In review: August 2021 Facts of the Day
A running log of past facts from August 2021.
You may need to click on some of the tweets to actually see the links!
2021-08-02:
Bob's Med Chem Fact of the Day:
Pitolisant en.wikipedia.org/wiki/Pitolisant is a histamine H3 receptor inverse agonist used for treatment of narcolepsy.
From its simple structure, it is a bit hard to see, but the discovery starting point was the strong natural agonist histamine itself
It's an interesting discovery story bpspubs.onlinelibrary.wiley.com/doi/10.1111/j.…
and don't be fooled, the final structure of Pitolisant belies all of the many, many analogs that had to be made to get from histamine to the drug. Many rapidly metabolized and toxic compounds encountered along the way.
In particular, it seems it was very hard to avoid bioavailability and/or toxicity issues with any of the synthesized imidazole derivatives.
2021-08-03:
Bob's Med Chem Fact of the Day:
Be careful to understand mechanism when ordering known compounds for repurposing screens or as positive controls for experiments.
Depending on mechanism, active form of the molecule may not be the approved drug structure (as in some prodrugs).
Essentially, you need to work to understand the details of the assays and the molecules.
And make sure you are helping out your biologist colleagues when sourcing compounds to understand these issues.
Here is a related example, if you are interested:
Bob the Grumpy Med Chemist @med_chemist
2021-08-04:
Bob's Med Chem Fact of the Day:
Sometimes you don't want an oral drug to be absorbed in the GI tract.
For example, Tenapanor en.wikipedia.org/wiki/Tenapanor has minimal systemic exposure, but thus can act primarily on its target, sodium-proton exchanger NHE3, in the intestinal tract.
By inhibiting sodium uptake in the GI tract, Tenapanor can help increase the water content and improve the symptoms of IBS. It also has been in the news lately for a recent FDA failure for CDK endpts.com/ardelyx-chops-…
Perhaps not surprisingly, one main side effect was diarrhea
2021-08-05:
Bob's Med Chem Fact of the Day:
"Fog fever" or "Cow Asthma" occurs when you move cattle too quickly from indoor feedlots to lush, protein rich pastures.
The reason for this connects back to a classic "structural alert" in med chem. ->
Essentially, the rumen of the cattle cannot adjust quickly enough to the protein rich grass with lots of tryptophan.
Bacteria in the rumen convert tryptophan-> indoleacetic acid->skatole (3-methylindole).
Skatole goes into circulation and goes to club cells in bronchioles. ->
There are a lot of CYPs in these cells that convert skatole to the reactive metabolite, 3-methyleneindolenine.
This reactive intermediate is then toxic to the cells in the lung. Emphysema, pulmonary edema, and often death are a result. ->
This is why 3-alkylindoles, which can form similar reactive metabolites, are often a structural alerts in medicinal chemistry.
sciencedirect.com/science/articl…
gives the examples below of where the reactive metabolite formation is observed
2021-08-06:
Bob's Med Chem Fact of the Day:
It's worth looking into the routes used to make HTS hits.
E.g. in this paper 100s of hits were made using transition metal in last step. Metal impurities created false positives via oxidation of cysteine.
This also explains why a fresh re-synthesis of hits is important and it's helpful to understand if impurities could affect assay.
Fortunately, counter-screens are possible too. In the example, they added EDTA to assay to chelate trace metals.
2021-08-09:
Bob's Med Chem Fact of the Day:
In vivo rodent PK studies are usually not prohibitively expensive. Use them before efficacy studies!
Don't do what this paper did: 1 (noisy) uM hit, no proper hit validation, 0 analogs, no in vitro/in vivo ADME, straight into mouse efficacy model
I will not link to the specific paper, because I believe there are likely many like it. But...
Seeing the names of NIH and National Cancer Institute grants in the acknowledgments is maddening.
Here is a good thread from @KRHornberger some time back. As more info gets out there, I hope eventually the excuse that "they just didn't no any better" gets less and less viable.
Friends don't let friends make these mistakes!
Keith Hornberger @KRHornberger
2021-08-10:
Bob's Med Chem Fact of the Day:
Here's one of my favorite court cases on the definition of "chemical similarity," which also has great advice for all comp-chemists:
- understand the methods you use
- LOOK AT YOUR MOLECULES before you send them to me
casetext.com/case/us-v-brow…
Sad that a US judge did a better job than some reviewers. Some other points from the case:
- Methods have domains of applicability
- Most methods are still missing clear things humans can understand
- Using a computational method doesn't automatically make you more "objective"
The number of times I look at output from models, point out very strange results or substructures, and realize the modeler did not look at the output in depth is alarming...
2021-08-11:
Bob's Med Chem Fact of the Day:
Benzodiazepines show large species differences in metabolism. Thus, dogs often need higher doses.
In the US, this can trigger algorithms that look for drug abusers since dog owner names are linked to the prescription.
wired.com/story/opioid-d…
Benzodiazepines can be quite complicated from a pharmacokinetics perspective because they often have many active metabolites of varying activity and half life. As if CNS drugs weren't hard enough to understand already...
2021-08-12:
Bob's Med Chem Fact of the Day:
Varenicline was originally discovered as an improvement of Cytisine, which dates back to the 1800s and was smoked by WW2 soldiers as a cheap tobacco substitute.
Also a good demonstration of old school CNS drug discovery->
endpts.com/first-generic-…
The original J Med Chem paper is quite short and worth a read. But this figure shows a lot of the guesses sometimes made in structure-free discovery. Sometimes they work!
Can have a lot of ? and ~= reasoning in these projects
Despite Chantix being effective, still worth knowing that most people who try to quit fail. I'm lucky to never have been interested in tobacco, but several members of my family have struggled to quit for the majority of their lives.
2021-08-13:
Bob's Med Chem Fact of the Day:
Today's fact is about the "HPE method" of drug discovery, which has been used extensively to find drugs in industry, but is perhaps less known to outsiders.
Benchmarks of the method are scarce, but I can confirm it has led to approved drugs. ->
In the HPE method, everyone on the project contributes their ideas and designs for new molecules, then the Highest Paid Employee (HPE) on the project ignores most of them and chooses his/her own favorites.
The technique is old and dates back to the beginnings of drug discovery.
Notably, the method requires minimal computational resources, and it can be understood by most chemists even without an extensive mathematical background.
Even so, I've seen very little published in the literature on the methodology.
2021-08-16:
Bob's Med Chem Fact of the Day:
Be careful when reasoning about 3D in 2D. Proteins don't think about substructures the way you do.
Nice example from @PracticalFrag blog:
Compounds 5 and 24 bind quite differently. 2nd pic gives a 2D idea of the shift. Not as pretty a picture
And the blog gives a nice 3D overlay, which is of course better than any 2D drawing
See the whole writeup here practicalfragments.blogspot.com/2021/08/fragme…
Central lesson is don't just do substructure matching in your head without being skeptical that 3D changes things.
2021-08-17:
Bob's Med Chem Fact of they Day:
Sometimes I think we forget how important it is to understand the synthetic tractability of chemical starting points in the modern era.
Paul Janssen was very aware of this when starting out, and it led to many, many drugs.
2021-08-18:
Bob's Med Chem Fact of the Day:
Helpful experiment to determine if CYP inhibition is likely from heme ligation or nonspecific hydrophobic recognition.
Add Methyl at the 2-position of pyridine
If less inhibition -> likely heme ligation
If ~ same inhibition -> likely nonspecific
This experiment was cleverly used in pubs.acs.org/doi/10.1021/ac…
where they were actually targeting another CYP (w/ heme ligation) and wanted more selectivity over CYP3A4.
By understanding that the interaction with CYP3A4 was mostly nonspecific, lipophilicity driven, they then could focus on designing less lipophilic molecules to improve selectivity.
2021-08-19:
Bob's Med Chem Fact of the Day
Drugs have unique target product profiles (TPP), with different safety and PK needs.
Thus, don't blindly use the line "this functional group is in approved drugs" in discussions, without understanding of respective TPPs, synthetic challenges, etc.
2021-08-22:
Bob's Med Chem Fact of the Day:
This is a poster I believe should sometimes be hung up for all to see at certain meetings. (Or set as a zoom background, I suppose, in this modern era)
See forbes.com/sites/davidsha… for a popular account of the Ibrutinib story.
2021-08-23:
Bob's Med Chem Fact of the Day:
Interesting approach here to reduce lipophilicity and nonselective metabolism issues but maintain penetration of BBB:
Add pyridine Ns at sterically hindered spots to limit polar SA. Would've liked to see measured and not cLogP, to prove it though
2021-08-24:
Bob's Med Chem Fact of the Day:
Usually not a huge fan of deuterium, but here's an interesting use case to slow down aldehyde oxidase metabolism resulting in toxic metabolites. Couldn't just flank the quinoline N with Me/F because constrained by H bond. pubs.acs.org/doi/10.1021/ac…
Should also note that they did not go with this strategy in the candidate and instead scaffold hopped to volitinib as shown in the image above from
I actually got to this example from reading about VX-984, which also has an interesting use of deuterium flanking pyridyl nitrogens
For the synthesis fans out there, here's how they got the D on the rings:
References are pubs.acs.org/doi/10.1021/ac… and US20130281431A1
2021-08-25:
Bob's Med Chem Fact of the Day:
ACS spends ~10 Million USD per year on abstracting services, which should tell you how bad the state of data in our field is.
Every time I read a patent or paper, I can't believe that blocks of text are still the method of sharing info.
Source is projects.propublica.org/nonprofits/org…
And with discussion of membership today, one reason I am not too excited about ACS is how little they have done to advance status quo, and I think their ownership of scifinder and journals may make changing status quo against their interests.
I believe one of the best ways to advance science and chemistry across the world is to make data more available and searchable. Unfortunately, I do not think we have done that very well.
Improving the readability of patents does not seem to be in many people's interests either.
Here's a sample task: go get a list of every molecule that was made in a specific drug discovery patent. While doing this task, ask yourself if the world need be this way.
2021-08-26:
Bob's Med Chem Fact of the Day:
Old school med-chemists don't like benzamidines b/c their charged and basic nature can cause poor absorption.
Read how the Berotralstat inventors learned from past anticoagulant projects to find a suitable replacement in
Avoralstat was the first generation compound (with a bezamidine and an acid) that unfortunately flopped due to poor oral exposure. Berotralstat is now approved and showed much improved oral bioavailability
2021-08-29:
Bob’s Med Chem Fact of the Day:
Gabapentin is a popular anticonvulsant with a simple and elegant design: take GABA (en.wikipedia.org/wiki/Gamma-Ami…) and add a 3,3-pentyl disubstitution (thus “pentin” in name) to increase blood–brain barrier penetration.
Reality is a bit different ->
Gabapentin is not actually a GABAmimetic, meaning it does not act like GABA and instead works by a much more complicated mechanism.
However, it does better penetrate the BBB, though by active transport. And, most importantly, it does help with seizures.
Lastly, should be noted that Gabapentin is not metabolized in humans, and is not protein bound.
Since it undergoes no hepatic metabolism, there is a smaller chance of drug-drug interactions, which is a win for patients often on other drugs.
Read more at
My dog received gabapentin for post-surgery recovery a bit ago. In dogs, it's also used as a general pain reliever.
Luckily the surgery was minor, and he is continuing to be a good boy. He is happy gabapentin works, and does not care that about the elegance of any hypothesis.
2021-08-30:
Bob's Med Chem Fact of the Day:
Even if you have a total synthesis PhD, it is not below you to do amide couplings + reductive aminations if it gets a drug to patients more quickly.
Another reminder that ~0 patients care about or even recognize "chemical elegance" the way you do
Similarly, even if you have a PhD in the fanciest of molecular simulation methods, it is not below you to admit that the method does not work well on the current system.
Knowing what isn't going to work can help move the project more quickly.
And even if you have a PhD in the newest and greatest med-chem technique to drug hard proteins, it is not below you to admit that the current target product profile of a project can be achieved via a much simpler, older method.
2021-08-31:
Bob's Med Chem Fact of the Day:
Worth remembering that not all small molecule drugs are alike, Lokelma en.wikipedia.org/wiki/Sodium_zi… is used to treat high blood potassium.
It is a zirconium silicate that mimics potassium channels and binds K+ cations in the blood.
