References and Notes 24. D. Konz, A. Klens, K. Schorgendorfer, M. A. Marahiel, 48. Y. Sun et al., Chem. Biol. 10, 431 (2003).
1. D. E. Cane, C. T. Walsh, C. Khosla, Science 282, 63 Chem. Biol. 4, 927 (1997). 49. H. J. Kwon et al., Science 297, 1327 (2002).
(1998). 25. N. J. Hillson, C. T. Walsh, Biochemistry 42, 766 (2003). 50. A. M. van Wageningen et al., Chem. Biol. 5, 155 (1998).
2. S. Omura et al., Proc. Natl. Acad. Sci. U.S.A. 98, 26. S. A. Sieber, M. A. Marahiel, J. Bacteriol. 185, 7036 51. J. Recktenwald et al., Microbiology 148, 1105 (2002).
12215 (2001). (2003). 52. J. Pootoolal et al., Proc. Natl. Acad. Sci. U.S.A. 99,
3. H. B. Bode, B. Bethe, R. Hofs, A. Zeeck, ChemBioChem 27. T. Brautaset et al., Chem. Biol. 7, 395 (2000). 8962 (2002).
3, 619 (2002). 28. C. N. Boddy, T. L. Schneider, K. Hotta, C. T. Walsh, C. 53. M. Sosio, S. Stinchi, F. Beltrametti, A. Lazzarini, S.
4. Y. Tang, T. S. Lee, C. Khosla, PLoS Biol., 2, 227 (2004). Khosla, J. Am. Chem. Soc. 125, 3428 (2003). Donadio, Chem. Biol. 10, 541 (2003).
5. H. D. Mootz, D. Schwarzer, M. A. Marahiel, ChemBio- 29. J. W. Trauger, R. M. Kohli, H. D. Mootz, M. A. Marahiel, 54. M. Sosio et al., Microbiology 150, 95 (2004).
Chem 3, 490 (2002). C. T. Walsh, Nature 407, 215 (2000). 55. D. Bischoff et al., Angew. Chem. Int. Ed. Engl. 40,
6. B. Shen, Curr. Opin. Chem. Biol. 7, 285 (2003). 30. J. W. Trauger, R. M. Kohli, C. T. Walsh, Biochemistry 4688 (2001).
7. C. T. Walsh et al., Curr. Opin. Chem. Biol. 5, 525 40, 7092 (2001). 56. A. Holtzel et al., J. Antibiot. 55, 571 (2002).
(2001). 31. R. M. Kohli, J. W. Trauger, D. Scwarzer, M. A. Marahiel, 57. Z. A. Hughes-Thomas, C. B. W. Stark, I. U. Bohm, J.
8. T. A. Keating et al., ChemBioChem 2, 99 (2001). C. T. Walsh, Biochemistry 40, 7099 (2001). Staunton, P. F. Leadlay, Angew. Chem. Int. Ed. Engl.
9. R. M. Kohli, C. T. Walsh, Chem. Commun. 2003, 293, 32. R. M. Kohli, C. T. Walsh, M. D. Burkart, Nature 418, 42, 4475 (2003).
(2003). 658 (2002). 58. W. Liu et al., Proc. Natl. Acad. Sci. U.S.A. 100, 11959
10. R. Pieper, G. Luo, D. E. Cane, C. Khosla, Nature 378, 33. R. M. Kohli, M. D. Burke, J. Tao, C. T. Walsh, J. Am. (2003).
263 (1995). Chem. Soc. 125, 7160 (2003). 59. J. L. Hansen et al., Mol. Cell 10, 117 (2002).
11. H. M. Patel, C. T. Walsh, Biochemistry 40, 9023 (2001). 34. S. A. Sieber, J. Tao, C. T. Walsh, M. A. Marahiel, 60. D. Hoffmeister et al., Chem. Biol. 7, 821 (2000).
12. D. A. Miller, L. Luo, N. Hillson, T. A. Keating, C. T. Angew. Chem. Int. Ed. Engl. 43, 493 (2004). 61. C. Mendez, J. A. Salas, Trends Biotechnol. 19, 449
Walsh, Chem. Biol. 9, 333 (2002). 35. H. D. Mootz et al., J. Am. Chem. Soc. 124, 10980 (2001).
13. B. A. Pfeifer, S. J. Admiraal, H. Gramajo, D. E. Cane, C. (2002). 62. L. L. Remsing et al., J. Am. Chem. Soc. 124, 1606
Khosla, Science 291, 1790 (2001). 36. Z. Hojati et al., Chem. Biol. 9, 1175 (2002). (2002).
14. B. A. Pﬁeifer, C. C. Wang, C. T. Walsh, C. Khosla, Appl. 37. S. Weist et al., Angew. Chem. Int. Ed. Engl. 41, 3383 63. ¨
D. Hoffmeister, G. Drager, K. Ichinose, J. Rohr, A.
Environ. Microbiol. 69, 6698 (2003). (2002). Bechthold, J. Am. Chem. Soc. 125, 4678 (2003).
15. C. Khosla, P. B. Harbury, Nature 409, 247 (2001). 38. B. Wilkinson et al., Chem. Biol. 7, 111 (2000). 64. H. C. Losey et al., Biochemistry 40, 4745 (2001).
16. N. Wu, S. Y. Tsuji, D. E. Cane, C. Khosla, J. Am. Chem. 39. Y. J. Yoon et al., Chem. Biol. 9, 203 (2002). 65. H. C. Losey et al., Chem. Biol. 9, 1305 (2002).
Soc. 123, 6465 (2001). 40. Y.-Q. Cheng, G.-L. Tang, B. Shen, Proc. Natl. Acad. 66. C. Walsh, C. L. Freel Meyers, H. C. Losey, J. Med.
17. S. E. O’Connor, C. T. Walsh, F. Liu, Angew. Chem. Int. Sci. U.S.A. 100, 3149 (2003). Chem. 46, 3425 (2003).
Ed. Engl. 42, 3917 (2003). 41. D. Panda, K. DeLuca, D. Williams, M. A. Jordan, L. 67. A. Li, J. Piel, Chem. Biol. 9, 1017 (2002).
18. T. A. Keating, C. G. Marshall, C. T. Walsh, Biochemistry Wilson, Proc. Natl. Acad. Sci. U.S.A. 95, 9313 (1998). 68. H. Chen et al., Biochemistry 40, 11651 (2001).
39, 15522 (2000). 42. P. R. August et al., Chem. Biol. 5, 69 (1998). 69. F. Pojer, S.-M. Li, L. Heide, Microbiology 148, 3901
19. T. L. Schneider, B. Shen, C. T. Walsh, Biochemistry 42, 43. T.-W. Yu et al., Proc. Natl. Acad. Sci. U.S.A. 99, 7968 (2002).
9722 (2003). (2002). 70. H. Chen, C. T. Walsh, Chem. Biol. 8, 301 (2001).
20. K. Shin-ya et al., J. Am. Chem. Soc. 123, 1262 (2001). 44. A. Rascher et al., FEMS Microbiol. Lett. 218, 223 71. M. G. Thomas, M. D. Burkart, C. T. Walsh, Chem. Biol.
21. R. Jansen, H. Irschik, H. Reichenbach, V. Wray, G. (2003). 9, 171 (2002).
Hoﬂe, Liebigs Ann. Chem. 759 (1994).
¨ 45. B. K. Hubbard, C. T. Walsh, Angew. Chem. Int. Ed. 72. C. R. Hutchinson, Proc. Natl. Acad. Sci. U.S.A. 100,
22. R. S. Roy, A. M. Gehring, J. C. Milne, P. J. Belshaw, C. T. Engl. 42, 730 (2003). 3010 (2003).
Walsh, Nat. Prod. Rep. 16, 249 (1999). 46. K. C. Nicolaou, C. N. Boddy, S. Brase, N. Winssinger, 73. Supported by NIH grants GM 20011, GM 49338, AI
23. L. Du, C. Sanchez, M. Chen, D. J. Edwards, B. Shen, Angew. Chem. Int. Ed. Engl. 38, 2096 (1999). 42738. Special thanks to D. Vosburg for preparation
Chem. Biol. 7, 623 (2000). 47. M. Oliynyk et al., Mol. Microbiol. 49, 1179 (2003). of artwork.
Organic Chemistry in Drug Discovery
Malcolm MacCoss1* and Thomas A. Baillie2
The role played by organic chemistry in the pharmaceutical industry continues to be In the recent past, the usual flow of informa-
one of the main drivers in the drug discovery process. However, the precise nature tion that was generated regarding any new com-
of that role is undergoing a visible change, not only because of the new synthetic pound prepared in the laboratory of a drug dis-
methods and technologies now available to the synthetic and medicinal chemist, but covery company followed a paradigm similar to
also in several key areas, particularly in drug metabolism and chemical toxicology, as that shown in Fig. 1. This scheme was driven by
chemists deal with the ever more rapid turnaround of testing data that inﬂuences the need to get the initial information on a com-
their day-to-day decisions. pound first, before deciding whether its proper-
ties met appropriate criteria before moving onto
Numerous changes are now occurring in the scientific advances in synthetic techniques the next evaluation step. Such a linear sequence
pharmaceutical industry, not just in the way and new technologies for rational drug de- of events, although sparing of the number of
that the industry is perceived, but also in the sign, combinatorial chemistry, automated compounds taken down the pathway, often
rapid expansion of biomedical and scientific synthesis, and compound purification and meant that a considerable amount of time passed
knowledge, which affects the way science is identification. In addition, with the advent of (several weeks) before it was known whether a
practiced in the industry. The recent changes high-throughput screening (HTS), we are particular change in a molecule was in fact a use-
in the way that synthetic chemistry is prac- now faced with many targets being screened ful transformation, or whether it was a potency-
ticed in this environment center around new and many hits being evaluated. However, enhancing change in the primary in vitro assay
success in this arena still requires skilled but was perhaps a liability in a downstream
medicinal chemists making the correct choic- evaluation. Thus, the delay in getting appropriate
Department of Basic Chemistry, Merck Research Lab- es, often with insight gleaned from interac- feedback to the synthetic chemist meant that
oratories, 126 East Lincoln Avenue, Rahway, NJ
07065, USA. 2Department of Drug Metabolism, Merck
tions with computational chemists and struc- decisions about which molecules to prepare in
Research Laboratories, Sumneytown Pike, West Point, tural biologists, about which “hits” (1) are the next round of synthesis were not guided by
PA 19486, USA. likely to play out as true “lead” (1) structures input from downstream data. With the advent of
*To whom correspondence should be addressed. E- that will meet the plethora of hurdles that any faster synthetic technologies, including advances
mail: email@example.com drug candidate must surmount. in nuclear magnetic resonance (NMR) methods,
1810 19 MARCH 2004 VOL 303 SCIENCE www.sciencemag.org
rapid separations, and automated syntheses, the that exhibits genetic polymorphism (poten- teristics are taken into account in arriving at this
cycle time for synthetic manipulation of analogs tially leading to large individual variability in key decision, which requires considerable ex-
has decreased dramatically. In addition, in the drug pharmacokinetics and clinical response perience and sound judgment on the part of
same time frame, advances have been made where metabolism is the major route of clear- the group of senior scientists collectively
in the ability to assay compounds, both in ance). Moreover, if the therapeutic target re- charged with this responsibility.
vitro and in vivo, at a much greater speed sides within the central nervous system This new paradigm has led to a different
than was previously possible, and so the (CNS), it becomes important to determine type of decision-making by chemistry group
current paradigm has shifted toward that whether the structural series of interest serve leaders. As noted above, the results from a
shown in Fig. 2, where it is now feasible to as substrates for the efflux transporter P- preliminary evaluation of the pharmacologi-
generate a tremendous amount of relevant glycoprotein and thereby are denied access to cal, pharmacokinetic, metabolic, and toxico-
data on a newly synthesized compound brain tissue in vivo. By obtaining such infor- logical profile of a series of molecules usu-
within 1 week of its initial preparation. This mation in the discovery phase, potentially ally will expose any serious deficits that
process allows for a much better-informed serious liabilities in a given structural se- would hinder or even preclude successful
set of decisions, as one considers the next ries become evident at the outset, and in- development of a drug candidate. As a result,
round of molecules that need to be prepared. formed decisions can be made accordingly such “flawed” compounds, or sometimes en-
It should be stressed that an awareness of to redirect chemistry efforts. tire structural series, are dropped from further
the potential downstream obstacles to suc- The chemist also needs to be conversant consideration, and development resources are
cessful drug development is an important with issues of toxicology, given conserved as a result.
consideration in the chemist’s decision- that the primary cause of failure In the majority of cases, how-
making process. Based on a rationalization of of drug candidates in early de- ever, there is no single factor that
experimental and computational approaches, velopment continues to be pre- would lead to the exclusion of a
Lipinski et al. presented the “rule of five” in clinical toxicity. Although the molecule from further consider-
the mid 1990s, which is an excellent working potential for genotoxicity can be ation, and the decision to advance
hypothesis for predicting good druglike prop- assessed directly through a num- a given compound needs to be
erties in new compounds (2, 3). Thus, close ber of in vitro assays, the same based on a critical assessment of
attention needs to be paid to molecular does not hold true for end-organ the relative attributes and poten-
weights, as well as to the physicochemical toxicities (such as drug-induced tial liabilities of that molecule.
properties of lead molecules, such as lipo- liver damage) or immune- Admittedly, the availability of
philicity (logP) and aqueous solubility (which mediated toxicities (idiosyncrat- more, rather than less, informa-
will affect oral bioavailability and the feasi- ic reactions) (4). However, tion on each drug candidate can
bility of generating a parenteral formulation), based on the premise that some introduce an element of ambigu-
together with animal pharmacokinetics, (but certainly not all) drug- ity into the chemist’s decision-
which can be extrapolated with caution to related adverse events appear to making process. For instance, if a
predict corresponding behavior in humans. be mediated by a chemically re- structural change leads to in-
The latter is particularly important in provid- active, electrophilic metabolite creased potency in the lead bio-
ing some assurance that the candidate drug or metabolites, as opposed to the chemical assay, but the com-
molecule will exhibit linear pharmacokinetics parent drug itself, it may be ar- pound is less orally bioavailable
in humans, with appropriate dose size and gued that the generation of such in a rodent, has more activity on a
elimination characteristics for the intended electrophiles is an undesirable biochemical counterscreen, and is
route and frequency of drug administration. feature of any drug candidate. less potent in a toxicity assay,
Preliminary absorption, distribution, me- By means of appropriate in vitro then the decision to continue ex-
tabolism, and excretion (ADME) studies of “trapping” experiments and as- ploring that avenue is less clear.
lead compounds in animal species also pro- sessments of covalent binding of Of course, all knowledge is
vide information on routes of clearance (such lead drug candidates to protein, useful and so the ongoing de-
as renal, biliary, or metabolic), which is help- both in vitro and in vivo, it is tailed compilation of structure-
ful in guiding the selection of compounds that usually possible for the medici- activity relationships (SARs)
exhibit a balance between elimination path- nal chemist, working closely across many assays is already
ways and thus would not be unduly depen- with colleagues in drug metabo- helping our understanding of
dent on a single organ for excretion. At the lism, to identify routes of meta- what types of functionality are
same time, in vitro data are provided on the bolic activation and, through responsible for binding to various
interaction of drug candidates with human appropriate structural modifica- CYPs, cardiac ion channels,
cytochrome P-450 (CYP) enzymes, so that tion, to minimize this potential Fig. 1. Linear path to drug transporters, nuclear receptors re-
CYP inhibitors and inducers are identified at liability (5). Moreover, before candidate, with numerous sponsible for CYP induction, etc.
an early stage, and due consideration is given selection of a lead compound for feedback loops designed In fact, the medicinal chemist has
to provide information for
to the attendant risk that such candidates may development, information also target selection in the always had to make judgments
cause drug-drug interactions in the clinic. In will be available from in vivo next round of synthesis. regarding such data, but in the
cases where oxidative metabolism by CYP studies in animals aimed at as- current environment the task is to
enzymes is likely to be an important mecha- sessing selected off-target phar- make such decisions rapidly and
nism of drug clearance in humans, it is pref- macological activities of the compound of in- to know how to weigh the data as they come in
erable to have contributions from multiple terest, including effects on the CNS and cardio- from different sources. These decisions can, of
isoforms, as opposed to a single CYP (again, vascular systems. It is true that different phar- course, also be influenced by the nature of the
to minimize the potential for drug-drug inter- maceutical companies generate and weigh the target itself, because a tolerance for a particu-
actions), whereas it is particularly undesir- above types of information to different extents lar toxicity might well be different for diseases
able for metabolism to be catalyzed solely by in selecting lead candidates for progression into with such different profiles as obesity versus a
an enzyme, such as CYP2D6 or CYP2C19, development. At Merck, all of the above charac- particular cancer.
www.sciencemag.org SCIENCE VOL 303 19 MARCH 2004 1811
methods (14). Such rapidly obtained information
on newly synthesized compounds is one of the
most important factors in the quest to shorten the
times from lead molecules to drug candidates.
One must constantly be aware that the rapid
synthesis of large numbers of molecules that are
laden with ADME, physical property, or toxico-
logical shortcomings may provide intriguing hits
or leads, but they may not shorten the time to the
elaboration of such a hit into a drug candidate. In
fact, as noted above, most experienced medicinal
chemists would prefer to start in a structural
series that has inherently good ADME proper-
Fig. 2. Nonlinear time-optimized path to drug candidate, with numerous feedback loops designed ties, albeit with poor potency on the target
to provide optimal information on the next round of synthesis. receptor, and then set about improving the po-
tency on the target, rather than working in the
At Merck, as at several other pharmaceuti- institutional knowledge of medicinal chemistry, other direction (starting with a potent molecule
cal companies, we have found that the most but now guided by more information, as that requires modification to optimize ADME
fruitful approach to the selection of new drug depicted in Fig. 2. This approach has led to moreand toxicological properties, which requires op-
candidates is to identify the key issues of a lead outsourcing of research medicinal chemistry timization of several, often opposing, structural
compound, based on early screening data, and than was common practice a few years parameters within the predefined tight structure-
then to focus on minimizing these deficiencies ago (11–13). activity boundaries required for potency), al-
by informed chemical intervention, bearing in It should be noted that pharmaceutical com- though the history of drug discovery is replete
mind SAR data for the pharmacological target. panies have sample collections filled with mol- with examples of both. A good recent example
For example, potent CYP inhibition in a lead ecules that were prepared many years ago for old of this situation from our laboratories has been
compound may be localized to a single func- discovery programs. Even if those molecules did the development of the orally active substance P
tional group, which may then be replaced by a not advance the program for which they were antagonist EMEND (aprepitant) (15) (Fig. 3).
noninhibitory substituent. Likewise, CYP in- initially made, they were designed at the time by Merck and many other companies have
duction (for example, through activation of medicinal chemists in the hope of interacting worked in this area for many years. The field
the nuclear transcription factor PXR), metab- with some type of proteinaceous domain (such was stimulated in 1991 by the discovery of CP-
olism to a reactive electrophile, or unwanted as an enzyme, heterotrimeric G protein–coupled 96,345 by Pfizer scientists, which showed that a
cardiovascular activities (for example, ion receptor, ion channel, etc). It is not unusual, potent subnanomolar small molecule could se-
channel activity that may lead to adverse therefore, for these molecules to be the startinglectively antagonize substance P at the NK-1
cardiac effects in vivo, such as that reflected point of new medicinal chemistry programs receptor (16). However, because of the difficulty
by prolongation of the QT interval on an when they show up as hits in a new HTS screen. in advancing structurally related molecules
electrocardiogram) (6) may be traced to spe- Thus, because of the rapid synthetic cycle times,through the drug development process, presum-
cific structural motifs that can be successfully a medium-sized group of medicinal chemists can ably due largely to off-target activities, metabo-
engineered out of the lead structure. This now advance several different lead classes at thelism issues, and the need to penetrate the CNS, it
multidisciplinary approach to drug discovery, same time and thus potentially shorten the time- took more than a decade before a small molecule
with organic chemistry serving as the corner- was identified that had the appropriate properties
lines for developing a hit or lead into a true drug
stone of the process, is far removed from the to be a drug, and EMEND was launched by
candidate. Usually, it is not clear at the start of
linear paradigm of former years (Fig. 1). a project what the downstream toxicological, Merck in 2003 for the treatment of both acute
Thus, while many new technologies such as and delayed-phase chemotherapy-induced nau-
metabolic, or off-target liabilities of a particular
combinatorial chemistry, rapid analog synthesis, sea and vomiting. Based on our experiences and
lead class are likely to be, and so different structural
automated synthesis, open access liquid chroma- classes can now be investigated simultaneously toknowing the large number of other companies
tography mass spectrometry, and high-speed au- allow for data-driven decisions. working in this area, it is very likely that tens of
tomated high-performance liquid chromatogra- When experienced medicinal chemists are thousands of molecules have been prepared in
phy (to name but a few) are now affecting asked to reflect on why various
medicinal chemistry, their main effect has been programs were advanced more
to shorten the cycle time of synthetic operations. quickly than others, they will
This, in turn, has led to a profound difference in invariably agree that it was be-
the way in which a medicinal chemistry project cause of the nature and quality
progresses through the system. Different compa- of the starting hit or lead. One
nies have embraced these new technologies in of the most difficult properties
different ways (7–10). For instance, some invest- to build into a newly discov-
ed heavily in the mid-1990s in combinatorial ered lead molecule is the de-
chemistry and made this technology a key driver sired pharmacokinetic (PK)
of their efforts to discover new leads and to profile, particularly in the case
expand their existing sample collections, partic- of orally dosed compounds. In
ularly when traditional sources of compounds recent years, the resources
failed to deliver new leads. Others have used available for early PK evalua-
these technologies in appropriate projects and tions in rodents have been in-
have forged alliances with smaller companies creased, both for single com-
that specialize in such efforts, thus freeing up pounds and, where appropriate,
their internal operations to use their historical with the use of cassette dosing Fig. 3. Structures of CP-96,345 and EMEND (aprepitant).
1812 19 MARCH 2004 VOL 303 SCIENCE www.sciencemag.org
the past decade as substance P antagonists, a important issues facing discovery medicinal “lead” is deﬁned as a structure that has been derived
from an early “hit” and, although still not fully opti-
large percentage of which likely exhibited sub- chemistry today: the continuing need for excel- mized, has been shown to have some appropriate char-
nanomolar potency at the NK-1 receptor, but lent synthetic chemists. In large pharmaceutical acteristics to be a precursor of a drug entity. Often a
only one has made it to market. This example companies, the drug discovery process is driven good lead will have shown some proof-of-concept ac-
tivity in an in vivo pharmacological model, but will likely
highlights the difficulty of and resources needed by multidisciplinary teams made up of the very not have been fully optimized for pharmacokinetic
to optimize the ancillary properties of potent best experts in each discipline, and chemistry is properties or undesirable off-target activities.
inhibitors/antagonists so that they can become one key element in this. These teams have ready 2. C. A. Lipinski, F. Lombardo, B. W. Dominy, P. J. Feeney,
safe, viable medicines. access to experts in other areas of biomedical Adv. Drug Delivery Rev. 23, 3 (1997).
3. In the discovery setting, the rule of ﬁve (2) predicts
In the drug discovery process, we must also science, and although chemists often end up as that poor absorption or permeation of drugs is more
be cognizant of the interrelatedness of academic, group leaders of discovery efforts, that usually likely when a drug molecule possesses either (i) more
government, and industrial research in the devel- occurs after much experience has been gained in than 5 hydrogen bond donors, (ii) 10 hydrogen bond
acceptors, (iii) a molecular weight greater than 500,
opment of new drug entities. Despite large re- the drug discovery process. The recent advances or (iv) a calculated logP greater than 5.
search budgets, the biomedical research carried discussed above have put more tools in the chem- 4. J. Uetrecht, Drug Discov. Today 8, 832 (2003).
out by pharmaceutical companies still represents ist’s toolkit, but in order to use these tools effec- 5. D. C. Evans, A. P. Watt, D. A. Nicoll-Grifﬁth, T. A.
only a small percentage of the overall worldwide tively, it invariably comes down to the ability to Baillie, Chem. Res. Toxicol. 17, 3 (2004).
6. See (17) for an excellent review of the cardiovascular
research effort on diseases and approaches to make the absolutely “correct” molecule in a timely effects manifested by QT interval prolongation and
their treatment. Academic and government lab- and cost-effective manner. This process requires the evaluation of drug candidates for this parameter.
oratories, funded with public monies, often pro- the very best organic chemistry skills, and we must 7. See the cover story in Drug Discov. Dev. 6, 30 (2003).
vide much basic research and fundamental in- continue to provide funding in the university sys- 8. T. Koppal, Drug Discov. Dev. 6, 59 (2003).
sight into diseases that can direct researchers tem for training in these core skill sets to chemists 9. A. DePalma, Drug Discov. Dev. 5, 50 (2002).
10. A. DePalma, Drug Discov. Dev. 6, 51 (2003).
toward novel ways of attacking diseases. How- in their graduate and postdoctoral studies if we are
11. M. McCoy, J.-F. Tremblay, Chem. Eng. News 81, 15 (2003).
ever, they are rarely organized (nor is it their to continue to provide the very best in medicines 12. A.M. Rouhi, Chem. Eng. News 81, 75 (2003).
mission) to embrace the drug discovery process for what is becoming an aging population. 13. T. Koppal, Drug Discov. Dev. 6, 22 (2003).
in the multidisciplinary fashion outlined above 14. W. A. Korfmacher et al., Rapid Commun. Mass Spec-
that is the modern paradigm by which new hits trom. 15, 335 (2001).
References and Notes
1. In this discussion, a “hit” is deﬁned as a nonoptimized 15. J. J. Hale et al., J. Med. Chem. 41, 4607 (1998).
or leads are first identified and then get trans-
structure obtained from some screening process on a 16. R. M. Snider et al., Science 251, 435 (1991).
formed into new viable medicines. All of the target protein. It is often a very weak binder and is likely 17. R. Netzer, A. Ebneth, U. Bischoff, O. Pongs, Drug
above discussion speaks to one of the most to have a nonoptimized pharmacokinetic proﬁle. A Discov. Today 6, 78 (2001).
The Many Roles of Computation in
William L. Jorgensen
An overview is given on the diverse uses of computational chemistry in drug discovery. tor or enzyme. Molecular libraries are
Particular emphasis is placed on virtual screening, de novo design, evaluation of drug- screened, and the resulting leads are opti-
likeness, and advanced methods for determining protein-ligand binding. mized in a cycle that features design, syn-
thesis and assaying of numerous analogs,
“Is there really a case where a drug that’s suggests misunderstanding and oversimpli- and animal studies. Crystal structure deter-
on the market was designed by a comput- fication of the drug discovery process. mination for complexes of some analogs
er?” When asked this, I invoke the profes- First, it is the rare case today when an with the biomolecular target is often possi-
sorial mantra (“All questions are good unmodified natural product like taxol be- ble, which enables “structure-based drug
questions.”), while sensing that the desired comes a drug. It is also inconceivable that a design” (SBDD) and the efficient optimi-
answer is “no”. Then, the inquisitor could human with or without computational tools zation of leads. The success of SBDD is well
go back to the lab with the reassurance that could propose a single chemical structure documented (1, 2); it has contributed to the
his or her choice to avoid learning about that ends up as a drug; there are far too introduction of 50 compounds into clinical
computational chemistry remains wise. The many hurdles and subtleties along the way. trials and to numerous drug approvals. Min-
reality is that the use of computers and Most drugs now arise through discovery imally, the role of computation here is in the
computational methods permeates all as- programs that begin with identification of structure refinement using simulated anneal-
pects of drug discovery today. Those who a biomolecular target of potential thera- ing (3), development of the underlying molec-
are most proficient with the computational peutic value through biological studies in- ular mechanics (MM) force fields, structure
tools have the advantage for delivering new cluding, for example, analysis of mice display, and building and MM evaluation of
drug candidates more quickly and at lower with gene knockouts. A multidisciplinary analogs. All top pharmaceutical companies
cost than their competitors. project team is then assembled with the have substantial structural biology and com-
However, the phrasing of the question goal of finding clinical candidates, i.e., putational chemistry groups that are inter-
druglike compounds that are ready for hu- twined and participate on the project teams.
Department of Chemistry, Yale University, New
man clinical trials, which typically selec- There is usually much “tweaking” to-
Haven, CT 06520-8107, USA. E-mail: william. tively bind to the molecular target and in- ward the end of the preclinical period of
firstname.lastname@example.org terfere either with its activity as a recep- drug discovery when a series of compounds
www.sciencemag.org SCIENCE VOL 303 19 MARCH 2004 1813