LETTERS TO NATURE
`
`22. Johnson, C. K. ORTEP Manual (Oak Ridge National Laboratory, Tennessee, 1965).
`23. Bjorkman, P. J. et al. Nature 329, 512-518 (1987).
`24. Zhang, W ., Young, A. C. M., lmarai, M., Nathenson, S. G. & Sacchettini, J. C. Proc. natn. Acad Sci.
`USA 89, 8403-8407 (1992).
`
`ACKNOWLEDGEMENTS. We acknowledge the contributions made by T. P. J. Garrett, M. A. Saper, and
`P. J. Bjorkman to the determination of the structure of HLA·Aw68 to 2.6 A reso1ution11. We thank
`J. Gorga for supervising the protein purifications and A. Haykov, K. Svenson and R. Crouse for technical
`assistance. H .• c.G. is a Cancer Research Institute fellow. J.L.S. and D.C.W. acknowledge support from
`the NIH. D.C.W. is supported by the Howard Hughes Medical Institute.
`
`The three-dimensional structure
`of an intact monoclonal
`antibody for canine lymphoma
`Lisa J. Harris, Steven B. Larson, Karl W. Hasel*,
`John Day, Aaron Greenwood & Alexander McPherson
`
`Department of Biochemistry, University of California Riverside, Riverside,
`California 92521, USA
`* lmmunopharmaceutics Inc., 11011 Via Frontera, San Diego,
`California 92127, USA
`
`CRYSTAL structures of Fab antibody fragments determined by
`X-ray diffraction characteristically feature four-domain, JJ-barrel
`3
`arrangements1
`• A human antibody Fe fragment has also been
`-
`found to have four /3-barrel domains4
`• The structures of a few
`intact antibodies have been solveds.-s: in two myeloma proteins,
`the flexible hinge regions that connect the Fe to the Fab segments
`were deleted5
`6 so the molecules were non-functional, structurally
`•
`restrained, T-shaped antibodies; a third antibody, Kol, had no
`hinge residues missing but the Fe region was sufficiently disordered
`that it was not possible to relate its disposition accurately with
`respect to the Fab components7
`8
`• Here we report the structure at
`•
`3.5 A resolution of an IgG2a antitumour monoclonal antibody
`which contains an intact hinge region and was solved in a triclinic
`crystal by molecular replacement using known Fe and Fab frag(cid:173)
`ments. The antibody is asymmetric, reflecting its dynamic charac(cid:173)
`ter. There are two local, apparently independent, dyads in the
`molecule. One relates the heavy chains in the Fe, the other relates
`the constant domains of the Fahs. The variable domains are not
`related by this 2-fold axis because of the different Fab elbow
`angles of 159° and 143°. The Fe has assumed an asymmetric,
`oblique orientation with respect to loosely tethered yet almost
`collinear Fahs. Our study enables the two antigen-binding seg(cid:173)
`ments as well as the Fe portion of a functional molecule to be
`visualized and illustrates the flexibility of these immune response
`proteins.
`The specific murine antibody described here reacts with cells
`of canine lymphoma9
`, the most common haemopoietic tumour
`in the dog, which resembles human non-Hodgkin's lymphoma.
`This antibody can participate in antibody-dependent cellular
`cytotoxicity as well as complement-dependent cytolysis10 and
`is used as an anticancer therapeutic' 1 by veterinarians. The
`immunoglobin crystallizes from a low concentration of poly(cid:173)
`ethylene glycol at slightly alkaline pH (ref. 12).
`The structure of the triclinic crystals, having one entire anti(cid:173)
`body as the asymmetric unit, was solved using the method of
`molecular replacement as implemented in the programs MER(cid:173)
`LOT13 for rotation functions and XPLOR14' 15 for translation
`functions. By virtue of the triclinic cell there were no symmetry
`constraints on the molecule, immediately suggesting that the
`antibody has an asymmetric conformation. Molecular probe
`coordinates for Fab fragments and the Fe fragment were
`obtained from the Brookhaven Data Bank 16
`. Sometimes cross(cid:173)
`rotation 17 searches were done with probes representing one third
`of the asymmetric unit, and in other cases the search probes
`comprised only one sixth of the asymmetric unit. Convincing
`
`369
`
`FIG. 4 The 'antigenic' surface composed of Np 91-99 peptide (orange) and
`MHC atoms (blue, conserved; red, polymorphic; light blue, not conserved or
`polymorphic). N-terminal of peptide is to the left, a 1 domain a-helix top, a 2
`domain a-helix bottom. Figure generated with RASTER3D. (Of the 12 poly(cid:173)
`morphic residues facing into the binding site, 8 contact the peptide directly
`(9, 45, 66, 70, 74, 77, 95, 116) and four do not (67, 97, 114, 156), but 11
`of the 12 (66 excluded) are nevertheless completely buried by the bound
`peptide. These polymorphic positions must therefore, as anticipated23
`, have
`their primary effect on T-cell recognition of HLA-Aw68 through the choice
`of peptides that can bind. Of the six polymorphic residues that face more
`directly toward solvent (62, 65, 69, 76, 80, 163), four also contact the
`peptide (62, 69, 80, 163) but all have atoms accessible to direct recognition
`by the TCR and therefore represent polymorphism recognizable by TCRs in
`the presence or absence of peptide.)
`
`may not be a major factor in the creation of novel antigenic
`surfaces recognized by T cells. On the basis of the number of
`atomic contacts, Np 91-99 appears to be bound to HLA-Aw68
`predominantly by two main features of the MHC molecule: (1)
`conserved MHC residues hydrogen bond to the peptide termini;
`(2) polymorphic MHC residues bury the two 'anchor' peptide
`side chains. Although both of these sets of interactions would
`also provide for the peptide-dependent stabilization of the MHC
`molecule, only the peptide termini binding sites are conserved
`in class I histocompatibility antigen sequences. The overall mode
`of peptide binding observed here seems to be a general mechan(cid:173)
`ism for class I MHC presentation now visualized in three human
`5
`alleles and one murine allele: HLA-B273
`, HLA-Aw68 10
`, HLA(cid:173)
`-
`0
`A212 and H-2Kb(refs 13, 24).
`
`Received 4 August; accepted 2 October 1992.
`1. Townsend, A. & Bodmer, H. A Rev. lmmun. 7, 601-624 (1989).
`2. Brodsky, F. M. & Guagliardi, L. E. A Rev. lmmun. 9, 707-744 (1991).
`3. Madden, D. R., Gorga, J. C., Strominger, J. L. & Wiley, D. C. Nature 253, 321-325 (1991).
`4. Jardetzky, T. S., Lane, W. s .. Robinson, R. A., Madden, D. R. & Wiley, D. C. Nature 253, 326-329
`(1991).
`5. Madden, D. R., Gorga, J. C., Strominger. J. L. & Wiley, D. C. Cell 70, 1035-1048 (1992).
`6. Turner, M. J. et al. J biol. Chem. 250, 4512-4519 (1975).
`7. Bjorkman, P. J., Strominger, J. L. & Wiley, D. C. J molec. Biol. 188, 205-210 (1986).
`B. Silver, M. L., Parker, K. C. & Wiley, D. C. Nature 350, 619-622 (1991).
`9. Cerundolo, V., Tse, A.G. O., Salter, R. 0., Parham, P. & Townsend, A. Proc. R Soc. 244, 169-177
`(1991)
`10. Guo, H·C., Jardetzky, T. s .. Lane, W. S., Strominger, J. L. & Wiley, D. C. Nature 360, 364-366 (1992).
`11. Garrett. T. P. J.. Saper, M.A .. Bjorkman, P. J., Strominger. J. L. & Wiley, D. C. Nature 342, 692-696
`(1989).
`12. Saper, M. A., Bjorkman, P. J. & Wiley, D. C. J. molec. Biol. 219, 277-319 (1991).
`13. Fremont, D. H., Matsumura, M. Stura, E. A., Peterson, P.A. & Wilson, I. A. Science 257, 919-927
`(1992).
`14. Teng, T-Y. J appl. Crystallogr. 23, 387-391 (1990).
`15. Durbin, R. M. et al. Science 232, 1127-1132 (1986).
`16. Blum. M .. Metcalf. P., Harrison, S. C. & Wiley, D. C. J appl. Crystal/ogr. 20, 235-242 (1987).
`17. Fox, G. C. & Holmes, K. C. Acta Crystallogr. A34, 517 (1966).
`18. Brunger, A. T. XPLOR (version 2.1) Yale University, New Haven. (1990).
`19. Silver, M. L. thesis, Harvard University (1992).
`20. Jones, T. A. J app/. Crystallogr. 11, 268-272 (1978).
`21. Parham, P. et al. Proc natn. Acad. Sci. US.A 85, 4005-4009 (1988).
`
`NATURE · VOL 360 · 26 NOVEMBER 1992
`
`(cid:141) (cid:21)(cid:29)(cid:29)(cid:22) (cid:50)(cid:69)(cid:88)(cid:89)(cid:86)(cid:73)(cid:4)(cid:4)(cid:52)(cid:89)(cid:70)(cid:80)(cid:77)(cid:87)(cid:76)(cid:77)(cid:82)(cid:75) (cid:43)(cid:86)(cid:83)(cid:89)(cid:84)
`
`1 of 4
`
`BI Exhibit 1108
`
`

`

`FIG. 1 Ribbon representation of the structure of the murine antibody against
`canine lymphoma determined by X-ray analysis of the triclinic crystals. The
`heavy chains are shown in yellow and blue, the light chains in red. The Fe
`stem of the molecule projects towards the viewer and assumes an asym(cid:173)
`metric, oblique orientation with respect to the Fabs. This orientation illus(cid:173)
`trates the large difference in hinge angles of about 65° and 115°. The local
`dyad relating the heavy chains of the Fe is that dyad indicated by the primary
`solution of the self-rotation function. Fab2 is viewed along the axis through
`the switch peptides. Fab2 has an elbow angle of 143°, in contrast to Fab1,
`which has an elbow of 159°. Twenty-three residues in each heavy chain,
`comprising the hinge regions seen here, were built into the model with
`5
`idealized geometry using both FROD029 and XPLOR14
`. These residues
`were missing from the fragment models taken from the Brookhaven Data
`Bank.
`
`·1
`
`LETTERS TO NATURE
`
`FIG. 2 a, stereo diagram of the monocloncal
`antibody viewed perpendicular to the approxi(cid:173)
`mate 2-fold axis relating the constant domains
`of the Fabs. This dyad was that indicated by
`the secondary solution of the self-rotation
`function. Apparent here is the difference in
`the two elbow angles and the consequent
`failure of the variable domains to maintain this
`relationship. Also apparent in this view is the
`failure of the Fe dyad to intersect the 2-fold
`axis relating the constant domains of the Fabs.
`Both symmetry axes are apparently indepen(cid:173)
`dent local dyads. b, Stereo diagram of the lgG2a
`antibody showing the region between the CH2
`domains. In the human lgGl Fe fragment4
`,
`carbohydrate was located in this area between
`the two CH2 domains and is probably in a
`similar location in this antibody. No attempt
`has yet been made to include the carbohydrate
`component in the model. It can also be seen
`here that the dyad axis of the Fe does not
`intersect the approximate long axis of the
`Fabs. Colour coding is the same as in Fig. 1.
`
`370
`
`NATURE · VOL 360 · 26 NOVEMBER 1992
`
`(cid:141) (cid:21)(cid:29)(cid:29)(cid:22) (cid:50)(cid:69)(cid:88)(cid:89)(cid:86)(cid:73)(cid:4)(cid:4)(cid:52)(cid:89)(cid:70)(cid:80)(cid:77)(cid:87)(cid:76)(cid:77)(cid:82)(cid:75) (cid:43)(cid:86)(cid:83)(cid:89)(cid:84)
`
`2 of 4
`
`BI Exhibit 1108
`
`

`

`LETTERS TO NATURE
`
`TABLE 1 Crystallographic data, structure solution and refinement
`
`Crystal data: Space group Pl; a=66.39A, b=77.34A, c=101.42A, a=87.6°, ,8=92.6°, y=97.5°, Z=1; resolution -3.0A. Data collection: SDMS
`(Xuong-Hamlin) detectors, Rigaku Ru-200 source, frame size 0.12°, counting time 60-120 s. Total observations, 114,867; at 3.5 A unique reflections,
`24,808; 98.7% complete; Rsym=0.098. After F/u=4.0 cutoff, unique reflections at 3.5A=20,964
`
`Operation
`
`Self-rotation function
`
`Fe rotation function
`(1) Search probe entire Fe
`(2) CH3 domains
`(3) CH2 domains
`
`Fab1 rotation function
`(1) Search probes were constant domains of 7 different Fabs
`(2) Probes were variable domains of 7 different Fabs
`(3) An intact Fab1 constructed according to above results
`
`Fab2 rotation function
`Range of Fab models with elbows of 120°-180° constructed using
`XPLOR by altering Fab1 elbow every 5°
`
`Translation function
`(1) Fe fixed, Fab1 moving
`(2) Fe fixed, Fab2 moving
`(3) Fab1 fixed, Fab2 moving
`(4) Fab1 and Fab2 fixed, Fe moving
`
`Refinement
`(1) Rigid body with twelve ,13-barrel domains; 3.5-12 A
`(2) Powell minimization and simulated annealing after insertion of
`correct amino-acid sequence (occupancy of 46 hinge residues set to
`zero); 3.5-8 A
`
`Result
`Two consistent dyad solutions in several resolution ranges
`
`Two peaks related by pseudodyad; r.m.s.*=6.03 and 5.54
`same solutions as (1); r.m.s. = 5.66 and 5.44
`same solutions as (1); r.m.s. = 3.41 and 3.14
`
`Exceptional peak with constant domains of HYHEL-5; r.m.s. =6.42
`Solution optimal for variable domains of McPC603; r.m.s.=4.35
`Outstanding solution consistent with (1) and (2); r.m.s.=7.98
`
`Unambiguous solution for probe with elbow angle 140°; r.m.s. =6.27
`
`A self-consistent set of solutions
`from all searches (1)-(4);
`cct =0.157-0.249 for 4-8 A
`resolution
`
`R=0.386
`R=0.188
`r.m.s. deviations
`Bonds
`Angles
`Dihedrals
`lmpropers
`
`cct=0.529
`cct=0.876
`
`0.011 A
`4.038°
`28.597°
`0.686°
`
`* All r.m.s. values are stated for 4-8 A resolution searches. t cc, Correlation coefficient.
`
`solutions were frequently found even with the probes represent(cid:173)
`ing one sixth of the antibody. Rotation function solutions were
`based on the human Fe fragment4
`, the constant domains of Fab
`HYHEL-5 (ref. 18), and the variable domains of Fab McPC603
`(ref. 19), including the hypervariable regions. Translation
`searches were performed to determine the relative distances
`between the three portions of the molecule, the two Fabs and
`the Fe (Table 1 ).
`The antibody structure was assembled according to essentially
`independent, but internally consistent, molecular replacement
`results that ultimately yielded a model fully consistent with the
`
`stereochemistry of an intact antibody. The packing revealed
`good complementarity of surfaces without interpenetration. Lat(cid:173)
`tice contacts immobilize all segments of the molecule to permit
`visualization of both Fabs as well as the Fe.
`The structure of the antibody is shown in Figs 1 and 2. Its
`most prominent features are: (1) There is an approximate 2-fold
`axis relating the heavy chains of the Fe portion of the molecule.
`The dyad deviates particularly for the CH2 domains; (2) the
`disposition of the Fe with respect to the Fab portions is quite
`oblique; (3) the hinge angle between the Fe and Fabl is approxi(cid:173)
`mately 65° and for Fab2 about 115°; (4) the long axes of the
`
`FIG. 3 Stereo diagram of the packing
`of four antibodies in the triclinic cell,
`each of a different colour, showing the
`intricate network of intermolecular con(cid:173)
`tacts that stabilizes the conformation
`of the molecule. The Fe segments,
`which lie more or less along the longest
`immobilized by
`body diagonal, are
`multiple Fab contacts, suggesting why
`the Fe in these crystals is ordered. The
`constant domains of an Fab2 of one
`molecule insert in the elbow region of
`Fab1 of a different antibody molecule
`to fix the dispositions of the Fabs. Not(cid:173)
`able initial exceptions to the otherwise
`acceptable packing were three hyper(cid:173)
`variable loops protruding from the vari(cid:173)
`able domains of Fab McPC603. When
`the correct sequence for the canine
`lymphoma antibody was examined, it
`was apparent
`that
`the offending
`residues corresponded to deletions in
`the latter molecule. Thus, when the cor(cid:173)
`rect amino-acid sequence was sub(cid:173)
`stituted, virtually all of the packing
`exceptions were eliminated, as shown
`in this view.
`
`NATURE · VOL 360 · 26 NOVEMBER 1992
`
`371
`
`(cid:141) (cid:21)(cid:29)(cid:29)(cid:22) (cid:50)(cid:69)(cid:88)(cid:89)(cid:86)(cid:73)(cid:4)(cid:4)(cid:52)(cid:89)(cid:70)(cid:80)(cid:77)(cid:87)(cid:76)(cid:77)(cid:82)(cid:75) (cid:43)(cid:86)(cid:83)(cid:89)(cid:84)
`
`3 of 4
`
`BI Exhibit 1108
`
`

`

`LETTERS TO NATURE
`
`two Fabs are almost collinear; thus, there is an approximate
`long axis running through the entire Fab assembly. The angle
`between the Fabs is 170 ± 2°, and the Fab axes are offset by 9 A;
`(5) Fabl has an elbow angle of 159°, and Fab2 has an elbow
`of 143°. These elbow angles are near the middle of the range of
`values observed for other Fabs 1
`; (6) the constant domains of
`Fabl and Fab2 are related by a near exact dyad axis of symmetry.
`The variable domains are not so related because of the difference
`in elbow angles of the Fabs; (7) the dyad of the Fe is at an
`angle of about 120° with that dyad relating constant Fab
`domains; (8) the Fe 2-fold axis does not intersect the dyad
`relating the constant domains of the Fabs, nor does it intersect
`the approximate long axis of the Fabs; (9) in the crystal, all
`segments share extensive interfaces which severely restrict the
`dispositions of neighbours. The contacts, illustrated in Fig. 3,
`presumably stabilize this particular conformation.
`The asymmetric conformation, observed in these crystals of
`the antitumour antibody, should probably not be considered as
`a static structure which is maintained in solution. The structure
`probably represents only one of many possible transient confor(cid:173)
`mations. The unique structure is a product of the intrinsic
`flexibility of the antibody and the lattice interactions that stabil(cid:173)
`ize this particular distribution of domains. Indeed, electron
`23
`, fluorescence polarization24
`25 and previous X(cid:173)
`microscopy20
`-
`'
`ray crystallographic studies5·26·27 have provided extensive
`evidence for a wide range of conformations based on segmental
`flexibility.
`The structure we present is instructive in that it illustrates the
`nature and extent of this structural variability, or dynamic range,
`which is inherent in the antibody. The Fabs are loosely tethered
`to a mobile Fe. Each Fab can assume its own elbow angle as
`its environment or function requires. Somewhat unexpected is
`the fact that, were the elbow angles the same, the Fabs would
`be related by an almost exact 2-fold axis that is quite independent
`of the Fe. It is the disposition of the Fe that disrupts the overall
`symmetry of the molecule. This is in keeping with the disorder,
`or multiple orientations of the Fe, observed in the Kol antibody
`structure7·8.
`The hinge polypeptides are not really hinges, but rather they
`are tethers that allow the Fab components to drift from the Fe
`to bind antigen or potentially allow the Fe to move in such a
`way to trigger effector functions, such as the activation of com(cid:173)
`plement25·28. The connecting polypeptides give the Fabs the
`freedom to move and twist so as to align hypervariable regions
`with antigenic sites on large, immobile carriers, in this case
`tumour cells. The crystal structure visually demonstrates that
`the antibody is an assembly of units possessing a high degree
`of flexibility, a molecule suited to the task of scavenging foreign
`0
`objects or activating a cell lysis system.
`
`Received 3 August accepted 9 October 1992.
`
`1. Wilson, I. A., Rini, J.M., Fremont, D. H., Fieser, G. G. & Stura, E. A. Meth. Enzym. 203, 153-176 (1991).
`2. Alzari, P. M., Lascombe, M.·B. & Poljak, R. J. A. Rev. lmmun. &, 555-580 (1988).
`3. Davies, D.R., Padlan, E. A. & Sheriff, S. A. Rev. Biochem. 59, 439-473 (1990).
`4. Deisenhofer, J. Biochemistry 20, 2361-2370 (1981).
`5. Silverton, E. W., Navia, M. A. & Davies, D. R. Proc. natn. Acad. Sci. U.S.A. 74, 5140-5144 (1977).
`6. Rajan, S. S. et al. Mo/. lmmun. 20, 797-799 (1983).
`7. Colman, P. M., Deisenhofer. J., Huber, R. & Palm, W. 1 molec. Biol. 100, 257~-282 (1976).
`8. Marquart, M., Deisenhofer. J., Huber, R. & Palm, W. 1 molec. Biol. 141, 369-391 (1980).
`9. Steplewski, z .. Jeglum, K. A., Rosales, C. & Weintraub, N. Cancer /mmun. lmmunother. 24, 197 -201
`(1987).
`10. Rosales, C., Jeglum, K. A., Obroci<a, M. & Steplewski, Z. Cell. lmmun. 115, 420-428 (1988).
`11. Jeglum, K. A. Proc. Seventh ACVIM Forum, San Diego (922-925) (Lippincott, Hagerstown, Maryland,
`1989).
`12. Larson, S., Day, J.,Greenwood, A., Skaletsky, E. & McPherson, A.1 molec Biol. 222, 17-19 (1991).
`13. Fitzgerald, P. M. D. 1 app/, Crystallogr. 21, 273-276 (1988).
`14. Brunger, A. T., Kuriyan, J. & Karplus, M. Science 235, 458-460 (1987).
`15. Brunger, A. T. A. Rev. phys. Chem. 42, 197-223 (1991).
`16. Bernstein, F. C. et al. 1 molec. Biol. 112, 535-542 (1977).
`17. Crowther, R. A. in The Molecular Replacement Method (ed. Rossman, M. G.) 173-178 (Gordon
`and Breach, New York, 1972).
`18. Sheriff, S. et al. Proc. natn. Acad. Sci. U.S.A. 64, 8075-8079 (1987).
`19. Satow, Y., Cohen, G. H., Padlan. E. A. & Davies, D. R. J. mo/ec. Biol. 190, 593-604 (1986).
`20. Wrigley, N. G., Brown, E. 8 & Skehel, J. J 1 molec Biol.169, 771-774 (1983)
`21. Roux, K. H. Eur. 1 lmmun. 14, 459-464 (1984).
`22. Lamy, J, et al. Biochemistry 24, 5532-5542 (1985).
`
`372
`
`23. Wade, R. H., Taveau, J. C. & Lamy, J. N. 1 molec. Biol. 206, 349-356 (1989).
`24. Schneider, W. P., Wensel, T. G., Stryer, L. & Oi, V. T. Proc. natn. Acad. Sci. U.S.A. 85, 2509-2513
`(1988).
`25. Dangl, J. L. et al. EMBO 11,1989-1994 (1988).
`26. Huber. R., Deisenhofer. J., Colman, P. M. & Matsushima, M. Nature 264, 415-420 (1976).
`27. Amzel, L. M. & Poljak, R. J. A. Rev. Biochem. 48, 961-997 (1979).
`28. Oi, V. T. et al. Nature 301, 136-140 (1984).
`29. Jones. T. A. in Computational Crystallography (ed. Sayre, D.) 307 -317 (Oxford University Press,
`New Yori<, 1982).
`
`ACKNOWLEDGEMENTS. This research was supported by grants from the NIH. from the NSF and from
`NASA, as well as by the lmmunopharmaceutics Company of San Diego. We thank the San Diego
`Supercomputer Facillty for time on the Cray Y-MP.
`
`RETRACTION
`
`Identification by anti-idiotype
`antibodies of an intracellular
`membrane protein that recognizes a
`mammalian endoplasmic reticulum
`retention signal
`D. Vaux, J. Tooze & S. Fuller
`
`Nature 345, 495-502 (1990)
`
`OuR further characterization of the M, 72,000 (72K) protein
`has shown that the data in Fig. 2 of our paper are erroneous
`and not repeatable. We retract the statement that the 72K protein
`is an integral membrane protein. The present evidence is con(cid:173)
`sistent with this protein being associated with the intermediate
`compartment. We also withdraw our speculation concerning its
`0
`function.
`
`GUIDE TO AUTHORS
`
`PLEASE follo"' these guidelines so that your manuscript may he
`hand!edexpc<litinusly
`,'\'ar~rr i> an international journal covering all the sciences. Con(cid:173)
`lributors shollld th1'refore hear in mind tho~e readers who work in
`01her field> and \Rose for whom English is a second language, and
`writeclearlyandsimply,avoidingunnecessaryted1nica\1erminology
`Space 111 the jour~a! is limited, making cvmpetition for publication
`>e,·erc. Bre,·ity is ~ighlyvalued. One printed p;igeof Namre, without
`in1erruptions, contains about 1,300 wvrds.
`Manuscripts are selected for publication according to editorial
`ar;.1essment of their .suitability and repons from independent referees
`They can be se nt 10 London or Washington and should be addressed
`lotheEditor.Manusnipt•maybedealtwilhineitheroffice,depending
`on the >ubject matter, and wilt whne necessary he sent between offices
`hyovernightcourier.Highprioritycannotbegiventopresubmission
`em1uiries;inurge ntcasestheycanbemadeintheformofaone-page
`fn. All m~nuseriph are acknowledged on receipt but fewer than half
`aresentforreview.Thosethatarenotreviewedare retuniedasrapidly
`as pqssible SO that they may be submitted elsewhere without delay
`Contributors may suggest reviewen; limited requests for the exclusion
`ofspl:cillcreviewersareusuallyhceded_Manuscriptsareusuallysent
`to two or thru re•·iewer~, chosen for their expertise rather than th eir
`geographica!loca1ion.Manuscriptsacceptedforpublicationarety~­
`~el from the London office
`Nawrerequemauthorstodep<:isitsequenceandcrystallographic
`data in the databases thal exist for this purpose, and to mention
`av;;iilabilityofthesedata
`Once a manuscript is accepted for publication, contributors will
`receive proofs in about 4wee ks. Nalflre's•taffwilledit manuscripls
`wi!h a view to hrevily and clarity, so contributor. should check proofs
`carefuUy. Manuscriptsaregcnerallypublished2-3weeksafterreceipt
`ofcorrectedprools.Nar .. redoesnotexactpagecharges.Conlributors
`receive a reprint order form with their proofs: reprint order.; are
`proce>sed aft er the manuscrip1 i1 puhlished and payment recei,·ed
`
`categories of paper
`Revl- .tides sllrvey recent developments in a field. Mo;\ are com(cid:173)
`missioned but su~estions are welcome in the form of a one-page
`synopsis addressed to the ReviewsCoordinarnr. length is negotiable
`in advance
`Artldes ;ire research report.I whose conclusions are of general interest
`and which arc sufficiently roi.lnded to he a sub.1tantial advance in
`understanding. The y should not have more than 3,000 words of text
`(not including figure legend•) or more than six <fop!ay items and
`;houldnotoc<'upymorcth<infivepagesof NulflrC.
`Articles start with a heading of 50-!W words written 10 adv ertise
`th eir con1.,nt in general terms, to which editors will pay particular
`auention. The heading doe~ not u'ually contain numbers, abbrev i(cid:173)
`ations or mea>urements. The introduction to the study iscon!ll.\ned
`in the first two w three paragraphs of the art1de, which al•o briefly
`summarize i[.S rewlts and •mplications. Arlides have fewer 1han ~()
`reference• and rruiy contain a re .. - short subheadings
`l.ttel't are short reports of outs1anding n,wel findings ,.·hose implica(cid:173)
`tions are general and important enough to be ofinteres\ to tho;;c
`outside tho:: field. Letters should ha~e l.OOOorfewcrwordsoftextand
`follrorfcwerdisp!ayilcms.Thcfirstparagraphde.~cribes,innotmore
`than t50wordsandwithoutchcuseofabhreviat1on.1,thebackgmund,
`ratiunale and chief conclusion• of the 'rndy for the particlllar t>enefit
`of non.r;pccialls1 reader•. L<:ttcr.1 do not have ~uhheadmg•and .•hould
`con1ainfrwer1han 30refcrcnces
`~yrid<ltdea! with issue> in, or arising from, r~search tha!
`arc a!'o of inleresl to reade rs outside reseirn;h. Some are commissioned
`but suggestions can be made to the Commentary Editor in the form
`<>fanne-f>il!IXWnOp'>i'
`
`Newwt ..i VI-• .tlclet inform non-specialist reader.1 about new
`scientific ~dvances, sometimes in the form of a conference rep<:irt
`Mostarecommissionedbutprop<:isalscanhemadeinadvancetothe
`News and Views Edi1or
`Scl.ntlflcComl~is for discussion oftopic~l.1cientificm;;i!\ers,
`including those published in Nalflre,andformisceUaneouscontribu·
`lions. Priority i• given to le\ler>of fewer than 500 words
`Preparation of manuscripts
`All manu1cripts.1hould he typed, douhle-sp aced,on onesideof1hc
`paper only. An original and four copies are required, each accom(cid:173)
`panied by artwork. If photographs aro::included, five sets of originals
`are required; for line drawing•, one •Cl of origmals and four good(cid:173)
`quality photocopies are acceptable Reference lists. figure legends and
`lllhle.; shoold all be on sepa rate sheets, all or which should be doubl e(cid:173)
`sp.ac"d and numbered. Three copies of re levam manu~cripts in press
`orsubmittedforpublicatione!sewhcrcshouldbeincludedwitheach
`copy of a submiued manu~cripl, and clearly marked as such. Five
`copies of revised and rnubmilted manuscripts, l;ibelled with their
`manu.1crip1riumbersarcrequircd,logetherwithfivecopiesofaleHcr
`detailing the changes made
`Tttletarehriefandsimple. Activcverbs,nilmerical valL!es,abbrevi(cid:173)
`atjonsand punctuation are to be avoided. Titles should contain one
`or two key words for indexing purposes
`Artworll should be marked individually and clearly with the author's
`name arid, when known, the manu,cript number. Ideally, no figure
`should be larger than 2~ by 22crn. Figures with severa l parts a re to
`beavoidctl and are permitted only if the partsarec!osely related,
`eitherexperimentallyorlogica!!y.Unlelteredorig;nalsofphotographs
`should be provided.Suggestions forcnverillustrations,withcaplionr;
`and labelled with the manuscript number, are welcome. Original
`artworkisretumerlwhenamanuscriptcannotbepublished
`Protein.lm1deotidesequencesshould ideallybeinthelhree-letter
`and not the sing\e-le1tcr code for amino acids. One column width of
`Na1urtc;inaccommodate20aminoacidsor60bascpairs
`COlow .rtwortL A charge of £500 P"f page is made as a contribution
`10wardsthc cost of reproducing colour figures. lnahility to pay these
`costs wil! not prevent publica1ion of essential colour figures ir the
`circumstances are explained. Proofs of colvur artwork may be sent to
`contributorsundersepmatecovcrfrom th eirga!!eyproors
`flcure fe&9fld• should not exceed}()() words and ideally should be
`shorter. The figure is described first,then.briefly,lhe method. Refer·
`encetoamethodpub\ishedelsewhereispreferabletoafulldescription
`Mcthods1renotdescrihedinthete.xt
`~•arenumberedsequentiallyastheyapP"<Hinthetut,
`fo\!owedbythuseintabl es and finallyhythoseinfigurelegends. Only
`papers published or in the press are numhered and included in the
`referenceli;t.Allotherformsofrefercnceshouldbeci1edinthetexl
`asapersonalcommunication,manuscriptsubmille<Jorinpreparation
`Text is not included in reference li•t•- Refere nces are ahhreviated
`according to the World Li.H of Sdemific Pmodica/J (Bunerworths,
`London, 1963-65). The first and last page numbers are i11cluded:
`referenc.: (O hooks sh<1uld include publisher, place and date
`Abbrevlllllon1, symbols, un•ts and greck !e1tcrs should be identified
`tbe firot time they are used. Acronyms ~hould he avoided when<over
`po~siblc and, if used, defined. Footnotes are not used in th<o 1ext
`Acknowledr;Mnmlll are hrief ~nd app;:ar after the reference li.1t; gr;;int
`andcon11ihution numbers are not allowed
`Supplementary Information i~ material rele vant to Articles or l~tten
`which c~~not. fM lack. of .1pa~'c, be publi.,hed in full, but which '~
`a•·ailablefrom Nuwreon request
`Subml11ion. Manuscripc• can he •ent to the Editor ac 4 !.iule Es;e•
`Street, London WC2R 3LF, LK or at 1234 National Prcs.1 Building,
`Washingtcm. DC 20045, CSA. Manuscripts or proof> •ent hy air
`courier lo London ohould be declared a.1 ·m~nuscrip1';' and 'value SS
`to prevenl 1he imrmi1inn nf impnrl dut1· and •alue-addcd lax
`
`NATURE · VOL 360 · 26 NOVEMBER 1992
`
`(cid:141) (cid:21)(cid:29)(cid:29)(cid:22) (cid:50)(cid:69)(cid:88)(cid:89)(cid:86)(cid:73)(cid:4)(cid:4)(cid:52)(cid:89)(cid:70)(cid:80)(cid:77)(cid:87)(cid:76)(cid:77)(cid:82)(cid:75) (cid:43)(cid:86)(cid:83)(cid:89)(cid:84)
`
`4 of 4
`
`BI Exhibit 1108
`
`