e
`
`CIENCE
`
`15 NOVBMBBll 1991
`VoL 254 • PAGBS 909-1o8o
`
`$6.00
`
`AMBIUCAN
`AssociATION POll THE
`ADVANCBM.BNT OF
`SciENCB
`
`NOV 19 1991
`
`Voyages of
`the Mind:
`
`90LiS
`
`IH
`
`--
`
`

`
`AMERICAN
`AsSOCIATION FOR THE
`ADVANCEMENT OF
`SCIENCE
`
`SciENCE
`
`ISSN 0036·8Cl75
`IS NOVEMBER 1991
`VOLUME 2.54
`NUMBER 5034
`
`915 This Week in Sciena
`
`917 The Land of the Dammed
`
`918
`
`Scientific Misconduct Investigations: Who Should Conduct Them?: D. F. KLEIN •
`Phylogeny and Diversity: C. I<MJEWSKJ • Extinction "Hot Spots": N. MYERS • Toxic
`Waste Cleanup: D. W. WOLF AND L. E. GR£ER • PCBs in the Environment:
`P. BOFFETTA AND H. VAINIO; G. w. GRJBBLE
`
`927 Brighter prospects for solar telescope; planning an Antarctic evacuation; etc.
`
`928 No Meeting of the Minds on Asbestos • Consensus Report Draws Fire From Both
`Extremes • "'Third Wave": Roiling the Waters
`Scientific Sleuths Solve a Murder Mystery
`931
`932 HUGO Takes on Role as Marriage Broker
`933 Hughes Investigators Rile NlH Peer Reviewers
`934 Briefings: Questions Raised on Math Rankings • Rating University R&D (cont.) •
`Catching Some (Cosmic) Rays • Students Thwart USDA Pest Plan • An Ice Cap on the
`H onest Planet? • Primatologist Band Together • Relaying Science to the People
`
`936 How Long Is the Human Life-span?
`938 The Sound of One Dune Booming
`939 Putting Einstein ro the Test-in Space
`941 AIDS: The Evolution of an Infection
`942 Extinction Potpourri: KiUers and Victims: Where to Run From a Mass Extinction? • Plate
`Tectonics as a Driver of Evolution • Looks Like the Yucatan Holds a Killer Crater
`
`944 What Next in the GaUo Case?
`946
`John Crcwdson: Science Journalist as Investigator • Report Card on Crewdson's Repomng
`
`953 Receptor·lnduccd Confonnation Change of the Immunosuppressant Cyclosporin A:
`K. WOniJuCH, B. VON FR£YBERG, C. WEBER, G. WIDER, R. TRABER, H. W IDMER,
`w. BRAUN
`954 Rusting of the Lock and Key Model for PI'O(ein-Ligand Binding: W. L. JoRGENSEN
`
`959 Changes in the West Antarctic Icc Sheet: R. B. ALLEY AND I. M. WHILLANS
`963 Antigenic Diversity Thresholds and the Development of AIDS: M.A. NOWAK,
`R. M. ANDERSoN, A. R. McLEAN, T. F. W. WoLFs, J. GouDSMIT, R. M. MAY
`
`970 Electronic Correlation Effects and Superconductivity in Doped FuUercnes:
`s. CHAKRAVARTY, M.P. GELFAND, s. I<JVELSON
`Protein Hydration in Aqueous Solution: G. O'rnNG, E. LlEP!NSH, K. WOTHRJCH
`
`974
`
`981 X-ray Damage to CF3C02-Tctrninated Organic Monolayers on Si/Au: Principal Effect of
`Electrons: P. E. LAIBINIS, R. L. GRAHAM, H. A. BlEBUYCX, G. M. WHTffiSIDES
`Scanning Tunneling Microscopy of Galena ( 1 00) Surface Oxidation and Sorption of
`Aqueous Gold: C. M. EGGLESTON AND M. F. HOCHEllA, )R.
`
`983
`
`LPtt t·r·:-
`
`s('ic·rH·,.s,·o1H ·
`\c'\\ :- ~x <:orrtmt· rrt
`
`.
`
`\ rt i<' lc·:-
`
`912
`
`•
`
`•
`
`-lor the Adv•ocato..nt of Science, 1333 H s-. NW, WMhlng1on, DC 20005. Sec:ond-dass pos1age
`oates on req.-. -to phcolocopy ITI8Ierial for Internal or personal use under cir~ not hl1lng
`
`SCIENCE ~SSN 0036-t075) Ia publlahecl WMicly on Frk18y, except the !at WMI<In o.c.mber, by the ArlleriQn
`
`(publlcallon No. 484460) paid at Washington, DC. and additianal mailing ollices. Copvright 0 1991 by the American
`Assodation for 1he 1\dv"""""""' of Science. The 1i11e SCIENCE Is a registered trademallc of !he AAAS. DotMstic
`lndMdual membership and subsct1>liOn (51 issues): $82 ($47 allocated 1o S~Jbsct1lti0n). DotMstic lnstitulional
`Slbecripcion (51 Issues): $150. Foreigl pos1age extra: Mexloo. CwibbeM (...tace mail) S50: Olher OOUi'ilriet (all asslst
`~ S95. F'orst class. airmail. s1udenl and emeritus raes on request. c..lalian rates Mh GST IMiilable upon
`r~. GST #1254 88122. C li8nge of.-..: dow 6 weeks. gMng old and new acb'esses and 1 1-<lgil aooounc
`number. P-
`Send change of adchss 1o Science, P.O. Box 2033, Marion, OH ~. Single copy
`: $6.00 pet Issue prepaid indudes SOJrface postage: GUde 1o Biotechnology PI'Oducls and 1~. $20. Blllk
`-
`
`users registered wllh !he
`wf!hln the lair use prolllsions Of !he Copyright Ad Is granted by AAAS 10 libnwi<!s and -
`Copyright Clearance Center (CCC) Transactional Reporting Seovlce, provided thai !he base fee of$ 1 pet OOf1'l plus
`$0.10 per page Is paid direclly to CCC. 27 Congrets Street. Salem, Massachusetts 01970. The lden1ifica1ion code for
`Science Is 0036-8075/83 $1 + .1 0. Science Is Indexed In !he Reader's Guld6 ro PeriodlcBI Ulerature and In several
`specialized Indexes.
`The Aol*ican Association for 11>e Advancement of Science was founded In 1848 and lncorporaled In 1874. Its~
`are to fiMther !he WOI1< of scientists. to facilitate cooperation among !hem. to foster scientific freedom and responsibii!y. to
`improve !he efec:IM!ness of science in !he promcCion of hunan - . . . to acMn:e education In sdenc:e. and to
`Increase public Ul >Cierstal >Clio >Q and appt8Cialion of the impor1anoe and promise of the methods of sdenc:e in ..........
`~
`
`

`
`10 is located on the opposite side of the ring plane. For the polypeptide
`backbone the strucrura.l rearrangement upon binding to the receptor is
`reminiscent of the inversion of a glove:, whereby the hydrophobic
`exterior edges formed by theN-methyl groups in free CsA are replaced
`by a polar surface of amide protons and carbonyl oxygens.
`The conformation of the receptor-bound CsA indicates the
`possibility of hydrogen bonding with the receptor protein. For the
`amide proton of Abu2 the presence of hydrogen bonding is directly
`supporred by the observation of slowed exchange with the solvent
`(10). Hydrogen bonding is also compatible with 1H-1H nuclear
`Overhauser effects (NOEs) observed between CYP and the residues
`1 to 3 and 9 to 11 ofCsA (10, 11). The dramatic global rearrangement
`of the polypeptide backbone conformation in CsA, which enables
`recognition by hydrogen bonding, is particularly remarkable for a
`cyclic compound, which has a greatly reduced accessible confor(cid:173)
`mation space when compared with a corresponding linear poly(cid:173)
`peptide. Considering the imponant role of molecular modeling in
`drug design, it is instructive to note that molecular dynamics
`calculations that starred from the crystal structure or the solution
`structure of free CsA and that used different potential functions to
`represent the solvent (15) gave no indication of an imminent
`major conformation change away from the starting conformation.
`Thus CsA may well end up as a textbook case to illustrate the
`imponance of experimental studies with both free and receptor(cid:173)
`bound effector molecules for understanding structure-function
`correlations as a guide to improved molecular design.
`The NMR investigations of receptor-bound CsA were performed
`with combined use of isotope-labeling and heteronuclear NMR
`experiments. In an unlabeled system the large number of protons
`from the receptor protein would interfere with the observation of
`the resonance lines of the ligand. However, binary complexes are
`ideally suited for studies with efficient labeling schemes, because the
`two components can be labeled separately with 13C or 15N before
`complex formation and subsequently combined with unlabeled
`panner molecules. Suitably chosen heteronuclear editing schemes
`(16, 17) can then be used to separate the 1H NMR lines of the two
`molecules in the complex. In particular, use of the so-called heter(cid:173)
`onuclear half-filters (17, 18) in 2D 1H NMR spectra represents a
`valid alternative to the use of three or higher dimensional experi(cid:173)
`ments for improved resolution in such systems. An intrinsic advan-
`
`tage is that the 2D 1H -1H NMR spectra can be recorded with
`sensitivity and digital resolution comparable to those of correspond(cid:173)
`ing conventional 2D 1 H NMR spectra. With 1 3C-Iabeled CsA a
`double-half-filter technique was particularly helpful (19) because it
`produced different subspectra that contained either exclusively in(cid:173)
`tramolecular NOE cross peaks between different protons of CsA or
`different protons of CYP or exclusively intermolecular NOE cross
`peaks relating protons of CsA with protons of CYP. These tech(cid:173)
`niques are generally applicable with binary or multicomponent
`molecular assemblies, primarily in systems with very stable receptor(cid:173)
`effector complexes, and represent an attractive avenue for the use of
`NMR in conjunction with projects on drug design.
`
`REFERENCES
`
`I. ). F. Borel, Ed., C iclosporin (Karger, Basel, 1986).
`2. S. L. Schreiber, S cietlle 251, 283 (1991).
`3. H. R. Loosli eta/. , Helv. C him. Acta 68, 682 (1985).
`4. H. Kessler, M. Kock, Th. Wein, M. Gehrke, ibid. 73, 1818 (1990).
`5. R. E. Handschumacher, M. W. Harding, J. R.ice, R. ) . Orugge, D . W. Speicher,
`S cittlle 226, 544 ( 1984 ).
`6. G. Fischer, B. Wittmann-Liebold, K. Lang, T. Kicfhaber, F. X. Schmid, Nature
`337, 476 (1989); N. Takahashi, T. Hayano, M . Suzuki, ibid. , p. 473.
`7. A. ) . Koletsky, M. W. Harding, R. E. Handschumacher,J. l mmunol. 137, 1054
`(1986); M. Kawamukai, H. Matsuda, W. Fuji, R. Utsumi, T. Komano, ].
`& cteriol. 171, 4525 (1989).
`8. S. W. Michnick, M. K. Rosen, T. ). Wandless, M. Karplus, S. L. Schreiber, Stieru:t
`252,836 (1991); J. M. Moore, 0. A. Pcatrie, M. ). Fitzgibbon, J. A. Thomson,
`Nature 351, 248 (1991).
`9. G. 0. VanDuyne, R. F. Standaert, P. A. Karplus, S. L. Schreiber, ). Clardy, Scietlle
`252, 839 (1991).
`10. C. Weber tt a/., BiO<hemistry 30, 6563 (1991).
`II. S. W. Fesik ttal., ibid., p. 6575.
`12. K. Wuthrich, C. Spitzfaden, K. Memmert, H . Widmer, G. Widcr, FEBS Lett. 285,
`237 (1991).
`13. ). Kallen et at. , Nature 353, 276 (1991).
`14. R. M. Wenger, Angew. Chern. Int. Ed. Engl. 24, 77 (1985).
`15. ). Lautz, H . Kessler, R. Kaptein, W. F. van Gunsteren,J . C omput. Aided M ol. Des.
`1, 219 (1987); ). Lautz, H . Kessler, H . P. Weber, R. M . Wenger, W. F. van
`GWlSteren, Biopolymers 29, 1969 (1990).
`16. R. H. Griffey and A. G. Redfield, Q. Rev. Biophys. 19, 51 (1987).
`17. S. W. Fesik, Nature 332, 865 (1988).
`18. G. Otting, H . Sc.nn, G. Wagner, K. Wuthrich,]. Magn. Reson. 70, 500 (1986);
`G. Otting and K. Wuthrich, Q. Rev. Biophys. 23, 39 (1990); S. W. Fesik, J. R.
`Luly, J. W. Erickson, C. Abad-Zapatero, Biochemistry 27, 8297 (1988); G. Wider,
`C. Weber, K. Wuthrich,]. Am. C hern. $0<. 113, 4676 (1991).
`19. G. Otting and K. Wuthrich,} . Magn. Reson. 85, 586 (1989); G. Wider, C. Weber,
`R. Tr.~ber, H. Widmer, K. Wuthrich, }. A m. Chern. SO<. 112, 9015 (1990).
`
`Rusting of the Lock and Key Model for
`Protein-Ligand Binding
`
`Wn.LIAM L. JORGENSEN
`
`A TOMIC-LEVEL KNOWLEDGE OF THE GEOMETRIES OF PRO(cid:173)
`
`tein-ligand complexes has only been accumulating since the
`mid-1970s. About 50 x-ray structures have now been
`determined for peptides or proteins bound to enzymes or antibod(cid:173)
`ies. The traditional notion of rigid lock and key complementarity
`received suppon from the early and numerous studies of complexes
`of proteolytic enzymes with small protein inhibitors (1) and from
`the first example of an antibody-protein complex (2). However, the
`idea has become increasingly challenged.
`In fact, conformational changes for enzymes upon ligand uptake
`are well known and range from modest loop motions to hinge
`bending (3). The prototypical case of strong binding, streptavidin-
`
`Department of Chemistry, Yale University, New Haven, Cf 0651 I.
`
`954
`
`biotin, involves adjustments to streptavidin that include a loop flip
`(4), and the bear hug applied by human immunodeficiency virus
`type 1 (lllV-1) protease to a peptide inhibitor is a striking example
`of large-scale domain motions (5).
`Recently, the effects of binding on ligand structure have received
`increased attention. Cases of profound conformational change have
`been provided by the determination of the structures of the immu(cid:173)
`nosuppressive agents FK506 and cyclosporin A (CsA) complexed
`with their cytosolic binding proteins FKBP and cyclophilin. X-ray
`structures have been reported for the uncomplexed drugs and the
`FKBP-FK506 complex (6), and the structure of CsA bound to
`cyclophilin has been determined by two groups using multidimen(cid:173)
`sional nuclear magnetic resonance (NMR) techniques (7).
`Both binding proteins are peptidyl-prolyl-cis-trans isomerases and
`have been shown to interfere with T cell signaling upon forming
`ternary complexes with their respective immunosuppressive agents
`and the protein phosphatase calcineurin (8). FK506 is a macrocyclic
`organic molecule with a critical a-keto-homoprolyl subunit (upper
`right pan of the structure) that acts as a transition-state surrogate
`(9), and CsA is a cyclic undecapeptide. Binding leads to substantial
`conformational change for FK506, including cis-trans isomerization
`
`SCIENCE, VOL. 254
`
`

`
`Rllpamycln
`
`FK506 unbound
`
`FK506 bound
`
`Cyelospor1n A
`
`of the amide bond, concomitant repositioning of the homoproline
`ring, and inward disposition of the pyranose ring (top) toward the
`macrocycle. For CsA, the unbound structure has essentially been
`rumed inside out to reach the bound conformation; the four
`intramolecular hydrogen bonds are lost, the 9,10-peptide bond
`isomerizcs from cis to trans, and two sets of four side chains switch
`sides of the ring. These examples show that flexible ligands can
`undergo substantial geometrical distortions to achieve a suitable
`binding conformation.
`For variety, narure has also provided rapamycin, an immunosup(cid:173)
`pre. ... ~anr that binds to FKBP and is closely related strucrurally to
`FK506, particularly in the C1-C14 binding domain. T he bound and
`unbound strucrures of rapamycin are virrually identical, and the
`distortion of FK506 on binding to FK.BP yields a binding domain
`that is superimposable on the rapamycin strucrure (10). The implicit
`greater preorganization of rapamycin might be thought to provide
`greatly enhanced binding relative to FK506; in fact, it only amounts
`to a factor of 2 in the binding constants.
`Although these examples are striking, they are not unprecedented.
`The binding of lysozyme by a F •b fragment requires negligible
`distortion of either component (2). However, subsequent investi(cid:173)
`gations suggest that this is not general. Complexation of the viral
`antigen neuraminidase by a F ab fragment was found to displace
`some Ca positions in a binding loop by more than 1 A from their
`locations in the free enzyme (11). More recently, an x-ray strucrure
`has been reported for a F ab complex with a 19- amino acid homolog
`of the C helix of myohemerythrin (12). Unbound in water, the
`peptide shows no stable secondary strucrure, but the NH2-terminal
`region of the peptide forms a type II ~ rurn when bound to the
`antibody. Another salient example is provided by the expulsion of
`the heme from myoglobin upon binding to an antibody to apomyo(cid:173)
`globin ( 13). Furthermore, the isolated S-peptide of ribonuclease A
`shows no helicity at 25°C but regains fuU helical character upon
`binding in the S-protein (14). In fact, most secondary strucrure in
`proteins can be considered to arise from bindinglike interactions
`with the remainder of the protein.
`
`CsAunbound
`
`CsAbound
`
`These examples confirm the reasonable expectation that flexible
`molecules distort to form optimal interactions with binding part(cid:173)
`ners. A practical consequence is the frustration that will often
`accompany attempts to design drugs by analogy to the structures of
`flexible, unbound active substances.
`
`REFERENCES
`
`I. D. M. Blow, A«. Chnn. Res. 9, 145 ( 1976); R . Huber and W. Bode, ibid. 11, 114
`(1978); S. J. Hubbard, S. F. Campbell, J. M. Thornton,]. Mol. Bioi. 220,507
`( 1991).
`2. A. G. Amit, R. A. Mariuzza, S. E. V. Phillips, R. J. Poljak, Scimc~ 233, 747
`( 1986).
`3. W. S. Bennett and R. Huber, CRC Cri1. Rw. Bi«hm>. 15, 291 ( 1984).
`4. P. C. Weber, D. H. Ohlendorf, j . j . Wendoloslc.i, F. R. Salemme, S<icru:~ 243,85
`( 1989).
`5. M. Miller n al., ibid. 246, 1149 (1989).
`6. G. D. Van Duyne, R. F. Standaert, P. A. Karplus, S. L. Schreiber, J. Clardy, iUd.
`252,839 ( 1991).
`7. C. Weber ~~ a/., BiO<hrnli.slry 30, 6563 (1991); S. W. Fcsik tl al., ibid., p. 6574.
`8. J. Liu, J. D. Fanner, Jr., J. Friedman, I. Weissman, S. L. Schrcibe.r, 011 66, 807
`(1991).
`9. M. K. Rosen, R. F. Standacrt, A. Galat, M. Nahtsuk.a, S. L. Schreiber, sn..u:~
`248, 863 {1990).
`10. G. D. Van Duync, R. F. Srandacrt, S. L. Schreiber, J. Qardy, ). Am. Chmt. S«.
`113,7433 ( 1991).
`II. P.M. Colman~~ al., Na1ur~ 326, 358 ( 1987).
`12. R. L. Stanfield, T . M. Fieser, R. A. Lerner, I. A. Wilson, Sci~ru:~ 248,712 (1990).
`13. M. ). Crumpton, Bi«h~m.J. 100, 223 (1966).
`14. P. S. Kim, A. Bierzynslc.i, R. L. Baldwin,]. Mol. Bioi. 162, 187 ( 1982).
`
`IS NOVEMBER 1991
`
`PERSPECTIVES 955