At neutral pH, the circular dichroism (CD) and optical rotatory dispersion (ORD) of oc-crystallin between 200 and 590 rnp are characteristic of the 0 conformation. The Moffitt parameter bo value obtained from the ORD between 313 and 589 rnp also indicates the absence of a-helix. Most of the data suggest that the /3 form may represent from 35 to 63% of the protein polypeptide chain with the remainder in random conformation. cz-Crystallin deaggregates to an i& of approximately 7 x lo4 when the pH

more »

... is raised from 8 to 11.7 (1). The corresponding CD and ORD indicate an increase in random structure as well as the appearance of approximately 5% a-helix. Reaggregation occurs when the pH is returned to pH 8.0 (1). The 216 rnp CD negative band of the 0 structure reappears and the ORD spectra are also similar to the original preparations although certain small irreversible changes were noted. One of t.he major structural proteins in the lens, cz-crystallin, has a weight average molecular weight, M" of 1 x lo6 and has been shown to be composed of aggregates of subunits with an M, of 20,000 held together by noncovalent forces (1). In order to gain a greater insight into structure of the ar-crystallin aggregate and the changes in conformation which occur as a result of deaggregation at alkaline pH (l),' a study of the rotatory dispersion and circular dichroism of this protein has been undertaken. Shown in Fig. 1 is the CD2 spectrum of c+ crystallin between 200 and 250 rnp at pH 8.0. The spectrum is similar to that found for poly-L-lysine heated at pH 11 (4) indicating a single negative band with a minimum at 216 mp. This spectrum is characteristic of the pleated sheet fi structure of the polypeptide chain (3-8). The mean reduced molar ellipticity at 216 mp, 0'216 for ar-crystallin is approximately -2,940 deg cm2 per decimole, comparable to values found for y-globulin and al-acid-glycoprotein by Sarkar and Doty (4). These workers interpreted data similar to ours to beindicative of p structure. The absence of minima at 208 rnp and 223 rnp suggests that the a-helical form is not present. Although the dichroism below 200 rnp is not reliable enough to warrant presentation, it clearly becomes positive with a peak around 195 rnp. This interpretation is supported by the ORD data shown in Fig. 2 which indicate a positive Cotton effect centered around 198 mp. According to the theoretical calculations of Pysh (5) the strong positive band at 195 to 198 rnp is attributable to the antiparallel arrangement of the p conformation, However, on the basis of the recent work of Rosenheck and Sommer (6) , the data can be interpreted to reflect either parallel or a mixture of parallel and antiparallel arrangements. The ORD spectrum of ac-crystallin between 198 and 250 rnp at pH 8.0 (Fig. 2) indicates a peak at 204 rnp and a trough at around 230 rnp. This data is again comparable to the result obtained with y-globulin, cY1-acid-glycoprotein (4, 9) as well as with silk fibroin (8) and human carbonic anhydrases (10, 11). These characteristics are probably caused by the a -7r* and n -?r* transitions of the p structure in the polypeptide backbone (5, 6). Although there are 5 tyrosyl and 2 tryptophanyl residues in oc-crystallinl examination of the ORD between 250 and 280 rnp does not reveal any bands comparable in magnitude to those observed below 250 mp. Further studies are t,herefore required before useful conclusions can be made concerning possible contributions of these chromophores. The ORD of oc-crystallin between 313 and 589 rnp at pH 8.0 can be fitted reasonably well into the Moffitt equation (12) , assuming X0 = 212 rnp. The parameters a0 and b. were found to be -290 + 20 and 6 f 15, respectively. The bo value suggests that there is very little or no a-helix in the backbone and is in agreement with the absence of the two-band feature in the CD of the protein3 Although the general features of the ORD and CD of o(crystallin are in excellent agreement with those of the p form of poly-L-lysine, the magnitudes of the Cotton effects differ considerably. For instance, the 0'216 = -2,940 for the protein while that for poly-r-lysine heated at pH 11 is -14,400. Likewise, m'230 = -3,500 and rn'205 = +6,500 for the protein, whereas they are -6,300 and $23,000, respectively, for the model compound. Thus, the ratio of the absolute values of m'205:m'230 for the protein (6,500:3,500) is different from that of the poly-L-lysine heated at pH 11 (23,000:6,300). This discrepancy can possibly be accounted for by assuming contributions by asymmetrical side chain chromophores. However, the ORD of poly-r-lysine and lysine-tyrosine copolymers (4) are almost identical, both in the ultraviolet and visible regions. Furthermore, the work with L-cysteine methyl ester and Xmethylcysteine and 20 L-amino acids (13, 14) suggests that Cotton effects in the 220 rnp region are attributable to the car-boxy1 groups. Recently, Rosenberg (15) has shown that the rotatory strength associated with the aromatic transition is lOO-fold or more weaker than the R -a* transition of the peptide bond. Thus, it is unlikely that asymmetrical side chain interactions are contributing significantly to the optical activity of the protein, reported in this communication. It should also be noted that ar-crystallinl contains approximately 0.5% car-3Similar conclusions have been reached on the basis of rotatory dispersion and infrared studies upon wxystallin by Dr. Elton Katz (personal communication). 3234 by guest on March 22, 2020 http://www.jbc.org/ Downloaded from bohydrate.