Reigniting pharmaceutical innovation through holistic drug ...

From the turn of the 20th century, pharma-cognosy and ethnopharmacology combinedwith anecdotal clinical evidence accumulat-

ed over centuries of hands-on knowledge from pri-mordial disease management practices, albeit withuncertain outcomes, formed the basis for the devel-opment of drugs. Reverse pharmacology, deeprooted in traditional medicine, laid the foundationfor the emergence and evolution of modern drugdiscovery approaches as a highly formal and regi-mented science. Advances in genomics, assay andcombinatorial chemistry technologies, informaticsand robotics led to increased screening operationssignificantly compared with traditional discoverymethods. The pharmaceutical industry-driven high

and ultra-high throughput workflows enabledscreening of millions of compounds to identify hitcandidates for lead development. Despite billionsof dollars spent on R&D, only a fraction of themolecules identified from the screening operationsfind their way into clinical trials. Difficulties inpredicting safety profiles, redundancies and effica-cies across a genetically-diverse patient populationcontribute to the high attrition rates in clinical tri-als. Success in clinical trials was found to be affect-ed, among other reasons, by the poor design ofclinical trials, selected clinical end-points unsuitedfor the desired outcome and lack of evaluation ofpharmacogenomics of the patient populationselected for the indication. In addition, while a

By Dr Anuradha Roy,Professor BhushanPatwardhan and Dr RathnamChaguturu

Drug Discovery World Summer 2016 45

Drug Discovery

Reigniting pharmaceuticalinnovation through holisticdrug targeting

Modern drug discovery approaches take too long, are too expensive, have toomany clinical failures and uncertain outcomes. There are many reasons for thisunsustainable business model, but primarily, the approaches are notcomprehensively holistic. Secondly, none of the pharmaceutical companiesopenly share the reasons for the failure of their clinical candidates in real timeto effectively navigate the ‘industry’ from committing the same mistakes. It istime for the pharmaceutical industry to embrace, metaphorically speaking, acommunity-driven ‘Wikipedia’ or ‘Waze’-type shared-knowledge, openly-accessible innovation model to harvest data and create a crowd-sourced pathtowards a safer and faster road to the discovery and development of life-savingmedicines. This may be a bitter pill for Pharma to swallow, but one that oughtto be given serious consideration. The time is now for a paradigm shift towardsmulti-target-network polypharmacology drugs exalting symphonic or concertperformance with occasional soloists to reignite pharmaceutical innovation.

majority of the marketed drugs was approvedbased on experimental data supporting single tar-get selectivity, recent clinical and basic studies haveunambiguously shown that many marketed drugslack absolute selectivity.

Paradigm shiftDespite high failure rates of compounds in clinicalsettings, target-based drug discovery has contribut-ed to drugs approved for several indications. Anoverall evaluation of the FDA-approved newmolecular entities reveal that the majority of drugsclustered into previously known classes, andencompasses a limited number of molecular targetsand diseases. A recent analysis of the DrugBankdatabase identified 435 targets modulated by 989drugs and include GPCRs, ligand gated ion-chan-nels, receptor tyrosine kinases and certain otherclasses of enzymes1. The analysis also highlightedthe fact that only half of the marketed drugs showhigher order and number of drug-target interac-tions within a very limited proteome-scape. The

vast majority of FDA approved drugs target dis-eases affecting large populations such as cancers,infectious diseases or cardiovascular diseases,which promises a good return on R&D invest-ment; and hence pursued extensively by bigPharma. Rare diseases (which affect small popula-tions) or neglected diseases (more prevalent indeveloping countries) have largely been ignoredbecause of low profitability for the pharmaceuticalindustry and low affordability for the poor. Marketsize decisions as well as business portfolios are alsoknown to influence indication selection, popula-tion sizes and end-points in clinical trials. The lastdecade witnessed patent expirations, drug recallsand toxicity-driven withdrawals, all of which neg-atively impacted pharma R&D productivity.Despite significant advances in our knowledge,safety from prolonged use of drugs in the post-FDA approval period remains uncertain.Treatment of complex diseases with single or smallcombination therapies, while effective in the shortterm or at early stages of a disease, continues to beinsufficient in mitigating advanced or recurrentdisease progression. Loss of responsiveness due tolong-term administration of single agents isattributable to the robustness of the redundantmolecular functions within biological networks.

Many diseases are complex, heterogeneous andmultifactorial, have several phenotypes, variablerisk factors and responses that are also influencedby genetic variations, age, gender and environ-mental factors such as diet, microbiome andlifestyle choices. The ever too often seen resistanceto continued use of drugs and loss in treatmentefficacy are generally mediated by networkrobustness (signalling pathway redundancy andcrosstalk) causing compensatory or counter-targetactivities or neutralising action. The overall lowproductivity and innovation of new drug classes,as well as safety- and efficacy-driven drug recalls,begs for a fundamental paradigm shift in currentdrug discovery approaches.

Druggability of genomeOf the 19,000 human protein coding genes pre-dicted from human genome sequence analysis,only about 10% are ‘druggable’ by in silico anal-ysis2. The majority of the ‘products’ from genome,such as the proteins or RNA, still remain function-ally unclassified and the expansion of functionali-ty may be further influenced by epigenomic eventsas well as post-transcriptional or post-translation-al regulation3. The challenge lies in generatinghigh quality data for target identification, drugga-bility via siRNA or chemical perturbation in

46 Drug Discovery World Summer 2016

Drug Discovery

healthy and disease states, confounded by the exis-tence of functional and physical networks betweenmolecular players in the same or other signallingpathways. The hub proteins such as p53, p27,BRCA1, ubiquitin and calmodulin can have fiveor more interactions with other proteins, and mayeither exhibit a single transient interaction at anygiven time or participate in several interactionssimultaneously. The cross-talk and interdepen-dence of biological networks adds another level ofcomplexity in responses to perturbations withchemicals, biologics as well as other factors.Systems analysis of large networks indicates thattargeting upstream events, hub proteins or redun-dancies will potentially be more impactful inquantitative and qualitative efficiency of cellularresponse. In addition to proteins and networks,targeting RNA druggability is still under-utilised.All known RNA motifs and their small moleculebinders, compiled in the Inforna database4, areattractive to interrogate and yield lead moleculeswith potential to be validated in disease systems,as exemplified by the evidence drawn from theselective binding of 5’-azido-neomycin B toDrosha and Dicer processing sites of miR-525, amicroRNA overexpressed in liver cancer5. A tar-get-agnostic approach can thus be effectively usedto identify in vivo modulators of oncogenic andother disease relevant microRNAs.

Comprehensive target biologyStrong correlative evidence between a target or tar-gets and a disease is ever more critical for the suc-cess of rationale-driven drug discovery. The successof designing well-tailored assays for primary andsecondary drug discovery screens, prediction oftoxicity profiles, biomarker profiling and ultimatepatient population-specific responses are all depen-dent on how much information, both accurate andreproducible, is available for the disease-relevant,therapeutic target6. It is essential to characterise atarget in both normal and diseased states for itsfunctional domain(s) redundancy, alternativelyspliced forms, subcellular localisation, tissue-spe-cific protein and RNA expression profiles,RNA/protein half-life, transcriptional/post-transla-tional regulation and effect of dominant negativemutant isoforms. Both genetic and chemical vali-dation helps define druggability and translationalpotential of a target. The same target may be mod-ulated in several distinct cancer types or in severalseemingly unrelated disorders, and as a result willhelp expand the indications or design treatments inclinical trials. An array of tools has become avail-able in recent years for target identification, char-

acterisation and validation. These tools include,but are not limited to, protein and RNA microar-rays, bioinformatics-driven protein interactionmaps, signalling pathways, phenotypic changes byfunctional studies using genetic-based technologies(RNA interference, knockdowns, overexpression,genomic mutations using clustered regularly inter-spaced short palindromic repeats (CRISPR-Cas),and mouse, zebrafish or Caenorhabditis elegansmodels). The exhaustive information pertaining tothe therapeutically-relevant drug target can facili-tate efficient design of assays or selection of appro-priate panel of cell-lines in targeted or phenotypicdrug discovery approaches to identify compoundsthat can serve as chemical tools to dissect the path-ways of interest.

Hit to lead strategiesOnce screen actives or hits are identified, it isessential to experimentally determine the mecha-nism of action, efficacy and safety of the scaffoldsand target on-off occupancy and engagementtime7. The direct binding of the hit molecules tocellular targets can be determined using mass spec-trometry on protein samples from cellular thermalshift assays, protein arrays, or compound-affinitychromatography. The binding affinities to specificand off-targets, binding site architecture, co-crys-tallography studies also provide invaluable infor-mation on understanding the chemical species.From the start scaffolds may be tested in assaysquantifying potency, selectivity, specificity,lipophilicity, molecular mass and early ADME(absorption, distribution, metabolism and excre-tion). Collaborative effort between AstraZeneca,Eli Lilly, GlaxoSmithKline and Pfizer to under-stand the reasons behind drug candidate attrition8

indicated a strong statistical significance betweenthe physicochemical properties of compounds andsafety failure rates in clinical trials, although norationale emerged for predicting which compoundswill be successful in clinical trials. Viable leadsshare the properties of high efficacy and potencyagainst a specific target, bind the target directly atthe same cellular site where the target is expressed,have low off-target effects, no activity againstundesirable players such as hERG channels andacceptable toxicity profiles. Target engagement oftime and activity at the target tissue/organ requiresdesigning pharmacokinetics/pharmacodynamics(PK/PD) modelling in both preclinical and clinicalmodels. Preclinical drug development that canaccurately predict clinical response of drugs caninfluence the success rate of lead compounds in latestage drug discovery.

Drug Discovery World Summer 2016 47

Drug Discovery

Relevant animal modelsThere is a growing awareness of the limitations ofthe true translation potential of clinical researchfrom mouse models; a case in point being that theaverage rate of successful translation from animalmodels to clinical cancer trials is less than 8%.Differences in genetics and physiology of mouseand humans add to issues in translational rele-vance. Cancer models involving mouse subcuta-neous xenografts may or may not translate well inlater stages of drug discovery. Mouse xenograftmodels from primary tumour-derived patient cellswith distinct driver mutations are often used to val-idate the clinical translatability of drug candidates,but are inadequate models for tumour initiationand progression. Several potential alternativesinclude using CRISPR-Cas technology to modifythe mouse genome specifically and use of large ani-mal models (rabbits, dogs, pigs, goats, sheep andnon-human primates) to closely replicate thehuman physiology and disease.

Exploiting pan-omicsCumulative knowledge acquired in recent yearshas led to an increased emphasis on designingtreatments that are more personalised, or clus-tered into subtypes based on information collated

from all sources and databases. The last decadewitnessed an explosion in data sets formetabolomics, transcriptomics, genomics, epige-nomics and proteomics etc. The high-throughput‘omics’ technologies applied to normal and dis-eased states have generated large volumes of dis-tinct types of omics data. Each type of data has itsown properties and requires specific analysismethods and tools. The emerging paradigm fordrug discovery underscores the value of integrat-ing the multi-omics data to build a completemodel of the processes in biological systems sincesignalling network cross-talks, functional biologi-cal redundancies and protein interactomes clearlyinfluence the drug activity and safety profiles. Amore holistic approach, designing drug discoveryprotocols that integrate the knowledge from tradi-tional medicine to mining omics datasets anddesigning smarter clinical trials, is certain toimprove the drug discovery process by providingnew and first-in-class drugs as well as in expand-ing the repertoire of global diseases targeted.

GenomicsAdvances in next-generation sequencing (NGS)platforms have led to exponential increase insequencing data from whole genomes (WGS),

48 Drug Discovery World Summer 2016

Drug Discovery

DNA coding exomes (WES), DNA variants (SNP,single nucleotide polymorphisms, deletions, inser-tions), DNS segments interacting with proteins(chromatin immunoprecipitation, Chl-IP),mRNAs (transcriptome), long non-coding RNA(lncRNA) and circular RNA (circRNA). Genome-wide association studies (GWAS) have the poten-tial to identify genetic variations, especially thesingle nucleotide polymorphisms, associated witha disease. Data on DNA variations is now avail-able for Crohn’s disease, Type 2 diabetes, prostatecancer, age-related macular degeneration, obesityand 50 other human diseases. The potential ofestablishing associations between disease andgenetic variations can ultimately influence diag-nostics and disease management.

PharmacogenomicsPrecision medicine can improve efficacy scores ofFDA-approved drugs or new compounds in clini-cal trials by integrating information from pharma-cogenomics. A genetic profile-based identificationof patient population subgroups is helpful in pre-dicting patient response to drug dosage, mecha-nism of action, minimising drug-related toxicityand adverse events. Pharmacogenomics data pro-vides information on genetic marker-driven effectson drug response, refractiveness of drug treat-ments, as well as functional effects of genomicvariants. The concept of pharmaco-metabolomics-aided pharmacogenomics helps combine the rolesof environment and gut microbiome interactionsin interpretation of especially immune mediateddisorders.

Epigenomics and transcriptomicsEpigenome encompasses all epigenetic mechanismsthat regulate chromatin via DNA methylation andhistone modifications (acetylation, methylation,phosphorylation, sumoylation, ubiquitination andproline isomerisation). Epigenetics mediates differ-ential gene expression, which may be development-,tissue- or cell-specific, or dysregulated in pathogen-esis. Numerous technologies, old and new, encom-passing high throughput DNA sequencing and chro-matin immunoprecipitation (ChIP), DNA microar-rays (ChIP-chip) are used to profile DNA/histonemodifications. Genome sequencing studies haveidentified epigenome hotspots in DNA from differ-ent cells and tissues. Epigenomics data fromEncyclopedia of DNA Elements (ENCODE) andThe NIH Roadmap Epigenomics projects, hashelped define functional DNA segments that eitherpossess specific chromatin structures or have proteinbinding sites or directly code for proteins or non-

coding RNAs in normal and diseased human tissuesand cells9. The epigenomic signature data from var-ious sources are accessible through IHEC(International Human Epigenome Consortium); andincreasingly, more comprehensive global or locus-specific genomic aberrations are being reported formany diseases including hypertension, asthma andpain. Application of epigenomics can positivelyimpact clinical therapeutics in multifactorial dis-eases and improve early cancer detection, prognosisand prediction of treatment responses (pharmaco-epigenetics).

BiomarkersGenomic, imaging, proteomic and metabolomicstechnologies have unravelled a plethora of novelgenetic/proteins which can serve as biomarkersthat can help predict disease susceptibility, diseasesubtypes and disease progression, as well as pre-dict response to a drug treatment10. If validatedreproducibly in disease, the biomarkers can beincorporated in preclinical and clinical workflows

Drug Discovery World Summer 2016 49

Drug Discovery

Biomarkers, challengesand opportunitiesBiomarkers are functional, reliable and measurable biochemical indicators in healthand disease. In recent years, biomarkers research has played a critical role in phar-maceutical R&D for creating groundbreaking therapies and companion products.Biomarkers use in such discoveries helped in the recognition, validation and thera-peutic intervention in diseases that were considered insurmountable just a fewdecades earlier; development of breast cancer drug Herceptin is such an example.Similarly, their widespread clinical use has enabled, along with high technology, devel-opment of ultra-sensitive detection devices and new instrumentation. In the lastthree decades, research for ever more complex drugs and biotechnology-derivedproducts has led to significant adaptation of ultrahigh-throughput automation, minia-turisation and data analytics with advanced computing. Coupling biomarkersresearch to drug discovery albeit presents a much higher complexity. Technicallychallenging detection modalities, needed to keep pace with emerging biology andchemistries, present immense challenges and contrasting opportunities. As mostbiomarkers work is carried out at the interface of basic biology, clinical and instru-mentation research, transitioning the laboratory-derived information to clinicalapplication adds another dimension to the problem. However, potential exists toleverage expertise from diverse disciplines and adapt techniques and methods toenhance productivity, reduce timelines and achieve compelling results. There is a con-stant need for innovation, with ever increasing cost pressures on pharmaceutical dis-covery in a globally challenging healthcare environment. An integrated approach fromthe outset thus adds a much-needed aspect to strengthening the traditional, lineardrug discovery for better clinical outcomes.

Krishna Kodukula, KK Biopharma

to address toxicity, safety, efficacy and patientselection. The safety and efficacy of investigationdrugs can be further improved by including genet-ic profiling of polymorphisms and drug-metabolising enzymes.

PharmacognosyIt is a study of ethnobiology, taxonomy, sample col-lection, extraction, isolation and defining chemicaland biological properties of naturally occurring,medicinally-active substances obtained from plants,microbes and animals, etc. Various herbal andmixed formulations containing minerals and metalshave been documented historically in folkmedicines. The natural products may be wholeorganisms, parts of a plant, animal, secondarymetabolites or extracts. The rich biodiversity in ter-restrial and marine organisms promises to offernovel cell permeable chemical scaffolds and novelbioactive compounds. Known bioactive natural

products include drugs against cancer, bacterial,protozoan and fungal infection and inflammation.The bioactivity evaluation of natural extracts is bothlabour and time-intensive, requiring several types ofextract preparation, reiterative bioassay guided frac-tionations and isolation of active scaffold (HPLC,Mass spec), target specificity and mechanism ofaction. In case of low active ingredient yields from asource, chromatographic fingerprinting is often per-formed to identify richer sources of the active com-pound from other related sources. The safety ofpublically-consumed botanicals, herbal and dietarysupplements has been addressed by EU hERGscreennetwork (http://hergscreen.univie.ac.at/), a databasefor cardiotoxic risk assessment based on ligand-based human Ether-à-go-go Related Gene (hERG)channel 3D pharmacophore models.

Natural products, the phoenixFrom the turn of the 20th century, natural prod-ucts have been a trusted source of a vast majorityof anti-cancer drugs and anti-infectious agents, andby far the richest source of novel molecular scaf-folds. Some famous examples derived from naturalproducts over the past 30 years include artemisinin(malaria), colchicine (gout), galanthamine (demen-tia) and paclitaxel (cancer). Curcumin is the mostcelebrated example and is currently in 26 clinicaltrials for a variety of indications. The therapeuticeffects of traditional medicine or formulationshave been documented in various civilisations andcontinue to be utilised for treatment of several dis-orders in many countries. Natural product discov-ery has largely been deprioritised due to complexi-ties in large-scale production, either naturalsources or via chemical synthesis, among others.Reverse pharmacology (clinic to laboratory) aimsto expand the drug discovery approaches toinclude lead identification from traditional formu-lations and extracts. Reverse pharmacology initi-ates from documented clinical and experimentalobservations, utilises modern preclinical and clini-cal approaches for active principal identificationand isolation, mechanism of action or target iden-tification as well as formal safety and toxicologyclinical study to identify viable leads.

Biologics revolutionRecombinant protein based therapies (monoclonalantibodies, growth factors, hormones, blood fac-tors, enzymes, vaccines, anticoagulants, fusionproteins) have been developed for various diseases.The recombinant drugs are produced in bacteria,yeast as well as mammalian cells. Biologics thatstimulate immune response against tumours such

50 Drug Discovery World Summer 2016

Drug Discovery

Natural products as drugleads: to be or not to be?Natural products (NPs), evolutionarily optimised by the living organisms for servingdifferent biological functions, and thus, with inherent diversity and complexity, are agood source of drug leads. However, NP-based drug discovery is rather neglected bythe pharmaceutical industry:• Typical high throughput screening platforms are not readily suitable for bioactivitydetection of plant extracts. • NPs are complex structures with multiple stereocentres. Isolation and unambigu-ous structure elucidation are time-consuming, and labour and cost-intensive.• Resupply of NPs is critical for hit-to-lead and preclinical studies. Total synthesis ofNPs with multiple chiral centres is rather challenging.• NPs are not necessarily ‘drug-like’, requiring structural modification for improvedsolubility and bioavailability.• NPs are not directly patentable and need significant pharma investment to modifythe structure and claim that the invention (therapeutic use) is ‘significantly different’from the natural source.The Western reductionist approach follows the ‘one target-one compound’

paradigm, however, such entities are not quite effective for multi-factorial diseases.The concept of polypharmacology and multi-targeted drugs has only recently begunto gain recognition. This paradigm shift has reopened the doors towards holisticmedication, and turned the attention towards the herbal universe. Such ‘natural’drugs have a complex chemical composition in a variety of quantities that allows syn-ergism between the active substances. Much integrated effort encompassingmetabolomics, chemical genomics, proteomics and network pharmacology isrequired for target identification and elucidating the mechanisms of multi-compo-nent/multi-target herbal medicines.

György Dorman, Target-Ex

as oncolytic viruses, antibodies and vaccines, aswell as T-cell mediated therapies hold greatpromise for tumour immunotherapy. Biologics dis-covery helps to diversify a company portfolio,shortens the path to discovery and gains fromrestricted competition from biosimilars and gener-ics. The increased pharma interest in biologics dis-covery is reflected in the 2015 FDA approval ofseveral first-in-class antibodies, proteins and hor-mones11. Though target specific, treatment withbiologics for chronic disorders are also known toresult in adverse events involving immune reac-tions and disorders, infections and tumours. Thelong-term treatment with biologics (eg Humira) isalso far more costly than the small molecule drugs.Unlike the generic small molecule drugs, whichbring down the treatment cost to patients, thebiosimilar drugs are still associated with high pro-duction costs and may not be as efficacious andsafe as the original biologic.

Stem cell therapyThe hype and hope offered by stem cell therapy areundeniable. Stem cells are being used in the treat-ment of blood and immune disorders, skin graftsor corneal damage repair (www.eurostemcell.org).Despite high optimism in the therapeutic potentialof the stem cells, the field faces numerous chal-lenges as developing new stem cell treatmentsrequires in-depth understanding of human cell lin-eages, cell markers, niche-dependent potency andprocesses controlling cell proliferation, differentia-tion and functional specialisation. Investigationaldrugs can be tested in stem cell models for theirability to repair damage in appropriate animalmodels under good laboratory practice, though theprogress has been rather slow. Cellular reprogram-ming efforts by genetic factors to generate inducedpluripotent- and lineage-specific stem cells fromsomatic cell types hold great promise for basicresearch as well as in clinical applications. A com-plementary, alternate approach that has beenequally successful involves the use of smallmolecules to maintain the self-renewal potential ofstem cells and thus target specific epigenetic pro-cesses, signalling pathways and the associated cel-lular processes. This approach is quite attractive,and holds promise for manipulating cell fate to adesired outcome.

Collaborative and open innovationPartnerships and collaborations across pharmaceu-tical companies and academia can help improvethe drug development process by reducing redun-dancy and a more judicious allocation of

resources12. Open innovation where ideas, data,failures, reagents and tools are shared between col-laborating partners can accelerate the early drugdiscovery research. Innovation from global collab-orations will support expansion of the drug discov-ery landscape. Open innovation has led to a boostin the research of neglected and rare diseases.Merck opened up its data and analysis tools toenable complex model building under SAGEbionetworks (http://sagebase.org/). In anotherform of compound collaboration, the Eli Lilly-PD2Initiative (https://pd2.lilly.com) seeks to testmolecules and promising compounds originatingfrom academic research in the Eli Lilly phenotypicdiscovery platform. The profiling data and the sec-ondary assay information are shared with the aca-demic researchers. Again, we reinforce that it istime for the pharmaceutical industry to adopt acommunity-driven ‘Wikipedia’ or ‘Waze’-typeshared-knowledge, openly accessible innovationmodel to harvest data and create a crowd-sourcedpath towards a safer and faster road to the discov-ery and development of life saving medicines.

Approaches for holistic drug targetingComprehensive disease models can be structuredto define the possible role of individual geneticloci, network interactions or post-transcriptionalevents that may contribute to complex diseasedevelopment and progression. With the advent ofpersonalised medicine, theoretically the sequenceinformation from a patient can be compared with-in the large datasets available from genomesequencing, transcriptome analysis, epigenomics,proteomics and metabolomics to create workingdisease models. The addition of phenome-wideassociation studies (PheWAS), which links diseasesto genetic variants, is a good starting point foridentifying new drug-disease pairs. The person-alised model can be used to define a treatment reg-imen based on all available FDA-approved drugs,known targets or pathways. In practice, there arelimitations in establishing a direct and specificcausal relationship between the experimental datasets and clinical disease progression. While there isan exponential increase in computational algo-rithms and deep machine learning protocols formining the databases, there is still not enough datavalidating the bioinformatics datasets with the ani-mal model/clinical data.

Biological systems are complex, circuitous andlabyrinthine. Diseases are even more complex. Adisease may more likely be an eventual phenotypicoutcome of a misguided pathway or network.Developing successful treatment strategies to

Drug Discovery World Summer 2016 51

Drug Discovery

ward-off complex diseases requires a comprehen-sive understanding of systems biology to integrateall available knowledge and fill gaps as needed.Expecting single drugs to cure diseases is thereforerather simplistic and/or unrealistic.

PolypharmacologyPolypharmacology is the property, norm ratherthan exception, of bioactive small molecules, natu-ral products and other chemical species to interactwith multiple targets/pathways in biological sys-tems. More than half of ‘single-target’ drugs wereshown to interact with more than five targets. Thepolypharmacology of molecules has wide implica-tions ranging from predicting harmful off-targeteffects to designing safe and efficacious drug com-binations or repositioning drugs for new indica-tions. Shared functional domains and binding sitesacross target families, binding promiscuity of min-imal scaffolds/fragments of chemical species aresome of the factors underlying polypharmacology.Identification of all possible interactions between achemical compound and the biological targets is achallenging task and predictions usually employintegrated approaches utilising bioinformatics min-ing of ‘omics’ datasets, molecular docking using x-

ray crystal structures or models, ligand-basedquantitative structure activity relationship (QSAR)similarity prediction of two- or three dimensionalfingerprints of small molecules, binding pocketsimilarity of proteins. The predictive power of vir-tual methods has limited utility unless validated byexperimental science in physiologically relevantsystems where for instance, the protein domainsare not inflexible and in binding sites/folds changewith conditions. Polypharmacology-driven drugdiscovery has the potential for selecting moleculesthat have higher efficacy, lower toxicity and lesspotential for synergistic effects compared withmultiple prescriptions administered simultaneous-ly. Polypharmacology that does not involve toxicpromiscuity of drugs is predicted to be beneficialfor the treatment of complex diseases and is ofgreat value in multi-target drug discovery. Thedrugs with adverse side-effects should be depriori-tised early in drug discovery while identifying thoseexhibiting potentiation of therapeutic activity byhitting multiple targets followed by characterisa-tion of pharmacological profiles.

Drug repositioningDrug repositioning or repurposing is one of severalendpoints of polypharmacological concept thatinvolves identification of new indications for FDA-approved drugs. Developing an approved drug fornovel indication is both time and cost-effective asthe safety, toxicity profiles, formulations and phar-macology of marketed drugs are already estab-lished and found satisfactory by the regulatoryagencies including the FDA. Drug repositioning isa financially viable approach that can help resur-rect not only the bioactive compounds that previ-ously failed clinical endpoints but could help reviveexpired patents. Approximately 9,000 approved orregistered drugs have been reported to be availablefor repurposing projects13. Drug repurposing hashistorically been anecdotal, based on clinical obser-vations and chance findings. The most cited exam-ple is the serendipitous repurposing of sildenafil forerectile dysfunction, a selective phosphodiesterase5 inhibitor, originally developed as an anti-hyper-tensive drug. In addition to chance discovery ofnew uses for old drugs, other strategies for drugrepurposing include experimental and in silico-based approaches. In the phenotypic screeningapproach, FDA-approved drugs are tested foractivity in assays for cell viability, migration, cas-pase activity, physiologically-relevant muscle cellcontractions, electrophysiology, etc. The activity ofdrugs is compared across a panel of cell lines rep-resentative of various stages of a disease or across

52 Drug Discovery World Summer 2016

Drug Discovery

Precision medicineThe landmark Precision Medicine Initiative, announced by US President Obama in2015, heralds an unprecedented paradigm shift in how the stakeholders –researchers, physicians, payers and patients – work together towards an individu-alised approach for disease treatment and prevention that takes into account indi-vidual variability in genes, environment and lifestyle for each person. It brings anunprecedented improvement in our understanding of underlying disease mecha-nisms and the detection and treatment options. The cohort programme’s goal is toextend precision medicine to all diseases, and in its current form it is US-centric.Precision medicine is truly translational in scope and integrates all allied biomedicaldisciplines including global clinical trials in an individualised, patient-centric way. Itdoes not mean, however, the discovery and development of drugs or medicaldevices, unique to any given patient. That would be exceedingly cost prohibitive andunrealistic to expect. The power of precision medicine is to bring forth the unique-ness of each patient for a particular disease at the molecular level and integrates theprecise treatment needed from the outset. Biomarkers and diagnostics play a criticalrole. The convergence of panomics and systems biology are integral to link all theinteracting pathways in addressing the heterogeneity of disease etiology and developan individualised, true combination drug therapy, much like the principle behind theherbal treatments. Routine practice of precision medicine in a clinical setting for alldiseases and all patients may or may not be realised in our lifetime, but it is a worthyapproach and a lofty goal for all nations to embrace at least in some form and workcollectively together to eradicate diseases that ravage mankind. This is the true holis-tic approach from ‘bench to bedside’.

cells derived from various cancers. The drugsactive in a selective cell or with pan-panel activityare further optimised for activity in combinationscreens against a set of standard of care drugs usedin clinical settings. The synergistic pair is tested inanimal models before it is advanced for furtherclinical investigation. In silico drug repositioningutilises bioinformatics tools to define interactionsbetween drugs, targets, text and literature miningof public databases. The PheWAS-driven associa-tion between genetic variants and disease can alsoserve as a starting hypothesis for defining newdrug-disease linkages. NIH lists 6,500 rare andneglected diseases (R&N) for which only 250treatments are available (www.ncats.nih.gov/trnd).The drug repositioning approach is a low-hangingfruit that can be exploited effectively to address thelow returns on investment and market-size issuesof neglected and rare diseases. With the increase ininformation on genetics and molecular pathwaysas well as availability of appropriate animal mod-els for rare and neglected diseases, the last decadehas seen an increase in the number of repurposeddrug approvals for R&N disorders. Examplesinclude: mitefosine, an anti-cancer alkylphospho-choline used for topical treatment of breast cancermetastasis repurposed for the treatment ofLeishmaniasis; arbaclofen, for cerebral palsy man-agement, is recommended for GABAergic treat-ment of Fragile X syndrome; losartan, an anti-hypertensive drug repurposed to alleviate thesymptoms in subset of population with Marfansyndrome, a connective tissue disorder whichaffects 1 in 5,000, and difluoromethylornithine(eflornithine, DFMO), an ornithine decarboxylaseinhibitor and anti-cancer drug, is being proposedfor the treatment of facial hirsuitism, neuroblas-toma as well as sleeping sickness. The drug repur-posing programmes are being accelerated througheffective collaborations and partnerships acrossacademia and industry. AstraZeneca, while pre-serving its patent rights, has provided access of itsdeprioritised compounds to researchers at MRC(UK) under the Mechanism of Disease Initiativeprogramme. The NCATS (NIH) in its TherapeuticDiscovery Pilot programme screens selected pro-jects using 58 compounds from eight top pharma-ceutical partnerships in an effort to reveal newmechanisms of action and potential indications.The progress in repositioning is mitigated by com-plex regulatory, legal, pricing and patenting issues.

Combination therapyPrescribing combinations of drugs for disease man-agement is a widely employed clinical practice, and

combinations are known to include two or moreactive ingredients against the same indication or aformulation of multiple distinct active ingredientsagainst distinct targets. The combinations couldtarget crosstalk between pathways that are activat-ed or repressed in disease settings. While clinicalobservations define combination prescriptions,experimental quantitative determination of drugcombination is more complex as the net effects aremathematically grouped under additive, synergisticor antagonistic based on fractional effects at vari-ous concentrations of the two drugs14. The in vitroprofiling is further complicated by translation ofthe effects in pharmacokinetic models where thetwo drugs may potentiate or reduce therapeuticefficacy via modulation of individual ADME char-acteristics. A large number of drug combinationscurated from 140,000 clinical studies have beencompiled into a Drug Combination database(DCDB; http://www.cls.zju.edu.cn/dcdb/). As withother approaches, profiling drug activity and side-effects, network crosstalk and modulation helpdesign effective new drug combinations. A largecombination study called the ComprehensiveUndermining of Survival Paths or (CUSP9) wasrationally designed for the treatment of recurrentglioblastoma (GBM)15. The drug combinationincludes nine FDA approved drugs that block theactivity of 17 distinct targets: artesuanate (PI3kinase, AKT, TLR2, TNF ), disulfiram (ALDH),captopril (ACE, AT1, MMP); celecoxib (carbonicanhydrases, COX), aprecipitant (NK-1 receptors),auranofin (thioreductase, STAT3, cathepsin B),itraconzole (P-gp efflux pump, BCRP, Hedgehog, 5Lipooxygease), ritonavir (AKT, mTOR, cyclin D3,proteasome) and sertraline (TCTP, AKT, mTOR).The side-effects and concentrations of individualdrugs were also carefully evaluated to minimiserisks of side-effects while at the same time targetingmultiple effectors of the disease. The HIV/AIDStreatment with HAART (Highly Active Anti-Retroviral Therapy) effectively combines two-nucleoside reverse transcriptase inhibitors lamivu-dine and zidovudine in presence of a proteaseinhibitor16. This combination was based on clini-cal data showing that lamivudine treatment alonetriggered resistance due to Met 184 Val (M184V)mutation in viral reverse transcriptase (RT) but theM184V mutant was still sensitive to zidovudineand a combination of the two RT inhibitors effec-tively suppressed the catalytic activity required forHIV replication. While several combination treat-ments are FDA approved, combination therapiesadd to the healthcare costs, require strict adherenceto regimens and the possibility of developing new

Drug Discovery World Summer 2016 53

Drug Discovery

adverse side-effects with time. Regulatory compli-ance also complicates combination synergistic ther-apies as well as drug repurposing especially if twocompeting pharmaceutical sources are involved.

Multi-target drug discoveryThere has been a recent shift in the drug-discoveryparadigm from single target-single drug to a multi-target drug discovery approach (MTDD) for theidentification of single compounds with activityagainst two or more targets. The MTDD approachhas several advantages over combination drugtherapies: (A) the pharmacokinetics and safety pro-files for single multi-target drugs are easier to eval-uate than predicting potential long term adverseeffects developing with combination drugs, and (B)a multi-target drug is guaranteed to interact withits targets in the same tissue/cell compartment and,unlike the combination drugs, will not haveadverse drug-drug interactions. Both experimentalscreening as well as virtual in silico approachesdeliberately design the screens to identify com-pounds that are active against multiple targets ofinterest. Several rational-based designs, computa-tional-based docking and virtual screeningapproaches are available for identifying drugs withmultiple functions. Several recent examples attestto the attractiveness of this approach, including: (i)successful linking of the ABL kinase inhibitorydomain of imatinib with the hydroxamic acid/ben-zamide (zinc binding) for anti HDAC activity todesign a fusion molecule inhibiting both ABL andHDACs17, and (ii) identification of a dualkinase/bromodomain inhibitor through virtualmining of >six million compounds in the E-molecules, wherein, of the 908 predicted EGFRinhibitors, 108 also docked to BRD4 co-crystalstructures, eight experimentally inhibited theBRD4 bromodomain-activity and one that inhibit-ed both BRD4 binding as well as EGFR kinaseactivity18.

Holistic insightsScience drives technology. Invention leads to inno-vation. Both scenarios are inherently and funda-mentally intertwined. For the betterment of human-ity, it is imperative upon us, the guardians, to seethat science-driven inventions ultimately lead totechnology-based innovations. The era of moderndrug discovery has gone through revolutionaryadvances, as outlined throughout this discussion,and it is time to coalesce around a broadly definedholistic approach, and to enthusiastically adopt acommunity-driven ‘Wikipedia’ or Waze’-type openinnovation model to create a crowd-sourced path

towards a safer and faster road to the discovery anddevelopment of life-saving medicines.

An overall analysis of drugs identified from tar-get-based drug discovery has revealed limitationsin drug innovation and proteome targeting.Despite high R&D expenditure, there is a lack ofsignificant improvement in therapeutic outcomesand overall patient survival. One of the main rea-sons for limited therapeutic efficacy of currenttherapies is partial understanding of the diseasepathophysiology and overall deficiency in devel-oping therapeutics targeting overlapping dysregu-lated pathways. Complex heterogeneous poly-genic diseases require more genetics-guided, fine-ly-tailored treatment regimens to maximise effec-tiveness. The panomics databases serve as valu-able tools for extracting diverse underlying diseasemechanisms, identifying genetic loci in disease andultimately revealing linkages between drugs andtargets that can serve as starting points for design-ing holistic approaches. The evolution of a highlyintegrated approach with minimal false positivesrequires quality control of data being deposited inpublic databases, optimisation of standard operat-ing protocols for data-mining, integration withpatient electronic medical records, supportive reg-ulatory protocols and collaborative efforts ofpatient groups, academia and industry in sharingnegative/positive datasets. With an increase in thenumber of studies identifying new and rare vari-ants in a wide spectrum of diseases, the process ofinquiry and data extraction is bound to evolveinto a highly optimised science with increaseddeliverables. The selection of the right combina-tion of targets for a disease of interest is critical inmulti-target drug discovery and requires goodunderstanding of target-disease associations, path-way-target-drug-disease relationships and adverseevents profiling. Systematic repurposing ofapproved drugs has the potential to bring a trans-formational therapy to regional breakouts of dis-eases such as Zika and Ebola in quick order.Finally, drug discovery will be even more holisticif it addresses the global health crisis, enablingpatient access to medicines and patient care forthe treatment and management of chronic, infec-tious, rare and neglected diseases. With the globalresources reforming health education, drug pro-duction costs, intellectual/regulatory systems,basic science and innovation and a renewed appre-ciation of traditional medicines as a rich source ofnovel scaffolds, we must embrace and empower adrug discovery paradigm that will truly be holis-tic19. The time is now for the drug discovery com-munity to focus efforts towards the discovery and

References1 Rask-Andersen, M, Almén,MS, Schiöth, HB. 2011. Trendsin the exploitation of noveldrug targets. Nat Rev DrugDiscov 10:579-90.2 Ezkurdia, I, Juan, D,Rodriguez, JM, Frankish, A,Diekhans, M et al. 2014.Multiple evidence strandssuggest that there may be asfew as 19 000 human protein-coding genes. HumanMolecular Genetics 23:5866-78.3 Hopkins, AL, Groom, CR.2002. The druggable genome.Nat Rev Drug Discov 1:727-30.4 Disney, MD, Winkelsas, AM,Velagapudi, SP, Southern, M,Fallahi, M, Childs-Disney, JL.2016. Inforna 2.0: A Platformfor the Sequence-BasedDesign of Small MoleculesTargeting Structured RNAs.ACS chemical biology.5 Childs-Disney, JL, Disney,MD. 2016. Approaches toValidate and Manipulate RNATargets with Small Moleculesin Cells. Annual review ofpharmacology and toxicology56:123-40.6 Anuradha, R, Byron, T, Peter,RM, Ashleigh, P, Rathnam, C.2009. Hit-to-Probe-to-LeadOptimization Strategies. InHandbook of Drug Screening,Second Edition:21-55: CRCPress. Number of 21-55 pp.7 Bunnage, ME, Gilbert, AM,Jones, LH, Hett, EC. 2015.Know your target, know yourmolecule. Nat Chem Biol11:368-72.8Waring, MJ, Arrowsmith, J,Leach, AR, Leeson, PD,Mandrell, S et al. 2015. Ananalysis of the attrition of drugcandidates from four majorpharmaceutical companies. NatRev Drug Discov 14:475-86.9 Mullard, A. 2015. TheRoadmap Epigenomics Projectopens new drug developmentavenues. Nat Rev Drug Discov14:223-5.10 Anderson, DC, Kodukula,K. 2014. Biomarkers inpharmacology and drugdiscovery. Biochemicalpharmacology 87:172-88.

Continued on page 55

54 Drug Discovery World Summer 2016

Drug Discovery

development of multi-target-network polyphar-macology drugs exalting symphonic or concertperformance with occasional soloists to reignitepharmaceutical innovation.

AcknowledgementsWe thank Drs György Dorman, Target-Ex, andKrishna Kodukula, KK Biopharma, for sharing theirvaluable insight on the relevance of ‘natural prod-ucts’ and ‘biomarkers’, respectively, in guiding drugdiscovery efforts. We are also immensely indebted tothe authors of our forthcoming book InnovativeApproaches in Drug Discovery: Ethnopharmaco-logy, Systems Biology and Holistic Targeting, forshaping our ideas; the viewpoints expressed hereinare our own, however. DDW

Dr Anuradha Roy is the Director, HighThroughput Screening Laboratory at theUniversity of Kansas, Lawrence, Kansas (USA).She has more than 20 years of academic andbiotech experience in executing HTS campaigns,and has managed drug discovery projects from tar-get identification, validation and hit to lead devel-opment. As director of the HTS core, her responsi-bilities include helping academic investigatorsacross state universities through the process ofdeveloping their target ideas into screen-readyassays for probe identification. Anuradha led vari-ous aspects of drug discovery at PTC Therapeuticsafter her postdoctoral work at Cleveland ClinicFoundation. She holds a doctorate inBiochemistry/Molecular Biology from JawaharlalNehru University (India).

Professor Bhushan Patwardhan, InterdisciplinarySchool of Health Sciences, Savitribai Phule PuneUniversity (India), is an internationally recognisedexpert on ethnopharmacology and integrativehealth. He brings a unique blend of industry-academia executive culture in advancing evidence-based Ayurveda. Professor Patwardhan is a Fellowof National Academy of Medical Sciences (India),Founder and Editor-in-Chief of the Journal ofAyurveda and Integrative Medicine, and serves onthe editorial boards of several journals. Bhushan isa recipient of the Parkhe Award for IndustrialExcellence, Dewang Mehta Award for educationalexcellence, and Sir Ram Nath Chopra Oration. Hereceived his PhD in Biochemistry from HaffkineInstitute (Mumbai, India).

Dr Rathnam Chaguturu is the Founder & CEO ofiDDPartners (Princeton Junction, NJ, USA), a

non-profit think-tank focused on pharmaceuticalinnovation, and most recently, Deputy Site Head,Center for Advanced Drug Research, SRIInternational. He has more than 35 years of expe-rience in academia and industry, managing newlead discovery projects and forging collaborativepartnerships with academia, disease foundations,non-profits and government agencies. He is theFounding President of the International ChemicalBiology Society, a Founding Member of the Societyfor Biomolecular Sciences, and Editor-in-Chief ofthe journal Combinatorial Chemistry and HighThroughput Screening. Rathnam passionatelyadvocates the need for innovation andentrepreneurship and the virtues of collaborativepartnerships in addressing the pharmaceuticalinnovation crisis, and aggressively warns the threatof scientific misconduct in biomedical sciences. Hereceived his PhD with an award-winning thesisfrom Sri Venkateswara University, Tirupati, India.

Continued from page 54

11 Morrison, C. 2016. Freshfrom the biotechpipeline[mdash]2015. NatBiotech 34:129-32.12 Chaguturu, R. 2014 (Ed).Collaborative Innovation inDrug Discovery: Strategies forPublic and PrivatePartnerships. Wiley.13 Huang, R, Southall, N,Wang, Y, Yasgar, A, Shinn, P et al.2011. The NCGCpharmaceutical collection: acomprehensive resource ofclinically approved drugsenabling repurposing andchemical genomics. Sci TranslMed 3:80ps16.14 Chou, TC. 2006. Theoreticalbasis, experimental design, andcomputerized simulation ofsynergism and antagonism indrug combination studies.Pharmacological reviews58:621-81.15 Kast, RE, Karpel-Massler, G,Halatsch, M-E. 2014. CUSP9*treatment protocol forrecurrent glioblastoma:aprepitant, artesunate,auranofin, captopril, celecoxib,disulfiram, itraconazole,ritonavir, sertraline augmentingcontinuous low dosetemozolomide. Oncotarget5:8052-82.16 Staszewski, S. 1995.Zidovudine and lamivudine:results of phase III studies.Journal of acquired immunedeficiency syndromes andhuman retrovirology : officialpublication of the InternationalRetrovirology Association 10Suppl 1:S57.17 Ganesan, A. 2016.Multitarget Drugs: anEpigenetic Epiphany.ChemMedChem.18 Allen, BK, Mehta, S, Ember,SW, Schonbrunn, E, Ayad, N,Schurer, SC. 2015. Large-ScaleComputational ScreeningIdentifies First in ClassMultitarget Inhibitor of EGFRKinase and BRD4. Scientificreports 5:16924.19 Patwardhan, B, Chaguturu,R. 2016 (Ed). InnovativeApproaches in DrugDiscovery:Ethnopharmacology, SystemsBiology and Holistic Targeting.Elsevier Science (in press).

Drug Discovery World Summer 2016 55

Drug Discovery