CRYSTAL STRUCTURE OF L-CYSTINE HYDROCHLORIDE
Aug. 5, 1958
3835
would allow the disulfide exchange to occur more BSA recovered after 15 sec. heating is obviously rapidly than if the molecules were forced apart by different from the native protein. The presence of electrostatic repulsions. a . small amount of a dimer which was formed The following sequence may be postulated for the during the heating could explain the increase in the heat denaturation of BSA: (1) hydrogen bonds are frictional properties as were observed with viscosity. ruptured and rearranged, and the protein molecules It is difficult to say whether or not the dimer would aggregate and precipitate in the form of a meta- have a lower optical rotation and increased suscepstable polymer of protein molecules linked by tibility to digestion by trypsin. A complete hydrogen and hydrophobic bonds, ( 2 ) a 30 sec. lag separation of the monomer from the dimer would follows during which no extensive intermolecular be necessary to see whether the monomer of redisulfide exchange takes place and (3) after 30 sec. covered-BSA was identical to native BSB. averyrapid intermolecular disulfide exchange occurs. Markus and KarushI4 have found that 011e The addition of acetic acid will dissolve and disulfide group of human serum albumin is more disperse the polymer formed in (1) but will not susceptible to reduction than are the rest. A single break the covalent bonds of the polymer formed in group, more reactive than the rest in BSX, would ( 3 ) . Mercuric ions tie up the sulfhydryl group explain the appearance of the dimer before the and prevent the formation of the disulfide polymer rapid disulfide exchange had started and would in ( 3 ) , thus the coagulum remains soluble in acetic also explain the frequent occurrence of small acid. The structure of BSA is such that, when a amounts of the dimer in preparations of BSh. sdution of the material in (1) is lyophilized, the (14) C Markus a n d F Karush, THIS JOIJHNAI., 79, 131 (1057) protein molecules are able to return to something similar to the native state. SEATTLE. WASH.
[ CONTRIUCTION FROM
THE DEPARTMENTS O F
BIOCIIEMISTRY AND
ilNATOhZY O F TIIE UNIVERSITY O F \\’ASHIN C I O X ]
The Crystal Structure of L-Cystine Hydrochloride1 9 u L. K. STEINRAUF,~ JUANITA PETERSON AND L. H.
JENSEN
RECEIVED FEBRUARY 3, 1955 T h e crystal structurc of cystiiic hydrochloride has been solved by the method of superposition aiid refined h y two-tliiiiensional FO and i l F syntheses. The configuration of the cystinyl group is t h e same as t h a t in N,N’-diglycyl-L-cystitle dihydrate but different from t h a t in hexagonal cystine. Bond lengths and angles are near expected values anti the angle hetween the planes determined by SS’C’ and by S’SC is 79.2’.
The importance of disulfide linkages in protein structure is well known and has been further demonstrated in recent work.3 Since the cystinyl group occurs frequently in proteins and since the S-S interaction makes an important contribution to the vector space representation of proteins, its configuration is important in X-ray diffraction studies and in model building. It is therefore a matter of some importance to examine the configuration of the cystinyl group under a wide range of conditions. Uakel and Hughes4have determined the configuration of the cystinyl group in a tripeptide X,N’diglycyl-L-cystine dihydrate, and recently Oughton and Harrison have reported the structure of hexagonal ~ y s t i n e . ~The present work is an investigation of the configuration of L-cystine as it crystallizes from aqueous hydrochloric acid, Experimental Crystals of L-cystine hydrochloridc, HOOC-CH(NH2. HC1)-CHz-S-S-CH2-CH(NH2.HCl)-COOH, were prepared by allowing a hydrochloric acid solution to evaporate iu a desiccator over phosphorus pentoxide. From oscillation and U’eissenberg photographs of the needle-like crystals ( 1 ) Presented a t t h e 4 t h International Congress of t h e International Union of Crystallography, Montreal, C a n a d a , July, 1957. (2) G a t e s a n d Crellin Laboratories of Chemistry, California I n s t i t u t e of Technology, Pasadena, California. (3) L. K. Steinrauf, Thesis, University of Washington, 1057. (4) Yakel a n d Hughes, i i c l a C r y s t , 7 , 291 (1054). ( 5 ) B. M. Oughton a n d P. M. Harrison, ibzd., 10, 479 (lY.57).
the following unit cell parameters were tlcterniiried: a0 = 18.61 A., bo = 5.25 B . , cg = 7.23 A., 0 = 103.6”. Systematic extinctions indicate the space group to be Cm, C% or C2/m. The density observed by flotation is 1.5‘20 g. cm.-’, calculated, 1.515 g. assuming 2 molecules per unit cell. The intensities of the X-ray reflections were measured from unidimensionally integrated LVeissenberg photographs with a microdensitometer tracing at right angles to the direction of integration by the camera. The areas under the peaks of the photometric tracings were measured with a planimeter and were assumed to be proportional t o the integrated intensities.5 Results indicate a precision of 2-37‘ ill F,olmeasured by this method.7
Determination of the Structure The space groups Cm, C2 and C2/n1 cannot be distinguished by the systematic extinctions of the X-ray reflections. Since these three space groups have a t least fourfold general positions, the two molecules of L-cystine hydrochloride must lie in special positions and have some symmetry elenient of the space group. Since the molecule of Lcystine cannot have symmetry m or i, the only one of the three space groups compatible with possible molecular symmetry is C2.8 This agrees with that found by Srinivasang and by Corsmit, Schuyff and Feil,lo and is also the same as was found for N,N’(6) L. H. Jenscn, THISJOURNAL, 76, 4663 (1954). (7) L. H. Jensen a n d E. C. Lingafelter, “Am. Cryst. Assoc., Ab. stracts,” 1956, p. 21. (8) L. K. Steinrauf a n d I,. H. Jensen. A C ~ CYYSI.. Q 9, 539 (1‘35.6). (9) R. Srinivasan, ibid., 9 , 1039 (1056). (10) A. I?. Corsmit, A. Schuyff a n d D. Feil, P w c . R o y . Soc. Ainrlerdam, 69, 470 (1956).
3836
L. K.
STEINRAUF,
J.
PETERSON AND
diglycyl-L-cystine dihydrate by Yakel and hug he^.^ The projection of the structure on the xz-plane was solved by the method of vector convergence,as proposed by Beevers and Robertson.ll As noted above, space group considerations require the molecule to have a twofold symmetry axis. Therefore the S-S bond, known to be present, i n p t be parallel to the xz-plane. A length of 2.0 A was assumed for this bond, and its orientation immediately determined by inspection of the Patterson projection on (OlO), Fig. l a . C
C.
C
f (b)
Fig. l.-((a) f J ( x , zj, cviitours a t arbitrary equal iiitcrvals of vector density-. (b) Superposition function. Solid circles, coiirdinates assumed froin superimposing molecular niodcl ; opeii circles, final atomic positions.
% . superposition function was then calculated with the origin moved to each of the assumed S atom positions. The function calculated was m PIl1(X,Z)
=
P ( N , , ZJ) j = 1
=
1‘ j h
ni
[lEi,o,j2 I
c
j = l
jCUS27r(lLXj
+
ZZj)]]
where Pm(x,z) is the sum of the superpositions of wt identical Patterson functions with the origin a t the points (xlrel), (x2,z?), etc., and where m = 2 for the case in point. The result of this superposition is shown in Fig. l b . Although resolution is poor, the general form of the molecule is clear. A set of structure factors using the atomic positions indicated by the solid circles i n Fig. lb was calculated. The reliability index, I
