NICELE: everyone gets an A+!

Since the invention of ligand efficiency (LE) more than two decades ago, scholars have been debating its strengths and weaknesses. And alternatives; we've written about LLE, LLE_AT, LELP, %LE, WTF, and more. Just last week we discussed CRE.

 

But all these metrics (with perhaps one exception) have a problem: they sometimes tell you that your molecule is not as good as you hoped. And who wants to hear that? 

 

Maybe you used cutting-edge AI technology to design your ligand, and then moved heaven and earth to make it. When you finally test it the result comes back - meh. But that doesn't mean the haters should win! 

 

You can't (or really shouldn’t) change the assay results, but you can change the scoring system, and now researchers at the University of Durak have come up with a new metric called NICELE (No Insulting Critical Evaluation Ligand Efficiency). Unlike xLE, the calculation is simple:

NICELE = LE + F, where F = (1.0 - LE)

Let's say your fragment comes in with a LE of 0.18 kcal/mol per non-hydrogen atom - not so good. That's OK, the NICELE is a much more impressive 1.0!

 

As a bonus, because LE is removed from the final calculation, NICELE is independent of standard state assumptions, so Dr. Saysno should embrace it.

 

Of course, NICELE makes it harder for those molecules with truly impressive ligand efficiencies to stand out, but they're a bunch of elitists anyway.  

Ligand reactivity efficiency (LRE)

As covalent drug discovery continues to rise, the demand for metrics to help guide lead optimization is increasing. Last year we discussed covalent ligand efficiency (CLE). In an open-access paper just published in J. Med. Chem., Benjamin Horning, Brian Cook, and colleagues at Vividion Therapeutics describe ligand reactivity efficiency (LRE). (Benjamin presented LRE at the DDC meeting in 2024.)

 

A key challenge when developing covalent ligands is maximizing specific reactivity towards the target of interest while minimizing intrinsic reactivity towards other proteins; the two types of reactivity are not the same, as we wrote about last year. For molecules that target cysteine residues, intrinsic reactivity is usually determined by assessing reactivity against the small molecule glutathione, which is abundant in cells.

 

For lead optimization more generally, a common metric is lipophilic efficiency (LLE or LipE, see here and here), in which the logP of a molecule is subtracted from the negative log of the IC₅₀ (pIC₅₀). More lipophilic molecules have higher logP values, so maximizing LLE helps to minimize increases in lipophilicity.

 

By analogy, the researchers defined LRE to help minimize increases in intrinsic reactivity. However, distinguishing specific from intrinsic activity is not necessarily straightforward. As we previously discussed, IC₅₀ alone is an inappropriate measurement for covalent inhibitors; the incubation time before the IC₅₀ is measured is an essential variable. The most rigorous value is k_inact/K_I, and although this ratio has been historically time-consuming to determine, we described an easier method earlier this year. Yet an even simpler measurement is the TE_{50(target, 1h)}, the concentration of compound necessary to label 50% of a target after one hour, which is a function of k_inact/K_I. The researchers thus defined LRE as:

 

    LRE = pTE_{50(target, 1h)} – pTE_{50(GSH, 1h)}

 

The variable in the second term, pTE_{50(GSH, 1h)}, is calculated from the reaction rate of the ligand with glutathione; intrinsically reactive ligands have higher rates.

 

In the case of LLE, values above 5 or 6 are generally considered acceptable for advanced leads, and the same is true for LRE. For example, a molecule with TE_{50(target, 1h)} = 10 nM and a (low) GSH reactivity of 0.01 M^-1s^-1 would have an LRE = 6.3. Also analogous to LLE, one can generate plots with pTE_{50(GSH, 1h)} on the x-axis and pTE_{50(target, 1h)} on the y-axis to assess whether LRE values are improving during a lead optimization campaign.

 

In my view, LRE is superior to previously discussed CLE because it explicitly considers the time component. A one hour incubation is practical; a ligand with k_inact/K_I = 10,000 M^-1s^-1 would have TE_{50(target, 1h)} = 19 nM. Also, LRE is more intuitive for medicinal chemists than CLE due to its similarity to LLE.

 

On the minus side, the researchers note that some of the assumptions break down for ligands with high non-covalent affinity (low K_I). Also, some folks may take issue with metrics that take the logarithm of a measurement that has units.

 

The researchers note another alternative metric, the reactivity enhancement factor (REF), which I briefly discussed here. REF is simply the ratio of the specific reactivity to the intrinsic reactivity, which is conceptually simpler to me than LRE. Nonetheless, the researchers state that LRE is commonly used at Vividion, which has put several covalent drugs into the clinic, so clearly it can be useful. Whether REF, LRE, or CLE, ultimately the choice of metric is less important than the ultimate goal: maximizing specific reactivity while minimizing intrinsic reactivity.

xLE: solving problems or missing the point?

Ligand efficiency (LE) has been discussed repeatedly and extensively on Practical Fragments, most recently in September. Two criticisms are its dependence on standard state and the observation that larger molecules frequently have lower ligand efficiencies than smaller molecules. In a just-published open-access ACS Med. Chem. Lett. paper, Hongtao Zhao proposes a new metric, xLE, to address these concerns.

 

LE is defined as the negative Gibbs free energy of binding (ΔG) divided by the number of non-hydrogen (or heavy) atoms, and of course ΔG is state-dependent. Standard state assumptions are 298K and 1M concentrations, choices that some people see as arbitrary since few biologically relevant molecules ever achieve concentrations near 1M. To remove the dependence on standard state, Zhao proposes to remove the translational entropy term of the unbound ligand from the free energy calculation.

 

Zhao also addresses the second criticism, that larger molecules often have lower ligand efficiencies. This phenomenon was observed in an (open-access) 1999 paper titled “the maximal affinity of ligands,” which found that, beyond a certain threshold, larger ligands do not have stronger affinities; there are very few femtomolar binders even among the largest small molecules. Thus, Zhao proposes attenuating the size dependence.

 

The new metric, xLE, is defined as follows:

 

xLE = (5.8 + 0.9*ln(M_w) – ΔG)/(a*N^α) - b

Where N is the number of non-hydrogen atoms, α is chosen to reduce size dependence, and a and b are “scaling variables.” He chooses α=0.2, a=10, and b=0.5, with little explanation.

 

To assess performance, Zhao examined nearly 14,000 measured affinities from PDBbind. When plotted by number of atoms, median affinity increased up to about 35 heavy atoms but then leveled off. Median LE values decreased sharply from 6 to 12 heavy atoms and then leveled off somewhere in the 20s. But median xLE values were consistent regardless of ligand size.

 

Zhao also examined LE and xLE changes for 175 successful fragment-to-lead studies from our annual series of J. Med. Chem. perspectives. LE decreased from fragment to lead for 48% of these, but xLE increased for all but a single pair.

 

And this, in my opinion, is a problem.

 

In the seminal 2004 paper, LE was proposed as "a simple ‘ready reckoner’, which could be used to assess the potential of a weak lead to be optimized into a potent, orally bio-available clinical candidate." The metric was particularly important before FBLD was widely accepted, when chemists were even less inclined to work on weak binders.

 

Here is the situation for which LE was devised. Imagine two molecules, compounds 1 and 2. The first has just 12 non-hydrogen atoms, a molecular weight of 160, and a modest 1 mM affinity for a target - similar to some fragments that have yielded clinical compounds. The second is much larger: 38 non-hydrogen atoms, a molecular weight of 500, and 10 µM affinity for the same target. Considering potency alone, compound 2 is the winner.

 

However, the LE for compound 1 is a respectable 0.34 kcal/mol/atom, while the LE for compound 2 is 0.18 kcal/mol/atom. So while a 10 µM HTS hit may initially look appealing, the LE suggests that this is an inefficient binder, and further optimization may require adding too much molecular weight to get to a desired low nanomolar affinity.

 

In contrast, the xLE values for both compounds are nearly identical, 0.38, and so this metric would not help a chemist prioritize which hit to pursue. In other words, xLE does not provide the insight for which LE was created. It might even lead to suboptimal choices. 

 

Moreover, unlike LE, xLE is non-intuitive. And finally, with three scaling or normalization factors, xLE is arguably even more arbitrary than a metric dependent on the widely-accepted definition of standard state.

 

Personally I find the practical applications of xLE limited, but I welcome your thoughts.

Covalent ligand efficiency

Ligand efficiency (LE) was proposed more than two decades ago as “a useful metric for lead selection.” The concept is simple: divide the binding energy of a ligand by the number of non-hydrogen (or heavy) atoms (HA). The higher the number, the higher the binding energy per atom, and thus the more “efficiently” the ligand binds to the protein. LE is particularly useful in fragment-based lead discovery when prioritizing among differently sized hits to ensure that small, weak molecules are not overlooked. While some have criticized the metric’s dependence on standard state, drug hunters have repeatedly found it to be useful, as we’ve discussed here, here, and here.

 

Irreversible covalent drugs are a horse of a different color - or perhaps a different species entirely. Because of their two-step mechanism, binding followed by bonding, time is an essential parameter, and the proper way to characterize them is with the ratio k_inact/K_I. Is it possible to develop a covalent ligand efficiency metric? This is the task that György Ferenczy and György Keserű at HUN-REN Research Centre for Natural Sciences and Budapest University of Technology and Economics set for themselves in a recent (open-access) Drug Discovery Today paper.

 

As we wrote just a couple months ago, an important distinction for covalent drugs is specific vs chemical reactivity: you want the first to be high and the second to be low. For cysteine-reactive molecules, this distinction is often assessed by measuring the rate of reaction with the abundant cellular thiol glutathione (GSH). The researchers sought to incorporate this parameter into their definition of covalent ligand efficiency (CLE) as follows:

 

CLE = LE – LE(GSH) = (-1.4*log₁₀(IC_50,t)/HA) - (1.4*log₁₀(k^2nd_sur*t)/HA)

Where IC_50,t is half maximal inhibitory concentration at time “t” and k^2nd_sur is the second-order rate constant of the ligand reacting with a surrogate nucleophile such as GSH.

 

The researchers cataloged multiple covalent modifiers from the literature. Some had reported glutathione reactivity data. For the rest, the researchers estimated these values based on analogs. They went on to calculate CLE values for the protein-ligand pairs. Laudably, all of these data are provided in the supplementary data.

 

So, how useful would CLE have been in prior lead discovery campaigns? The researchers calculated CLE values for the BFL1 covalent fragment hits we wrote about here. The potencies of the six reported fragment hits varied, reflected in k_inact/K_I values, from 0.7 to 9.5 M^-1s^-1. But their CLE values spanned a narrower range, from 0.08 to 0.12. The fragment that was successfully optimized was one of the most potent, with a k_inact/K_I of 7.5 M^-1s^-1, but had a CLE of just 0.09. If anything, CLE would have deprioritized this fragment, at odds with the stated goal that “CLE is designed to support compound priorization.”

 

As we discussed earlier this year, the researchers previously proposed that covalent fragments may need to be larger than reversible fragments. If this is true, then normalizing for size may be less important for covalent ligands than for noncovalent ones, which can be very small and weak yet still valuable. Indeed, the researchers’ analysis of covalent ligands from the literature shows a smaller range of CLE values than LE values.

 

The researchers acknowledge other oddities too: “there is no smooth transition from CLE to LE as the reactivity of ligands decreases. Moreover, CLE can take negative values for compounds with low affinity and high reactivity.”

 

But for me, the biggest liability is the fact that – unlike LE for reversible binders or k_inact/K_I – the value of CLE depends on the time the measurement was taken. (In the paper, the researchers use a 1 hour incubation, so I would propose the annotation CLE_1h.) This makes it difficult to compare CLE values taken at different time points.

 

The first word of this blog is “practical,” and I’m not convinced this adjective applies to CLE, though I applaud the effort. The popularity of LE spawned a cottage industry of other metrics, some of which we summarized in a 2011 post. I confess that I had nearly forgotten about some of them, but I think they were a useful way for the field to grapple with what characteristics mattered. As covalent drug discovery becomes increasingly popular, perhaps we will see a similar proliferation of metrics. (Indeed, we already wrote about another one here.) It will be interesting to revisit these a decade hence to see which ones have caught on.

Fragments vs KAT6A: ligand efficiency in action

Epigenetic writers such as histone acetyltransferases (HATs) control gene expression, which often goes awry in cancer. The gene encoding the lysine acetyltransferase KAT6A, for example, is amplified in multiple types of cancer, and its overexpression is associated with poor clinical outcomes for patients with ER⁺ breast cancer. A recent paper in Bioorg. Med. Chem. Lett. from Andrew Buesking and colleagues at Prelude Therapeutics reports a new series of inhibitors.

 

The researchers started by considering previously reported molecules such as PF-9363. Recognizing the importance of the central sulfonamide, they generated a library of 150 fragments containing this core and screened them in a biochemical assay. Taking a similar approach as other researchers reporting on a different epigenetic target we discussed last year, the Prelude team explicitly used ligand efficiency to call hits, setting the cutoff at > 0.3 kcal per mol per heavy atom.

 

A direct deconstruction of PF-9363 led to compound 6, but, seeking something novel, the researchers were drawn to compound 8, with similar ligand efficiency. Modifications were made to both of the terminal aromatic groups, leading to compound 13, which was co-crystallized with KAT6A. This led to further structure-guided modifications, with compound 25 being the most potent.

 

Unfortunately, further improvements in potency were not forthcoming, and the ligand efficiency had dropped steadily from both the initial fragment as well as PF-9363. The researchers conclude by stating that they “decided to pursue alternative approaches.” Still, this paper is a nice, concise example of using metrics both for pursuing a chemical series and, ultimately, discontinuing it.

Efficiency metrics in action for a bromodomain inhibitor

The metrics ligand efficiency (LE) and lipophilic ligand efficiency (LLE or LipE) are frequently used during fragment-to-lead optimization. A recent paper in J. Med. Chem. by Philip Humphreys and colleagues at GSK describes how they were useful in developing an “oral candidate quality” inhibitor of BET-family bromodomains.

 

Practical Fragments has written frequently about bromodomains, which bind to acetyl-lysine residues in histones to epigenetically modulate gene transcription. Some 17 bromodomain inhibitors have entered the clinic, of which at least three (pelabresib, PLX51107, and ABBV-744) came from fragments. GSK was an early pioneer in the field, and researchers there were interested in using fragments to develop a differentiated class of molecules that would inhibit all four members (BRD2, BRD3, BRD4, and BRDT) of the BET family, each of which contains two separate bromodomains designated as either BD1 or BD2.

 

GSK already had BRD4 BD1 binding data for 50,000 compounds, and these were analyzed to find molecules with LE>0.3 kcal/mol per atom that were structurally differentiated from known bromodomain binders. Compound 9 was quite potent and had high LE as well as respectable LipE. (As the researchers note, LE is “the more relevant metric” for fragments, with LipE becoming increasingly important during later optimization.) A crystal structure of this molecule superposed onto another bromodomain inhibitor suggested that adding a methyl group to fill a small pocket could boost affinity, and this was confirmed by compound (R)-10. This molecule showed cell activity and good permeability, although hepatocyte stability was poor, likely due to the two methoxy groups. Removing these led to compound 12, the most ligand-efficient compound that had been seen. (All values in the figure below are for binding to BRD4 BD1.)

 

Compound 12 mimics the N-acetyl lysine residue of the natural ligand, and previous research had revealed two additional regions of the bromodomain that could be targeted for enhanced affinity, the so-called “WPF shelf” and the “ZA channel.” Structure-based design was used to independently explore both areas, leading to compounds such as 24 and 31. In addition to assessing LE and LipE, the researchers paid close attention to other factors such as permeability. Virtually combining the best moieties that bind at the WPF shelf and ZA channel led to 770 potential molecules to make, which were winnowed down to just 40 on the basis of predicted lipophilicity (specifically chromLogD_pH7.4), molecular weight, and TPSA. The best of these were more extensively profiled, including in pharmacokinetic studies. I-BET432 emerged as the winner.

 

I-BET432 binds tightly to both bromodomains of the four BET family proteins and is at least 80-fold selective against two dozen other bromodomains. It shows excellent oral bioavailability in rats and dogs, does not inhibit hERG, is not mutagenic in an Ames test, and does not inhibit CYP3A4. The molecule is also clean in a panel of four dozen off-target proteins. Human oral dose predictions come in at 5-18 mg per day. A crystal structure of the molecule bound to BRD2 BD2 showed the expected binding pose, and that the two alcohol substituents may be forming an intra-molecular hydrogen bond, which could explain the high permeability.

 

This is a nice case study in metric-driven optimization. As the researchers note, I-BET432 “has the highest LipE (6.2) and LE (0.43) of the candidate quality GSK pan-BET inhibitors disclosed to date.” Although the molecule does not seem to have gone forward into development, the story is nonetheless worth reading to see how metrics can yield quality molecules.

Review of 2019 reviews

The year ends, and with it the awkward teenage phase of the twenty-first century. As we have done since 2012, we're using this last post of the year to highlight conferences and reviews over the previous twelve months.

There were some good events, including CHI’s Fourteenth Annual Fragment-based Drug Discovery meeting in San Diego in April, their Discovery on Target meeting in Boston in September, and the third Fragment-based Drug Design Down Under 2019 in Melbourne in November, which also saw the launch of the Centre for Fragment-Based Design. Our updated schedule of 2020 events will publish next week.

Turning to FBLD reviews, Martin Empting (Helmholtz-Institute for Pharmaceutical Research Saarland) and collaborators published a general overview in Molecules. This is a nice up-to-date summary, covering library design, methods to find, confirm, and rank fragments, and optimization approaches. It’s also open access so you can read it anywhere.

Targets

Protein-protein interactions can be particularly challenging drug targets, and these are covered in a Eur. J. Med. Chem. review by Dimitrios Tzalis (Taros Chemicals), Christian Ottmann (Technische Universiteit Eindhoven) and colleagues. The focus is on clinical compounds, and several of these – including venetoclax, ASTX660, mivebresib, onalespib – are discussed in detail. The article is particularly useful in discussing late-stage optimization of pharmacokinetic and pharmacodynamic properties. It also provides a nice summary of physicochemical properties for fragment hits and derived candidates.

Target selectivity is always important, and this is the focus of a review in Exp. Opin. Drug Disc. by Rainer Riedl and collaborators at the Zurich University of Applied Sciences and the Università degli Studi dell’Insubria. Although the broader topic is de novo drug design, fragment-based methods are prominent, and include case studies we’ve discussed on nNOS, pantothenate synthetase, and MMP-13.

In terms of specific targets, Fubao Huang, Kai Wang, and Jianhua Shen at the Shanghai Institute of Materia Medica provide an extensive review of lipoprotein-associated phospholipase A2 (Lp-PLA₂) in Med. Res. Rev. This serine hydrolase has been studied for four decades but – as the researchers note – “divergence seems to be ubiquitous among Lp-PLA₂ studies.” At least this is not for lack of good chemical tools, fragment-derived (see here, here, and here) and otherwise.

Methods

Although NMR has fallen behind crystallography in our latest poll, that is certainly not reflected in terms of reviews. In particular, ¹⁹F NMR is covered in three papers. CongBao Kang (A*STAR) manages to pack a lot (including 261 references!) into a concise review in Curr. Med. Chem. Topics include protein-observed ¹⁹F NMR, in which one or more fluorine atoms are introduced into a protein genetically, enzymatically, or chemically, as well as ligand-observed methods, in which fluorine-containing small molecules are directly observed or used as probes that are displaced by non-fluorine-containing molecules.

Protein-observed ¹⁹F NMR (PrOF NMR) is covered in Acc. Chem. Res. by William Pomerantz and colleagues at the University of Minnesota. Although the first example was published 45 years ago, only in the past few years has the technique been used for studying protein-ligand interactions. The researchers note that introducing fluorines into aromatic residues is ideal because they are relatively rare, simplifying interpretation, and overrepresented at protein-protein interactions, maximizing utility. Several case studies are described, and even proteins as large as 180 kDa are amenable to the technique.

Ligand-based fluorine NMR screening is simpler and more common than techniques that focus on proteins, and this topic is thoroughly reviewed by Claudio Dalvit (Lavis) and Anna Vulpetti (Novartis) in J. Med. Chem. After a section on theory, the researchers discuss library design, including a long section on quality control (which involves assessing solubility, purity, and aggregation of the molecule in a SPAM filter). Direct and competition-based screening approaches are covered in detail; for the latter, a new method for determining binding constants is provided. The paper concludes with more than a dozen case studies. Clearly much has changed in the ten years since I wondered “why fluorine-labeled fragments are not used more widely.” This perspective is a definitive guide to the topic.

Moving to less common methods for characterizing fragments, György Ferenczy and György Keserű (Research Center for Natural Sciences, Budapest) cover thermodynamic profiling in Expert Opin. Drug Disc. After discussing several case studies, they conclude that “thermodynamic quantities are not suitable endpoints for medicinal chemistry optimizations” due to the complexity of contributing factors. This is consistent with another recent paper on the subject (see here), though the information provided is still interesting for understanding molecular interactions.

And although you might have thought the 2017 VAPID publication was the last word on the limitations of ligand efficiency (LE), Pete Kenny has published a splenetic jeremiad on the topic in J. Cheminform. (see also his blog post on the topic, which includes a sea serpent). This is largely a retread of a 2014 article on the same topic (reviewed by Teddy in his inimitable manner here). Pete also describes a more complicated alternative to LE involving residuals, though unfortunately he provides no evidence that it provides more useful information. Pete is of course correct to remind us that metrics have limitations, but assertions that LE “should not even be considered to be a metric” are overwrought.

Chemistry

Two articles discuss virtual chemical libraries. In J. Med. Chem., W. Patrick Walters (Relay Therapeutics) describes efforts to measure, enumerate, and explore chemical space. He notes that false positives could quickly overwhelm a virtual screen of a hundred million molecules, but as we saw earlier this year, progress is being made. Indeed, Torsten Hoffmann (Taros Chemicals) and Marcus Gastreich (BioSolveIT) focus on navigating the vastness of chemical space in Drug Disc. Today. They note that the Enamine REAL Space is up to 3.8 billion commercially accessible compounds, more than double the number of stars in the Milky Way. But this pales in comparison to the 10²⁰ potential compounds in Merck’s MASSIV space. Just storing the chemical structures of these in compressed format would require 200,000 terabytes – and searching them exhaustively is beyond current technology.

Ratmir Derda and Simon Ng (University of Alberta) discuss “genetically encoded fragment-based discovery” in Curr. Opin. Chem. Biol. This involves starting with a known fragment that is then coupled to a library of peptides and screened to find tighter binders. The researchers provide a number of case studies, though adding even a small peptide to a fragment will generally have deleterious effects on ligand efficiency. And – Rybelsus not withstanding – oral delivery of peptides is challenging.

Finally, Vasanthanathan Poongavanam, Xinyong Liu, and Peng Zhang, and collaborators at Shandong University, University of Bonn, University of Southern Denmark, and K.U. Leuven review “recent strategic advances in medicinal chemistry” in J. Med. Chem. Among a wide range of topics from drug repurposing to antibody-recruiting molecules is a nice, up-to-date section on target-guided synthesis. As I opined a couple years ago, I still doubt whether this will ever be generally practical, but from an intellectual standpoint I’m happy to see work continue on the approach.

And with that, Practical Fragments says goodbye to the teens and wishes you all a happy new year. Thanks for reading and commenting. May 2020 bring wisdom, and progress.

Review of 2018 reviews

As 2018 recedes into history, we are using this last post of the year to do what we have done since 2012 – review notable events along with reviews we didn’t previously cover.

This was a busy year for meetings, starting in January with a FragNet event in Barcelona, then moving to San Diego in April for the annual CHI FBDD meeting. Boston saw an embarrassment of riches, from the first US-based NovAliX meeting, to a symposium on FBDD at the Fall ACS meeting, followed closely by a number of relevant talks at CHI’s Discovery on Target. Finally, the tenth anniversary of the renowned FBLD meeting returned to San Diego. Look for a schedule of 2019 events later this month.

If meetings were abundant, the same can be said for reviews.

Lead optimization

Writing in J. Med. Chem., Dean Brown and Jonas Boström (AstraZeneca) asked “where do recent small molecule clinical development candidates come from?” For three quarters of the 66 molecules published in J. Med. Chem. in 2016 and 2017 the answer is from known compounds or HTS, though fragments accounted for four examples. Although average molecular weight increased during lead optimization, lipophilicity did not, suggesting the importance of this parameter.

The importance of keeping lipophilicity in check is also emphasized by Robert Young (GlaxoSmithKline) and Paul Leeson (Paul Leeson Consulting) in a massive J. Med. Chem. treatise on lead optimization. Buttressed with dozens of examples, including several from FBLD, they show that the final molecule is usually among the most efficient (in terms of LE and LLE) in a given series, even when metrics were not explicitly used by the project team. Perhaps with pedants like Dr. Saysno in mind, they also emphasize the complexity of drug discovery, and note that “seeking optimum efficiencies and physicochemical properties are guiding principles and not rules.”

Lipophilic ligand efficiency (LLE) is also the focus of a paper in Bioorg. Med. Chem. by James Scott (AstraZeneca) and Michael Waring (Newcastle University). This is based largely on personal experiences and provides lots of helpful tips. Importantly, the researchers note that calculated lipophilicity values can differ dramatically from measured values, and go so far as to say that “this variation is sufficient to render LLEs derived from calculated values meaningless.”

Turning wholly to fragments, Chris Johnson and collaborators (including yours truly) from Astex, Carmot, Vrije Universiteit Amsterdam, and Novartis have published an analysis in J. Med. Chem. of fragment-to-lead success stories from last year. This review, the third in a series, also summarizes all 85 examples published between 2015 and 2017, confirming and expanding some of the trends we mentioned last year.

Targets

Two reviews focus on specific target classes. Bas Lamoree and Rod Hubbard (University of York) cover antibiotics in SLAS Discovery. After a nice, concise review of fragment-finding methods, the researchers discuss a number of case studies, many of which will be familiar to regular readers of this blog, including an early example of whole-cell screening.

David Bailey and collaborators from IOTA and University of Cambridge discuss cyclic nucleotide phosphodiesterases (PDEs) in J. Med. Chem. The researchers provide a good overview of the field, including mining the open database ChEMBL for fragment-sized inhibitors. As they point out, the first inhibitors discovered for these cell-signaling enzymes were fragment-sized, so it is no surprise that FBLD has been fruitful – see here for an example from earlier this year. Interestingly though, although at least six fragment-sized PDE inhibitor drugs have been approved, none of these were actually discovered using FBLD.

PDEs are an example of “ligandable” targets, for which small molecule modulators are readily discovered. In Drug Discovery Today, Sinisa Vukovic and David Huggins (University of Cambridge) discuss ligandability “in terms of the balance between effort and reward.” They use a published database of protein-ligand affinities to develop a metric, LIG_exp, for experimental ligandability, and also describe their computational metric, Solvaware, which is based on identifying clusters of water molecules binding weakly to a protein. Comparisons with experimental data and with other predictive metrics, such as FTMap, reveal that while the computational methods are useful, there is still room for improvement.

We have previously written about how target-guided synthesis methods such as dynamic combinatorial chemistry have – despite decades of research – yielded few truly novel, drug-like ligands. Is this because the targets chosen were simply not ligandable? In J. Med. Chem., Anna Hirsch and collaborators at the University of Groningen, the Helmholtz Institute for Pharmaceutical Research, and Saarland University review some (though by no means all) published examples and examine their computationally determined ligandability scores. There seems to be no difference between these targets and a set of traditional drug targets.

Finding fragments

Crystallography continues to be a key tool for FBLD: as we noted in the review of the 2017 literature, 21 of the 30 examples made use of a crystal structure of either the starting fragment or an analog, and only 3 projects didn’t use crystallography at all. That said, FBLD is possible without crystallography, as illustrated through multiple examples in a Cell Chem. Biol. review by Wolfgang Jahnke (Novartis), Ben Davis (Vernalis), and me (Carmot).

In the absence of a crystal structure, NMR is best suited for providing structural information, and this is the subject of a review in Molecules by Barak Akabayov and colleagues at Ben-Gurion University of the Negev. The researchers provide a nice summary of NMR screening methods and success stories within a broader history of FBLD. They also include an extensive list of fragment library providers as well as a discussion of virtual screening.

Speaking of virtual screening, three reviews cover this topic. In Methods Mol. Biol., Durai Sundar and colleagues at Indian Institute of Technology Delhi touch on a number of computational approaches for de novo ligand design, though the lack of structures sometimes makes it challenging to read. A broader, more visually appealing review is published in AAPS Journal by Yuemin Bian and Xiang-Qun Xie at University of Pittsburgh. In addition to an overview and case studies, the researchers also provide a nice table summarizing 15 different computational programs. One of these, SEED, is a main focus of a review in Eur. J. Med. Chem. by Jean-Rémy Marchand and Amedeo Caflisch (University of Zürich). The researchers describe how this docking program can be combined with X-ray crystallography (SEED2XR) to rapidly identify fragments; we highlighted an example with a bromodomain. Their ALTA protocol uses SEED to generate larger, more potent molecules, as we described for the kinase EphB4. The researchers note that together these protocols have led to about 200 protein-ligand crystal structures deposited in the PDB over the past five years.

Rounding out methods, Sten Ohlson and Minh-Dao Duong-Thi (Nanyang Technological University) provide a detailed how-to guide in Methods for performing weak affinity chromatography, and how this can be combined with mass spectrometry (WAC-MS), as we noted last year.

Chemistry

One drawback of some computational approaches for fragment optimization is that they do not consider synthetic accessibility. In Mol. Inform., Philippe Roche, Xavier Morelli, and collaborators at Aix-Marseille University and Institut Paoli-Calmettes focus on hit to lead approaches that do, and provide a handy table summarizing nearly a dozen computational methods. We highlighted one from the authors, DOTS, earlier this year.

DOTS is an example of using DOS, or diversity-oriented synthesis. In Front. Chem., David Spring and colleagues at University of Cambridge review recent applications of DOS for generating new fragments, some of which we recently highlighted. Only a couple examples of successfully screening these new fragments are described, but the authors note that this is likely to increase as virtual library screening continues to advance.

Perhaps the most productive fragment of all time is 7-azaindole, the origin of three fragment-derived clinical compounds. (The moiety appears in both approved FBLD-derived drugs, vemurafenib and venetoclax.) Takayuki Irie and Masaaki Sawa of Carna Biosciences devote their attention to this little bicycle in Chem. Pharm. Bull. The researchers count six clinical kinase inhibitors that contain 7-azaindole (not all from FBLD) as well as more than 100,000 disclosed compounds containing the fragment. More than 90 kinases have been targeted by molecules containing 7-azaindole, and the paper provides a list of 70 PDB structures of 37 different kinases bound to molecules containing the moiety.

Finally, in J. Med. Chem., Brian Raymer and Samit Bhattacharya (Pfizer) survey the universe of “lead-like” drugs. Among the most highly prescribed small molecule drugs, 36% have molecular weights below 300 Da. Only 28 of 174 drugs approved between 2011 and 2017 fall into this category, consistent with the increasing size of newer drugs. The researchers discuss 16 recently approved drugs, and find that 13 have very high ligand efficiencies (at least 0.4 kcal mol^-1 per heavy atom). As noted above, optimization often entails adding molecular weight by growing or linking, and the researchers suggest that alternative strategies such as conformational restriction and truncation also be investigated.

And with that, Practical Fragments wishes you a happy new year. Thanks for reading some of our 686 posts over the past decade plus, and please keep the comments coming!

Ligand efficiency invalidated!

Practical Fragments has had quite a few posts on ligand efficiency (see here, here, and here, for starters). Ligand efficiency (LE) is defined simply as the free energy of binding for a ligand divided by the number of heavy atoms in the ligand. One of the criticisms of LE is that the definition of free energy depends on the definition of standard state, which may be different on different planets. With the discovery of silicon-based life on Venus, this is no longer just an academic argument. Indeed, a recent paper in Venusian Analytical, Physical, & Inorganic Discoveries describes an excellent case study.

Professor Perelandra and colleagues at East Eistla University performed a crystallograhic fragment screen on the enzyme silica hydratase, which is essential for the life cycle of the viciously parasitic Crystalline Horde. Fragment 1 binds in the active site, and although it has low affinity, structure-guided medicinal chemistry rapidly led to compound 42, with low nM activity in vitro and good efficacy in a silicon resorption model.

Things get even more interesting when you calculate the ligand efficiency values. The Venusians define standard temperature and pressure very differently from us. More importantly, they don't believe that standard state concentration should be 1 M. Given the extreme conditions on their home world, they choose a standard state concentration of 10 M.

LE = - ΔG/HA

(where HA = number of non-hydrogen atoms)

Thus, LE_Venus = -RTln(K_D/[A]⁰)/HA

(where T = 737 K and [A]⁰ = 10 M)

Using our terracentric definitions, the (impressive) LE of the fragment hit stays roughly the same during optimization, suggesting that the medicinal chemists have done a good job. However, by Venusian standards, the LE decreases!

This rock-solid example shows that Dr. Saysno was right: ligand efficiency is arbitrary and should never be used – on Venus.

An interview with Dr. Saysno

Readers of a certain age may fondly remember the interviews with Dr. Noitall that used to enliven the pages of Science. Sadly, he died a few years back. But his cousin, Dr. Saysno, is still very much alive. Practical Fragments caught up with him at a recent conference in Shutka.

Practical Fragments (PF): Dr. Saysno, you've stated that experts should never be trusted.

Dr. Saysno (DS): Niels Bohr defined an expert as a person who has made all the mistakes that can be made in a very narrow field. If someone has made every possible mistake, how could you possibly trust them?

PF: But don't you think they may have learned from their mistakes?

DS: Balderdash! Hegel was right: the only thing we learn from history is that we learn nothing from history.

PF: What's your opinion of ligand efficiency (LE)?

DS: Ligand efficiency is an abomination! It's mathematically invalid!

Worse, determining the free energy of binding from a dissociation constant is not even wrong: if you change your definition of standard state, you can make ΔG° any positive or negative number you want. Just how relevant do you think your definition of standard state is on the surface of Venus? Or Pluto?

For the same reason, pH is utterly meaningless. You really ought to throw out your pH paper, not to mention your pH meters, since they all assume an arbitrary reference state.

PF: But what about all the researchers who find pH and LE useful?

DS: Usefulness is the last refuge of the scoundrel!

Look, many of the best selling drugs on the market are antibodies, and when you calculate their ligand efficiencies, they are close to zero. How can you have a metric that doesn't work on some of the most important drugs out there?

I only believe in equations that are universal and apply in all situations, unsullied by the physical world. Anything that involves standard states is just mumbo-jumbo.

PF: What do you think of pan-assay interference compounds, or PAINS?

DS: Now that’s a topic that really gets my blood boiling! PAINS were defined on the basis of just six assays. Six assays I tell you!!! [DS vigorously pounds his shoe on the desk.] Just because something hits six assays – or six hundred for that matter – doesn’t mean it will hit the six hundred and first!

PF: But aren't there some chemical substructures that are so generically reactive they should never be used in probes?

DS: Nothing is universal! All molecules are unique, like little snowflakes. If a compound comes up as a hit in your assay, by all means publish it as a chemical probe in the best possible journal, and try to encourage suppliers to start selling it so other people can use and cite your brilliant discovery.

No one has a right to criticize your molecule unless they test it against every single protein in the human body and show that it hits all of them.

When the revolution comes, the imperialist PAINS stooges will be swept into the dustbin of history along with the lackeys of ligand efficiency!

PF: So if you don't trust experts, you don't like metrics, and you can't make generalizations, how can we move forward in science short of deriving every result ourselves from first principles?

DS: That's simple: just ask me!

Fragments vs MKK3: modeling all the way to low nanomolar

The mitogen-activated protein kinase (MAPK) signaling pathway is a rich source of targets, particularly for inflammation. Within this cascade the p38 kinases have been heavily studied, but many of the inhibitors that entered the clinic derailed for various reasons, including efficacy. Thus, some groups have sought to block the pathway upstream of p38. A paper just published online in Bioorg. Med. Chem. Lett. by Steve Swann and colleagues at Takeda describes some of their efforts to accomplish this.

The researchers focused on MKK3 and to a lesser degree the related MKK6, both of which phosphorylate and activate p38. They began by screening their 11,012 fragments in a biochemical assay at 100 µM each. Hits were prioritized by estimating the IC₅₀ values and thus approximate ligand efficiency (LE) and lipophilic ligand efficiency values (LLE) for each compound that inhibited >30%. Of these, 93 gave LE ≥ 0.35 kcal/mol per heavy atom and LLE ≥ 4. (Incidentally, this seems like a perfectly reasonable use of metrics to triage a large number of compounds, and the speed and simplicity is a good counterargument to more complicated proposals.) Some hits were tested using full dose-response curves to determine actual IC₅₀ values and surface plasmon resonance assays to determine K_d values; compound 1 was particularly compelling.

Readers may recall that Takeda found this very same fragment as an inhibitor of BTK (a kinase in an unrelated pathway), and they used the compound/BTK crystal structure along with the published crystal structure of MKK6 to develop a binding model. In their pursuit of MKK3/6 inhibitors, the Takeda team performed biochemical screens of available related compounds. This led to compound 2, which modeling predicted would bind in a similar fashion. The binding model also suggested the possibility of picking up a hydrogen bond to a lysine residue, leading to the more potent compound 3. Further optimization led to compounds 4 and 6, both with low nanomolar potency against MKK3 and low micromolar or high nanomolar cell-based activity. Profiling these against a dozen other kinases within the p38 signaling pathway revealed good selectivity against all except MKK6.

This is a nice, concise paper that illustrates how modeling, even without direct structural information, can be used to advance a fragment to low nanomolar inhibitors, albeit in a well-studied class of targets. It is also another illustration that the same fragment can be used to develop completely different series. And finally, these molecules look promising as chemical probes and possibly drug leads; it will be fun to watch as more data are disclosed.

Group efficiency

Ligand efficiency (LE) is one of the more controversial topics we cover at Practical Fragments. One critic asserted – incorrectly – that it is mathematically invalid. Another has stated that it is “not even wrong,” because the metric is predicated on standard state conditions and thus "arbitrary". (As he acknowledges, this also applies to the value and even the sign of the Gibbs free energy for a reaction.) A related metric that has received less attention is group efficiency (GE). In a paper just published in ChemMedChem, Chris Abell and colleagues at the University of Cambridge use this to help them optimize pantothenate synthetase (Pts) inhibitors.

Ligand efficiency is defined simply as the free energy of binding divided by the number of non-hydrogen, or “heavy” atoms (often abbreviated as HAC for heavy atom count) in the ligand. (Geek notes: although the binding energy is negative, LE is expressed as a positive number, so LE = - ΔG / HAC. Also, on Practical Fragments, units are assumed to be kcal mol^-1 per heavy atom unless otherwise stated.)

Instead of focusing on a single ligand, group efficiency compares two ligands that differ by the presence or absence of a given group of atoms. To calculate GE, you simply subtract the ΔG values for the two ligands and divide by the number of heavy atoms in the group. For example, if you add a methyl group to your molecule and are lucky enough to get a 100-fold pop in potency, the methyl group has a group efficiency of 2.7 kcal mol^-1 per heavy atom.

The current paper chronicles lead discovery for Pts, a potential target for tuberculosis. Previous screening efforts followed by fragment growing and fragment linking had generated low micromolar and high nanomolar inhibitors. The researchers turned to group efficiency to improve their molecules further.

As expected from ligand deconstruction studies (see for example here, here, and here), different portions of a molecule are likely to have vastly different group efficiencies. Indeed, this turned out to be the case here: the acetate moiety had high group efficiency, whereas the pyridyl moiety had lower group efficiency. Thus, the researchers set out to replace the pyridyl with ten diverse substituents. Happily, one of these improved the dissociation constant to 200 nM as assessed by isothermal titration calorimetry of the fully elaborated molecule. Compound 11 also showed reasonable enzyme inhibition in a functional assay.

One potential problem with group efficiency is that it assumes the molecules being compared bind in a similar fashion, which is not always a safe assumption. In this case, the researchers obtained a crystal structure of compound 11 bound to the enzyme, which not only revealed that it binds similarly to compound 5, but also suggested that inserting a methylene may improve binding. The resulting compound 20 showed better activity in the inhibition assay, as well as activity against M. tuberculosis in a cell assay (though unfortunately the dissociation constant was not reported).

This paper offers a clear illustration of how group efficiency can be useful for prioritizing which portions of a molecule to change. In some cases, such as the example here, it makes sense to try to replace groups with low group efficiency. On the other hand, the core fragment may bind in a hot spot, and so just a slight tweak can dramatically boost potency. As with lead optimization in general, there are many paths – both to enlightenment and to perdition.

Fragment-Based Drug Discovery

This is the straight-to-the-point title of a new book published by the Royal Society of Chemistry, edited by Steven Howard (Astex) and Chris Abell (University of Cambridge). It is the second book on the topic published so far this year, and it is a testimony to the fecundity of the field that the two volumes have very little overlap.

After a brief forward by Harren Jhoti (Astex) and a preface by the editors, the book opens with two personal essays. The first, by me, is something of an apologia for Practical Fragments and the growing role of social media in science (and vice versa). If you’ve ever wondered how this blog got started or why it keeps going, this is where to find out. The second essay is by Martin Drysdale (Beatson Institute). Martin is a long-time practitioner of FBDD, dating back to his early days at Vernalis (when it was RiboTargets) and he tells a fun tale of “adventures and experiences.”

Chapter 1, by Chris Abell and Claudio Dagostin, is entitled “Different Flavours of Fragments.” With a broad overview of the field it makes a good introduction to the book. There are sections on fragment identification, including the idea of a screening cascade, as well as several case studies, some of which we’ve covered on Practical Fragments, including pantothenate synthetase, CYPs, RAD51, and riboswitches.

The next two chapters deal with two of the key fragment-finding methods. Chapter 2, by Tony Giannetti and collaborators at Genentech, GlaxoSmithKline, and SensiQ, covers surface plasmon resonance (SPR). This includes an extensive discussion of data processing and analysis, which is critical for improving the efficiency of the technique. Competition studies are also described, as are advances in hardware, notably those from SensiQ. This is a good complement to Tony's 2011 chapter.

Chapter 3, by Isabelle Krimm (Université de Lyon), provides a thorough description of NMR methods, both ligand-based (STD, WaterLOGSY, ILOE, etc) and protein-based (mostly HSQC). The chapter does a nice job of describing techniques in terms a non-specialist can understand while also providing practical tips on matters such as optimal protein size and concentration.

Chapter 4, by Ian Wall and colleagues at GlaxoSmithKline, provides an overview of FBLD from the viewpoint of computational chemists. The chapter includes some interesting tidbits, such as the observation that fragment hits that yield crystal structures tend to be less lipophilic but also contain a smaller fraction of sp³ atoms and more aromatic rings. The researchers note that the current fashion for “3D” fragments is yet to be experimentally validated. They also include accessible sections on modeling, druggability, and integrating fragment information into a broader medicinal chemistry program.

The remaining chapters focus on specific types of targets. Chapter 5, by Miles Congreve and Robert Cooke (both at Heptares) is devoted to G protein-coupled receptors (GPCRs). This includes descriptions of how to screen fragments against these membrane proteins using SPR, TINS, CE, thermal melts, and competition binding. It also includes a detailed case study of their β1 adrenergic receptor work (summarized here). Congreve and Cooke assert that, although many of the GPCR targets screened to date have been highly ligandable, technical challenges only now being addressed have caused this area of research to lag about a decade behind other targets. They predict a bright future.

Rod Hubbard (Vernalis and University of York) turns to protein-protein interactions in Chapter 6. After describing why these tend to be more challenging than most enzymes and covering some of the methods for finding and advancing fragments, he then presents several case studies, including FKBP (one of the first targets screened using SAR by NMR), Bcl-2 family members (including Bcl-x_L and Mcl-1), Ras, and BRCA2/RAD51. He concludes with a nice section on “general lessons,” which boils down to “patience, pragmatism, and integration.” As Teddy recently noted, this can lead to substantial rewards.

Allosteric ligands have potential advantages in terms of selectivity and addressing otherwise challenging targets, and in Chapter 7 Steven Howard (Astex) describes how fragments can play a role here. This includes how to establish functionality of putative allosteric binders, as well as case studies such as HIV-1 RT, FPPS, and HCV NS3. Astex researchers have recently stated that they find on average more than two ligand binding sites per protein, and this chapter includes a table listing these (including 5 binding sites each on bPKA-PKB and PKM2).

The longest chapter, by Christina Spry (Australian National University) and Anthony Coyne (University of Cambridge) describes fragment-based discovery of antibacterial compounds. After discussing some of the challenges, the authors report several in depth case studies including DNA gyrase, DNA ligase, CTX-M, AmpC, CYP121, and pantothenate synthetase, among others. At least one fragment-derived antibacterial agent entered the clinic; hopefully more will follow.

Chapter 9, by Iwan de Esch and colleagues at VU University Amsterdam, focuses on acetylcholine-binding proteins (AChBPs), both as surrogates for membrane-bound acetylcholine receptors and as well-behaved model proteins on which to hone techniques (see for example here, here, and here). Since AChBPs have evolved to bind fragment-sized acetylcholine, these proteins can bind tightly to small ligands; 14-atom epibatidine binds with picomolar affinity, for example, with a ligand efficiency approaching 1 kcal mol^-1 atom^-1.

And Chapter 10, by Chun-wa Chung and Paul Bamborough at GlaxoSmithKline, concisely covers epigenetics. Bromodomains are well-represented, including a table of ten examples (see for example here, here, here, here, here, and here). Happily, although some of these projects started from similar or identical fragments, the final molecules are quite divergent. However, the authors note that much less has been published on histone-modifying enzymes, such as demethylases and deacetylases, perhaps reflecting the challenges of achieving specificity with what are often metalloenzymes.

Finally, this is the 500^th post since Teddy founded Practical Fragments way back in the summer of 2008. Thanks for reading, and special thanks for commenting!