Novel Assay (SPR) Leads to Novel Allosteric Binding Site (nAChR)

I love this blog.  But sometimes it drags.  Finding new articles to blog about can be hard, there seems to be waves.  So, I love it when someone points one out to me.  This article was sent to me by someone associated with it.  We don't always blog those articles people point out to us, but this one has all the hallmarks: high visibility journal, good target, and "innovative" in the approach.  So, let's dive in and see what we have here.

Nicotinic receptors (nAChR) are pentameric, ligand-gated ion channels.  nAChR mutations are implicated in a wide range of neurological disorders and are the targets for a wide range of current drugs.  It is important to note that most of these drugs work through an allosteric site distant from the agonist binding site.  Most high resolution, structural data is from homologous molluscan Acetylcholine binding proteins.  The orthosteric site is located at an interface of a "principal" and "Complementary" subunit.  Ligand-binding induces conformational changes in the ligand binding site which are coupled to the ion opening.  There is little information on structural implications at allosteric sites.  The authors decided to address this unmet need with a chimeric human α7 ligand binding domain and AChBP which has 71% sequence similarity to the native protein, compared to 33% for the most commonly used target (Aplysia).  

The authors used SPR against a target with a blocked orthosteric site.  They did this one of two ways: 1. pre-incubation with a high affinity orthosteric ligand or 2. mixing each fragment with an orthosteric ligand of lower affinity.  This second approach (Figure 1) allows detection of fragments not competing for binding to the orthosteric site and were therefore potential allosteric ligands.  

Figure 1.  To distinguish allosteric binders from competitive binders using SPR spectroscopy we perfused each fragment alone (green triangle) or in combination with the competitive antagonist d-tubocurarine (black circle). In the case of an allosteric binder, the response units observed for the mixture of fragment + d-tubocurarine is close to the sum of fragment alone + d-tubocurarine alone (blue dashed line). No competition exists because the fragment and d-tubocurarine bind at distinct sites. (C) In the case of a competitive binder, the response units for the mixture of fragment + d-tubocurarine is lower than the sum of fragment alone + d-tubocurarine alone because both compounds compete for binding at the same site. (D) Example traces for fragment 4, which was identified as one of the allosteric binders in this study.

 In screening 3000 fragments, they found 300 putative allosteric binders.  Follow up, including dose-response, led to 24 fragments being selected for co-crystallization.  Crystal trials were set up using blocked nAChR and fragments that were soluble at 5-10mM; yielding in the end 5 crystal structures.  All five proved to be allosteric binders in three separate locations, including one never observed before (top pocket).  The most potent of these fragments had a IC50 of 34 uM, while the least potent was 400 uM. 

Figure 2.  Allosteric Binding Sites

This is a really nice piece of work.  I think the SPR assay is really clever and I would expect that many people will now be taking a similar approach to discovering allosteric sites in their targets.

Fragments vs DDR1/2: a chemical probe

Our last post was about the utility of chemical probes: small molecules with sufficient potency and selectivity to be able to address specific biological questions. A recent paper in ACS Med. Chem. Lett. by Chris Murray and colleagues at Astex describes an excellent example of finding a new probe for the discoidin domain receptors (DDR1 and DDR2). Previous publications had suggested a role for these receptor tyrosine kinases in certain types of lung cancer, but some of this work had relied on non-selective inhibitors.

The researchers started with a thermal shift assay of their 1500 fragment library against DDR1, followed by crystallographic screening, resulting in around 50 fragment-protein complexes. Not surprisingly most of the fragments bound in the hinge-binding region of the kinase, but around 10 bound in the so-called “back pocket”, with the protein in the inactive DFG-out conformation. As the researchers point out, it is rare to see fragments binding here.

Compound 1 was of particular interest. At first glance, it's not impressive: with 18 heavy atoms and MW > 250 Da, it is on the large end for an Astex fragment, and it had low activity in a biochemical assay. However, its binding mode revealed potential areas for improvements, and the methylene was an unusual feature in a back-pocket binder.

The first step in improving affinity was to add a hinge binder. This was done with the aid of an in-house program called AstexMerge, based on the program BREED, which superimposes a set of ligands. The user chooses a starting molecule, and the program tries to merge that with other molecules while taking account of bond angles and distances. This process led to the design of compound 2, and a few tweaks quickly led to compound 4.

Although compound 4 was potent against DDR1 and 2 and showed good cell-based activity in a phosphorylation assay, it did potently inhibit some other kinases too, most notably c-kit. That problem was fixed with further medicinal chemistry, notably addition of a methyl group to the previously mentioned methylene and replacement of the urea, leading to compound 9, which was potent, selective, and showed good pharmacokinetics in mice.

Although compound 4 was not completely selective, it was more so than some of the previously described molecules, so the researchers tested it in lung cancer cells and found that, despite the fact that it inhibited DDR2 phosphorylation, it showed no effect on cell proliferation. Thus, “the project was halted in favor of more attractive targets.”

Clearly the researchers didn't start out trying to disprove the role of DDR1/2 in squamous cell lung cancer, but their efforts will save others from pursuing the same course. The publication introduces some attractive chemical probes for interrogating the biology of these receptors; hopefully one of these molecules will be added to the Chemical Probes Portal as an alternative to the less-selective probes that have been used in previous studies. Who knows, perhaps someone will find another indication for which DDR1/2 inhibitors are just the ticket.

Introducing the Chemical Probes Portal

Chemical probes can be incredibly powerful reagents for understanding biology. A potent, selective, and cell-active modulator of a specific protein can be invaluable for figuring out what that protein actually does. Fragment-based methods can be effective at identifying these tool compounds, as we've described here and here.

Unfortunately, good chemical probes are difficult to discover, and scientists are left struggling with suboptimal reagents that hit multiple targets, often through pathological mechanisms. This leads to "pollution of the scientific literature," in Jonathan Baell's memorable phrasing. Despite our occasional PAINS Shaming, high-profile articles in C&EN and Nature, and even a dedicated blog, the problem continues. What is to be done?

Yesterday, a team of 53 authors from 46 academic and industrial organizations published a Commentary in Nature Chemical Biology entitled "The promise and peril of chemical probes" (see here for excellent coverage in Nature, here for Science's take, and here for In the Pipeline). This provides a good working definition for a chemical probe. According to the Structural Genomics Consortium, a chemical probe for epigenetics targets must have:

• Potency < 100 nM against the desired target
• >30-fold selectivity vs related targets
• On-target cell activity < 1 µM

It should also be profiled against a larger panel of potential off-targets, and a related inactive compound (such as a stereoisomer) should be available as a control.

After discussing examples of high-quality probes, the researchers turn their attention to what they term – rather charitably – "probes of lesser value:"

The continued use of these probes poses a major problem: tens of thousands of publications each year use them to generate research of suspect conclusions, at great cost to the taxpayer and other funders, to scientific careers and to the reliability of the scientific literature.

The authors then go on to describe best-practices. For example, even high-quality probes can give spurious results when used at high concentrations. As Paracelsus recognized five centuries ago, the dose makes the poison.

All of this is important, but as the authors acknowledge, it's been said before. What really differentiates the Commentary is the simultaneous launch of a companion web site, the Chemical Probes Portal. Its creators hope that this will lead to vigorous community discussion around questions such as:

Is there a probe for my target protein?

Which ones should I use?

How should I use this probe properly?

Is this probe suitable for use in animal models?

Currently the Portal lists just seven probes with links to references and descriptions of selectivity, solubility, and the like. All of these are “good probes,” but hopefully this will expand: the paper itself discusses the shortcomings of molecules such as staurosporine, chaetocin, obatoclax, and gossypol, and including them in the portal with detailed warnings would be valuable for the scientific community.

I hope this takes off. Understanding the natural world is hard enough even with well-behaved reagents and carefully controlled experiments. Practical Fragments will check back in a year or so to see how the site is doing. In the meantime, probe cautiously!

Covalent Inhibitor of KRas

So, Ras is big.  We keep on talking about it.  And sometimes we talk about the same work repeatedly.  This recent paper from AZ is a publication of work we have talked about here and here.  This follows on closely to work done by Vanderbilt and Genentech.  Those two papers were done using NMR and this one took a X-ray approach.  The AZ folks were taking a different approach to this PPI: stabilization of the interface.  They took 1160 fragments in pools of 4 and screened against HRas (homolog)-catalytic domain of SOS stable complex.  There were able to identify 3 bindings sites on HRas-SOS (Figure 1):

Figure 1.  HRas-SOS Complex.  HRas (Green), SOS (Blue), A: SOS binding site  (gold) (same as Vanderbilt), B SOS-Hras Interface binding site (Red) (same as Genentech), and C HRas covalent binding site (black).

 

Site A was the same site identified by the Vanderbilt group  Site B was the same as identified as Genentech.  However, the AZ compounds bound to both proteins at the interface.  Their initial hope was to use this site to stabilize the Ras-SOS interface.  Both of the fragments binding to these sites had their affinity determined by TROSY-HSQC NMR.    However, they were not potent enough to elicit a biological effect, which was not unexpected.  After several rounds of chemistry, they were not able to improve these fragments significantly, or even show that they actually stabilized the interface.  

Looking at the growing covalent literature, they hypothesized that an irreversible inhibitor may be the only way to inhibit GTPase activity, especially considering the pM affinity of GTP for Ras.  They identified Cys118R (conserved between HRas and KRas) as a potentially reactive sidechain proximal to the GDP binding site on Ras.  To go after this site covalently, AZ assembled a 400 fragment covalent library (Figure 2) and screened it by mass spectrometry.

Figure 2.  Chemotypes represented in AZ 400 fragment covalent library.

They chose the N-substituted maleimide was deemed "ideal"; other warheads were either insufficiently reactive or overly reactive.  Covalent modification of Cys118R by a fragment partially occludes the nucleotide binding site and potentially prevents the reorganization of the Cys118R loop, thus locking it into the catalytically inactive Ras-SOS complex.  Interestingly, their covalent compounds only inhibited catalytically activity when pre-incubated with Ras-GDP-SOS.  This supports the hypothesis that Cys118R becomes more accessible during SOS-mediated nucleotide exchange.  

This paper brings together several topics which I think are becoming hot: covalent fragments, mass spectrometry, and K-Ras. 

Fragments vs BTK: metrics in action

Sometimes the discussions over metrics, such as ligand efficiency, can devolve into exegesis: people get so worked up over details that they forget the big picture. A recent paper in J. Med. Chem. by Chris Smith and (former) colleagues at Takeda shows how metrics can be used productively in a fragment-to-lead program.

The researchers were interested in developing an inhibitor of Bruton’s Tyrosine Kinase (BTK) as a potential treatment for rheumatoid arthritis. This is the target of the approved anti-cancer drug ibrutinib, but ibrutinib is a covalent inhibitor, and the Takeda researchers were presumably concerned about the potential for toxicities to arise in a chronic, non-lethal indication. Many of the reported non-covalent BTK inhibitors are large and lipophilic, with consequently suboptimal pharmacokinetic properties. Thus, the team set out to design molecules with MW < 380 Da, < 29 non-hydrogen atoms (heavy atoms, or HA), and clogP ≤ 3.

The first step was a functional screen of Takeda's 11,098 fragment library, all with 11-19 HA, comfortably within the bounds of generally accepted fragment space. At 200 µM, 4.6% of the molecules gave at least 40% inhibition. Hits that confirmed by STD NMR were soaked into crystals of BTK, ultimately yielding 20 structures. Fragment 2 was chosen because of its high ligand efficiency, novelty, and the availability of suitable growth vectors.

Close examination of the structure suggested a fragment-growing approach. Throughout the process, the researchers kept a critical eye on molecular weight and lipophilicity. This effort led through a series of analogs to compound 11, with only 24 heavy atoms and clogP = 1.7. This molecule is potent in biochemical and cell-based assays and has excellent ligand efficiency as well as LLE (LipE). Moreover, it has good pharmacokinetic properties in mice, rats, and dogs, with measured oral bioavailability > 70% in all three species. Finally, compound 11 shows efficacy in a rat model of arthritis when dosed orally once per day.

Although compound 11 is selective over the closely related kinase LCK, unfortunately it is a double digit nanomolar inhibitor of oncology-related kinases such as TNK2, Aurora B, and SRC, which would probably be unacceptable in an arthritis drug. Nonetheless, this study is a lovely example of fragment-growing guided by a strict commitment to keeping molecular obesity at bay.

Merging Fragments for Matriptase

We often talk about methods here: how to screen, how to prosecute those actives, and everything in between.  This is one of those what you do with the actives posts.  In this paper, a group from Aurigene and Orion present their results on Matriptase.  There have been multiple reported matriptase inhibitors, small molecule and peptide based.  Previous work from this group showed compounds that were active in cell-based migration and invasion assays and in mice with tri-substituted pyridyls and benzene compounds.  For this work, they take a SBDD approach:  "structure divulges a trypsin-like S1 cavity, a small hydrophobic S2 subpocket, and a solvent exposed spacious S4 region."

In screening benzamidine fragments (MW less than 300) they found 2 actives, ~80uM (Figure 1).

Figure 1.  Benzamidine screening actives

These were modeled in to the active site and obviously the amidine moiety went into S1.  Cpd 1's benzene moiety went nicely into S4 while 2's piperidyl went into S1'.  S4 easily accepted the more hydrophobic napthyl instead of phenyl and then they decided to see if the napthyl compound and 2 could be "linked" and the beta carbon. [So, my first quibble here is that this is not really a linking approach; this is fragment merging. Linking involves modeling, SBDD, and discovery of different linkers.  Its very difficult to do without specialized methods.  What they did here was see huge spatial overlap of compounds and voila, "we can add something right here".]  Well, not surprisingly, this worked.  They describe their SAR around each pocket to pick the compounds to merge, go read it if that interests you. The did crystallize the merged compound and it confirmed the modeling.  The final compound showed activity in cell-based assays and in mice.  That's good.  

This work can be summarized as follows: if you have significant spatial overlap you have a very good chance of merging disparate moieties.  So, two things bother me here.  First, the actives 1 and 2 are mighty big for fragments (more than 22 HAC).  That's fine, tomato...to-mah-to.  The final compound ends up pretty honking big too (37 HAC).  What is really bothersome, at least to me, is the LE.  Both actives start well below 0.2 (for a protease!) and they never improve on it. Now, Pete may disagree, but metrics have a place in FBDD.  Does the LE metric in this case tell us anything? 

Fragments vs 53BP1

As we’ve noted (repeatedly), epigenetics is big. However, much of the focus has been on bromodomains, which recognize acetylated lysine residues. In a paper published earlier this year in ACS Chem. Biol., Lindsey James, Stephen Frye and collaborators at the University of North Carolina, the University of Texas, the Mayo Clinic, and the University of Toronto describe their efforts on a protein that recognizes methylated lysine residues (a Kme reader).

The protein 53BP1 is involved in DNA repair and could have anticancer potential. It recognizes a dimethylated lysine sidechain within a histone protein, so the researchers screened a set of molecules containing amines to mimic this moiety. They used an AlphaScreen assay, with each compound at 100 µM. This does not appear to have been a library of fragments (and unfortunately the number of compounds screened was not stated), but the most notable hit was the fragment-like UNC2170.

Although the affinity was modest, it was quite selective for 53BP1, showing no activity up to 500 µM against 9 other Kme readers. Since AlphaScreen assays can be prone to false positives (the original PAINS compounds were identified in this assay), the researchers tested their compound using ITC, which gave a dissociation constant of 22 µM, in good agreement with the AlphaScreen assay, though with unusual stoichiometry (more on that later).

Thus encouraged, the researchers set off to optimize their hit. Initially they tried modifications around the amine, but even changes as subtle as adding or removing a methyl group killed activity. Attempts to rigidify the propyl linker were also unsuccessful, and shortening it or lengthening it failed too. Replacing the amide with a sulfonamide or amine abolished activity. Most substitutions around the phenyl ring also gave dead compounds, though the bromine atom could be replaced with similarly hydrophobic moieties such as iodine, isopropyl, or trifluoromethyl. Many other analogs were made too, all to no avail. Though the text is measured, the frustration is palpable.

Ultimately the researchers were able to solve the crystal structure of the compound bound to 53BP1, which produced a surprise: one molecule of UNC2170 binds to two molecules of protein, making interactions with each. This explains the stoichiometry seen in the ITC data. It also explains the intolerance to substitutions, as “the ligand is encircled by both proteins,” with no room for modifications.

Happily, UNC2170 is highly cell permeable and non-toxic, and does show some modest activity in cell-based assays. Hopefully the researchers will ultimately find more potent compounds, though this may require a different approach. Indeed, another Kme reader also proved to be quite challenging, but was amenable to fragments. It would be fun to see whether an explicit fragment screen produces more tractable starting points against 53BP1.

Updated: fragment events in 2015 and 2016

It is hard to believe that the year is already half over, but there are still important events coming up, and 2016 is already starting to take shape!

2015

August 11-13: The OMICS Group is holding a conference entitled Drug Discovery & Designing in Frankfurt, Germany, with FBDD listed as a conference highlight.

September 22-24: Newly added! CHI's Thirteenth Annual Discovery on Target will be held in Boston again, and fragments play an important role in several tracks, including epigenetics and kinases. Also, Teddy and I will be teaching a short course on Targeting Protein-Protein Interactions on Monday, September 21.

December 15-17: More than 40 presentations. 8 countries. 3 days. One event:
The first-ever Pacifichem Symposium devoted to fragments.

The Pacifichem conferences are held only once every 5 years in Honolulu, Hawaii to bring together scientists from Pacific Rim countries including Australia, Canada, China, Japan, New Zealand, and the US. Registration is now open!

2016

February 21-24: Zing conferences is holding its inaugural Structure Based Drug Design Conference in Carlsbad, California. This looks like a cousin of last year's Caribbean meeting, so it should be a lot of fun.

April 19-22: CHI’s Eleventh Annual Fragment-Based Drug Discovery, the longest-running fragment event, will be held in San Diego. You can read impressions of this year's meeting here, here, and here; last year's meeting here and here; the 2013 meeting here and here; the 2012 meeting here; the 2011 meeting here; and 2010 here.

October 9-12: Finally, FBLD 2016 will be held in Boston, MA. This marks the sixth in an illustrious series of conferences organized by scientists for scientists, the last of which was in Basel in 2014. Surprisingly, this also seems to be the first dedicated fragment conference in Boston. You can read impressions of FBLD 2012, FBLD 2010, and FBLD 2009.

Know of anything else? Add it to the comments or let us know!