19F-NMR-guided fragment linking on BACE1

Fragment growing has been the dominant strategy of most of the recent posts involving lead optimization, consistent with our poll results. However, fragment linking can be powerful too, as illustrated by the recent approval of venetoclax, which was derived from fragment linking. A recent paper in J. Med. Chem. by Brad Jordan and colleagues at Amgen provides another nice case study.

Amgen researchers had previously used fragment growing to discover inhibitors of BACE1, an Alzheimer’s target which has been heavily tackled by fragments. However, the most potent molecules in the series also inhibited the related aspartic protease cathepsin D (CatD), which could cause serious side effects. The researchers sought to gain selectivity by building inhibitors to occupy the so-called S3subpocket of BACE1. To do so, they used ¹⁹F-NMR to find fragments that would bind to BACE1 in the presence of a “blocking compound” that filled most of the active site but not the S3subpocket. This led to the discovery of seven fragments, the most potent being compound 3. Interestingly, this fragment only bound in the presence of the blocking compound as assessed both by NMR and SPR. Also, it could be competed by a compound that binds in the S3subocket.

Having thus identified a fragment that bound in the presence of one of their inhibitors, the researchers used interligand NOE (ILOE) to determine how the two compounds bind relative to one another. This supported the idea that compound 3 binds in the S3subpocket, and also suggested how the fragment could be linked onto the existing lead series, exemplified by compound 5. Just four compounds were designed and synthesized, and all of them were more potent than either of the starting points, with compound 9 being the best. More importantly, this compound also proved to be ~2000-fold selective for BACE1 over CatD in enzymatic and cell-based assays.

Despite the excellent (high picomolar) affinity of compound 9 for BACE1, this is actually about 25-fold worse than would be predicted by a simplistic additivity of binding energies – a not uncommon occurrence when linking molecules. Still, with its combined used of multiple NMR techniques and structure-based design to solve a specificity challenge, this paper is worth perusing.

NMR poll results

The results of our latest poll are in – thanks to all who participated! Of the 119 people who responded to the first question, 87% said they use NMR for finding or validating fragments. Even if we assume that responses were biased towards NMR aficionados, big magnets are clearly popular.

The second question asked about specific NMR techniques. If everyone who said they used NMR in the first question also answered the second, this means the average user applies more than 3 different techniques; I’ll let Teddy weigh in to see whether this matches his experience.

One surprise for me was that, although many techniques are widely used, none are nearly universal; even the most popular methods seem to be used by just over half of respondents.

Among ligand-detected methods (blue in the figure), STD ranks at the top, with line-broadening, WaterLOGSY, and fluorine-based techniques all tied for second place.

Protein-detected methods (red in the figure) also appear quite healthy, with nearly as many respondents using ¹⁵N-HSQC/HMQC as STD.

Finally, 11 of you said you use "other" techniques. We didn't include TINS, even though it seems quite useful, because it is only available through the services of ZoBio. But what else is out there?

NMR poll!

Among fragment-finding techniques, nuclear magnetic resonance (NMR) ranks near the top. Protein-detected methods, like HSQC/HMQC-based SAR by NMR, helped usher in fragment-based drug discovery as a practical endeavor. More recently, ligand-detected methods such as line broadening (or CPMG), STD, and WaterLOGSY appear to have gained the edge. There are also more boutique methods, such as ILOE and spin labeling. And of course, some people proudly embrace their fluorine fetishism.

So what’s your favorite flavor? Now's your chance to weigh in on our latest poll (on the right). The first question asks whether you use NMR, and the second asks which methods you use. PLEASE ANSWER BOTH QUESTIONS - the free version of Polldaddy doesn't track individuals, so we need the answer to the first question to know the total number of respondents.

And, as always, your comments are welcome.

Tool Discovery Done Right

I have been on a bit of a jag lately pummeling academic "Drug discovery".  Dan recently hopped on, in his much more circumspect manner.  My big problem is calling something drug discovery that is not.  I may come off as harshly anti-academic; I am not.  I am pro-good science and even more for the right things in the proper place.  I think this one is one of those.  In this paper, the authors describe their use of fragments to find inhibitors of Mycobacteria.  Thiolactomycin is a natural product of beta-ketoacyl acpM synthase (KAS) from M. tuberculosis, but exhibits broad spectrum anti-biotic activity. While it has rather low potency against M. tuberculosis KAS (200uM, 0.26 LEAN), it has favorable physicochemical properties, low cytotoxicity, high bioavailability, and activity in animal infection models.  This makes it an excellent target for optimization, and considering the activity and lack of selectivity, lots of it!

The authors decided to use inter-ligand NOE to guide their efforts.  This is an approach that has been used previously, and in a case of cosmic togetherness, against a different M. tuberculosis target.  Amzingly, that work is NOT cited in this paper.   As pointed out, iLOE has some issues, aggregation and so on.  The authors do not seem worried by this and instead are interested in utilizing selective iLOE.  One thing the authors do NOT do is include any detergent in their samples.  They synthesized PK940 and used this for the NMR studies.  

They were unable to detect iLOEs with mixing times shorter than 500 ms (that's long!) using the standard 2D NOESY, so they used a selective experiment.  This allowed them to shorten the mixing time, increase sensitivity, and solve some other NMR based problems.  [N.B. They had to use 900 MHz to get good enough separation for some of the methyl resonances.  I guess you have to justify that humongous tin can somehow.]  This was able to generate the needed NOE data to put into a model to give them an idea on how to make compounds, suggesting elaboration of the thiolactone at the 3 or 4 position (A below).

They started to explore these hypotheses, and found that activities were all within 2x fold of the original.  So, unlike others, they decided this was flat SAR. Although they do make some conjectures that would seem to be easily testable (FOLLOW UP PAPER?). None of the compounds were significantly better than the lead and many in fact lost the "slow onset" binding that is considered important for good PD.  There is nothing really ground breaking here, but what I like about this paper is where it was published Journal of Biological Chemistry, not Journal of Medicinal Chemistry.  This is where these types of papers belong.  As I keep on saying, your computation is only as good as your follow up.  In this case, the computation is to support the NMR and not the other way around.

Getting misled by NMR: ILOE artifacts

We’ve pointed out potential pitfalls with crystallography (here, here, and here) as well as with biochemical screening, but NMR has escaped attention– until now.

NMR has of course been a mainstay of fragment discovery methods since the original SAR by NMR paper. There have been plenty of developments since, but one that is particularly intriguing relies on the “interligand nuclear Overhauser effect,” or ILOE. In the “SAR by ILOE” approach, a 2D NMR experiment is used to detect when two small molecule ligands bind to a protein next to one another. There are some attractive features of this method. First, only ligands that bind in relatively close proximity to each other will generate a signal, thereby allowing researchers to identify fragments close enough to allow productive linking. Second, the technique can be applied to proteins that are too large to study by other NMR methods. In fact, it can be used even in the complete absence of structure. So what’s the problem?

In a new paper in J. Am. Chem. Soc., Chris Abell and colleagues at the University of Cambridge applied the approach to pantothenate synthetase (PtS) from M. tuberculosis. They previously did rigorous fragment screening followed by both linking and growing on this enzyme, which we discussed last year. Initial NMR experiments with compounds 1 and 2 (see figure) in the presence of PtS showed strong ILOE signals; the problem was that signals were seen between all the protons of compound 1 and all the protons of compound 2. This suggests non-specific binding: if the two molecules were binding next to each other in a single orientation you would expect that some protons from compound 1 would be closer to some protons in compound 2 than others, and there would thus be differences in signal intensities.

Adding a methyl group to compound 1 to give compound 4 didn’t help. In fact, there were ILOE signals from both the methyl groups of compound 4 to all the aromatic protons of compound 2, again suggesting non-specific binding. Even more damning, adding the substrates ATP and pantoate failed to significantly diminish the ILOE signals as expected; because crystallography showed these fragments bind in the active site, they should have been readily displaced by substrates.


Reasoning that the hydrophobic nature of compound 4 might be causing it to aggregate at high concentrations, the researchers appended a carboxyl group to give compound 5. NMR experiments with this compound in the presence of compound 2 and the protein now revealed specific ILOE signals between the 2-methyl group of compound 5 and H2 of compound 2. Moreover, this signal could be competed by adding ATP and pantoate.

Happily, linking these two fragments together resulted in compound 6, which bound to the enzyme three orders of magnitude more tightly than either of the starting fragments. The compound was also well-behaved mechanistically, showing competitive inhibition with ATP, and a crystal structure revealed that it binds as expected given the structures of the individual fragments.

Overall then this is a success story. However, it does suggest that the ILOE method may be more prone to aggregation artifacts than other biophysical methods. In particular, had the researchers not been able to do competition experiments (if, for example, they did not have another small molecule inhibitor available) they would have had a harder time sorting things out. Also, the researchers actually had crystal structures of both compounds 1 and 2 bound to PtS, so it is not clear how valuable the ILOE data really were for linking. Still, the potential advantages of an NMR-based method that doesn’t require structure are appealing. Hopefully we will see more applications of SAR-by-ILOE, now that people are more aware of the dangers.