Principles That Govern the Folding of Protein Chains Author(S): Christian B

Principles that Govern the Folding of Protein Chains Author(s): Christian B. Anfinsen Source: Science, New Series, Vol. 181, No. 4096 (Jul. 20, 1973), pp. 223-230 Published by: American Association for the Advancement of Science Stable URL: http://www.jstor.org/stable/1736447 . Accessed: 16/10/2011 12:23

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

American Association for the Advancement of Science is collaborating with JSTOR to digitize, preserve and extend access to Science.

http://www.jstor.org 20 July 1973, Volume 181, Number 4096 ;CIENCE

exists under conditions similar to those for which it was selected-the so-called physiological state. After several years of study on the ribonuclease molecule it became clear Principles that Govern the to us, and to many others in the field of protein conformation, that proteins of Protein Chains devoid of restrictive disulfide bonds or Folding other covalent cross-linkages would make more convenient models for the Christian B. Anfinsen study of the thermodynamicand kinetic aspects of the nucleation, and subsequent pathways, of polypeptide chain folding. Much of what I will review deals with studies on the flexible and The telegram that I received from eight sulfhydryl groups to form four convenient staphylococcal nuclease the SwedishRoyal Academy of Sciences disulfide linkages take place.The origi- molecule, but I will first summarize specifically cites "... studies on ribo- nal observationsthat led to this conclu- some of the older background experi- nuclease, in particular the relationship sion were made together with my col- ments on bovine pancreaticribonuclease between the amino acid sequence and leagues Michael Sela and ]Fred White itself. the biologically active conformation. in 1956-1957 (3). These vvere, in ac- ..." The work that my colleagues and tuality, the beginnings of a long series I have carried out on the nature of of studies that rather vaguely aimed at Support for the the process that controls the folding of the eventual total synthesis of the pro- 'Thermodynamic Hypothesis" polypeptide chains into the unique tein. As we all know, Gutte and Merri- three-dimensionalstructures of proteins field (4) at the Rockefellrer Institute, An experiment that gave us a partic- was, indeed, strongly influenced by ob- and Ralph Hirschmanand his colleagues ular satisfaction in connection with servationson the ribonucleasemolecule. at the Merck Research Institute (5), the translation of information in the Many others, including Anson and have now accomplished this monumen- linear amino acid sequence into native Mirsky (1) in the 1930's and Lumry tal task. conformation involved the rearrange- and Eyring (2) in the 1950's, had ob- The studies on the renatturation of ment of so-called "scrambled"ribonu- served and discussed the reversibilityof fully denatured ribonucleaqse required clease (8). When the fully reduced pro- denaturationof proteins. However, the many supporting investigations (6-8) tein, with eight SH groups, is allowed to true elegance of this consequence of to establish,finally, the gene:rality which reoxidize under denaturing conditions natural selection was dramatized by we have occasionally called (9) the such as exist in a solution of 8 molar the ribonucleasework, since the refold- "thermodynamichypothesis " This hy- urea, a mixture of products is obtained ing of this molecule, after full denatura- pothesis states that the three-dimen- containing many or all of the possible tion by reductive cleavage of its four sional structure of a native3 protein in 105 isomeric disulfide bonded forms disulfide bonds (Fig. 1), required that its normal physiological miliieu (solvent, (schematicallyshown at the bottom right only 1 of the 105 possible pairings of pH, ionic strength, presencce of other of Fig. 2). This mixture is essentially components such as metal icons or pros- inactive-having on the order of 1 per- Copyright ? 1973 by the Nobel Foundation. thetic groups, temperature,and other) cent the activity of the native enzyme. The author is chief of the Laboratory of is the one in which the Gibbs free If the urea is removed and the Chemical Biology, National Institute of Arthritis, Metabolic and Digestive Diseases, National Insti- energy of the whole systenn is lowest; "scrambled" protein is exposed to a tutes of Health, Bethesda, Maryland 20014. This that that the native conformation is article is the lecture he delivered in Stockholm, is, small amount of a sulfhydryl group- Sweden, on 11 December 1972 when he received determined by the totalityr of inter- containing reagent such as mercapto- the Nobel Prize for Chemistry, a prize he shared with Stanford Moore and William H. Stein. It atomic interactions and hence by the ethanol, disulfide interchange takes is published here with the permission of the amino acid in a venenviron- and the mixture Nobel Foundation and will also be included in sequence, giv place, eventually is the complete volume of Les Prix Nobel en 1972 ment. In terms of naturaal selection converted into a homogeneous product, as well as in the series Nobel Lectures (in Eng- the of macJromolecules from lish) published by the Elsevier Publishing Com- through "design" indistinguishable native ribonu- pany, Amsterdam and New York. Dr. Moore's during evolution, this idea emphasized clease. This process is driven entirely and Dr. Stein's combined lecture appeared as a the fact that a single article in the 4 May issue of Science, page protein mo'lecule only by the free energy of conformationthat 458. makes stable, structural senise when it is gained in going to the stable, native 20 JULY 1973 223 structure. These experiments, inciden- discrepancy in rates would not have can be achieved. Model studies on tally, also make unlikely a process of been observed in the case of the fold- ribonucleasederivatives had shown that, obligatory, progressive folding during ing of structures that were not cross- when the intactness of the genetic mes- the elongation of the polypeptidechain, linked and, as discussed below, such sage representedby the linear sequence during biosynthesis, from the NH2- to motile proteins as staphylococcal nu- of the protein was tampered with by the COOH-terminus.The "scrambled" clease or myoglobin can undergo virtu- certain cleavages of the chain, or by protein appears to be essentially devoid ally complete renaturationin a few sec- deletions of amino acids at various of the various aspects of structuralregu- onds or less. points, the added disulfide interchange larity that characterizethe native mole- The disulfide interchange enzyme enzyme, in the course of its "probing," cule. subsequentlyserved as a useful tool for discovered this situation of thermody- A disturbingfactor in the kinetics of the examination of the thermodynamic namic instabilityand caused the random the process of renaturationof reduced stability of disulfide-bonded protein reshufflingof disulfide bonds with the ribonuclease,or of the "unscrambling" structures. This enzyme, having a mo- formation of an inactive cross-linked experiments described above, was the lecular weight of 42,000 and containing network of chains and chain fragments slowness of these processes, frequently three half-cystineresidues, one of which [see, for example (13)]. With two natur- hours in duration(7). It had been estab- must be in the SH form for activity ally occurring proteins, insulin and lished that the time required to syn- (12), appears to carry out its rearrang- chymotrypsin, the interchange enzyme thesize the chain of a protein like ing activities on a purely random basis. did, indeed, induce such a randomizing ribonuclease, containing 124 amino Thus, a protein whose disulfide Ibonds phenomenon (14). Chymotrypsin,con- acid residues, in the tissues of a higher have been deliberately broken and re- taining three disulfide-bondedchains, is organism would be approximately 2 formed in an incorrect way, need only known to be derived from a single- minutes (10). The discrepancybetween be exposed to the enzyme (with its es- chained precursor, chymotrypsinogen, the in vitro and in vivo rates led to the sential half-cystine residue in the pre- by excision of two internal bits of se- discovery of an enzyme system in the reduced, sulfhydryl form) and inter- quence. The elegant studies of Steiner endoplasmicreticulum of cells (particu- change of disulfide bonds occurs until and his colleagues subsequentlyshowed larly in those concerned with the secre- the native form of the protein substrate that insulin was also derived from a tion of extracellular, disulfide-bonded is reached. Presumably,disulfide bonds single-chained precursor, proinsulin proteins) which catalyzes the disulfide occupying solvent-exposed, or other (Fig. 3), which is converted to interchange reaction and which, when thermodynamically unfavorable posi- the two-chained form, in which added to solutions of reduced ribonu- tions, are constantly probed and pro- we normally find the active hor- clease or to protein containingrandom- gressively replaced by more favorable mone, by removal of a segment from ized disulfide bonds, catalyzed the half-cystine pairings, until the enzyme the middle of the precursorstrand after rapid formation of the correct, native can no longer contact bonds because of formation of the three disulfide bonds disulfide pairing in a period less than steric factors, or because no further net (15). In contrast, the multichainedim- the requisite 2 minutes (11). The above decrease in conformationalfree energy munoglobulins are not scrambled and inactivated by the enzyme, reflecting the fact that they are normal products of the disulfide bonding of four pre- formed polypeptide chains.

Factors Contributingto the Correct Folding of PolypeptideChains The results with the disulfide interchange enzyme discussed above sug- gested that the correct and unique translationof the genetic message for a particularprotein backboneis no longer possible when the linear informationhas been tampered with by deletion of amino acid residues.As with most rules, however, this one is susceptible to many exceptions. First, a number of proteins have been shown to undergo reversibledenaturation, including disulfide bond rupture and re-formation, after being shortened at either the NH2- or COOH terminus (16). Others may be cleaved into two (17-19), or even three, fragments which, although devoid of detectable structure alone in solution, recombine noncova- Fig. 1. The amino acid sequence of bovine pancreatic ribonuclease (50). through SCIENCE. VOL. 181 224 lent forces to yield biologically active in vitro structureswith physical properties very similar to those of the parent protein molecules. Richards and his colleagues discoveredthe first of these recom- Fig. 2. Schematic representation of the (17) reductive in 8M urea solu- ribonuclease which denaturation, bining systems, S, tion containing 2-mercaptoethanol, of a consists of a 20-residue fragment from disulfide-cross-linked protein. The con- the NH2terminal end held by a large version of the extended, denatured form number of noncovalent interactions to to a randomly cross-linked, "scrambled" the rest of the molecule in set of isomers is depicted at the lower which, turn, right. consists of 104 residues and all four of the disulfide bridges. The work by Wyckoff, Richards, and their associates on the three-dimensionalstructure of this two-fragment complex (20) and studies by Hofmann (21) and Scoffone (22) and their colleagues on semisynthetic analogs of this enzyme derivative are well known. Studies in our own laboratory (23) showed that the 20- residue "ribonuclease S peptide" fragment could be reduced by five residues at its COOH-terminuswithout loss of enzymic activity in the complex, or of its intrinsic stability in solution. Other examples of retentionof native structural "memory" have been found with complexing fragments of the staphylococcal nuclease molecule (18, 24). This calcium-dependent, RNA- and DNA-cleaving enzyme (Fig. 4) consists of 149 amino acids and is devoid of disulfide bridges and sulfhydryl groups (25). Although it exhibits considerable flexibility in solution, as evidenced by the ready exchange of labile hydrogen atoms in the interior of the molecule with solvent hydrogen atoms (26), only a very small fraction Fig. 3. The structureof porcineproinsulin (51). of the total populationdeviates from the intact, native format at any moment. Spectral and hydrodynamic measurements indicate marked stability up to 1 temperatures of approximately 55?C. NH2--AHSERT The protein is greatly stabilized, both against hydrogen exchange ,(26) and against digestionby proteolytic enzymes (27) when calcium ions and the inhibitory ligand, 3',5'-thymidinediphosphate (pdTp), are added. Trypsin, for example, only cleaves at very restricted positions-the loose NH2-terminalpor- tion of the chain and a loop of residues that protrudesout from the molecule as visualized by x-ray crystallography. Cleavage occurs betweenlysine residues 5 and 6 and, in the sequence -Pro-Lys- Lys-Gly- (residues47 through50) (28), between residues 48 and 49 or 49 and 50 The 149 (18). resulting fragments '(resi- SEGLU AP S ASN-COOH dues 6 to 48) and (49 to 149) or (50 to 149), are devoid of detectable structure Fig. 4. Covalent structureof the major extracellularnuclease of Staphylococcus in solution (29); however, as in the case aureus (25). 20 JULY 1973 225 nuclease T. The products, although con- taminated by sizable amounts of "mis- take sequences" that lack amino acid residues because of slight incomplete- ness of reaction during coupling, could be purified by ordinary chromatographic methods to a stage that permitted one to make definite conclusions about the relative of various O g importance compo- C g n3 a nents in the chain. Taking advantage of o c 0: the limited proteolysis that occurs when 4) nuclease is treated with trypsin in the presence of the stabilizing ligands, calcium and pdTp, Chaiken (35) was able - to digest away those aberrant synthetic zr.3d) molecules of (6 to 47) that did not lt form a stable complex with the large, native fragment (49 to 149). After digestion of the complex, chromatography on columns of phosphocellulose (Fig. 100 200 5) yielded samples of semisynthetic Volume of eluate(ml) nuclease T that were essentially indistinguishable from native nuclease T. 5. Isolation of semisynthetic nu- 3 320 340 360 380 400 Fig. For the enhancement of clease T on a phosphocellulose column Wavelengthof emission(nm) example, large fluorescence of the after "functional" purification by trypsin 6. Use of fluorescence measurements single tryptophan in the of calcium ions Fig. residue in nuclease at digestion presence to determine the relative hydrophobicity [located position and (35). thymidine-3'5'-diphosphate (presumably reflecting "nativeness"in the 140 in the fragment (50 to 149)] upon case of nuclease) of the molecular en- addition of the native fragment (6 to vironment of the single tryptophan residue 49) was also shown when, instead, in this protein (33, 35). of ribonuclease S, when these structure- synthetic (6 to 47) peptide isolated from T less fragments are mixed in stoichio- semisynthetic nuclease that had metric amounts, regeneration of activity been purified as described above was added (about 10 percent) and of native struc- that residues 45 through 49 are dispens- (Fig. 6). tural characteristics occurs (the com- able, the latter conclusion the result of The dispensaibility, or replaceability, of a number of residues to plex is called nuclease T). Nuclease T solid phase-synthetic studies (32) on the stability has now been shown (30) to be closely analogs of the fragment (6 to 47). of the nuclease T complex was established the isomorphous with native nuclease (31). Earlier studies by David Ontjes (33) by examining fluorescence, Thus the cleavages and deletions do not had established that the rapid and con- activity, and stability to enzymatic of a number of destroy the geometric "sense" of the venient solid-phase method developed digestion large semisyn- chain. Recently it was shown that resi- by Merrifield (34) for peptide synthesis thetic analogs (36), As is illustrated in due 149 may be removed by carboxy- could be applied to the synthesis of Fig. 7, interaction with the calcium nor- peptidase treatment of nuclease, and analogs of the fragment (6 to 47) of atom required for nuclease activity

Asp' 40

Fig. 7 (left). Amino acid residues in the sequence of nuclease that are of particular importance in the catalytic activity and binding of substrate and calcium ions (36). Fig. 8 (right). A schematic view of the three-dimensional structure of staphylococcal nuclease (31, 47) .

SCIENCE, VOL. 181 226 mally requires the participationof four 120 . ....I is, of course, precisely the conclusion dicarboxylic amino acids. Although the reached by Perutz and his colleagues activities of complexes containing syn- tl0 - * ,N Reducedviscosity (40), and by others (41) who have re- thetic (6 to 47) fragments in which one au A! , AMolar ellipticlty, 220nm viewed and correlatedthe data on vari- of these had been replaced with an s80 rous protein systems. Mutation and asparagine or glutamine residue were u$xso natural selection are permitted a high abolished one (with partial exception- ~~g~~~~~~ 60 -degree of freedom during the evolution asparagine at position 40), three- 0-g^~~~~~. *of species, or during accidental muta- dimensional structure and complex t 40 tion, but a limited number of residues, stability were retained for the most 8 destined to become involved in the in- part. Similarly, replacementof arginine ?'20 ternal, hydrophobic core of proteins, residue 35 with lysine yielded an inac- l :n must be carefully conserved (or at most tive complex, but neverthelessone with dq* 0 _ A replaced with other residues with a strong three-dimensional similarity to '' ' close similarity in bulk and hydropho- native nuclease T. 2 3 4 5 6 7 8 bicity). A second kind of complementing system of nuclease fragments (24) con- Fig. 9. Changesin reducedviscosity and sists of tryptic fragment (1 to 126) and molar ellipticity at 220 nm during the CooperativityRequired for a partially overlapping section of the acid-inducedtransition from native to denaturednuclease. [3 and Reduced Folding and Stabilityof Proteins (99 to U, sequence 149) prepared by viscosity; A and A, molar ellipticity at cyanogen bromide treatment of the 220 nm; 1] and A, measurementsmade The examples of noncovalent inter- native molecule (shown schematically during the addition of acid; I and A, action of complementing fragments of in Fig. 8). These two peptides form a measurementsmade duringthe additionof proteins quoted above give strong base (44). sup- complex with about 15 percent of the port to the idea of the essentiality of activity of nuclease itself, which is suf- cooperative interactions in the stability ficiently stable in the presence of pdTp of protein structure. As is the case in proteins from a very large of and calcium ions to exhibit remarkable variety any an sentence sources and the isolation of numerous language, incomplete resistanceto digestion by trypsin. Thus, frequentlyconveys There bacterial proteins after mutation of the only gibberish. many of the overlapping residues in appears to exist a very fine balance be- corresponding genes have made it the complex consisting of (1 to 126): tween native structure quite clear that considerable modifica- stable, protein (99 to 149), may be "trimmed"away and random, tion of protein be made biologically meaningless with the production of a derivative, (1 sequence may chains. without loss of function. In those cases polypeptide to 126): (111 to 149). Further degrada- A of the where studies of three- very good example inade- tion of each of the two components,the crystallographic of an dimensional structure have been quacy incomplete sequence comes former with carboxypeptidasesA and B made, from our observations on the nuclease the results indicate that the geometric and the latter with leucine aminopepti- fragment (1 to 126). This con- problem of nat- fragment dase, permits the preparation of (1 to "designing," through tains all of the residues that make ural selection, molecules that can sub- up 124): (114 to 149), which is as active the active center of nuclease. Neverthe- serve a particular functional need can and as structurally similar to native less, this fragment, about be solved in many ways. Only the representing nuclease (as evidenced by estimates of 85 percent of the total of nu- geometry of the protein and its active sequence hydrodynamic, spectral, and helical clease, exhibits about 0.12 site need be conserved, of only per- properties) as the parent, undegraded except, cent of the of the course, for such residues as activity native en- complex. A numberof syntheticanalogs actually zyme (42). The further addition of 23 participatein a unique way in a catalyt- of the (114 to 149) sequence have been residues during or the ad- ic or regulatorymechanism (38). Studies biosynthesis, prepared (37), which also exhibit ac- dition in vitro, of residues 99 to 149 as and "native" of model systems have led to similar tivity physical properties a complementing fragment (24), re- when added to to I will discuss conclusions. In our own work on ribo- ,(1 126). stores the stability required for activity below the nuclease, for example, it was shown manner in which these com- to this unfinishedgene translation. have been useful that fairly long chains of poly-DL- plexing fragments in The transition from incomplete, in- to the alanine could be attachedto eight of the devising experiments study pro- active enzyme, with random structure, cesses of nucleation and of eleven amino groups of the enzyme folding to competent enzyme, with unique and chains. without loss of enzyme activity (39). polypeptide stable structure, is clearly a delicately Furthermore,the polyalanylated enzyme balanced one. The sharpness of this could be converted to an extended transition may be emphasized ex- Mutabilityof Informationfor chain by reductionof the four by disulfide periments of the sort illustratedin bridges in 8M urea, and this dena- Fig. Chain Folding fully 9. Nuclease undergoes a dramatic tured material could then be reoxidized change from native globular structure Biological function to be to yield the active, folded appears correctly to-random disorientedpolypeptide over more a correlate of macromolecular substance. the starting Thus, chemistry a very narrow range of centered geometry than of chemical of the pH, detail. The protein could be greatly modi- at pH 3.9. The transition has the classic chemical and ap- crystallographic fied, and its capacity to refold after pearance of a work on the "two-stage" process- large number of abnormal denaturation seemed to be dependent either all native or all human on the var- denatured-and, hemoglobins, species only on internal residues and not those indeed, two-state mathematical treat- iants of cytochrome c, and on other on the to outside, exposed solvent. This ment has classically been employed to 20 JULY 1973 227 Table 1. Studies of the equilibrium between the peptide fragment (99 to 149) in its random of a requirement for cooperative sta- form [fragment (99 to 149)r] and in the form this fragment assumes in the native structure of bilization is because, in nuclease [fragment (99 to 149)j. Abbreviations: P, fragment (99 to 149); Ab, antibody; conf, high aqueous conformation; assoc, association; T, total; t, time. solution, ionic or hydrogen-bondedin- teractions would not be expected to K conf-- [AbP,] K.6' [Ab] [PT] compete effectively with interactions with solvent molecules and [Ab]total sites [PT] tl/2 [Ab]f re siies [Abib P.. a P,, anything (M) (M)(M) (sec) (AiM) (AM) Kcon (%) less than a sizable nucleus of interact- 0.076 0 18 0.076 0 ing amino acid side chains would prob- .076 0.65 20 .068 0.0080 2.20 X 10-4 0.022 ably have a very short lifetime. Further- .076 2.0 24 .057 .019 2.02 X 10-4 020 more, it is importantto stress that the .076 2.6 27 .051 .025 2.29 X 10-4 .023 amino acid sequences of polypeptide .076 7.8 35 .039 .037 1.47 X 10-4 .015 chains designed to be the fabric of pro- .076 6.5 33 .042 .034 1.51 X 10-4 .015 tein molecules only make functional Kconf = (2.0 ? 0.4) X 10-4 sense when they are in the three-dimensional arrangement that charac- terizes them in the native protein struc- describe such data. In actuality, it has phase is essentially temperature-inde- ture. It seems reasonableto suggest that been possible to show, by nuclear mag- pendent (and therefore possibly en- portions of a protein chain that can netic resonance and spectrophotometric- tropically driven) and the second tem- serve as nucleation sites for folding experiments (43) that one of the four perature-dependent. will be those that can "flicker" in and histidines and one tyrosine residue of out of the conformation that they oc- the seven in nuclease become disori- cupy in the final protein, and that they ented before the general and sudden Nucleation of Folding will form a relatively rigid structure, disintegration of organized structure. stabilizedby a set of cooperative inter- However, such evidences of a stepwise A chain of 149 amino acid residues actions. These nucleation centers, in denaturation and renaturation process with two rotatable bonds per residue, what we have termed their "native for- are certainly not typical of the bulk each bond probably having two or mat" (Fig. 10), might be expected to. of the cooperatively stabilized mole- three permissible or favored orienta- involve such potentially self-dependent cule. tions, would be able to assume on the substructures as helices, pleated sheets, The experiments in Fig. 9, involv- order of 4149 to 9149 different confor- or ibeta-bends. ing measurements of intrinsic viscosity mations in solution. The extreme ra- Unfortunately, the methods that de- and helix-dependent circular dichro- pidity of the refolding makes it essen- pend on hydrodynamicor spectralmea- ism, are typical of those obtained with tial that the process take place along a surements are not able to detect the most proteins. In the case of nuclease, limited number of "pathways," even presence of these infrequent and tran- not only is the transition from native when the statistics are severely restrict- sient nucleations. To detect the postu- to denatured molecule during transfer ed by the kinds of stereochemical lated "flickeringequilibria" and to de- from solution at pH 3.2 to 6.7 very ground rules that are implicit in a so- termine their probable lifetimes in so- abrupt, but the process of renaturation called Ramachandran plot. It becomes lution requires indirect methods that occurs over a very short time period. I necessary to postulate the existence of will record the brief appearance of in- shall not discuss these stop-flow kinetic a limited number of allowable initiating dividual "native format" molecules in experiments (44) in detail. In brief, the events in the folding process. Such the population under study. One such process can be shown to take place in events, generally referred to as nucle- method, recently used in our labora- at least two phases-an initial rapid ations, are most likely to occur in parts tory in a study of the folding of staphy- folding with a half-time of about 50 of the polypeptide chain that can par- lococcal nuclease and its fragments,em- milliseconds and a second, somewhat ticipate in conformational equilibria ploys specific antibodies against re- slower transformation with a half-time between random and cooperatively of about 200 milliseconds. The first stabilized arrangements. The likelihood

!I . I I I I

NATIVE Nl-----2 FORMAT .* ** C Fig. 10 (left). How protein chains * NATIVE ?** ,* HOOC FORMAT might fold (see the text for a dis- * .* cussion of this fairly reasonable, No Antibody. -' but 0 subjective proposal). Fig. tD 11 (right). Inhibition of nuclease usi activity by antibody to (99 to I1.20 Anti(99-149)r NATIVE NATIVE 149), and lack of inhibition by NH2-- FORMAT FORMAT -H I II antibody to (99 to 149), made against the peptide (99 to 149), Anti(99-149)n ,.**' in a random confor- presumably t 1i 11 mation (45). Nuclease 0.04#g i', r 0-- - RANDOM RANDOM - COO 0 2 3 4 NH2. 228a--- I Time (minutes) SCIENCE, VOL. 181 228 stricted portions of the amino acid sequence (45). Figure 8 depicts the three-dimensional pattern assumed by staphylococcal Fig. 12. Semilogarithmic nuclease in solution. Major features in- plot of activity against volving organized structure are the time for assays of 0.05 three-strandedantiparallel pleated sheet uAg of nuclease in the of: no approximately located between resi- presence O-O, ._ dues 12 and and the three antibody; --0, 6 t/g E 35, alpha- of antibody to (99 to helical regions between residues 54 to 126),a; l---l, 6 ,g of 67, 99 to 106, and 121 to 134. An- antibodyto (99 to 126)u 0o tibodies against specific regions of the plus 12 Alg of (99 to cm nuclease molecule were 149); and A-A, 6 /ug prepared by of to immunizationof with either antibody (99 to goats poly- 126)n plus 48 zg of (99 peptide fragments of the enzyme or by to 149). The dotted line injection of the intact, native protein representsone-half of the with subsequent fractionation of the initial activity. resulting antibody population on affin- ity chromatographycolumns consisting of agarose bearing the covalently at- tached peptide fragment of interest (45, 46). In the former manner there was prepared, for example, an di- antibody anti-(99 to 149), through a column of may be calculated from measurable rected against the residues polypeptide, Sepharose-immobilizedfragment (127 parameters. A series of values 99 to 149, known to exist in solution typical to 149). This fraction, which shown in Table 1, that as a random chain without the exten- antibody suggests approxi- forms an inactive and soluble complex mately 0.02 percent of fragment (99 to sive helicity that characterizesthis por- with nuclease, was used in the experi- 149) exists in the native format at any tion of the nuclease chain when pres- ments described below. The reaction moment. Such a value, although low, is ent as part of the intact enzyme. Such between anti-(99 to 149) and nuclease probablyvery large relative to the likeli- an antibody preparation is referred to could be shown by measurements of hood of a peptide fragment of a pro- as anti-(99 to 149) , the subscript in- changes in the kinetics of inhibition tein being found in its native format dicating the disordered state of the of enzyme activity (Fig. 12) to be ex- on the basis of chance alone. antigen. tremelyrapid, with k,,, 4.1 x l OM-1 Empirical considerationsof the When, on the other hand, a fraction large sec-1; koff,on the other hand, is small. amount of data now available on cor- of antiserum to native nuclease, iso- The system may be describedby two relations between sequence and three- lated on an agarose-nuclease column, simultaneous equilibria, the first con- dimensionalstructure (48), togetherwith was further fractionated on agarose- cerned with the "flickering" of frag- an increasing in the fragment(99 to 149), a fraction was ob- sophistication ment (99 to 149), which we shall term theoretical treatment of the tained which was specific for the se- energetics P, from random to native "format," of chain are (99 to 149), but polypeptide folding (49) quence presumablyonly and the second with the association of to make more when this bit of the beginning realistic the sequence occupied anti-(99 to which we shall idea of the a of "native format." This latter conclusion 126),,, term, priori prediction pro- Ab with P in its na- tein conformation. It is is based on the observation that the simply, fragment certain that tive format: that is, P,. major advances in the of latter fraction, termed anti-(99 to 149),, understanding cellular organization,and of the causes (the subscript n referring to the native Prr~-P,, K..rm=-[P.] and control of abnormalitiesin such or- format) exhibited a strong [P ] inhibitory ganization, will occur when we can effect on the enzymic activity of nu- Ab + P. AbP,,KR K.(., [AbP,,] predict, in advance, the three-dimen- clease (47) whereas anti-(99 to 149)r [P,,]Ab] sional phenotypic consequences of a was devoid of such an effect (see Fig. Ko,,=f [AbP,,] genetic message. 11). This conclusion was further sup- k ..... [Ab] [P,,] ported by the observationthat the con- The amount of unbound antibody References and Notes and formation-stabilizingligands, pdTp in the second equilibrium may 1. M. L. Anson, Advan. Protein Chem. 2, 361 calcium ions, showed a marked inhibi- be estimated from measurements (1945). effect on the of 2. R. Lumry and H. Eyring, J. Phys. Chem. 58, tory precipitability nu- of the kinetics of inactivation of 110 (1954); C. H. W. Hirs, S. Moore, W. H. clease by anti-(99 to 149), but had lit- the digestion of denatured DNA sub- Stein, J. Biol. Chem. 235, 633 (1960); J. T. tle Potts, A. Berger, J. Cooke, C. B. Anfinsen, effect, if any, on such precipitability strate by a standardamount of nuclease ibid. 237, 1851 (1962); D. G. Smyth, W. H. by anti-(1 to 149),, (45). This finding added to the incubated mix- Stein, S. Moore, ibid. 238, 227 (1963). previously 3. M. Sela, F. H. White, C. B. Anfinsen, Science reinforced the idea that many of the ture of fragment (99 to 149) and anti- 125, 691 (1957). determinants 4. B. Gutte and R. B. Merrifield, J. Biol. Chem. antigenic recognized by (9 to 126),,. If we make the assumption 246, 1922 (1971). the antibodies to fragments are present that the affinityof anti-(99 to 126), for 5. R. Hirschmann, R. F. Nutt, D. F. Veber, R. in the "unfolded" or A. Vitali, S. L. Varga, T. A. Jacob, F. W. only "nonnative" fragment (99 to 149) in its folded (P) Holly, R. G. Denkewalter, J. Amer. Chem. conformation of nuclease. A further form is the same as that determinedfor Soc. 91, 507 (1969). 6. F. H. White, Jr., and C. B. Anfinsen, Ann. subfractionation, yielding anti-(99 to this antigenic determinant in native N.Y. Acad. Sci. 81, 515 (1959); F. H. White, 126),,, was carried out by passage of nuclease, the value for the term Jr., J. Biol. Chem. 236, 1353 (1961). K,.,,f 7. C. B. Anfinsen, E. Haber, M. Sela, F. H. 20 JULY 1973 229 White, Jr., Proc. Nat. Acad. Sci, U.S.A. 47, 21. K. Hofmann, F. M. Finn, M. Linetti, J. 33. D. Ontjes and C. B. Anfinsen, ibid. 244, 6316 1309 (1961). Montibeller, G. Zanetti, J. Amer. Chem. Soc, (1969). 8. E. Haber and C. B. Anfinsen, J. Biol. Chem, 88, 3633 (1966). 34. R. B. Merrifield, Science 150, 178 (1965). 237, 1839 (1962). 22. E. Scoffone, R. Rocchi, F. Marchiori, L. 35. I. M. Chaiken, J. Biol. Chem. 246, 2948 9. C. J. Epstein, R. F. Goldberger, C. B. An- Moroder, A. Marzotto, A. M. Tamburro, (1971). finsen, Cold Spring Harbor Symp. Quant, ibid*. 89, 5450 (1967). 36. --- and C. B. Anfinsen, ibid., p. 2285. Biol. 28, 439 (1963). 23. J. T. Potts, Jr., D. M. Young, C. B. Anfinsen, 37. I. Parikh, L. Corley, C. B. Anfinsen, ibid., 10. H. M. Dintzis, Proc. Nat. Acad. Sci. U.S.A. J. Biol. Chem. 238, 2593 (1963). p. 7392. 47, 247 (1961); R. E. Canfield and C. B. An- 24. H. Taniuchi and C. B. Anfinsen, ibid. 246, 38. W. M. Fitch and E. Margoliash, Evol. BioL! finsen, Biochemistry 2, 1073 (1963). 2291 (1971). 4, 67 (1970). 11. R. F. Goldberger, C. J. Epstein, C. B. Anfin- 25. J. L. Cone, C. L. Cusumano, H. Taniuchi, 39. J. P. Cooke, C. B. Anfinsen, M. Sela, J. Biol. sen, J. Biol. Chem. 238, 628 (1963); P. Vene- C. B. Anfinsen, ibid., p. 3103; J. L. Bohnert Chem. 238. 2034 (196t. tianer and F. B. Straub, Biochim. Biophys. and H. Taniuchi, ibid. 247, 4557 (1972). 40. M. F. Perutz, J. C. Kendrew, H. C. Watson, Acta 67, 166 (1963). 26. A. N. Schechter, L. Moravek, C. B. Anfinsen, J. Mol. Bliol. 13, 669 tiyo). 12. F. DeLorenzo, R. F. Goldberger, E. Steers, ibid. 244, 4981 (1969). 41. C. J. Epstein, Nature 210, 25 (1966). D. J. Givol, C. B. Anfinsen, Biol. Chem. 27. H. Taniuchi, L. Moravek, C. B. Anfinsen, 42. D. Sachs, H. Taniuchi, A. N. Schechter, A. 241, 1562 (1966); S. Fuchs, F. DeLorenzo, C. ibid., p. 4600. Eastlake, unpublished work. B. ibid. 398 Anfinsen, 242, (1967). 28. The abbreviations for the amino acid residues 43. H. Epstein, A. N. Schechter, J. Cohen, Proc. 13. I. Kato and C. B. 5849 Nat. Sci. Anfinsen, ibid., 244, are: Ala, alanine; Arg, as- Acad. U.S.A. 68, 2042 (1971). arginine; Asp, 44. H. A. (1969). partic acid; Cys, glutamnic F. Epstein, N. Schechter, R. F. Chen, half-cystine; Glu, C. 14. G. Givol, F. DeLorenzo, R. F. Goldberger, acid; histidine; Ileu, iso- B. Anfinsen, J. Mol. Biol. 60, 499 (1971), Gly, glycine; His, 45. D. A. N A. C. B. Anfinsen, Proc. Nat. Acad. Sci. U.S.A. leucine; leucine; lysine; meth- H. Sachs, Schechter, Eastlake, Leu, Lsy, Met, B. Proc. Nat. Acad. Sci. 53, 766 (1965). ionine; Phe, Pro, proline; C. Anfinsen, U.S.A, phenylalanine; Ser, 3790 15. D. F. Steiner, Trans. N.Y. Acad. Sci. Ser. serine; Thr, threonine; Tyr, tyrosine; 69, (1972). Val, 46. - , J. Immunol. 1300 2 30, 60 (1967). valine; Glu-NH,, glutamine; Asp-NH2, as- 109, (1972). 47. - , 4268 16. C. B. Anfinsen, Dev. Biol. Suppl. 2, 1 (1968). Biochemistry 11, (1972). paragine. 48. C. B. Anfinsen and H. Advan. 17. F. M. Richards, Proc. Nat. Acad. Sci. U.S.A. H. Taniuchi and B. J. Scheraga, 29. C. Anfinsen, Biol. Prot. Chem., in preparation. 44, 162 (1958). Chem. 3864 244, (1969). 49. H. A. Scheraga, Chem. Rev. 71, 195 (1971). 18. H. Taniuchi, C. B. Anfinsen, A. Sodja, ibid. 30. H. Taniuchi, D. Davies, C. B, Anfinsen, 50. C. H. W. Hirs, S. Moore, W. H. Stein, J. 58, 1235 (1967). ibid. 247, 3362 (1972). Biol. Chem. 235, 633 (1960); J. T. Potts, A. 19. I. Kato and N. Tominaga, Fed. Eur. Biochem. 31. A. Arnone, C. J. Bier, F. A. Cotton, E. E. Berger, J. Cooke, C. B. Anfinsen, ibid., 237, Soc. Lett. 10, 313 (1970). Hazen, Jr., D. C. Richardson, J. S. Richard- 1851 (1962); D. G. Smyth, W. H. Stein, S. 20. H. W. Wyckoff, D. Tsernoglou, A. W. Han- son, A. Yonath, ibid. 246, 2302 (1971), Moore, ibid., 238, 227 (1963). son, J. R. Knox, B. Lee, F. M. Richards, J. 32. G. R. Sanchez, I. M. Chaiken, C. B. A- 51. R. E. Chance, R. M. Ellis, W, W. Bromer, Biol. Chem. 245, 305 (1970). finsen, ibid., in press, Science 161, 165 (1968).

and in other parts of the world and more than 1000 cases had been as- cribed to exposure to the ethylmercury homolog (la, 2). The earliest cases were due to occupational exposure fol- Methylmercury Poisoning in Iraq lowing the introduction of methylmercury compounds as antifungal seed An interuniversity report. dressing agents. In the 1950's, reports of poisonings from nonoccupational sources appeared in the literaturewith F. Bakir, S. F. Damluji, L. Amin-Zaki, M. Murtadha, increasing frequency. These included a few cases arising from the treatment A. Khalidi, N. Y. Al-Rawi, S. Tikriti, and H. I. Dhahir of fungal skin infections as well as (University of Baghdad) accidental and suicidal ingestion. Sev- T. W. Clarkson, J. C. Smith, and R. A. Doherty eral large incidents of poisoning have occurred in Pakistan, and Guate- (University of Rochester) Iraq, mala due to the ingestion of flour and wheat seed treated with methyl- and ethylmercury compounds. The fungi- cide ethylmercury-p-toluenesulfonani- Methylmercury and its short-chain and feet and in areas around the mouth lide was claimed to be responsible for homolog elicit characteristic toxic ef- (paresthesia), loss of coordination in two outbreaks in Iraq in 1956 and fects in man that differ from the toxic gait (ataxia), slurred speech (dysarth- 1960. effects of other mercury compounds ia), diminution of vision (concentric In the 1960 outbreak, an estimated (la, 2). The primary signs and symp- constriction of the visual field), and 1000 patients were affected by methyl- toms of methylmercury poisoning re- loss of hearing. Severe poisoning can mercury poisoning and 370 were ad- sult from damage to the nervous system cause blindness, coma, and death. mitted to hospital. In Guatemala,cases and are characterizedby loss of sensa- There is a latent period of weeks or that were originally thought to be viral tion at the extremities of the hands months between exposure to methyl- encephalitis were reported during the mercury and the development of wheat growing seasons of 1963, 1964, The first eight authors are members of the poisoning symptoms. Ataxia may sub- and 1965. Forty-five people were af- University of Baghdad, Baghdad, Iraq. Dr. Tikriti is also a member of the Directorate of Preventive sequently decrease but general recovery fected and 20 died. Methylmercurydi- Medicine, Ministry of Health, Iraq. Drs. Clarkson, is poor. Prenatal exposure to methyl- cyandiamide, used to treat the seed Smith, and Doherty are in the departments of radiation biology and biophysics, pharmacology mercury has resulted in mental retarda- wheat before distribution to farmers, and toxicology, and pediatrics, respectively, of tion with cerebral palsy. Prior to the was later established as the causative the University of Rochester School of Medicine and Dentistry, Rochester, New York 14642. See present outbreak in Iraq, between 200 agent. A similar outbreak occurred in (1) for an acknowledgment of the substantial con- and 300 cases of Pakistan in 1969. tributions that many others have made to this methylmercury study, poisoning had been reported in Iraq Despite the large number of people 230 SCIENCE, VOL. 181