Proceedings of the I1JATIONALACADEMY OF SCIENCES voiLune 47 - Number 9 - September 15,1961 THE KISETICS OF FORMATION OF NATIVE RIBONUCLEASE DCRISG OXIDATION OF THE REDUCED POLYPEPTIDE CHAIN BY C. B. AXFIXSEN, E. HABER,* M. SELA,~ AND F. HI WHITE, JR. LABORAMRT OF CELLFLAB PEYSIOLOQY MD METABOLISM, NATIONAL EEABT INSTITUTE, NATXONN. INSTITUTES OF EIEJUXE Communicated &y John T. Edsall, July S&1981 Bovine pancreatic ribonuclease is completely reduced by treatment with mer- captoethanol in 8 M urea to-yield a randomIy coiled poIypeptide chain containing eight cysteine residues.`-) Under optimal condit.ions of polypeptide concentration and pH, essent~inlly complete reformation of the disulfide bonds of the native enzyme occurs in the presence of molecular oxygen.** t From chemical and physical studies of the reformed enzyme, it may be concluded that the information for the correct pairing of half-cystine residues in disulfide linkage, and for the assumption of the nat.ive secondary and tertiary structures, is contained in the amino acid sequence itself. Preliminary to studies on the interactions involved in the refolding proceq and to establish the order of chemical events during the formation of active protein, we have followed the rates of disappearance of sulfhydryl groups, and of the appearance of the spectral properties characteristic of the native enzyme and its active derivatives. The appearance of increased positive optical rotation asso- ciated with secondary structure was also studied. The results rule out the se- quential formation of one active molecule after another. They suggest as a major possibility t.hat some disulfide bonds formed during the early stages of oxidation are not identical with those of the native protein but undergo rearrangement to yield the native configuration. .Ma&ria& and Melho&--Bovine pancreatic ribonuclease (RSaae) was purehaxd from the .Sigms Chemical Company, St. Louis, Missouri. Test samples of the lot employed in these studies ( jrR6OB-LXM. "chromatographic .grade") we:e subjected, before use, to chromatography4 on the cation exchanger, carbo~methylcellulos~ (Brown Paper Company, Lot 1069; 0.6 meq&n) and on IRC-50,` and were found to consist almost entirely of the major "A" peak with a smaller peak (approximately 55: of the major peak) running in the position of t,he usual "B" component .Ymino acid anal+ according to the procedure of Spackman, Stein and Moore' yielded values in good agreement with those reported for the commercial preparations that have been used for most of the structurtll studies on this enzyme. Reduced RSase was prepared by treatment of the native enzyme with mercaptoethanol in 8 Y urea followed by separation of the reduced chain from reagents on SephadexG25 (Pharmaria, rppsala. Lot 6-M) as described earlier.* The procedure employed for oxidation of fully reduced RSase consisted of adjustment of the cwcentration of the solution of reduced protein in 0.1 .M acetic acid that emerged from the 1309 1310 BIOCHEMISTRI-: AA-FI.l-SE.?' ET AL. PFUX. s. -4. P. Sephadex-G25 columns uith water to the desired protein concentration, followed by adjustment. to pH 8.2 with tris (hydroxymethyl) aminomethane (trip)-acetate buffer. Three large-scale oxidations were carried out, tao at 2.0 mg/ml and one at 0.1 mg/ml. Oxidation in t,wo experi- ments xu allowed to proceed at room temperatu.re (2%24oC) in open beakers without stirring. In a third esperiment (at 2.0 mg/ml). slow stirring was employed, without detectable differences in results except for a slight iocrea+e in the amount of turbidity produced. The soluble protein concentration, after as long as 24 hr, was still 95:; or more of the original concentration. During oxidat.ion, and particularly frequently during the early stages, samples were withdrawn, diluted to the desired concentrations with 0.W acetate buffer, pH 5.0, and assayed* for enzyme activity against ribonucleic acid prepared according to Crestfield, Smith, and Allen,O and against uridglic- 2',3'-cyclir phosphate (barium salt, &%hwarz BioResearch). The latter assays were carried out with the Gary Recording ultraviolet spectrophotometer as previously described.*. `0 Each analy- sis was compared with simultaneous controls run on native RSase golut,ions. The Cary spectrophotometer was also employed for the serial determination of the absorption spectra of the solutions, with the speed of reading adjusted to yield curves that permitted accurate estimation of the differences in absorbancy at various wavelengths. Protein concentrations were estimated from the same spectra at 275 mp where R?;ase and reduced R?;ase have the same molar extinction (E = 9,200). Opt.ieal rotatory measurements were made with a Rudolph precision ultraviolet polarimeter, model 80, equipped a&h the Rudolph photometric polarimeter attachment and an oscillating polarizer prism. The measurements were carried out at a aavelength of 366 rnp, using a mercury lamp as the light source. At this wavelength, the changes in specific optical rot&ion were suffi- cient.ly large to permit reasonable accuracy with 20 cm polarimetric tubes even with relatively di- lute solutions of protein (2 mg/ml). The specific optical rotation for native RSase at t.his wave length was - 288", and for fully reduced RXase, - 372'. In one experiment, where oxidation was carried out on a solution of reduced RX- at a concentration of 0.1 mg/mI, samples of the reoxi- dat,ion mixture were taken at various times, acidified with acetic acid to approximately pH 4, and lyophilized. The resulting material, when diiolved in water at a concentration of 2 mg,/ml, gave optical rotations in excellent agreement with corresponding samples taken at similar times of oxidation in other experimenti at higher protein concentrations. Most samples taken during the latter stages of oxidation were somevmhat turbid and required brief centrifugat.ion at approxi- mately 15,CKJO g before measurement. The extent of reduction and the kinetica of SH diaappeerance during oxidation were determined both by titration with pchloromercuribenaoaW' and by reaction with I-C&iodoacetic acid of known specific radioactivity followed by determination of radioactivity on the alkylated pr+ t&n.* Aliquota of the protein solution (approximately 2 mg protein in 1.0 ml were added to 1.0 ml 1.0 M t&-acetate buffer, pH 8.5 containing 10 pmoles C,, iodoacetic acid (0.2 rcurie) and the alkylation was allowed to proceed for 10 min. After thii time, excess iodoacetic acid was immobii by the addition of an excess (5 d) of mercaptoethanol. After reaction for 15 min, t.he alkylated protein was separated from the various reagents, by passage of the sample through a (1 X 40) cm column of Sephadex G25 in 0.1 M acetic acid, and the radioactivity was subsequently counted in the Packard Scintillation counter in the presence of a water-miscible phosphor." Pro- tein concentration in the samples counted was calculated from absorbancy measurements at 275 mu. Results.-The data in Figure 1 summarize the changes in the sulfhydryl group content, specific apt,ical rotation and enzyme activity during oxidation of t,he SH groups of fully reduced RNase under the conditions presented in the experimental section. Final values obtained, expressed as percentages of the differences between fully reduced RXase and native RSaze, are shown at the right. of the figure. These values indicate essentially complete restoration of the nat.ive properties. A final value for t.he activity oft.he oxidation mixture toward the synt.hetic substrate is not given since this value was quite variable (generally 60-iO% of the native level) due, possibly, to inhibition by components of the solution. - This possibility hris not been investigated as part of the present study. It. has been found, however, VOL. 17, 1961 BIOCHEMISTRY: AiVFINSEiV ET AL. 1311 that on chromatography of samples of reformed RSase, the major component is identical to that present in commercial samples of the enzyme and possesses completely normal activity toward both RXA and synt.hetic subskates.* The most striking phenomenon observed in these studies is the marked lag phase before enzymatic activity appears, during which &&d the sulfhydryl titer and the specific optical rotation change along a curve similar to that of a first-order reaction. The presence of this lag phase (various explanations for which are offered in the discussion) immediately rules out the possibility that disulfide bonds are formed io such a way that detectable amounts of complete, native molecules are produced in a one-by-one fashion. 2 .( z 60- a I H & 40- g r'."b/- 100 206 300 400 ii30 600. i200 1 I TiME OF REOXIDATION (MINUTES) es, during the oxidation of reduced ribonuclease, in titration with pchloromercuribenaoate ( o ) Changes in spectral properties during oxidation (from one typic&' &riment) are summarized in Table 1 and conform, in general, to earlier o&serv&o&s on the relationship between activity and spectral properties. Sat% R&se L&O&S an absor&ion maximum at S27i.3 rnF, which is also shown by,.`such &ve deriva- tives", 1( as RSase-S, carboxypeptidase-treated RSase rind I&&me w&i& has been modified by addition of polyalanine side chains of mGider&le length on available amino groups. Is On the other hand, inactive derivatives still posstkiog intact SS bonds (pepsin-inactivated RSase, methylated R&au, eta.) have shown a maximum at 56 mp.la. Ld Both oxidized and reduced RSsse show still another absorption mnsimum, at 5.3 mb, which appears to be characteristic of the chain devoid of disutfide cross-links. - , In the present studies (see Table 1) & positi df the mnximum'&anges from 275 rnr to 276 mp during the lag period p&r to the'%ppearance of en i _ T&T, tiv$y, ti% and then increases roughly in parallel with the appearance of enzym ~1 %a 1312 BIOCHEMISTRY: ANFIKSEN ET AL. Paoc. N. A. s. ThI?, min 720 177 gj 800 Kative RXase TABLE 1 SPEerau. CECAKGES Wmd?nptb of maximum absorption, mp 275.4 255.0 Et.; 27614 276.8 277.5 61181 value of 277.5 mp. The presence of a maximum at 276 LIP is, incidentally, aLso characteristic of the randomly cross-linked derivative obtained during oxidation in urea, guanidine, and other disorienting reagents'* and may be characteristic of tyrosine interactions of a nonspecific sort, occurring w-hen the specific interaction of the phenolic hydroxyl groups is prevented. The reaction was also followed at 287 W, as &is is the peak of the difference spectrum between R&se and inactive RSase derivatives. Although the data are not included in Figure 1, a direct correlation becomes evident between change in activity and change in molar extinction at 287 m~1 when these parameters are plotted as functionsof the time of oxidation, assuming as final and init,ial values the absorptions at 287 nqr of nat.ive Rh'ase and of the inactive form of the enzyme with %hifted spectrum," having a maximum at 276 rnp. &cuss&.-During the formation of native RKase from the reduced chain, sulfhydryl groups might be converted to disulfide bonds by one of two general mechanisms, t.he first involving the i&al pairing of the correct half-cystine residues and the second, of random pairing with subsequent reshuffling to yield the nat.ive arrangement. A. Correct pairing: (1) The process might proceed by a "on&y-one" nwchuniam in which each molecuIe of reduced RNase is rapidly oxidized to the native, active form without the accumulation of significant quantities of partial oxidat,ion pro+ ucts. The formation of the first bond would, in this system, greatly &ease the likelihood that a second bond would form in the same molecule over the probabilit,y that a first bond would form in another molecule. (2) The rates of formation of successive &u&de bonds might be suEGently different that, at any time, the ntly of me kind; e.g., one SS bond with . Although the correct yme activity would first the f.inal, fourth, bond (if all four are indeed anism No. 2 would involve rates of oxidation are essenttiy llre fame. This variation would ally oxidized molecules; uctidy would possibly earlier, should some molecules possess activity. (4) A situation might be bonds has been formed but of Y 2 orwct secondrrry and d "native" spectral nt of noncovalent interactions to VOL. 47, 1961 BIOCHE~UISTRY: A~FINSE?~ ET AL. 1313 B. Random pairing: Pairing might be random in the initial stages of reaction and activity might first appear after reshtiing, through d&tide interchange, to yield the native configui=ation. The results presented in this paper make it possible' to rule out mechanism Al since only traces of enzyme activity and of the native spectral characteristics have appeared even after the disappearance of approximately half of t.he total SH group content of the reduced protein. Alternatives A2 and A3 are somewhat more difficult to rule out, particularly without a careful examination of the chemical nature of the disulfide bridges that have formed at various times by techniques similar to those used in establishing the disulfide bridges of the native enzyme.16 However, on a purely kinetic basis, these mechanisms seem unlikely since, when three-fourths of the SH groups have disappeared, enzyme activity has risen to more than half of t.he theoretical value. Mechanisms A2 and A3 would require that no activity appear before three of the four final SS bonds were complete in all the molecules present (unless partially oxidized molecules with only three SS bonds reformed possess some activity). Mechanism A4 permits a lag in the appearance of activity even if the conditions of mechanism Al obtain. Nevertheless, this mechanism is somewhat improbable on the basis of the known characteristics of rearrangement of the secondary and tertiary structures of native RSase after distortion in solutions of 8 M urea. It has been shown, for example, that the modifications in spectral and optical rotatory properties of RSase occurring in urea solutions are entirely reversible upon dilution to lower urea concentrations or upon dialysis.14* I' Further, the time required for reversal is relatively short, and certainly only a small fraction of the time involved in the 1-2-hour lag period found in the present experiments. The rapid refolding of the urea-distorted configuration of native RSase by polyanions" (presumably including ribonucleic acid itself) yields an active enzyme, also mili- tates against aberrant three dimensional structure as a major rea.son for t.he exist- ence of a lag phase in adtivity in the present esperiments on the reformation of SS bonds. Mechanism B is consistent with the kinetic data presented here and becomes even more probable when other available informat,ion is taken into account. .Ls mentioned above, oxidation of the SH groups of reduced RXase may be carried out in the presence of various agents t.hat cause the formaCon of enzymatically inactive molecules, for which the typical spectrum of RSase, with a maximum at 277.3 rnp, is not obtained. When such randomly cross-linked molecules are incubated under the conditions employed for oxidations as described ti the present paper, no detectable changes occur. Hoffever, the addition of SH compounds (including reduced RSase) under conditions known to favor disulfide interchange,IB induces the rearrangement of the molecule to yield an active pr+uct in high yield, possess- ing the physical properties of native RSa.se.lg The requirement. for SH catalysis supports the idea that the inactive materials are inactive because of random pairing of half-cystine residues and that, in the case of the lag in appearance of enzyme acti\-ity in the present experiments, a similar situation obtains. It is tentatively concluded, therefore, that oxidgtion of SH groups in this system occurs initially through relatively random formation of SS bonds with subsequent rearrangement taking place under the influence of disuffide interchange driven by 1314 BIOGHEhfISTRY: ARNOh- ET AL. PROC. N. A. S. thermodynamic forces toward the most probable form? native ribonuclease. Some of the less likely possibilities mentioned above can only be rigorously excluded upon completion of current experiments on the nature of the pairing of half-cyst.ine residues during t,he lag phase. The authors wish to thank Mrs. Juanita Cooke for her expert assistance in many of these experiments. * Present address: Ma9sarhusetts General Hospital, Boston 14, Massachusetts. t On leave of absence from Tbe Weizmann Institute of S&ence, Rehovoth, Israel. 1 Sela, M., F. H. White, Jr., and C. B. .4nfinsen, Biochim. et Btiphys. .4cta, 31,417 (1959). 0 White, F. H., Jr., .I. Biol. Chem., 236,1353 (1961). 1 Anfinsen, C. B., and E. Haber, ibirl., 236,1361(1961). 4 Aqvist, S. E.G., and C. B. Anlinsen, ibid., 234,1112 (1959). 6 Peterson, E. .4., and H. -4. Sober, J. .4m. Chem. Sot., 78, i51 (1956). 6 Hirs, C. H. u'., 8. Moore, and W. H. Stein, .I. Bid. Chem., 200,493 (1953). 7 Spa&man, D. H., W. H. Stein, and S. Moore, .4nai. C'hem., 30,llQO (1958). * Anlinsen, C. B., R. R. Redfield, K. L. Choate, J. Page, and IV. R. Carroll, J. BioI. Chem., 207,201(1954). 9 Crestfield, -4. M., Smith, K. C., and F. W. Allen, ibid., 216, 185 (1956). 10 Richards, F. M., Compt. rd. tnw. lub. CarLdwg, Ser. Chim., 29,315 (1955). 11 Boyer, P. D., J. tlm. Chem. Sot., 76,4331(1954). 12 Bray, G. A., Anal. B&&-m., 1,279 (1960). Ia Sela, M., and C. B. Anfmsen, B&him. et Biophys. A&a, 24,229 (195i). fi( S&, -M., C. B. Anfinsen, and W. F. Harrington, ibid., 26,502 (1957). 1` Hater, E., M. Se& and C. B. Anfinsen, Federation Proc., 20, Part I, 215 ( 1961). 1` Jkperimental work is now in progress to determine the nature of the pairing of haIf-cystine residuea at various times during the early stages of reoxidation. Preliminary results indicate that pairing is quite random and that, "incorrectly" formed bonds are present. 17 Harrington, W. F., and J. A. Schellman, Compl. rend. trav. lab. Cadsberg, Ser. Chim., 30, 2t (1956). * Sluyterman, L. A. A., BGchim. d Biophys. Ada, 48,429 (1961). ** Haber, E., and C. B. Anfinsen, unpublished data. PHOTOSYNTHETIC PHOSPHORYLATION AND MOLECC:LAR OX YGEK* BY DAXIEL I. ARSON,~ hf. L~sAD.~, F. R. WN~TLEY, H. Y. TSEJIMOTO, D. 0. IV&L, AND A A. HORTOS DEPARTKENT OF CELL PHYSIOLOaY, UNIVERSITY OF CALIFORNIA, BERKELEY Oxygen and photosynthesis were first. linked about, 200 years ago when both were discovered almost simultaneously. The earliest. concept of photosynthesis was that. of planetary ventilat,ion in which illuminated plants exchanged CO* or "bad air" for 02 or "vital air" (see historical re\-iew-I). A mechanism for this gas exchange was proposed in 1796 by Ingenhousz.* Green plants, he suggested, absorb from "carbonic acid in the sunshine, the carbon, throwing out at that time the oxygen alone, and keeping the carbon to itself as nourishment." ". For over a hundred years afterward, the view that CO2 assimilat,ion always in- volved a liberation of oxygen gas was so firmly entrenched that it was even ex- tended to the dark CO, assimilation by chemosynt,hetic bacteria.*+ The idea