Available online at www.sciencedirect.

com

Enhancements of screening collections to address areas of

unmet medical need: an industry perspective
David H Drewry1 and Ricardo Macarron2

The past 20 years have witnessed an impressive expansion of Failure of a screen to identify a chemical starting point
the ‘drug space’; defined as the intersection of the Medicinal can be simplified to one of two factors: the target itself is
Chemistry space and the Biologically Active space relevant in un-druggable, (unable to be modulated appropriately by a
the quest for new treatments for disease. Despite the success small molecule), or the screen did not test the correct
of known lead discovery tactics, areas of unmet medical need compounds (yet). There is a very interesting body of work
are often linked to challenging or novel targets and are poorly that deals with methodologies to quantify the druggabil-
served by current screening collections. A successful strategy ity of targets [5–22]. These methodologies have the
to fill the gaps is to diversify the approaches taken in the potential to help prioritize targets of interest, bias screen-
enhancement of screening collections. Possible strategies ing efforts toward the druggable, and may be valuable in
include investments through proven methods, exploring areas the risk management of a discovery portfolio.
of chemical space previously neglected (e.g. hydrophilic
compounds, natural product mimics), and applying tactics to The goal is to identify molecules for all targets of interest
the lead discovery process that are complementary to HTS by expanding current medicinal chemistry space with
(e.g. fragment based screening or multidisciplinary team efforts improved screening collections (Figure 1). In this review
to tackle new target classes). we touch on some strategies that can be utilized to move
Addresses toward this goal (Figure 2). We advocate for a multi-
1
Molecular Discovery Research, GlaxoSmithKline, 5 Moore Drive, pronged approach to improve the chances of success.
Research Triangle Park, NC 27709, United States
2
Molecular Discovery Research, GlaxoSmithKline, 1250 S Collegeville
Rd, Collegeville, PA 19426, United States
‘Drugging’ targets of therapeutic interest
Knockout technology suggests that 10% of human genes
Corresponding author: Drewry, David H ([email protected]) are associated to diseases [23]. This method was cited by a
seminal paper from Hopkins and Groom [10], and esti-
Current Opinion in Chemical Biology 2010, 14:289–298
mated that there are 3000 disease modifying human
genes. The actual number of genes modulating disease
This review comes from a themed issue on may be higher when the caveats of knockout technologies
Molecular Diversity are considered. Additionally, many human proteins are
Edited by Lisa A. Marcaurelle and Michael A. Foley
involved in the so-called interactome, estimated to
Available online 21st April 2010 involve more than half a million contacts that may present
opportunities for therapies [24]. Finally, exploitation of
1367-5931/$ – see front matter
new modes of action or understanding of new regulatory
# 2010 Elsevier Ltd. All rights reserved.
paths can open new avenues for the use in medicine of old
DOI 10.1016/j.cbpa.2010.03.024 targets (Box 1).

With fewer than 1000 targets with known small molecule

effectors [10,25], there is a clear gap between the number
Introduction of potential therapeutic targets and the ones that have
Scientists working in lead discovery have the best chance already been successfully paired with small molecule
of meeting unmet medical needs by identifying lead ligands (‘drugged’). Most analysis on the subject of
molecules for all targets with biological validation, and druggability employ a circular argument that looks at
by identifying tool molecules to explore promising un- the properties of drugged targets and marketed drugs
validated biological targets. Drug discovery chemical space to understand what other targets are druggable and to
has been recently expanded to include proteins, peptides, predict what the next generation of drugs will look like.
and ribonucleic acids (referred to as biologicals); however, There is no room for innovation if we simply imitate
this review will focus on traditional small molecule agents. previous success. One example of this constraint is the
conclusion that druggable proteins contain highly hydro-
In practice, starting points are not found for all targets in phobic pockets [26]. Most small molecules synthesized to
High-Throughput Screening (HTS); [1,2]. Not surpris- date are indeed hydrophobic, but we cannot rule out that
ingly, screening yields different results based on target more hydrophilic drugs are possible and therefore, targets
class [3,4]. There are target types that almost always find with less hydrophobic pockets may also be druggable. A
leads, target types that rarely find leads, and target types recent review of carbohydrate based drugs supports this
that we do not know enough about to guess the outcome. line of thought [27].

www.sciencedirect.com Current Opinion in Chemical Biology 2010, 14:289–298

290 Molecular Diversity

Figure 1 Box 1 Emerging Targets in Biomedical Research

 RNA-protein interactions
 Protein–protein interactions
 GPCR kinases
 Arrestins
 Oligomerization of GPCRs
 Allosteric protein kinase inhibitors
 Protein kinase activators
 Transporters
 Epigenetic targets
 Stem cell differentiation and modulation

the same and other target classes should have a greater

chance of success in HTS. This assumption is loosely
related to recent industrial experience at Novartis [4]
Schematic representation of the goal of enhancement of screening and GlaxoSmithKline (GSK) [3], though some exceptions
collections for pharmaceutical research. Chemical Space represents the are noteworthy. The success rate at Novartis is quite high
immense universe of possible molecules. Bioactive space represents all
for proteases and low for nuclear receptors compared to
those molecules, known and unknown, which bind effectively to any
biological target. Drug space is only a small region at the intersection of the GSK experience. This is probably a reflection of the
the Medicinal Chemistry space (small molecules ever made) with the different history and enhancement strategy for respective
Bioactive space, as it only includes those bioactive small molecules that HTS collections.
are bioavailable, safe and efficacious in treating disease.
The traditional classification of targets is based on
Unfortunately, there are few publications on success rate sequence similarity. It has been suggested that ‘cavity
per target class in HTS [3,4,28]. Table 1 is based on the similarity’ is a better predictor of druggability [30]. That
map of pharmacological space published by Paolini et al. might explain the poor predictability of success in HTS,
[29]. Target classes most likely to bind ligands within with occasional failures among most successful target

Figure 2

Schematic summary of tactics available to explore new chemical space in the quest for new drug space and thus expansion of the druggable genome.
Biologicals (dotted line) are out of scope on this review, which is focused on small molecules (solid lines).

Current Opinion in Chemical Biology 2010, 14:289–298 www.sciencedirect.com

Enhancements of screening collections to address areas of unmet medical need: an industry perspective Drewry and Macarron 291

Table 1

Tractability of target classes in HTS extrapolated from the pharmacological space mapping by Paolini et al. [29]

Target family Intra-family connectivity Overall connectivity Relative HTS tractability

Aminergic GPCRs High High Very High
Protein Kinases High High Very High
Peptide GPCRs Medium High High
GPCRs Class A-others Medium Medium High
PDEs Medium Medium High
Metalloproteases Medium Medium High
Nuclear hormone receptors Medium Medium High
Oxidoreductases Low Medium Medium
Hydrolases Low Medium Medium
Ion Channels-others Low Medium Medium
Ion Channels-Ligand Gated Low Medium Medium
Serine Proteases Low Low Low
GPCRs Class C Low Low Low
Transferases Low Low Low
GPCRs Class B Low Low Low
Kinases-others Low Low Low
Cysteine Proteases Low Low Low
Aspartyl Proteases Low Low Low

The data were derived from Supplementary Table 1 [29].

Connectivity is defined from Gkl, the sum of the strength of the links between targets in two target families based on the potency and number of
common ligands. High connectivity indicates values of Gkl > 100, Medium between 100 and 20, Low < 20.

classes (although this is often related to selectivity or has to interrogate the biological assays. There has been
patentability issues in developing leads out of HTS hits as much written about what a good collection should look like
described by Bender et al. [4]) and occasional successes [33–35]. In general terms, it appears simple. One would
in challenging families, like protein–protein interactions. like a compound collection that contains diverse scaffolds
decorated with diverse functionalities [36,37]. The collec-
Some challenging targets not so long ago became tract- tion should avoid compounds that are unstable [38–41],
able for HTS owing to recent advances in assay technol- impure [42], promiscuous [43–47], overly complicated
ogies and/or enhancement of chemical collections. For [48,49], have undesirable functionalities [50,51], or that
instance, Family B and C GPCRs had no success in HTS cause assay interference. In practice, the situation is more
in GSK for many years [3] but since 2006 four out of four complex. Chemical space is vast, and any collection of a few
screened have been successful in delivering leads by million molecules cannot represent the diversity of the
accessing new chemical series and new assay technol- possible. There are trillions of molecules we can make and
ogies. In the antibacterial arena, a poor success rate for millions of molecules that we can buy [52]. The question is:
HTS campaigns was observed in GSK over the period what kinds of molecules should populate the set?
1995–2001 [31]. Our current collection, enhanced over
the years with some of the tactics discussed below, is Compounds from medicinal chemistry programs
producing hits of interest for many targets in this area. Traditionally, a major source of these diverse scaffolds in
a company’s collection is the on-going and historical
The aim of a screening collection should be to render medicinal chemistry efforts of the company. As such,
multiple chemotypes that bind with visible potency to these scaffolds represent chemotypes useful in past or
every druggable pocket in every protein of therapeutic present areas of biological interest. It makes a lot of sense
interest. This ideal collection of chemicals would still fail (and it is empirically proven particularly for narrowly
to render hits for un-druggable targets, but as discussed connected target classes such as kinases) that these com-
above these might be a smaller proportion of the expand- pounds may be of particular utility in new projects that are
ing universe of disease associated targets than currently in related areas to the old projects and are thus valuable
predicted (90% according to [10], 75% according to [32]). for screening. Additionally, if the diverse scaffolds that
arise from this variety of projects display diverse func-
In the next sections we will describe how to create this ideal tionalities, then it is both possible and not uncommon to
HTS collection and other tactics to render the challenging find starting points for completely unrelated biological
targets of today into the therapeutics of tomorrow. targets. A potential downside to a collection built from
past efforts is that there may not be appropriate diversity
Continually enhance the HTS compound deck for future important disease targets (Box 1) [53]. Another
The foundation of success for identifying a tool or a lead for potential issue with these historical compounds from lead
any target (given robust assays) is the set of compounds one optimization programs is that many of them are drug-like

www.sciencedirect.com Current Opinion in Chemical Biology 2010, 14:289–298

292 Molecular Diversity

rather than lead-like [54]. These molecules are being As high-throughput screening of synthetic libraries became
screened against new targets and we are relying on the method of choice for lead finding, companies moved
serendipity to identify a new hit. Lead-like compounds, away from natural product chemistry and screening. Hope
by virtue of their lower molecular weights and lower was placed on synthetic libraries that tended to be rela-
complexity than drug-like molecules, are better suited tively simple (so they can be efficiently prepared) and
for the lead optimization process and offer a greater reside in a constrained property space. Incorporating some
probability of finding a productive binding mode (hit) of the features and complexity of natural products into
though typically at lower potency [48]. screening collections might increase HTS success.

Compounds purchased to enhance the collection Natural products and biologically inspired molecules
Another source of chemical diversity for screening is the have a great track record
purchase of compounds from the large number of vendors A recent review indicates that 34% of all small molecule
around the globe [55]. The scientists involved in this NCEs (new chemical entities) approved between 1981 and
process have robust methods to add compounds that are mid 2006 are either natural products or semi-synthetic
different from the current collection and do not have the derivatives [73]. In addition to these past successes, there
undesirable properties mentioned earlier [56]. An are over 100 natural product derived compounds in clinical
example of creating a screening set for academic use trials for a variety of indications in a range of therapeutic
through purchase was recently described [57]. This pro- areas [74,75]. Another recent report posits that the fact that
cess is often an exercise of ‘filtering out’. Start with an HTS campaigns find any hits at all is because the screening
enormous list of commercially available molecules and sets are heavily biased toward biogenic molecules [76].
filter out compounds that are too expensive similar to These authors suggest that it would be prudent to increase
what is in the collection with reactive functional groups, this bias, perhaps by adding sets of molecules around
non-lead-like, and so on. Purchase of compounds in this scaffolds that resemble biogenic scaffolds underrepre-
way can be a cost effective and efficient means to enhance sented in screening collections. In related work, Dobson
the lead-like nature of a screening collection, counter et al. compared drugs and library compounds to endogen-
balancing the more drug-like molecules often made in ous metabolites and found that drugs are significantly more
lead optimization programs. Snowden and Green illus- similar to the metabolites than library compounds are
trate how this ‘purposeful diversity’ can lead to starting [77,78]. Thus ‘metabolite-likeness’ in a screening set
points for medicinal chemistry programs [58]. A restric- can lead to drug discovery success and can be used in
tion here is that we are limited to what is available, and we compound design. A recent study looking at hit rates from
are not really choosing what we want, but rather excluding HTS campaigns at Novartis suggests that in this collection,
what we do not want. natural products are both the most diverse class (compared
with traditional synthetic and combinatorial molecules)
A complementary method to the ‘filter out’ method is the and show the highest hit rate [79].
‘select in’ method (exemplified for kinase inhibitors in Ref.
[57]). It is common practice for computational chemists to Natural products explore different and valuable
use a knowledge base and make selection based purchases chemical space
for specific targets or families of targets [36,59–66]. A number of groups have compared natural products to
This valuable approach serves to balance exclusion drug molecules and library compounds [80–84]. Although
methodology and ensures that knowledge around targets different methodologies have been used on a variety of
a company is interested in is being utilized to expand the different data sets, many of the take home messages are
screening collection in relevant directions. In fact, many of the same. Compared to drugs and combinatorial chem-
the small molecule vendors do their own design work based istry molecules, natural products have the following
on literature data and offer compound sets targeting, for characteristics:
example, proteases, kinases, ion channels and GPCRs.
Ertl et al. describe a natural product likeness score [67].  More rigidity.
Methodology such as this can also be used to ‘select in’  More fused, bridged, and spiro ring systems (as
compounds with desired features, such as being natural opposed to rings separated by single or double bonds).
product like.  A lower percentage of aromatic atoms.
 An increased number of chiral centers.
Infuse the collection with more biologically inspired  A higher carbon–oxygen bond count.
compounds  A lower carbon–nitrogen bond count.
Chemists have been inspired by the molecules of nature  More scaffold diversity.
since organic synthesis began, and man has been using
natural products to treat disease for thousands of years [68]. Dobson has noted that metabolites are more hydrophilic
Many of our modern drugs are natural products, or have than drugs and library compounds [75]. It is suggested
been derived from or inspired by natural products [69–72]. that mimicking this sort of polarity may allow access to

Current Opinion in Chemical Biology 2010, 14:289–298 www.sciencedirect.com

Enhancements of screening collections to address areas of unmet medical need: an industry perspective Drewry and Macarron 293

transporters to facilitate uptake into cells. Ganesan [98–102,103], schemes that give analogs around particular
looked at the properties of 24 natural products that led natural products of interest [104,105], or mimic elements
to an approved drug and found that half of them did not of protein structure such as helices [106–111] and
comply with the rule of 5 but were successful none the turns.[112–115]. All of these approaches produce what
less [85]. Those that did not comply maintained a low can be called biologically inspired molecules. Practitioners
log P and fewer than 5 H-bond donors. Exploration of of diversity oriented synthesis (DOS) design and imple-
relatively rigid aliphatic cores that may have higher ment synthetic sequences that efficiently generate both
molecular weight, yet remain in the low log P regime scaffold and stereochemical diversity. Biology oriented
and have some stereochemical complexity may be a synthesis (BIOS) focuses on synthesis of libraries of mol-
prudent avenue for collection enhancement. ecules based on privileged, biologically active natural
products, particularly those that may provide hits for
Natural products are privileged structures multiple proteins based on protein structural similarity
It has been suggested that many natural products are clustering (PSSC). A screening collection poised for
privileged structures: changes around a given scaffold success should increase the number of molecules that
leads to structures that can bind to a variety of targets. incorporate the positive features of natural products
For example the indolizidine and beta carboline natural (rigidity, complexity, chirality, scaffold diversity, increased
product scaffolds each have more than 20 different aliphatic content, increased oxygen content) and avoid the
reported activity types in the MDDR (MDL Drug Data primary downside of traditional natural product screening:
Report TM Database) [86]. Natural products are made by poor synthetic tractability. Some recent reviews highlight
proteins to interact with proteins, and can thus be con- the success in this area and suggest the future is bright for
sidered as scaffolds that are pre-validated by nature [87]. incorporating these sorts of compounds into screening
A great deal of work has been done to explore the scaffold collections [116,117–120].
space of natural products and to suggest ways forward for
drug discovery. A system called ChemGPS-NP has been Invest in strategies complementary to HTS
developed and used to evaluate bioactive medicinal If our aspiration is to identify small molecule starting
chemistry space and natural product space [88,89]. points for all targets of interest, then, where appropriate,
Two possible applications are the identification of natural targets should be attacked with different approaches. The
products that are neighbors to known drugs for the previous sections have focused on enhancement of HTS
purpose of lead hopping and the identification of natural collections to improve success rates, and now we will
product scaffolds in regions of the bioactivity space with briefly touch on three complementary tactics.
low compound density. Schneider and Grabowski have
computationally analyzed the scaffold diversity of natural Fragment screening
product databases to suggest scaffolds that could be Fragment screening is a complementary approach to
exploited for combinatorial chemistry [90]. The Wald- high-throughput screening that has gained a lot of traction
mann group has published a great deal on ways to exploit in the pharmaceutical industry [121–127]. A successful
the intersection between natural product space and bio- fragment screen provides a project team with weak but
logical space [91–95]. They developed a tool called efficiently binding small molecules that exploit key inter-
Structural Classification of Natural Products (SCONP) actions with the target of interest. These hits, free from
that organizes scaffolds in a hierarchical tree like arrange- the encumbrance of complexity, can bind in an ideal
ment that helps chart the chemical space explored by conformation, and chemists can use structural information
nature. These scaffold trees can be used to help chemists to grow the molecules into more potent leads and candi-
design sets of molecules around biologically relevant dates. The approach has been successful for a range of
scaffolds. This concept has been carried further recently projects, including challenging targets such as phospha-
with the disclosure of a tool called ‘Scaffold Hunter’ that tases and protein–protein interactions. For example, frag-
can be used to interactively navigate chemical and bio- ment screening was used to identify a novel small
logical space, and can be used to get ideas for new molecule binding site on the protein survivin, a potential
synthetic directions based on scaffold hierarchy [96,97]. cancer target [128]. The strategy is not limited to protein
targets, and early work to identify fragments that bind to
Synthesis of biologically inspired compounds RNA has been described [129,130]. Fragment screening
One reason the pharma industry has shied away from in its various guises has been validated as a valuable tool
natural products is the synthetic challenge they represent. in the drug discovery arsenal, with at least 13 molecules in
Infusing a compound collection with biologically inspired clinical trials that emanated from fragment-based
compounds should be done with downstream workflow in approaches [131].
mind, which means utilizing robust chemistry schemes
that can lead to efficient analog synthesis around a hit Encoded library screening
molecule. Recent literature is replete with examples Clark et al. recently published work on the design and
of clever synthetic schemes that give scaffold diversity synthesis of DNA-encoded small molecule libraries [132].

www.sciencedirect.com Current Opinion in Chemical Biology 2010, 14:289–298

294 Molecular Diversity

This technology facilitates the screening of extremely these targets will help us understand novel biology and
large libraries (108–1010 compounds) and can be used to teach us if the targets have potential for drug discovery
identify families of molecules that bind to proteins of [148].
interest. The molecules can be efficiently made,
screened, and the results deconvoluted. Based on the Conclusions
efficiency of the process and the large numbers of com- Much work has been done to improve screening collec-
pounds that can be interrogated, this innovative approach tions over the years. Current lead discovery efforts have
should prove extremely useful as a complement to HTS. greatly expanded the list of druggable biological pro-
cesses in the past decade. Despite this recent success,
Challenge teams of scientists to explore new there remain many validated targets associated with dis-
and ‘un-druggable areas’ eases with poor or no treatment that are yet to be drugged.
Another strategy is to assemble teams of scientists to Additionally, emerging targets demand that we continu-
explore new or difficult opportunities in rich veins of ously explore the depth and boundaries of medicinal
science that have potential to impact human health in chemistry space. Current advances in the field continue
broad ways. An example from recent history is the study on successful paths (feeding from known bioactive mat-
of kinase inhibitors. When dozens of new kinases were ter, addition of new diversity with defined chemical
identified in the 80 s and 90 s as key signal transducers, it quality) and look back to natural products for inspiration
was generally thought that kinases were not tractable as and overlooked properties of interest. Importantly, com-
drug targets because of their remarkable homology in plementing HTS with fragment based discovery and
their most druggable pocket (ATP binding) combined other tactics are key drivers in the quest for drugs for
with the high ATP concentration in the cytoplasm. both new targets and so-called un-druggable targets.
Gradually, knowledge grew around small molecule che- Finally, multidisciplinary teams are crucial trail blazers
motypes that inhibited kinases, and X-ray crystallography in this fascinating exploration of chemistry and biology.
allowed snap-shots of multiple active sites. It became
clear that although all the kinases bound ATP, there were Acknowledgements
conserved and non-conserved regions in the active site The authors thank many colleagues in GSK for helpful discussions, in
particular Ian Churcher, Darren Green, Bob Hertzberg, and Karen Lackey.
that could be exploited for potency and selectivity [133].
The discovery of even modestly selective tools by pio-
References and recommended reading
neering HTS work at Parke Davis, then allowed for Papers of particular interest, published within the annual period of
greater understanding of signaling pathways and biology. review, have been highlighted as:
Today, there are multiple kinase inhibitors approved for
 of special interest
use as drugs that impact human health in a positive way.  of outstanding interest

Protein–protein interactions are widely considered diffi-

1. Triggle DJ: The chemist as astronaut: searching for biologically
cult, or un-druggable, and it is generally accepted that  useful space in the chemical universe. Biochem Pharmacol
modulating them offers great promise for the treatment of 2009, 78:217-223.
Interesting discussion of strategies to access new biologically interesting
disease [134–137,138,139,140]. Biologicals (e.g. anti- compounds; the author points to the prevalent reductionist approach in
bodies) and natural products (e.g. rapamycin analogs) drug discovery as a key factor that needs to be addressed in the search
are already successful therapeutics in this important area. for new medicines.
Because small molecules have advantages as drugs, scien- 2. Jacoby E, Mozzarelli A: Chemogenomic strategies to expand
 the bioactive chemical space. Curr Med Chem 2009,
tists continue to delve into the nature of protein–protein 16:4374-4381.
interactions and look for ways to modulate them with This review highlights several different strategies to make new biologically
relevant molecules.
small molecules. This work has led to a growing set of
‘clues’ in the literature suggesting ways forward, and 3. Macarron R: Critical review of the role of HTS in drug discovery.
Drug Discov Today 2006, 11:277-279.
several recent papers summarize some of the recent
success and identify some common themes for effective 4. Bender A, Bojanic D, Davies JW, Crisman TJ, Mikhailov D,
 Scheiber J, Jenkins JL, Deng Z, Hill WAG, Popov M et al.: Which
‘protein mimicry’ [141,142,143–146]. A report by Wells aspects of HTS are empirically correlated with downstream
and McLendon analyzes recent examples of small mol- success? Curr Opin Drug Discov Dev 2008, 11:327-337.
This is a very nice review that discusses factors that influence the
ecule blockers of protein–protein interactions and pre- downstream success of HTS campaigns.
sents arguments to dispel some misconceptions about the
5. Imming P, Sinning C, Meyer A: Drugs, their targets and the
possibilities of drug discovery in this area [147]. nature and number of drug targets. Nat Rev Drug Discov 2006,
5:821-834.
Another possibility for addressing un-met medical need is 6. Agüero F, Al-Lazikani B, Aslett M, Berriman M, Buckner FS,
to work on new targets that offer rich possibilities, but Campbell RK, Carmona S, Carruthers IM, Chan AWE, Chen F</
ET-AL>: Genomic-scale prioritization of drug targets: the TDR
biological understanding is low and chemical tractability Targets database. Nat Rev Drug Discov 2008, 7:900-907.
is unknown. Efforts to identify chemical probes (potent, 7. Landry Y, Gies JP: Drugs and their molecular targets: an
selective, and cell permeable molecules) that modulate updated overview. Fundam Clin Pharmacol 2008, 22:1-18.

Current Opinion in Chemical Biology 2010, 14:289–298 www.sciencedirect.com

Enhancements of screening collections to address areas of unmet medical need: an industry perspective Drewry and Macarron 295

8. Overington JP, Al-Lazikani B, Hopkins AL: How many drug Good summary of the importance of protein cavity similarity to under-
targets are there? Nat Rev Drug Discov 2006, 5:993-996. stand cross-reactivity of ligands.
9. Fuller JC, Burgoyne NJ, Jackson RM: Predicting druggable 31. Payne DJ, Gwynn MN, Holmes DJ, Pompliano DL: Drugs for bad
binding sites at the protein–protein interface. Drug Discov bugs: confronting the challenges of antibacterial discovery.
Today 2009, 14:155-161. Nat Rev Drug Discov 2007, 6:29-40.
10. Hopkins AL, Groom CR: The druggable genome. Nat Rev Drug 32. Dixon SJ, Stockwell BR: Identifying druggable disease-modifying
Discov 2002, 1:727-730. gene products. Curr Opin Chem Biol 2009, 13:549-555.
11. Plewczynski D, Rychlewski L: Meta-basic estimates the size of 33. Nielsen TE, Schreiber SL: Towards the optimal screening
druggable human genome. J Mol Model 2009, 15:695-699. collection: a synthesis strategy. Angew Chem Int Ed 2008,
47:48-56.
12. Russ AP, Lampel S: The druggable genome: an update. Drug
Discov Today 2005, 10:1607-1610. 34. Irwin JJ: How good is your screening library? Curr Opin Chem
Biol 2006, 10:352-356.
13. Schalon C, Surgand JS, Kellenberger E, Rognan D: A simple and
fuzzy method to align and compare druggable ligand-binding 35. Jacoby E, Schuffenhauer A, Popov M, Azzaoui K, Havill B,
sites. Proteins Struct Funct Genet 2008, 71:1755-1778. Schopfer U, Engeloch C, Stanek J, Acklin P, Rigollier P</ET-AL>:
Key aspects of the Novartis compound collection
14. Egner U, Hillig RC: A structural biology view of target enhancement project for the compilation of a comprehensive
drugability. Expert Opin Drug Discov 2008, 3:391-401. chemogenomics drug discovery screening collection. Curr
15. Cheng AC: Predicting selectivity and druggability in drug Top Med Chem 2005, 5:397-411.
discovery. Annu Rep Comput Chem 2008, 4:23-37. 36. Akritopoulou-Zanze I, Hajduk PJ: Kinase-targeted libraries: the
16. Cheng AC, Coleman RG, Smyth KT, Cao Q, Soulard P, Caffrey DR, design and synthesis of novel, potent, and selective kinase
Salzberg AC, Huang ES: Structure-based maximal affinity inhibitors. Drug Discov Today 2009, 14:291-297.
model predicts small-molecule druggability. Nat Biotechnol 37. Sauer WHB, Schwarz MK: Molecular shape diversity of
2007, 25:71-75. combinatorial libraries: a prerequisite for broad bioactivity. J
17. Gupta A, Gupta AK, Seshadri K: Structural models in the Chem Inf Comput Sci 2003, 43:987-1003.
assessment of protein druggability based on HTS data. J 38. Matson SL, Chatterjee M, Stock DA, Leet JE, Dumas EA,
Comput Aided Mol Des 2009, 23:583-592. Ferrante CD, Monahan WE, Cook LS, Watson J, Cloutier NJ et al.:
18. Hajduk PJ, Huth JR, Tse C: Predicting protein druggability. Drug Best practices in compound management for preserving
Discov Today 2005, 10:1675-1682. compound integrity and accurately providing samples for
assays. J Biomol Screen 2009, 14:476-484.
19. Hajduk PJ, Huth JR, Fesik SW: Druggability indices for protein
targets derived from NMr-based screening data. J Med Chem 39. Bowes S, Sun D, Kaffashan A, Zeng C, Chuaqui C, Hronowski X,
2005, 48:2518-2525. Buko A, Zhang X, Josiah S: Quality assessment and analysis of
biogen idec compound library. J Biomol Screen 2006, 11:828-835.
20. Halgren TA: Identifying and characterizing binding sites and
assessing druggability. J Chem Inf Model 2009, 49:377-389. 40. Zitha-Bovens E, Maas P, Wife D, Tijhuis J, Hu QN, Kleinöder T,
Gasteiger J: COMDECOM: predicting the lifetime of screening
21. Seco J, Luque FJ, Barril X: Binding site detection and compounds in DMSO solution. J Biomol Screen 2009,
druggability index from first principles. J Med Chem 2009, 14:557-565.
52:2363-2371.
41. Blaxill Z, Holland-Crimmin S, Lifely R: Stability through the ages:
22. Sugaya N, Ikeda K: Assessing the druggability of protein– the GSK experience. J Biomol Screen 2009, 14:547-556.
protein interactions by a supervised machine-learning
method. BMC Bioinform 2009, 10:263. 42. Lane SJ, Eggleston DS, Brinded KA, Hollerton JC, Taylor NL,
Readshaw SA: Defining and maintaining a high quality
23. Walke DW, Han C, Shaw J, Wann E, Zambrowicz B, Sands A: In screening collection: the GSK experience. Drug Discov Today
vivo drug target discovery: identifying the best targets from 2006, 11:267-272.
the genome. Curr Opin Biotechnol 2001, 12:626-631.
43. Han L, Wang Y, Bryant SH: A survey of across-target bioactivity
24. Stumpf MPH, Thorne T, De Silva E, Stewart R, Hyeong JA, results of small molecules in PubChem. Bioinformatics 2009,
Lappe M, Wiuf C: Estimating the size of the human interactome. 25:2251-2255.
Proc Natl Acad Sci USA 2008, 105:6959-6964.
44. Peters JU, Schnider P, Mattei P, Kansy M: Pharmacological
25. Zhu F, Han L, Zheng C, Xie B, Tammi MT, Yang S, Wei Y, Chen Y: promiscuity: dependence on compound properties and target
What are next generation innovative therapeutic targets? Clues specificity in a set of recent Roche compounds.
from genetic, structural, physicochemical, and systems profiles ChemMedChem 2009, 4:680-686.
of successful targets. J Pharmacol Exp Ther 2009, 330:304-315.
45. Feng BY, Simeonov A, Jadhav A, Babaoglu K, Inglese J,
26. Bakheet TM, Doig AJ: Properties and identification of human Shoichet BK, Austin CP: A high-throughput screen for
protein drug targets. Bioinformatics 2009, 25:451-457. aggregation-based inhibition in a large compound library.
J Med Chem 2007, 50:2385-2390.
27. Ernst B, Magnani JL: From carbohydrate leads to glycomimetic
drugs. Nat Rev Drug Discov 2009, 8:661-677. 46. Shoichet BK: Screening in a spirit haunted world. Drug Discov
Today 2006, 11:607-615.
28. Spencer RW: High-throughput screening of historic
collections: observations on file size, biological targets, and 47. Coan KED, Maltby DA, Burlingame AL, Shoichet BK:
file diversity. Biotechnol Bioeng 1998, 61:61-67. Promiscuous aggregate-based nhibitors promote enzyme
unfolding. J Med Chem 2009, 52:2067-2075.
29. Paolini GV, Shapland RHB, Van Hoorn WP, Mason JS, Hopkins AL:
 Global mapping of pharmacological space. Nat Biotechnol 48. Hann MM, Leach AR, Harper G: Molecular complexity and its
2006, 24:805-815. impact on the probability of finding leads for drug discovery.
In this work, the scientists integrated a number of data sources to show J Chem Inf Comput Sci 2001, 41:856-864.
which human targets chemical tools and drugs have been identified for,
and created a framework for making connections between molecules, 49. Schuffenhauer A, Brown N, Selzer P, Ertl P, Jacoby E: Relationships
their properties, and proteins. It would be great to update the dataset to between molecular complexity, biological activity, and
cope with progress in the field. structural diversity. J Chem Inf Model 2006, 46:525-535.
30. Weskamp N, Hüllermeier E, Klebe G: Merging chemical and 50. Davis AM, Keeling DJ, Steele J, Tomkinson NP, Tinker AC:
 biological space: structural mapping of enzyme binding Components of successful lead generation. Curr Top Med
pocket space. Proteins 2009, 76:317-330. Chem 2005, 5:421-439.

www.sciencedirect.com Current Opinion in Chemical Biology 2010, 14:289–298

296 Molecular Diversity

51. Axerio-Cilies P, Castañeda IP, Mirza A, Reynisson J: Investigation 71. Butler MS: The role of natural product chemistry in drug
of the incidence of ‘‘undesirable’’ molecular moieties for high- discovery. J Nat Prod 2004, 67:2141-2153.
throughput screening compound libraries in marketed drug
compounds. Eur J Med Chem 2009, 44:1128-1134. 72. Cragg GM, Grothaus PG, Newman DJ: Impact of natural
products on developing new anti-cancer agents. Chem Rev
52. Monge A, Arrault A, Marot C, Morin-Allory L: Managing, profiling 2009, 109:3012-3043.
and analyzing a library of 2.6 million compounds gathered
from 32 chemical providers. Mol Divers 2006, 10:389-403. 73. Newman DJ, Cragg GM: Natural products as sources of new
 drugs over the last 25 years. J Nat Prod 2007, 70:461-477.
53. Schreiber SL: Organic chemistry: molecular diversity by A comprehensive review of the source of drugs affirming natural products
design. Nature 2009, 457:153-154. as a very special source of unique starting points for drug discovery.
54. Rishton GM: Molecular diversity in the context of leadlikeness: 74. Butler MS: Natural products to drugs: natural product-derived
 compound properties that enable effective biochemical compounds in clinical trials. Nat Prod Rep 2008, 25:475-516.
screening. Curr Opin Chem Biol 2008, 12:340-351.
Good description of the concept of lead-likeness, especially as it relates 75. Harvey AL: Natural products in drug discovery. Drug Discov
to various types of biological screening. Today 2008, 13:894-901.

55. Verheij HJ: Leadlikeness and structural diversity of synthetic 76. Hert J, Irwin JJ, Laggner C, Keiser MJ, Shoichet BK: Quantifying
screening libraries. Mol Divers 2006, 10:377-388.  biogenic bias in screening libraries. Nat Chem Biol 2009,
5:479-483.
56. Harper G, Pickett SD, Green DVS: Design of a compound Introduces an interesting approach to quantify the similarity of screening
screening collection for use in high throughput screening. sets to biogenic molecules (metabolites and natural products), leading to
Comb Chem High Throughput Screen 2004, 7:63-70. the suggestion that pushing this bias even further will enhance success.

57. Brenk R, Schipani A, James D, Krasowski A, Gilbert IH, Frearson J, 77. Gupta S, Aires-de-Sousa J: Comparing the chemical spaces of
Wyatt PG: Lessons learnt from assembling screening libraries metabolites and available chemicals: models of metabolite-
for drug discovery for neglected diseases. ChemMedChem likeness. Mol Divers 2007, 11:23-36.
2008, 3:435-444.
78. Dobson PD, Patel Y, Kell DB: ’Metabolite-likeness’ as a criterion
58. Snowden M, Green DVS: The impact of diversity-based, high-  in the design and selection of pharmaceutical drug libraries.
 throughput screening on drug discovery: ‘‘Chance favours the Drug Discov Today 2009, 14:31-40.
prepared mind’’. Curr Opin Drug Discov Dev 2008, 11:553-558. This paper compares the similarity of known drugs and library com-
The utility and impact of enhancing compound collections with ‘purpose- pounds to naturally occurring metabolites. Based on the descriptors used
ful diversity’ is described. here, drugs are more similar to metabolites than library compounds are,
and this information can be used to narrow the search space as we strive
59. Lowrie JF, Delisle RK, Hobbs DW, Diller DJ: The different to make biologically active compounds.
strategies for designing GPCR and kinase targeted libraries.
Comb Chem High Throughput Screen 2004, 7:495-510. 79. Sukuru SCK, Jenkins JL, Beckwith REJ, Scheiber J, Bender A,
Mikhailov D, Davies JW, Glick M: Plate-based diversity selection
60. Gregori-Puigjané E, Mestres J: Coverage and bias in chemical based on empirical HTS data to enhance the number
library design. Curr Opin Chem Biol 2008, 12:359-365. of hits and their chemical diversity. J Biomol Screen 2009,
14:690-699.
61. Stewart EL, Brown PJ, Bentley JA, Willson TM: Selection,
application, and validation of a set of molecular descriptors for 80. Singh N, Guha R, Giulianotti MA, Pinilla C, Houghten RA, Medina-
nuclear receptor ligands. Comb Chem High Throughput Screen Franco JL: Chemoinformatic analysis of combinatorial
2004, 7:407-412. libraries, drugs, natural products, and molecular libraries
Small Molecule Repository. J Chem Inf Model 2009,
62. Aronov AM, McClain B, Moody CS, Murcko MA: Kinase-likeness 49:1010-1024.
and kinase-privileged fragments: toward virtual
polypharmacology. J Med Chem 2008, 51:1214-1222. 81. Feher M, Schmidt JM: Property distributions: differences
between drugs, natural products, and molecules from
63. Deanda F, Stewart EL, Reno MJ, Drewry DH: Kinase-targeted combinatorial chemistry. J Chem Inf Comput Sci 2003,
library design through the application of the pharmprint 43:218-227.
methodology. J Chem Inf Model 2008, 48:2395-2403.
82. Shelat AA, Guy RK: The interdependence between screening
64. Decornez H, Gulyás-Forró A, Papp Á, Szabó M, Sármay G, Hajdú I, methods and screening libraries. Curr Opin Chem Biol 2007,
Cseh S, Dormán G, Kitchen DB: Design, selection, and 11:244-251.
evaluation of a general kinase-focused library. ChemMedChem
2009, 4:1273-1278. 83. Shelat AA, Guy RK: Scaffold composition and biological
relevance of screening libraries. Nat Chem Biol 2007, 3:442-446.
65. Gozalbes R, Simon L, Froloff N, Sartori E, Monteils C, Baudelle R:
Development and experimental validation of a docking 84. Grabowski K, Baringhaus KH, Schneider G: Scaffold diversity of
strategy for the generation of kinase-targeted libraries. J Med natural products: inspiration for combinatorial library design.
Chem 2008, 51:3124-3132. Nat Prod Rep 2008, 25:892-904.

66. Sun D, Chuaqui C, Deng Z, Bowes S, Chin D, Singh J, Cullen P, 85. Ganesan A: The impact of natural products upon modern drug
Hankins G, Lee WC, Donnelly J et al.: A kinase-focused  discovery. Curr Opin Chem Biol 2008, 12:306-317.
compound collection: compilation and screening strategy. This paper presents a detailed look at the properties of 24 drugs that
Chem Biol Drug Des 2006, 67:385-394. came from natural products, and highlights the importance of log P and
the number of H-bond donors. Ganesan presents strong arguments for
67. Ertl P, Roggo S, Schuffenhauer A: Natural product-likeness continued efforts in natural product research as a means for identifying
score and its application for prioritization of compound valuable medicines.
libraries. J Chem Inf Model 2008, 48:68-74.
86. Beghyn T, Deprez-Poulain R, Willand N, Folleas B, Deprez B:
68. Ji HF, Li XJ, Zhang HY: Natural products and drug discovery: Natural compounds: leads or ideas? Bioinspired molecules for
can thousands of years of ancient medical knowledge drug discovery. Chem Biol Drug Des 2008, 72:3-15.
lead us to new and powerful drug combinations in
the fight against cancer and dementia? EMBO Rep 2009, 87. Balamurugan R, Dekker FJ, Waldmann H: Design of compound
10:194-200. libraries based on natural product scaffolds and protein
structure similarity clustering (PSSC). Mol Biosyst 2005,
69. Li JWH, Vederas JC: Drug discovery and natural products: end 1:36-45.
of an era or an endless frontier? Science 2009, 325:161-165.
88. Rosén J, Gottfries J, Muresan S, Backlund A, Oprea TI: Novel
70. Chin YW, Balunas MJ, Chai HB, Kinghorn AD: Drug discovery  chemical space exploration via natural products. J Med Chem
from natural sources. AAPS J 2006, 8:E239-E253. 2009, 52:1953-1962.

Current Opinion in Chemical Biology 2010, 14:289–298 www.sciencedirect.com

Enhancements of screening collections to address areas of unmet medical need: an industry perspective Drewry and Macarron 297

This paper contains an analysis of the chemical space of drugs and 107. Biros SM, Moisan L, Mann E, Carella A, Zhai D, Reed JC, Rebek J:
natural products using the ChemGPS-NP tool and features identification Heterocyclic a-helix mimetics for targeting protein–protein
of low density regions, regions of intersection, and natural products with interactions. Bioorg Med Chem Lett 2007, 17:4641-4645.
similarity (using these property-based descriptors) to known drugs.
108. Haridas V: From peptides to non–peptide alpha-helix inducers
89. Larsson J, Gottfries J, Muresan S, Backlund A: ChemGPS-NP: and mimetics. Eur J Org Chem 2009, 30:5112-5128.
tuned for navigation in biologically relevant chemical space. J
Nat Prod 2007, 70:789-794. 109. Moisan L, Odermatt S, Gombosuren N, Carella A, Rebek J:
Synthesis of an oxazole-pyrrole-piperazine scaffold as an a-
90. Grabowski K, Schneider G: Properties and architecture of drugs helix mimetic. Eur J Org Chem 2008, 10:1673-1676.
and natural products revisited. Curr Chem Biol 2007, 1:115-127.
110. Restorp P, Rebek J: Synthesis of a-helix mimetics with four
91. Dekker FJ, Koch MA, Waldmann H: Protein structure similarity side-chains. Bioorg Med Chem Lett 2008, 18:5909-5911.
clustering (PSSC) and natural product structure as inspiration
sources for drug development and chemical genomics. Curr 111. Saraogi I, Hamilton AD: a-Helix mimetics as inhibitors of
Opin Chem Biol 2005, 9:232-239. protein–protein interactions. Biochem Soc Trans 2008,
36:1414-1417.
92. Koch MA, Waldmann H: Protein structure similarity clustering
and natural product structure as guiding principles in drug 112. Lomlim L, Einsiedel J, Heinemann FW, Meyer K, Gmeiner P:
discovery. Drug Discov Today 2005, 10:471-483. Proline derived spirobarbiturates as highly effective a-turn
mimetics incorporating polar and functionalizable constraint
93. Koch MA, Wittenberg LO, Basu S, Jeyaraj DA, Gourzoulidou E, elements. J Org Chem 2008, 73:3608-3611.
Reinecke K, Odermatt A, Waldmann H: Compound library
development guided by protein structure similarity clustering 113. Hirschmann RF, Nicolaou KC, Angeles AR, Chen JS, Smith AB III:
and natural product structure. Proc Natl Acad Sci USA 2004, The b-D-glucose scaffold as a b-turn mimetic. Acc Chem Res
101:16721-16726. 2009, 42:1511-1520.

94. Arve L, Voigt T, Waldmann H: Charting biological and chemical 114. Sun H, Nikolovska-Coleska Z, Yang CY, Qian D, Lu J, Qiu S, Bai L,
space: PSSC and SCONP as guiding principles for the Peng Y, Cai Q, Wang S: Design of small-molecule peptidic
development of compound collections based on natural and nonpeptidic smac mimetics. Acc Chem Res 2008,
product scaffolds. QSAR Comb Sci 2006, 25:449-456. 41:1264-1277.

95. Koch MA, Schuffenhauer A, Scheck M, Wetzel S, Casaulta M, 115. Che Y, Marshall GR: Privileged scaffolds targeting reverse-turn
Odermatt A, Ertl P, Weldmann H: Charting biologically relevant and helix recognition. Expert Opin Ther Targets 2008,
chemical space: a structural classification of natural products 12:101-114.
(SCONP). Proc Natl Acad Sci USA 2005, 102:17272-17277.
116 . Kumar K, Waldmann H: Synthesis of natural product inspired
96. Wetzel S, Klein K, Renner S, Rauh D, Oprea TI, Mutzel P,  compound collections. Angew Chem Int Ed 2009, 48:3224-3242.
Waldmann H: Interactive exploration of chemical space with This review gives numerous examples of strategies used to synthesize
Scaffold Hunter. Nat Chem Biol 2009, 5:581-583. sets of natural product inspired compounds, and demonstrates the
feasibility of infusing screening collections with more complicated mole-
97. Renner S, Van Otterlo WAL, Dominguez Seoane M, cules inspired by nature.
Möcklinghoff S, Hofmann B, Wetzel S, Schuffenhauer A, Ertl P,
Oprea TI, Steinhilber D et al.: Bioactivity-guided mapping and 117. Shaw JT: Naturally diverse: highlights in versatile synthetic
navigation of chemical space. Nat Chem Biol 2009, 5:585-592. methods enabling target- and diversity-oriented synthesis.
Nat Prod Rep 2009, 26:11-26.
98. Sunderhaus JD, Martin SF: Applications of multicomponent
reactions to the synthesis of diverse heterocyclic scaffolds. 118. Peuchmaur M, Wong YS: Expanding the chemical space in
Chem Eur J 2009, 15:1300-1308. practice: diversity-oriented synthesis. Comb Chem High
Throughput Screen 2008, 11:587-601.
99. Thomas GL, Spandl RJ, Glansdorp FG, Welch M, Bender A,
Cockfield J, Lindsay JA, Bryant C, Brown DF, Loiseleur O et al.: 119. Cordier C, Morton D, Murrison S, Nelson A, O’Leary-Steele C:
Anti-MRSA agent discovery using diversity-oriented Natural products as an inspiration in the diversity-oriented
synthesis. Angew Chem Int Ed 2008, 47:2808-2812. synthesis of bioactive compound libraries. Nat Prod Rep 2008,
25:719-737.
100. DiMicco S, Vitale R, Pellechia M, Rega MF, Riva R, Basso A,
Bifulco G: Identification of lead compounds as antagonists of 120. Galloway WRJD, Bender A, Welch M, Spring DR: The discovery of
protein Bcl-xL with a diversity-oriented multidisciplinary antibacterial agents using diversity-oriented synthesis. Chem
approach. J Med Chem 2009, 52:7856-7867. Commun 2009, 18:2446-2462.
101. Zhou A, Rayabarapu D, Hanson PR: ‘‘Click, click, cyclize’’: a dos 121. Albert JS, Blomberg N, Breeze AL, Brown AJH, Burrows JN,
approach to sultams utilizing vinyl sulfonamide linchpins. Org Edwards PD, Folmer RHA, Geschwindner S, Griffen EJ, Kenny PW
Lett 2009, 11:531-534. et al.: An integrated approach to fragment-based lead
generation: philosophy, strategy and case studies from
102. Waldmann H, Kühn M, Liu W, Kumar K: Reagent-controlled AstraZeneca’s drug discovery programmes. Curr Top Med
domino synthesis of skeletally-diverse compound collections. Chem 2007, 7:1600-1629.
Chem Commun 2008, 10:1211-1213.
122. Alex AA, Flocco MM: Fragment-based drug discovery: what has
103. Morton D, Leach S, Cordier C, Warriner S, Nelson A: Synthesis of it achieved so far? Curr Top Med Chem 2007, 7:1544-1567.
 natural product-like molecules with over eighty distinct
scaffolds. Angew Chem Int Ed 2009, 48:104-109. 123. Chen IJ, Hubbard RE: Lessons for fragment library design:
This paper presents an elegant example of diversity oriented synthesis analysis of output from multiple screening campaigns. J
and the power of the ring closing metathesis reaction. Comput Aided Mol Des 2009, 23:603-620.
104. Kombarov R, Altieri A, Genis D, Kirpichenok M, Kochubey V, 124. Congreve M, Chessari G, Tisi D, Woodhead AJ: Recent
Rakitina N, Titarenko Z: BioCores: identification of a drug/ developments in fragment-based drug discovery. J Med Chem
natural product-based privileged structural motif for small- 2008, 51:3661-3680.
molecule lead discovery. Mol Divers 2010, 14:193-200.
125. Eitner K, Koch U: From fragment screening to potent binders:
105. Marcaurelle LA, Johannes C, Yohannes D, Tillotson BP, Mann D: strategies for fragment-to-lead evolution. Mini Rev Med Chem
Diversity-oriented synthesis of a cytisine-inspired pyridone 2009, 9:956-961.
library leading to the discovery of novel inhibitors of Bcl-2.
Bioorg Med Chem Lett 2009, 19:2500-2503. 126. Neumann L, Ritscher A, Müller G, Hafenbradl D: Fragment-based
lead generation: identification of seed fragments by a highly
106. Garner J, Harding MM: Design and synthesis of a-helical efficient fragment screening technology. J Comput Aided Mol
peptides and mimetics. Org Biomol Chem 2007, 5:3577-3585. Des 2009, 23:501-511.

www.sciencedirect.com Current Opinion in Chemical Biology 2010, 14:289–298

298 Molecular Diversity

127. Warr WA: Fragment-based drug discovery. J Comput Aided Mol This paper presents examples of recent successes in inhibiting protein–
Des 2009, 23:453-458. protein interactions.
128. Wendt MD, Sun C, Kunzer A, Sauer D, Sarris K, Hoff E, Yu L, 139. Villoutreix BO, Bastard K, Sperandio O, Fahraeus R, Poyet JL,
Nettesheim DG, Chen J, Jin S et al.: Discovery of a novel small Calvo F, Déprez B, Miteva MA: In silico-in vitro screening of
molecule binding site of human survivin. Bioorg Med Chem Lett protein–protein interactions: towards the next generation of
2007, 17:3122-3129. therapeutics. Curr Pharm Biotechnol 2008, 9:103-122.
129. Bodoor K, Boyapati V, Gopu V, Boisdore M, Allam K, Miller J, 140. Wilson AJ: Inhibition of protein–protein interactions using
Treleaven WD, Weldeghiorghis T, Aboul-ela F: Design and  designed molecules. Chem Soc Rev 2009, 38:3289-3300.
implementation of an ribonucleic acid (RNA) directed This review highlights strategies pursued to inhibit protein–protein inter-
fragment library. J Med Chem 2009, 52:3753-3761. actions.
130. Schüller A, Suhartono M, Fechner U, Tanrikulu Y, Breitung S, 141. Fry DC: Protein–protein interactions as targets for small
Scheffer U, Göbel MW, Schneider G: The concept of template- molecule drug discovery. Biopolym Pept Sci Sect 2006,
based de novo design from drug-derived molecular fragments 84:535-552.
and its application to TAR RNA. J Comput Aided Mol Des 2008,
22:59-68. 142 . Fry DC: Drug-like inhibitors of protein–protein interactions: a
 structural examination of effective protein mimicry. Curr
131. Chessari G, Woodhead AJ: From fragment to clinical Protein Pept Sci 2008, 9:240-247.
candidate—a historical perspective. Drug Discov Today 2009, In this review of recent successful cases of protein–protein interactions,
14:668-675. the author extracts some common traits that may prove valuable in
132. Clark MA, Acharya RA, Arico-Muendel CC, Belyanskaya SL, protein–protein inhibitor design.
Benjamin DR, Carlson NR, Centrella PA, Chiu CH, Creaser SP, 143. Higueruelo AP, Schreyer A, Bickerton GRJ, Pitt WR, Groom CR,
Cuozzo JW et al.: Design, synthesis and selection of DNA- Blundell TL: Atomic interactions and profile of small molecules
encoded small-molecule libraries. Nat Chem Biol 2009, disrupting protein–protein interfaces: the TIMBAL database.
5:647-654. Chem Biol Drug Des 2009, 74:457-467.
133. Liao JJL: Molecular recognition of protein kinase binding 144. Neugebauer A, Hartmann RW, Klein CD: Prediction of protein–
pockets for design of potent and selective kinase inhibitors. J protein interaction inhibitors by chemoinformatics and
Med Chem 2007, 50:409-424. machine learning methods. J Med Chem 2007, 50:4665-4668.
134. Zinzalla G, Thurston DE: Targeting protein–protein interactions
145. Keskin O, Gursoy A, Ma B, Nussinov R: Principles of protein–
for therapeutic intervention: a challenge for the future. Future
protein interactions: what are the preferred ways for proteins
Med Chem 2009, 1:65-93.
to interact? Chem Rev 2008, 108:1225-1244.
135. Murray JK, Gellman SH: Targeting protein–protein interactions:
lessons from p53/MDM2. Biopolym Pept Sci Sect 2007, 146. Ma B, Nussinov R: Trp/Met/Phe hot spots in protein–protein
88:657-686. interactions: potential targets in drug design. Curr Top Med
Chem 2007, 7:999-1005.
136. Verdine GL, Walensky LD: The challenge of drugging
undruggable targets in cancer: lessons learned from targeting 147 . Wells JA, McClendon CL: Reaching for high-hanging fruit in
BCL-2 family members. Clin Cancer Res 2007, 13:7264-7270.  drug discovery at protein–protein interfaces. Nature 2007,
450:1001-1009.
137. Arkin MR, Whitty A: The road less traveled: modulating signal This article succinctly presents data to help dispel some myths about
transduction enzymes by inhibiting their protein–protein disrupting protein–protein interactions.
interactions. Curr Opin Chem Biol 2009, 13:284-290.
148. Edwards AM, Bountra C, Kerr DJ, Willson TM: Open access
138. Berg T: Small-molecule inhibitors of protein–protein chemical and clinical probes to support drug discovery. Nat
 interactions. Curr Opin Drug Discov Dev 2008, 11:666-674. Chem Biol 2009, 5:436-440.

Current Opinion in Chemical Biology 2010, 14:289–298 www.sciencedirect.com