. I don’t think I’ll be synthesizing fluorosulfonic acid starting from hydrofluoric acid any time soon, either. I did that one as an undergrad, too, if you can believe that – this guy must have had confidence in me, which I’m not at all sure was warranted by the evidence at hand. Nor do I foresee any need to make Fremy’s Salt from scratch again. (You can see someone else do it here, though - the internet amazes me sometimes). And if I never do another reaction
, and you can guess her point from that title:
is still applicable, over thirty years later:
as a general review of the field, for anyone who's thinking about it. We'll see how much use I get out of my current setup, but for now, I'm happily blasting away with the ultraviolet. . .
. Many votes also were cast for Camille Wermuth's 
. For getting up to speed, several readers recommend Graham Patrick's 
. And an older text that has some fans is Richard Silverman's 
.
by Neal Anderson and 
by Stan Lee (no, not that Stan Lee) and Graham Robinson.
by Rick Ng and also in Walter Sneader's 
.
.
and 
by Nogrady and Weaver.
by Kerns and Di. For getting up to speed in this area, there's 
by Donald Birkett.
. Since all of us make a fair number of poisons (as we eventually discover), it's worth a look.
. I haven't seen it personally, but if you're interested in microwave work, it looks worthwhile.
