Ed.pdf

Volume 87 Number 7 10 October 2004
EDITORIAL
Drug Discovery: Myth and Reality
An editor penning an editorial column has a wonderful cline and erythromycin among them. Once ‘leads’ have advantage; the hurdles of referees and sub-editors do not been identified, the task of moving forward requires care- stand as obstacles before the article is published. For a ful pharmacological and toxicological evaluation, before regular columnist, the most difficult task is to find a topic the clinical trials can begin. But, as the drug industry has to write about. Then comes the almost equally formidable matured and as the regulatory controls have become more task of gathering at least a smattering of facts. Finally, stringent, the process of discovering promising new mole- there is the threatening spectre of the deadline for the cular entities has become more difficult, more time con- printers. Committed, as I am, to writing a fortnightly col- suming and significantly more expensive. umn for this journal, I am often rescued in the difficult Over the last decade or so, the advances in genome se- task of finding a subject, by papers sent to me by friends, quencing, protein analysis and computational biology have colleagues and, most importantly, readers whom I have been spectacular, giving rise to the expectation that the never met. A few days ago a friend sent me a marvel- new technologies will hasten the process of drug discov- lously provocative editorial by Gerry Higgs in Drug Dis- ery. ‘High throughput’ is a magical prefix that promises covery Today (2004, 9, 727). The polemic entitled ‘Mole-
to make chemical analysis and pharmacological screen- cular Genetics: the Emperor’s Clothes of Drug Discovery’ ing a matter for robots; raising hopes that automation will was a frontal attack on the prevailing view ‘that knowl- indeed aid innovation in the search for new pharmaceuticals. edge of the human genome sequence when harnessed to Combinatorial synthesis, a favourite with pharmaceutical automated high-throughput technologies will result in R&D managers, coupled with automated high throughput drugs being made more easily’. Higgs was challenging a screening has often been touted as the driving technology view expressed in the same journal by Peter Goodfellow, of future drug discovery. With genes and pharmaceutical an academic geneticist-turned-manager of drug research targets appearing in profusion, it seemed only a short step at GlaxoSmithKline (Drug Discovery Today, 2003, 8,
to screening tens of thousands of ligands to select new 1017). Goodfellow presents an extremely optimistic view of ‘lead’ molecules. These ‘leads’ would then be optimized the impact of high throughput screens (HTS) and high by computational scientists using ‘docking’ techniques, throughput (HT) chemistry, on the traditionally slow pro- which would reveal the best fit of low molecular weight cess of drug discovery. The drug discovery cycle is the ligands to macromolecular receptors. This ‘in silico’ (an process where a promising laboratory finding is conver- infelicitous phrase) approach would presumably save medi- ted into a marketable product; an uncertain process that cinal chemists years of toil; quickly directing their ener- resembles a game of snakes and ladders, potential prod- gies towards producing optimal molecules. The new ‘para- ucts often failing in the final stages of clinical trials. The digm of drug discovery’ has been relentlessly propagated timescales involved are immensely long, estimates of at innumerable conferences and in articles strewn across 10–15 years being common. Indeed, the gulf between the the sprawling literature of chemistry, biology, medicine laboratory and the clinic can be formidable. The main- and computer science. The biological revolution has stays of the conventional drug industry have been medi- spawned many new disciplines, genomics, proteomics, cinal chemistry and microbiology. While the chemists metabolomics, systems biology and bioinformatics. The have toiled ceaselessly producing molecules, structures confluence of these areas is expected to accelerate the patiently modified in incremental fashion, the microbio- drive towards new drugs. Enthusiasts for the new tech- logists have provided both antimicrobial screens and an nologies have looked into the future and announced that inexhaustible supply of culture filtrates, which provide a ‘pharmacogenomics’, a term that promises customized drugs treasure trove of new molecules. The history of antibiotic for individuals, is just around the corner. The rhetoric of research is replete with wonderful molecules that emerged drug discovery research has been raised to fever pitch. It from microbial cultures, penicillin, streptomycin, tetracy- is in this context that Higgs’ assault on molecular gene- CURRENT SCIENCE, VOL. 87, NO. 7, 10 OCTOBER 2004 tics and its role in drug discovery makes interesting read- and 1960s were spearheaded by the work of Gertrude ing. Indeed, over the last couple of years even the US Elion and George Hitchings at Burroughs Welcome and Food and Drug Administration (US FDA) notes a fall in James Black at King’s College, London, who between the number of new molecules entering the regulatory them were responsible for developing a large number of pipeline. This decline appears at odds with the optimistic drugs, many of which are widely used today. To Elion projections of the proponents of the new technologies. and Hitchings goes the credit for introducing azathio- With the timescales for discovery lengthening and the prine, the first immunosuppressive agent, allopurinol for costs of innovation rising (one estimate that I came across gout, pyrimethamine for malaria and trimethoprim for was as high as $ 800 million as the cost for introducing a bacterial infections. Later, Elion went on to introduce the new drug), the game of mergers and acquisitions has be- antiviral, acyclovir using the principle of exploiting dif- come an integral part of the pharmaceutical industry. ferences in nucleic acid metabolism between target and In his essay, Higgs questions the assumption ‘that study- host; an approach that would later be followed in deve- ing the genome at a molecular level will reveal new targets loping the AIDS drug, azidothymidine (AZT). Black used for drug intervention and that molecular biology will pro- an incisive approach, relying on a deep understanding of vide the relevant tools for indentifying new drugs. The physiology, to discover propranolol, the ‘beta-blocker’ surprisingly poor success rate of this approach suggests used for heart disease and cimetidine used for treating that these assumptions should be questioned’ (Drug Dis- gastric ulcers. Black, Hitchings and Elion received the covery Today, 2004, 9, 727). I particularly liked his emphatic
1988 Nobel Prize for Physiology or Medicine; one of the dismissal of ‘the assumption that the best new medicines rare instances where drug discovery has been honoured in will be the most potent and selective against a particular target’. Higgs argues that this view is based on a ‘misun- In reading Goodfellow’s optimistic view of the future derstanding of how drugs work’. He supports his thesis of drug discovery and Higgs’ characterization of the new by pointing out that ‘some of the most successful medi- technologies as a case of the ‘Emperor’s new clothes’, I cines are remarkably weak or non-selective. For example, was struck by the evident tensions between the classical aspirin, ibuprofen and cimetidine are blockbuster drugs disciplines of pharmacology, physiology and medicinal with potencies in the micromolar range’. The reference chemistry on one hand and the new offshoots of molecu- here is to the hunt for ‘nanomolar’ receptor ligands, mole- lar and cellular biology on the other. Higgs notes that the cules that seek their targets with remarkably high affini- ‘large pharmaceutical companies are increasingly depen- ties. Higgs drives his point home by pointing out that dent on old products that have a rapidly expiring patent ‘drugs such as fluticasone and budenoside, which are widely life. Alarmingly, the direction of discovery research is used to treat asthma, belong to a class of anti-inflamma- often governed by scientists who are unfamiliar with the tory steroids, that are so non-selective that nobody is origin of these drugs. The industry needs to rediscover quite sure how they work. In the high throughput screen the discipline of deductive pharmacology. Above all, it regimes of today, these drugs would not even be rated as needs to promote clinical pharmacology’. Ironically, al- “hits” ’. Higgs goes on to question the validity of geneti- though ‘drug discovery’ is a widely trumpeted activity, cally engineered cells as disease models and argues that pharmacology is a low-profile discipline with none of the gene ‘knock outs’ are ‘highly vulnerable to the creation glamour associated with modern biology or medicine. of misleading artefacts’. In stating ‘that the reductionist Curiously, in his Nobel lecture, James Black advanced a structure-based approach of molecular biology is a poor view on a discipline that he termed as analytical pharma- starting place for drug discovery’, Higgs clearly argues cology: ‘What we are allowed to see of a new molecule’s for a greater emphasis on approaches that move forward properties is totally dependent on the techniques of bio- from clinical investigations. To bolster his case, Higgs assay that we use. The prismatic qualities of an assay cites the case of infliximab, an anti-inflammatory cyto- distort our own views in obscure ways and degrees. Our kine used for treatment of rheumatoid arthritis, developed only defence lies in restless improvement in technique over the period 1989–1998. Here the initial leads in semi- and experimental design in the hope that collimation of nal work by Ravinder Maini and Marc Feldman at the several techniques will improve the reliability of our vi- Kennedy Institute, Imperial College, London came from sion. We would make changes self-consciously today, but work with inflamed tissues from patients. Maini and Feldman received the 2003 Lasker Award for this achieve- Maybe some of the classical disciplines that have con- tributed to drug discovery in the past, may well do so again The history of drug research over a period of a century, in the future. In drug discovery research today, it is be- since Paul Ehrlich introduced the concept of chemothera- coming increasingly hard to distinguish between myth peutic agents, is a marvellous record of accomplishment. The great victories of the antibiotic era were often seren- dipitous; penicillin, streptomycin and tetracyclines fol- lowed in quick succession. The successes of the 1950s CURRENT SCIENCE, VOL. 87, NO. 7, 10 OCTOBER 2004

Source: http://mbu.iisc.ernet.in/~pbgrp/10oct04.pdf