29 January 2013

Fragment merging for Mcl-1

One of the most heroic examples of fragment-based drug discovery is navitoclax (ABT-263), which blocks the anti-apoptotic proteins Bcl-xL and Bcl-2 from binding to their partner proteins. This Abbott (AbbVie?) compound is in Phase 1 and 2 clinical trials for a variety of cancers. Abbott has also reported inhibitors of Bcl-2 that don’t inhibit Bcl-xL. However, many cancer cells are unfazed by inhibitors of Bcl-2 and Bcl-xL because they can instead rely on another protein, Mcl-1. Thus, ABT-263 can be overcome when cancer cells overexpress Mcl-1. Previously, Mcl-1 had been considered by many to be a “Teflon target.” Happily, it has now been successfully tackled with fragments.

The work, published recently in J. Med. Chem., was led by Stephen Fesik, now at Vanderbilt University. Fesik was one of the inventors of the SAR by NMR technique that led to navitoclax, and in this case the team used a similar approach, screening a fairly large fragment library (> 13,800 compounds) in pools of 12 using 1H-15N HMQC NMR. This produced 132 hits, of which two chemical classes were pursued.

One chemical class, exemplified by compound 2, consisted of 6,5-fused heterocyclic carboxylic acids, while another class, exemplified by compound 17, consisted of hydrophobic aromatic groups separated by a linker from a (usually) anionic substituent. NOE-guided fragment docking indicated that these compounds bind in similar but non-overlapping regions of Mcl-1, suggesting a fragment-merging approach.

Indeed, merging the compounds led to nanomolar binders such as compounds 60 and 53, which were also completely selective against Bcl-xL and more than 15-fold selective against Bcl-2. Crystal structures of these molecules bound to Mcl-1 confirmed the binding hypothesis. A number of additional analogs were synthesized; pleasingly, the SAR of the isolated fragments generally translated to the merged compounds.

This is a beautiful example of FBLD in academia. Of course, there is still a long way to go: there is a large and disconcerting disconnect between biochemical and cell-based potency for many reported Bcl-family inhibitors, and the lack of cell data here suggests that the same may hold true for Mcl-1. Still, it is nice to see that a venerable technique can succeed against this challenging protein.

24 January 2013

News and Updates

I am not sure how many of you follow the discussion in the LinkedIn FBDD group (Dan and I try to cross post as much as possible), but Ben Davis started a discussion based on a status update I had (how meta and 21st century of us).  How many commercially available fragment libraries come with the associated 1H spectrum (for NMR screening).  I only know of Maybridge's collection having 1H spectra.  However, I would think most companies would have the spectrum as part of their QC (or I hope they would).  

That leads into the second point of discussion: how do you QC your collection?  I would think LC-MS and NMR are a minimum.  But, what do people do for solubility?  The old stick-it-in-solution-and-see-if-it-craps-out or something more "science-y"?  

Lastly, I just received word this morning in my Inbox that Infarmatik has closed up shop.  I had heard it as a rumor, but now its real. 

21 January 2013

STD-SPR smackdown

Once you’ve established a library and chosen a target, the first step in FBLD is performing a fragment screen. There are lots of ways to do this, and since each method has its pros and cons it is best to use more than one. A good illustration of why this is important has just been published in J. Biomol. Screen. Results were also discussed last November at FBDD Down Under.

Two separate research groups were both interested in the core domain of HIV-1 integrase (IN). They both purchased 500-compound fragment libraries from Maybridge, though since they were purchased about six months apart they contained only 455 compounds in common. One group screened pools of 10 fragments by STD-NMR to identify 84 hits, of which 62 confirmed as single compounds both by STD-NMR and 15N-HSQC NMR. All of these were soaked into crystals of IN, resulting in 15 co-complexes.

The second group used SPR to screen each compound individually; compounds that showed a significantly stronger signal binding to IN than to a reference protein were confirmed by doing full dose-response curves. 16 hits were taken into crystallography, resulting in 6 co-structures, and another 3 gave ambiguous electron density.

The problem, as shown in the figure, is that there was no overlap between the confirmed NMR hits and the SPR hits, or between the crystallographically confirmed fragments!
To try to understand this discrepancy, the researchers re-tested the SPR hits by NMR, and the crystallographically confirmed NMR hits by SPR. The two assays were originally run under slightly different buffer and pH conditions, but these seemed not to be a significant factor. Eight of the 15 crystallographically-confirmed NMR hits did show activity in the SPR assay, but also hit the reference protein, so had not been taken forward. Another five had technical issues in the SPR screen (DMSO mismatches); only two showed no binding by SPR.

Five of the crystallographically confirmed SPR hits were retested by STD-NMR, though at a lower concentration (0.3 mM) than the original screen (1 mM) due to solubility issues. Four of these gave good signals, while the fifth produced a weaker signal that could only be detected at the pH of the original SPR screen. The reason the others may not have been detected initially could be because of competition in the original pooled NMR screen: with a 17% hit-rate, many pools probably contained multiple binders.

The title of the paper is “Parallel screening of low molecular weight fragment libraries: Do differences in methodology affect hit identification?” Clearly the answer is yes. Nonetheless, it is important to note that, at the end of the day, this may not matter so much. As the researchers observe:

We find that despite using different approaches with little overlap of initial hits, both approaches identified binding sites… that provided a basis for fragment-based lead discovery and further lead development.

In other words, no matter what technique you use, as long as you have a tractable target and you’re careful (and a little bit lucky) you’ll be able to find useful fragments.

14 January 2013

Poll Results - Hurray for Diversity

In our latest poll, we asked what kind of libraries people like, giving three options:
  • I like a maximal diverse library (SAR comes from follow up)    
  • I like diversity, but not at the expense of SAR (follow up is easier with some SAR)              
  • My target is teflon so any active fragment is welcome news.   
 60% of respondents like a maximally diverse library, 31% like diversity with some SAR, and 8% work of teflon targets, so any hit matter is welcome.  

The way I read this is that 60% of people don't consider the screen done when the first results come in.  In my eyes, the screen is over when there are actives identified with testable SAR hypotheses.  This is probably just my bias of having lived in a very resource constrained environment where follow up to a screen was a second serving of resources.  To me, this is great news; companies that are doing fragment screening are invested and not giving short shrift to these efforts. 

I would be curious to hear in the comments how people develop SAR with a maximally diverse library.  Do you just pick every available fragment that has the same central core and evaluate all possible side chains?  Would you apply a similarity cutoff of 0.9 or something?  How many compounds do you follow up with per active fragment?

10 January 2013

Fragment events in 2013


As far as we know there are only a few fragment-heavy events this year, all in the first half, but please leave a comment if you know of anything else.

March 4-5: Fragments 2013, the 4th RSC-BMCS Fragment-based Drug Discovery meeting, will be held at the Harwell Science and Innovation Campus near Oxford, UK. There is also a pre-conference training course on Sunday, March 3. Abstracts for posters are being accepted through January 31, with a special invitation to graduate students and postdocs.

March 19-20: Select Biosciences is holding its Discovery Chemistry Congress in Munich, Germany, with a full two days devoted to fragment-based lead discovery.

April 16-18: Cambridge Healthtech Institute’s Eighth Annual Fragment-Based Drug Discovery will be held in San Diego. You can read impressions of last year's meeting here, the 2011 meeting here, and 2010 here. Also, on April 15, Teddy and I will teach a short course on FBDD. Rumor has it this meeting will be moving to Boston in 2014, so if you're looking for an excuse to visit San Diego don't wait!

June 19-21: Cambridge Healthtech Institute’s Thirteenth Annual Structure-Based Drug Design will be held in Boston, with several talks on FBLD.

08 January 2013

Looking for trouble

Anyone who runs a fragment screen, especially for the first time, is likely to encounter problems. Large companies with sophisticated screening groups generally have a wealth of experience and procedures for dealing with false positives, but smaller organizations and academic labs can all too easily get lured into blind alleys. To help folks avoid these, Ben Davis and I are putting together a mini-review that summarizes the problems that can arise. While there are plenty of examples in the literature, we are also interested in hearing from you about artifacts and other problems that you’ve encountered but either not gotten around to publishing or decided against doing so. Feel free to leave comments here, anonymously if you wish, or email fbldproblems@gmail.com. Thanks – and may all your hits confirm!

02 January 2013

Fragments in the clinic: 2013 edition

It’s been more than two years since Practical Fragments updated its list of fragment-derived compounds in the clinic, and a lot has changed since then – mostly for the better. The latest list is inspired by a fantastic news article in Nature Review Drug Discovery that quotes a wide range of fragment-practitioners and outside experts. It’s a fun, fast read, so definitely check it out. It also includes a handy table of late-stage fragment-derived clinical compounds, their ClogPs, and their molecular weights, along with those of the initial fragment hits.

The list below borrows from this table and also includes molecules from other sources, whether or not they are still in development (indeed, some of the originator companies no longer exist). Those listed as still active in clinicaltrials.gov or company websites are in bold, and those that have been covered in Practical Fragments are hyperlinked to the relevant post.


Vemurafenib (PLX4032)        Plexxikon         B-Raf(V600E) inhibitor

Phase 2/3

MK-8931                                Merck              BACE1 inhibitor

Phase 2

AT13387                                 Astex              HSP90 inhibitor
AT7519                                   Astex              CDK1,2,4,5 inhibitor
AT9283                                   Astex              Aurora, Janus kinase 2 inhibitor
AUY922                         Vernalis/Novartis      HSP90 inhibitor
Indeglitazar                             Plexxikon         pan-PPAR agonist
Linifanib (ABT 869)                Abbott             VEGF & PDGFR inhibitor
LY2886721                             Lilly                 BACE1 inhibitor
LY517717                        Lilly/Protherics          FXa inhibitor
Navitoclax (ABT 263)              Abbott             Bcl-2/Bcl-xL inhibitor
PLX3397                                 Plexxikon        FMS, KIT, and FLT-3-ITD inhibitor

Phase 1

ABT-518                                 Abbott             MMP-2 & 9 inhibitor
ABT-737                                 Abbott             Bcl-2/Bcl-xL inhibitor
AZD3839                                AstraZeneca     BACE1 inhibitor
AZD5363                        AstraZeneca/Astex  AKT inhibitor
DG-051                                  deCODE            LTA4H inhibitor
IC-776                                   Lilly/ICOS         LFA-1 inhibitor
JNJ-42756493                     J&J/Astex         FGFr inhibitor
LEE011                             Novartis/Astex      CDK4 inhibitor
LP-261                                   Locus               Tubulin binder
LY2811376                              Lilly                 BACE1 inhibitor
PLX5568                                 Plexxikon         kinase inhibitor
SGX-393                                 SGX                 Bcr-Abl inhibitor
SGX-523                                 SGX                 Met inhibitor
SNS-314                                 Sunesis            Aurora inhibitor

There are some interesting trends, such as the number of BACE1 inhibitors – a fact the Nat Rev Drug Disc piece also notes. This has been an immensely difficult target, so it’s nice to see fragment-based approaches deliver compounds to the clinic. Whether BACE1 inhibitors will ultimately prove useful for treating Alzheimer’s disease remains to be seen, but at least FBLD has provided the tools to test this hypothesis.

The current list contains 26 clinical-stage drugs but is certainly incomplete, particularly in Phase I. If you know of any others (and can mention them!) please leave a comment.