3-Galactosidase: Immunological Activity of Ribosome-Bound,

Transcription

1 Proc., Nt. Acd. Sci. USA Vol. 69, No. 2, pp , Februry Glctosidse: Immunologicl Activity of Ribosome-Bound, Growing Polypeptide Chins (immune hemolysis inhibition/ntiserum/puromycin/protein conformtion/chin folding) JOYCE HAMLIN* AND IRVING ZABIN Deprtment of Biologicl Chemistry, School of Medicine nd Moleculr Biology Institute, UCLA, Los Angeles, Cliforni Communicted by Emil L. Smith, November 12, 1971 ABSTRACT Ribosomes crrying nscent chins of fbglctosidse were prepred by disruption of Escherichi coli in detergent-free buffer of high slt concentrtion, followed by purifiction on discontinuous sucrose grdient. Assy by the method of immune hemolysis inhibition with nti-,b-glctosidse indicted tht considerble mounts of ntibody were bound by the growing chins. Much of the crossrecting mteril could be relesed from the ribosomes by tretment with puromycin. The bility to bind nti-,f-glctosidse ws completely destroyed when ribosomes were heted t 60'C. At very erly times fter induction, well before the ppernce of ctive enyme, crossrecting mteril could be demonstrted on ribosomes; this finding correlted with the ppernce of n mino-terminl frgment of j3-glctosidse. Thus, growing chins of,-glctosidse must begin to fold before their relese from the ribosome. Abbrevition: IPTG, isopropyl-3-d-thioglctoside. * Present ddress: Deprtment of Pthology, University of Cliforni School of Medicine, Irvine, Clif. 412 Under some experimentl conditions, nscent peptides longer thn 30 or 35 residues re susceptible to ttck by proteolytic enymes (1), finding tht implies tht only the growing, crboxyl-terminl portion of nscent peptides is physiclly shielded from the milieu by the ribosome itself. This observtion suggests tht s the polypeptide elongtes, successively longer portions of the chin will be subject to interction with solvent, nd folding my occur. It is of considerble interest to determine whether ny conformtion ssumed by growing polypeptide my contribute to the three-dimensionl structure of the finished protein. Studies by Cook nd Koshlnd (2) on the renturtion of severl multisubunit enymes indicte tht in the presence of other proteins (such s would obtin in the cell), completely uncoiled polypeptides my find it difficult to reestblish the conformtion of the ntive, multimeric enyme, suggesting tht polymerition would be more efficient if, before relese from the ribosome, the individul polypeptides were to ssume conformtion cpble of ccepting the other subunits of the protein. There is some evidence for this view in studies (3-6) in which low levels of n enymticlly-ctive form of 0- glctosidse were demonstrted to be trnsiently ssocited with the ribosome. In ddition, work on the in vitro synthesis of IgG indictes tht ctive ntibody molecules my be formed on the ribosome while t lest one of the polypeptides of this multisubunit protein is still being synthesied (7). Severl lbortories hve exmined indirectly the question of conformtion of nscent polypeptides. Tniuchi nd Anfinsen (8) hve demonstrted tht fter enymic removl of portions of the crboxyl-terminus of Stphylococcus nuclese, these derivtives demonstrte none of the physicl or immunologicl properties of the ntive molecule, implying tht in this smll protein the entire mino-terminl sequence is required before ny significnt degree of folding cn occur. Similr observtions were reported for sperm-whle myoglobin (9). These results were tken s evidence ginst folding of nscent polypeptides. Immunologicl studies in this lbortory demonstrted tht premturely terminted polypeptides of j3-glctosidse produced by nonsense-mutnt strins of Escherichi coli crossrect with ntibodies prepred ginst the ntive enyme (10, 11). Anlogous to certin of the derivtives of nuclese nd ribonuclese, nd to growing polypeptide chins, these nonsense-mutnt proteins re missing significnt portions of the crboxyl-terminl sequence of l3-glctosidse (up to 60% of the entire protein in some cses). The loss of crossrecting bility upon heting or sodium dodecyl sulphte tretment indictes tht the ntigenicity of these premturely-terminted polypeptides is conformtion dependent. Here, we test directly whether nscent polypeptides of 13-glctosidse, ttched through their crboxyl termini to ribosomes, crossrect with ntibody prepred ginst the ntive enyme, thus demonstrting conformtionl similrities to the completed protein. In contrst to other studies in which ntibodies were used to demonstrte low levels of presumbly completed proteins ttched to ribosomes, these experiments were designed to mesure quntittively the binding of nti-,3-glctosidse to incomplete chins. MATERIALS AND METHODS Growth of Bcteril Strins nd Preprtion of Cell Extrcts. The wild-type E. coli strin, Hfr 3000, the ochre mutnt X90, the mber mutnt NG 200, nd the lc deletion strin RV were generously supplied by Dr. Frncois Jcob. The "hyper" strin E 203 (12) ws obtined from Dr. Aron Novick nd J 90 nd M 15 were from Dr. Jonthn R. Beckwith. Cultures were grown with ertion t 37 C in either tryptone broth (13) or in medium 63 (14) contining 1.0% glycerol s crbon source, nd were induced with 1 mm #-D-isopropylthioglctoside (IPTG) (Cyclo Chemicl Corp.) for 4 min during the exponentil phse of growth, unless otherwise noted. Cells were hrvested in n equl volume of ice, wshed with ribosome buffer [0.01 M Tris HCl (ph 7.4)-0.15 M KCl I Mg cette] nd the pellet ws stored t -40 C. Cells were thwed, suspended in n pproprite smll volume of ribosome

2 Proc. Nt. Acd. Sci. USA 69 (1972) buffer, nd (with the exception of J 90) broken in n Aminco French pressure cell t 18,000 psi. After incubtion t 4VC for 15 min with 10,Ag/ml of DNse (Worthington Biochemicl Corp.), the cler extrct ws obtined by centrifugtion t 27,000 X g for 15 min, nd usully contined mg of protein/ml of extrct, s determined by the method of Lowry etl. (15). Strin J 90, which contined n integrted X-qs80dlc region with temperture-sensitive X-repressor, ws lso used. Repliction of the defective phge genome ws induced by three cycles of temperture shift to 45 C, followed by 10-min incubtions with ertion t 40 C. The culture ws finlly incubted t 37 C for 22 min, nd induced with 1 mm IPTG. The cells were hrvested by centrifugtion with ice, nd the pellet ws froen nd stored t -40 C. Upon thwing, the cells lysed spontneously, nd the cler extrct ws obtined s bove. Preprtion of Ribosomes. Ribosomes were obtined by lyering 2-3 ml of cler extrct onto discontinuous sucrose grdient prepred by lyering 5 ml of 30% sucrose over 2.5 ml of 60% sucrose in 10-ml polyllomer tubes. Sucrose solutions were prepred in ribosome buffer. The grdients were centrifuged in Spinco 40 rotor for 2.3 hr t 39,000 rpm in Spinco model L265B preprtive ultrcentrifuge. Frctions were collected nd red for bsorbnce t 260 inm; the ribosoml pek ws pooled nd dilyed ginst ribosome buffer. Sedimenttion vlues were kindly determined (16) by Dougls M. Brown. Immune Hemolysis Inhibition. The ssy for immunologicl crossrection is modified from Arquill et l. (17, 18). Indictor cells were prepred by coupling,-glctosidse to the surfce of erythrocytes with dio compound. Sheep erythrocytes (Dvis Lbortories, Inc.) were wshed twice with cold, buffered sline [0.01 M sodium phosphte (ph 7.4)-0.15 M NCl], nd 0.1 ml of the viscous pellet ws mixed with 1.6 mg of,b-glctosidse in 1.5 ml of 0.11 M sodium phosphte (ph 7.4). The mixture ws shken vigorously t room temperture, nd 0.35 ml of freshly diluted solution of the coupling gent bis-diobenidine [1:15 in 0.11 M sodium phosphte buffer (ph 7.4)] (18, 19) ws dded immeditely. The cells were mixed for 10 min nd wshed twice in cold Veronl-buffered sline, nd were resuspended in volume of 0.15% bovine serum lbumin in this buffer such tht 0.1 ml of the cell suspension in 1.4 ml of H20 gve n bsorbnce of 1.2 t 414 nm. Rbbit nti-fl-glctosidse ws kindly prepred (11) by Dr. Eli Sercr, nd ws heted for 30 min t 56 C in order to destroy endogenous complement. Guine pig complement ws purchsed s lyophilied powder from Hylnd Lbortories, Los Angeles. It ws diluted in Veronl- NCl plus lbumin to concentrtion of 50 C'H50 units/ml (20) nd stored t - 40 C. Ribosomes (usully mg in 0.1) ml were incubted with 0.1 ml of severl different mounts of nti-f3-glctosidse for 15 min t 370C with gentle gittion. At the end of this incubtion period, 0.1 ml of fl-glctosidse-coted indictor cells nd 0.1 ml of the complement solution were dded, nd the incubtion t 370C ws continued for 30 min with vigorous gittion. To ech tube ws dded 1 ml of cold Veronl-NCl, the tubes were centrifuged t 6000 X g, nd the dsorbnce resulting from cell lysis nd hemoglobin relese ws mesured t 414 nm. Immunologicl Activity of Growing j3-glctosidse 413 RESULTS Cell Disruption nd Purifiction of Ribosomes. Severl stndrd techniques of cell disruption nd ribosome purifiction were tested. In ll cses, concentrted suspensions obtined by these methods were cloudy nd difficult to ssy in the spectrophotometer, nd interfered significntly with severl immunossys. A stisfctory method consisted of cell disruption by the French press in high-slt buffer contining no detergent, followed by purifiction of ribosomes on discontinuous sucrose grdient. For the specil cse of strin J 90, the step involving the French press ws omitted. A typicl frctiontion of the cler extrct by this method is shown in Fig. 1. No precutions were tken to prevent degrdtion of mrna by endogenous nucleses; 95% of the mteril in the pek ws determined to hve sedimenttion vlue of 70 in the nlyticl ultrcentrifuge. For wild-type cultures induced for 4 min t 370C, the ssocited fl-glctosidse enyme ctivity of the pooled 70S ribosoml frction fter seprtion on the grdient ws 2-6 units of fl-glctosidse per A260 unit (66,ug) of ribosomes. This ctivity is comprble to or lower thn tht fter 4-5 cycles of differentil centrifugtion (1), nd is below the sensitivity of the immunologicl ssy. When strin J 90 ws induced for 4 min t 370C, the j3-glctosidse ctivity ws roughly 4-fold higher. E C4 w 0 6I m FRACTION NUMBER FIG. 1. Purifiction of ribosomes on discontinuous sucrose grdient. A cler cell extrct obtined from 250 ml of log-phse culture of E. coli strin J 90 ws lyered onto the grdient nd centrifuged s described in the text. Frctions were collected, nd the reltive enyme ctivity nd bsorbnce t 260 nm were determined. The verticl dshed line represent-s the originl position of the meniscus between the 60 nd 30% sucrose. Frctions 1-6 were pooled nd dilyed ginst ribosome buffer; ny cloudiness ws removed by centrifugtion t 27,000 X g for 15 min. The clrified mteril ws used directly s the source of nscent fl-glctosidse polypeptides. In this prticulr experiment, considerble mount of hevy polysoml mteril ws recovered, s well s the 70S mteril in the pek t the 60-30% boundry. I- w mu -I mu U

3 414 Biochemistry: Hmlin nd Zbin Immune Hemolysis Inhibition. The ssy depends upon the quntittion of free ntibody remining fter incubtion of jb-glctosidse or crossrecting mteril with known, stndrd concentrtions of nti-j3-glctosidse. Fig. 2 presents n experiment in which the binding of different quntities of nti-f-glctosidse to three different mounts of j3- glctosidse ws determined in the presence of,-glctosidse-free ribosomes purified from E. coli strin RV, which hs deletion of the entire lc region. These curves were virtully identicl to the one obtined when the sme experiment ws done with the ribosomes omitted, indicting tht ribosomes do not interfere in the ssy. Binding of Anti-P-Glctosidse to Growing P.-Glctosidse. To test whether growing P-glctosidse chins ttched to ribosomes bind nti-j-glctosidse, ribosomes were purified from strin J 90, which produces 4- to 5-times more P-glctosidse thn wild-type strin fter het shock nd induction. The experiment illustrted in Fig. 3 indictes tht ribosomes from J 90 bind significnt mount of nti-j3-glctosidse. E c V w1 IC w RECIPROCAL ANTIBODY DILUTION FIG. 2. Estimtion of ntibody binding to,5-glctosidse in the presence of lc deletion ribosomes. Three concentrtions of j3-glctosidse (GZ) in the presence of 0.15 mg of lc-deletion (RV) ribosomes were incubted with severl dilutions of nti-,@- glctosidse. The free ntibody remining ws then mesured by its bility to lyse fixed number of P-glctosidse-coted erythrocytes in the presence of complement. The hemoglobin relesed upon lysis of the erythrocytes ws determined spectrophotometriclly t 414 nm, nd the bsorbnce ws plotted ginst the respective ntibody dilution. The difference in re between the 100% ntibody curve (lc deletion ribosomes only) nd the curves obtined when the ntibody dilutions were first incubted with three concentrtions of g5-glctosidse is proportionl to the mount of ntibody bound. Added Reltive % GZ curve Anti-GZ in jg re bound I A A DIUTiON 4.2.4B 20 Noo REe AL ATOD DILUTIO - FIG. 3. Binding of nti-g6-glctosidse to ribosomes purified from n induced culture of J 90. Ribosomes were purified from n induced culture of J 90 nd from the lc deletion strin, RV Mg of ech type of ribosome ws tested for the bility to bind nti-3-glctosidse in the immune hemolysis inhibition ssy. Slightly more thn 70% of the ntibody ws bound by 0.15 mg of ribosomes. The control smple contined the sme mount of ribosomes prepred from the deletion strin RV. From Fig. 2, this mount of ntibody binding corresponds to 1 tsg of P-glctosidse. When the J 90 ribosomes were ssyed for ssocited enyme ctivity by the stndrd O-nitrophenyl- B-D-glctoside ssy, vlue of 0.15 Mg of,3-glctosidse per 0.15 mg of ribosomes ws obtined. Thus, 85% of the crossrectivity detected on J 90 ribosomes ws due to nscent polypeptides of j3-glctosidse. Puromycin Relese of Crossrecting Mteril. If the crossrecting mteril ssocited with strin J 90 ribosomes ws due to nscent f3-glctosidse chins ttched through trna to the ribosome, this mteril should be relesed by tretment with puromycin (21). Accordingly, prt of n extrct prepred from n induced culture of J 90 ws incubted with puromycin in order to promote relese of peptidyl-trna from the ribosomes. The culture ws pulse lbeled with [14C]leucine before hrvest; the relese of rdioctive mteril from the ribosomes in typicl experiment is demonstrted in Fig. 4A. When tested for their bility to bind nti-j-glctosidse in the immune ssy, the puromycin-treted ribosomes hd lost roughly 50% of the crossrecting mteril detectble on untreted ribosomes (Fig. 4B). This result further indicted tht nscent chins, while in the process of elongtion, bind to ntibody. When ribosomes prepred from n induced culture of J 90 were heted to 60'C for 30 min, no crossrecting mteril could be detected, indicting tht the ntigenic determinnts recognied by nti-j9-glctosidse re strictly conformtion dependent. Number of Nscent Chins on Ribosomes. To mesure the number of P3-glctosidse chins bound to ribosomes, n ssy ws used tht detects segment of the polypeptide chins locted very ner the mino terminus. This segment, which is relesed upon CNBr clevge, cn be mesured by its bility to restore #-glctosidse enyme ctivity to extrcts of M 15, deletion mutnt tht produces protein lcking tht region of the polypeptide chint. Since growing peptide chins re synthesied from the mino to the crboxyl t Lin, S. & Zbin, I., in preprtion. Proc. Nt. Acd. Sci. USA 69 (1972)

4 'Proc. Nt. Acd. Sci. USA 69 (1972). Immunologicl Activity of Growing p-glctosidse 415 I 'I 3 0V.0 it FACTION NUMBER E * ~~.s~~j90 + PUROMYCIN u RECIPtOCAL ANTOODY DILTION FIG. 4. Puromycin relese of crossrecting mteril from J 90 ribosomes. (A) Strin J 90 ws induced with IPTG for 4 mm nd pulsed with 12.5 uci/ml of [14C]leucine for 20 see before hrvest on ice. An extrct ws prepred nd prt of it ws incubted for 15 min t 350C with 250 jug/ml of puromycin nd 20,gM GTP. An identicl control without puromycin ws included, nd n liquot of ech tube ws lyered onto liner 5-20% sucrose grdients nd centrifuged for 6 hr t 24,000 rpm in SW25.1 rotor. (B) A second liquot of ech smple ws pplied to the discontinuous sucrose grdient, nd the purified ribosomes obtined were tested for crossrecting mteril by inhibition of immune hemolysis. end, ll chins, except for very short ones, will contin the complementing peptide. When 1 mg of induced J 90 ribosomes (2.4 X 1014 ribosomes) ws treted with CNBr nd subsequently ssyed for the number of complementing peptides (with jb-glctosidse *s stndrd), vlue equivlent to 10lg of p-glctosidse ws obtined. Since 10 g of fl-glctosidse corresponds to 4.4 X 101" monomers, this result implies tht 4.4 X 10"l/ 2.4 X 1014 or 18% of the ribosomes crry nscent P-glctosidse chin. Kinetics of Appernce of Ribosome-Bound Crossrecting Mteril. Though unlikely in view of the dt presented so fr, it cn be rgued tht the crossrecting mteril ssocited with ribosomes tht is relesed by puromycin cn be lrgely ccounted for by polymeric form of j3-glctosidse, one monomer of which is still in the process of chin elongtion. In order to test this possibility, ribosomes from strin E 203 were isolted t very erly times fter induction nd exmined for their bility to crossrect with nti-#- glctosidse. E 203 produces high ctivities of wild-type enyme upon induction with IPTG, while bckground levels of enyme t very erly times fter induction of J 90 vried considerbly. Anlyses of enyme ctivity, of the number of growing chins, nd of ntibody binding to E 203 ribosomes re presented in Fig. 5. Both the number of growing chins, s determined by the CNBr-complementtion ssy, nd the crossrecting mteril ssocited With the ribosomes begn to increse immeditely fter induction. This result indictes tht the crossrection ' ssocited with the ribosomes t time intervls under see is due to incomplete, nscent P-glctosidse chins, " nd not to the presence of highly crossrective polymeric t' structures, since the first increse in enyme ctivity does not occur until 90 see fter induction. It is lso of interest tht trnsltion (ppernce of the complementing peptide) begins within see fter induction. A similr result ws obtined by Morrison nd Zipser (22). DISCUSSION It ws demonstrted previously tht premturely-terminted polypeptidd chins of jb-glctosidse produced by strins contining nonsense codons in the Z gene of the lc operon crossrect with ntibodies prepred ginst the ntive, polymeric enyme (10, 11). Since t lest four of these mutnt proteins exist in the monomeric stte under the conditions of ssy (10), it seemed resonble to ssume tht mny of the ntigenis determinnts present in the wild-type enyme re formed in the monomer before polymerition, possibly while the nscent chin is still being elongted. We show tht ribosomes crrying nscent chins of P-glctosidse bind ntibody prepred ginst the wild-type enyme. Most of the crossrecting mteril ssocited with ribosomes ws relesed 0 1 I I IN SECONDS AFTER ADNG IPTO FIG. 5. Rte of ppernce of enyme, growing chins, nd crossrecting mteril in strin E 203. A culture of E 203 ws induced with 1 mm IPTG t time ero, nd smples were dded to n equl volume of ice-cold 0.04 M sodium ide t the times indicted. Extrcts were prepred in the French press nd the ribosomes were purified. Ech ribosoml preprtion ws ssyed for ssocited enyme ctivity (c), the number of nscent, glctosidse chins ws determined by the CNBr ssy (), nd crossrection with nti-f-glctosidse ws mesured by inhibition of immune hemolysis (b). 0 U -2

5 416 Biochemistry: Hmlin nd Zbin by tretment with puromycin. This mteril could lso be detected on ribosomes isolted from cultures t very erly times fter induction with IPTG, well before the ppernce of incresed quntities of ctive enyme. In ddition, ll detectble crossrecting mteril ssocited with ribosomes from induced cultures ws completely destroyed by heting to 60'C. We ssume tht the ribosome itself hs no effect on the folding of nscent chins, nd therefore conclude tht most of the binding of ntibody to ribosomes cn be ttributed to conformtionl similrities between nscent #-glctosidse polypeptides nd the completed enyme. It is possible to compre the reltive proportion of ntibody bound to nscent chins nd to completed #-glctosidse ssocited with J 90 ribosomes. As determined in the complementtion ssy, ech milligrm of ribosomes crries bout 4.4 X 1013 nscent chins. By enyme ssy, which detects only ntive f3-glctosidse, ech milligrm of ribosomes contins 1 jug or 0.4 X 101' completed chins. Therefore, on molr bsis, the nscent chins comprise 90% of the totl. Only 15% of the binding of ntibody to J 90 ribosomes could be ccounted for by completed, enymticlly-ctive,- glctosidse; thus, 85% of the ntibody binding detected on J 90 ribosomes could be ttributed to nscent chins. These vlues suggest tht incomplete, ribosome-bound polypeptide chins bind considerble quntity of ntibody s compred to equivlent molr mounts of completed f3-glctosidse. The results of experiments reported here gree well with erlier work in which enymticlly-ctive,,3-glctosidselike mteril ws demonstrted to be trnsiently ssocited with ribosomes of induced cultures of E. coli. It ws concluded tht polymeric form of,3-glctosidse ttched to the ribosome is norml intermedite in the biosynthesis of the enyme, since the removl of inducer promoted the relese of the frction (4). The immunologicl studies presented here demonstrte tht this low ctivity of polymeric, enymticlly-ctive,b-glctosidse ccounts for only minor frction of the immunologicl crossrection observed with nti-,bglctosidse. The inbility of Stphylococcus nuclese or sperm-whle myoglobin to fold fter removl of portion of the crboxylterminl end of the molecule seems to indicte tht smll proteins require the entire primry sequence to stbilie the conformtion of the protein s whole (8, 9). In protein s lrge s,-glctosidse, there my be severl locl centers of folding (23), such tht the bsence of portion of the polypeptide chin would not disrupt the entire conformtion Proc. Nt. Acd. Sci. USA 69 (1972) chieved by the completed protein. Experiments reported here nd elsewhere support this hypothesis, nd suggest tht nscent polypeptides begin to ssume conformtion tht is, by virtue of crossrection with nti-j3-glctosidse, similr to tht ttined by the completed, functionl protein. This investigtion ws supported in prt by grnts AI nd GM-1531 from the United Sttes Public Helth Service nd ws prt of disserttion submitted by J. Hmlin s prtil fulfillment for the degree of Doctor of Philosophy. 1. Mlkin, L. I. & Rich, A. (1967) J. Mol. Biol. 26, Cook, R. A. & Koshlnd, D. E. (1967) Proc. Nt. Acd. Sci. USA 64, Cowie, D. B., Spiegelmn, S., Roberts, R. B. & Duerksen, J. D. (1961) Proc. Ntl. Acd. Sci. USA 47, Zipser, D. (1963) J. Mol. Biol. 7, Zipser, D. & Perrin, D. (1963) Cold Spring Hrbor Symp. Qunt. Biol. 28, Kiho, Y. & Rich, A. (1964) Proc. Nt. Acd. Sci. USA 51, Askons, B. A. & Willimson, A. R. (1967) Cold Spring Hrbor Symp. Qunt. Biol. 32, Tniuchi, H. & Anfinsen, C. B. (1969) J. Biol. Chem. 244, Atssi, M. Z. & Singhl, R. P. (1970) J. Biol. Chem. 245, Fowler, A. V. & Zbin, I. (1966) Science 154, Fowler, A. V. & Zbin, I. (1968) J. Mol. Biol. 33, Horiuchi, T., Tomiw, J. & Novick, A. (1962) Biochim. Biophys. Act 55, Gilbert, W. & Muller-Hill, B. (1966) Proc. Nt. Acd. Sci. USA 56, Herenberg, L. A. (1959) Biochim. Biophys. Act 31, Lowry, 0. H., Rosebrough, N. J., Frr, A. L. & Rndll, R. J. (1951) J. Biol. Cfiem. 193, Zomely, C. E., Roberts, S., Peche, S. & Brown, D. M. (1971) J. Biol. Chem. 246, Arquill, E. R., Bromer, W. W. & Mercol, D. (1969) Dibetes 18, Arquill, E. R., Hmlin, J., Hmshige, S. & Miller, A. (1964) J. Exp. Med. 120, Kbt, E. A. & Myer, M. M. (1961) in Experimentl Immunochemistry (C. C Thoms, Springfield, Ill.), 2nd ed., pp Kbt, E. A. & Myer, M. M. (1961) in Experimentl Immunochemistry (C. C Thoms, Springfield, Ill.), 2nd ed., pp Nthns, D. (1967) in Antiobiotics, eds. Gottlieb, 0. & Shw, P. (Springer-Verlg, Berlin), Vol. 1, pp Morrison, S. L. & Zipser, D. (1970) J. Mol. Biol. 50, Goldberg, M. (1969) J. Mol. Biol. 46,