EFFECT OF PROTEOLYTIC ENZYMES ON SEDIMENTATIONPROPERTIES OF RIBONUCLEOPROTEIN PARTICLES FROM

HEART MUSCLE*

BY MURRAY RABINOWITZ, t RADOVAN ZAK, T BARRY BELLER,OLIVER RAMPERSAD, AND IRA G. WOOLt

DEPARTMENTS OF MEDICINE, BIOCHEMISTRY, AND PHYSIOLOGY,AND THE ARGONNE CANCER RESEARCH HOSPITAL,§ UNIVERSITY OF CHICAGO

Communicated by Dwight J. Ingle, October 19, 1964

There is much support for the view that the functional unit of protein synthesisis an aggregate of ribosomes held together by a strand of messenger RNA.1-6The concept derives from the observation that radioactive amino acids are pref-erentially incorporated into ribosomal aggregates (termed polyribosomes, poly-somes, or ergosomes) and that treatment of the ribosomal aggregates with smallamounts of ribonuclease leads to the formation of single ribosomes.2-5 In astudy7-9 of the characteristics of ribonucleoprotein particles from rat heart muscleit was found that the particles when analyzed on linear sucrose gradients gave nodistinct 70-80S peak, nor were there more than small amounts of intermediate-sized aggregates present; instead, almost all the material sedimented in a pellet ofS value greater than 400. The sedimentation pattern was not affected by ribo-nuclease treatment of the ribonucleoprotein particles prior to analysis, but chymo-trypsin did increase the amount of 70-80S material. The latter observation sug-gested the possibility that in muscle the functional unit for protein synthesis mightbe a very large assemblage of ribonucleoprotein particles held together by protein,perhaps the nascent peptide being synthesized on the ribosomes, as well as bymessenger RNA. Indeed, it seemed possible that in muscle more than one ribo-some might be contributing to the synthesis of a single protein, an interpretationthat accords with the suggestion that for the synthesis of structural protein of highmolecular weight, or of enzymes composed of several subunits, polymerization ofthe polypeptide units occurs on the ribosomal assemblage. 10

It is the purpose of this paper to describe the results of controlled proteolytic di-gestion of ribonucleoprotein particles isolated from heart muscle on their sedimen-tation behavior in linear sucrose gradients. Treatment of ribonucleoprotein parti-cles, labeled with radioactive leucine, with small amounts of chymotrypsin or tryp-sin has been found to lead to the disruption of the rapidly sedimenting material intosmaller particles of 70-80S and, pari passu, by the release of radioactive materialderived from the nascent peptide. We believe the finding to have significance forthe mechanism of the synthesis of the protein structural units of the myofibril.

Experimental Procedures.-Ribonucleoprotein particles were labeled by incubation of mincedheart muscle from 14-18-day chick embryos in 5 vol of Eagle's minimal essential culture medium(spinner),I1 from which the C"2-leucine had been omitted; the medium contained 40 juc/ml of H3-leucine (2.5 c/mmole). Incubation was for 15 min at 37°. After incubation the ribonucleoproteinparticles were prepared by either of two procedures: treatment of the 105,000 X g sedimentfrom a whole tissue homogenate with deoxycholate; or from the microsomal fraction. Bothmethods have been described in detail.9 The ribonucleoprotein particles were suspended in buffer(0.05 M Tris-HCl, pH 7.4; 0.005 M MgCl2; 0.1 M KCI) and an aliquot incubated for 30 min atO with or without the concentration of enzyme indicated in the figures. Liver ribosomes wereprepared from microsomes in the customary manner.'2 The ribonucleoprotein particles were

1353

Dow

nloa

ded

by g

uest

on

June

11,

202

1

1354 BIOCHEMISTRY: RABINOWITZ ET AL. PROC. N. A. S.

immediately analyzed on linear gradients made with 2.1 ml of 15% sucrose and 2.1 ml of 30%sucrose; 0.5 ml of 2.5M sucrose was layered at the bottom of the gradient. The sucrose solutionscontained 0.005 M MgCl2; 0.05 M Tris-HCl, pH 7.4; 0.1 M KC1. On the top of each gradientwas layered 0.2 ml of the ribosomal suspension; in any single experiment great care was takenthat exactly the same amount of ribosomal RNA was layered on each gradient. The gradientswere centrifuged at 25,000 rpm for 2 hr at 40 in the no. SW-39 rotor of the model L Spinco cen-trifuge. After centrifugation the bottom of the tube was punctured and 10-drop fractions werecollected. To each fraction was added 1.0 ml of water and the optical density at 260 mg deter-mined in a Beckman model DU spectrophotometer; an 0.8-ml sample from each fraction wasadded to a dioxane-containing scintillation fluid and the radioactivity measured in a PackardTriCarb liquid scintillation counter.The crystalline enzymes, ribonuclease, chymotrypsin, chymotrypsin inactivated with diiso-

propyl phosphofluoridate (DFP-chymotrypsin), and trypsin were purchased from WorthingtonBiochemical Corp.; the enzymes with the exception of the DFP-chymotrypsin were chromato-graphically homogeneous.

Results.-Effect of proteolytic digestion on ribonucleoprotein aggregates: The sedi-mentation pattern (Fig. la) obtained after density gradient centrifugation of ribo-nucleoprotein particles from heart muscle of chick embryos differed in several re-spects from that seen with ribosomes from other tissues and also from that of theribonucleoprotein particles from rat heart muscle previously studied.9 There wasa small peak at 70-80S and heavier material (100-250S) that may be presumed tobe aggregates of two or more ribosomes, but a significant portion (usually more thana half) of the ribonucleoprotein material appeared either in the 2.5 M sucrose"step" at the bottom of the gradient or as a pellet at the bottom of the centrifugetube. While the specific activity of the heavier material tended to be greater thanthat of the 70-80S peak, it is equally true that labeling of the 70-80S ribonucleo-protein particles was always observed. The absence of a distinct dichotomy be-tween the optical density and the radioactivity makes it uncertain as to whetherthe material on the gradient is polysomes, at least as polysomes are usually defined(in general, polysomes have a greater specific activity than single ribosomes forthey contain the newly synthesized polypeptide2-5). The congruent distributionof radioactivity and of ultraviolet absorption may be the result of the fracture froma much larger assemblage, of single ribosomes, and of ribosomal aggregates ofvarious sizes, during the relatively vigorous isolation procedure that must be em-ployed with muscle. However, there is evidence that monomers as well as aggre-gates of all sizes are able to incorporate amino acid into protein provided they areattached to a strand of messenger RNA,"3-"5 and the present results (Fig. la) canbe fitted to that interpretation.The anomalous sedimentation behavior of ribonucleoprotein particles from heart

muscle seems not to be merely a reflection of the manner of their isolation, for similarpatterns were obtained if the particles were prepared by treatment of the 105,000X g sediment from a whole tissue homogenate, or if the particles were preparedfrom the microsomal fraction in the conventional way. Nor did increasing the de-oxycholate concentration to 1 per cent alter the sedimentation characteristics,making it less likely that the results were due to contamination with microsomes orthe lipoprotein of the endoplasmic reticulum.Treatment of the ribonucleoprotein preparation with 0.5 jug of ribonuclease at

00 for 30 min leads to an increase in both the optical density and the radioactivityof the 70-80S peak as a result of transfer of material from the region of the 2.5 M

Dow

nloa

ded

by g

uest

on

June

11,

202

1

VOL. 52, 1964 BIOCHEMISTRY: RABINOWITZ ET AL. 1355

RIBONUCLEASE -05p9 CONTROL A 270

.360 .360

230 ~~~~~~~~~~~~~~~~~~~~230320 _320

190 190

.280 _ 170 280 170

.240 120 .240 120

0~~~~~~~~~~~06 U~~~~~~~~~~~~~~~~~~~~~~~~~as

lllll, l~~~~~~~~~~~~~Iao

* *

.120 W ~~~~~~~~~~~~~.1201 6

.040020 .0 20

I ~~~~I C*MTOM1p ROUMESE0-5/L9 D 270 CHyMOTRYPSIN-1069 C 270

d6 360 290

FIG.~~~~~~~~~~~~91.Surs'rdetaayi90ioulorti atcesfo er uceo hc

320u1o0 3280-i240~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~§

1900

J20~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~7

The -atce eepeae yMto 17 280f9)-nnuae o 0mi t0 iho ihu

060~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~2

.040 120 240

lyzd0 was1.82.0(Atoto-teoooulortiatilswr nuaeda °fr3

0~~~~~~~~~~~~~~~2

prior2oanalyss.(B)Rioa so0 .040 a0wit.(C) Chymotrypsi incubated for 30 0w 120 I (i

FIG.easeand1.-Sucrosegradient ancubals ofrbono0ate0i particl0

g~~~~~~~~~~~~~~~fo hear muscle ofchickembryos:effect ofdribonucleaeandlhymtrypsin.lb. RIbnucleorotein particlstowerhabeledobyincuatio ofo mincedhertomusl fothr 15ssumsath37nmdia contatinin oth-euie (40dl c/mI).Thenpaticle waereapreare bytcmpethde (ref.9) ahnd incbaedfior30mmat0 wicethrato witou

increased~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~o10fo0

fo.0rat44Gea0aewa ae to enuethttesaeaonto4ioncepoti a

zed s 12.( Cotrothe ribonucleoprotein partice were iub 00 for 3mpIGorto analysis.(B) R ibonuclease;incbated for 30 mmpatic0 wfthr5eag ofsriboncea.ic(bc Cymotrypesioptinculbaed for chymotrypsn DRibo-uepeaoisnearyrtice were labeled

nu prtcleasewandchmorypsin; Mto 1(e. nincubated for 30 min at00 with05p frbnces andwth10

sucrose "step" andos fromieth pellth (Fig su)os Intepmarkd contrastge ato what0ispobsevdfor polysom4Gestfaromwa othker tisesureshat tedsaggegatiount of thoneorpidlyi sed-mnigmterial wasnotyeiec complethe evenrwennyrthe26ribnuceascI)ontentrateion wasainretaased10-ol. BRioulaeinuaefo30mna 'wt .ugorbnceseTreamentroftherincbonucleoprotein preparaitio wit 10g of chymotrypsin.D atb

00vefor30moleasalsotroma inereaisuesin the oncentreation of thefast-sdlymenting

fraction; the optical density of the 70-80S peak is nearly twice that of untreated

Dow

nloa

ded

by g

uest

on

June

11,

202

1

1356 BIOCHEMISTRY: RABINOWITZ ET AL. PROC. N. A. S.

controls (Fig. 1c). The material appearing on the gradient after chymotrypsintreatment seems to derive from the pellet, for in this case the optical density of the2.5 M sucrose "step" is unchanged. Once again disaggregation is not near beingcomplete. Chymotryptic digestion strips most of the leucine radioactivity from theribonucleoprotein particles; the radioactivity appears at the top of the gradient butis not associated with a detectable increase in the absorbancy at 260 or 280 m/A.The great specific activity of the material that is released suggests that it is thenascent protein that is preferentially affected by proteolytic digestion with chymo-trypsin, although the possibility is not ruled out that the enzyme is also causingthe hydrolysis of some nonlabeled protein present in small amounts.The influence of ribonuclease and chymotrypsin would appear to be additive, for

if the ribonucleoprotein particles are treated with both enzymes simultaneously,most of the material layered on the gradient sediments in a single symmetrical peakof 70-80S (Fig. ld); the preponderance of the radioactivity is found now either inthe 70-80S region or at the top of the gradient.The influence of the chymotrypsin is not due to contamination with ribonuclease,

for assay of the proteolytic enzyme for ribonuclease activity failed to reveal anyin circumstances where as little as 0.01 per cent contamination would have beendetected. Nor is the influence of chymotrypsin to be accounted for by the activa-tion of latent ribosomal ribonuclease, for incubation of ribonucleoprotein particleswith 50 ,ug of the chymotrypsin led to no detectable increase in ribonuclease activity(Table 1). These observations accord with the finding that the changes in theoptical density and radioactivity patterns of the ribonucleoprotein particles aftertreatment with chymotrypsin are quite distinct from those seen after treatmentwith ribonuclease (cf. Fig. lb and c). What is more, it is the proteolytic activityof the chymotrypsin that is responsible for the disaggregation of the ribonucleo-protein assemblages rather than a nonspecific protein effect, for treatment of theparticles with diisopropylfluorophosphate-chymotrypsin (DFP-chymotrypsin)followed by density gradient analysis revealed no change in either the optical den-sity or the radioactivity pattern (Fig. 2).The influence of proteolysis on the sedimentation behavior of ribonucleoprotein

aggregates from heart muscle is not confined to chymotrypsin, for trypsin is at leastequally effective in leading to disruption of the rapidly sedimenting material (Fig.3a, b); indeed, a larger portion of the material that ordinarily appears in the 2.5 M

TABLE 1EFFECT OF CHYMOTRYPSIN ON LATENT RIBONUCLEASE ACTIVITY OF RIBONUCLEOPROTEIN

PARTICLESTime of AO.D.-260mj-

incubation Heart LiverTemp. (IC) (min) Control Chymotrypsin Control Chymotrypsin

0 5 0.006 0.001 -0.015 0.00910 0.006 -0.001 0.009 -0.00330 0.006 -0.006 0.001 0.00160 0.015 0.006 0.038 0.016

37 10 0.030 0.025 0.036 0.03830 0.105 0.086 0.077 0.09060 0.155 0.160 0.150 0.183120 0.293 0.201 0.220 0.229

The assay system contained in 1 ml: 150 psg of ribosomal RNA; 1 mg E. coli RNA; 0.005 M MgCl2; 0.05M Tris-HCl, pH 7.4; 0.1 M KCl; and 50 pg of chymotrypsin or an equal volume of water. After the timeindicated, 0.2 ml of the reaction mixture was withdrawn and added to 2.8 ml of ice-cold 10% perchloric acidin 50% ethanol, kept at 00 for 10 min. centrifuged, and the supernatant read at 260 m;p. The AO.D.-260mp1 is the increment in absorption of acid-soluble material at 260 mu due to hydrolysis of substrate.'

Dow

nloa

ded

by g

uest

on

June

11,

202

1

VOL. 52, 1964 BIOCHEMISTRY: RABINOWITZ ET AL. 1357

- T XBr T 2

A T5 I A°aita0 'g20O '04'0 200 *0

c2 s s 5 41> & 5 e 1020 S1 06XD020 9i

FIG. 2.-Sucrose gradient analysis of ribonucleoprotein particles from heart muscle of chickembryos: effect of chymotrypsin and DFP-chymotrypsin. The conditions were exactly asdescribed in Fig. 1. The absorbancy ratio (260 mMs280mM) of the ribonucleoprotein particlesanalyzed was 1.82. (A) Control; the ribonucleoprotein particles were incubated at 00 for 30

mmprior to analysis. (B) Chymotrypsin; incubated for 30 mmn at 00 with 10 ,ug of chymotrypsin.(C) DFP-chymotrypsin; incubated for 30mq at 00 with 10 ,sg of DFP-chymotrypsin.

sucrose "step" is disaggregated by trypsin than by chymotrypsin. The effect oftrypsin on the sedimentation characteristics of ribonucleoprotein particles is over-come by the addition of a threefold excess of soybean trypsin inhibitor (Fig. 3c), anamount of inhibitor that was as d toPdcrease activity of the enzyme by 95 percent. The inhibitor itself was without effect on the absorbancy profile.A similar effect of proteolytic digestion was observed with ribonucleoprotein

particles from liver (results not shown). The sedimentation pattern of ribosomesprepared from the microsome fraction from the liver of rats revealed a 70-80Speak and more rapidly sedimenting aggregates. Ribonuclease treatment resultedin the breakdown of the larger aggregates and an increase in the amount of materialsedimenting at 70-80S. Chymotrypsin treatment of the ribosomes had a verysimilar effect leading also to an increase in the amount of material in the 7080Speak; however, somewhat more of the heavy material (100o250S) persisted. Theresults obtained with ribonucleoprotein aggregates from liver are, then, qualitativelysimilar to those obtained with like preparations from heart muscle.

CG4TROL ~~~~~~~~~~~TRYPSIN-102qTY0NI,9RPt40002 4TR3p

300 _ A 3 300

5- 2402240IS ISI .A 0

0~~~~~~~~~~~~~~.6

Thbe, Nmbw* Tub 19_A Tub PhI111

FIG. 3.-Sucrose gradient analysis of ribonucleoprotein particles from heart muscle of chickembryos: effect of trypsin. The conditions were as described in Fig. 1. The absorbancy ratio(260 mM/280 Ma) of the ribonucleoprotein particles analyzed was 1.73. (A) Control; the ribo-nucleoprotein particles were incubated at 00 for 30 min prior to analysis. (B) Trypsin; incubatedfor 30 min at 00 with 10 jg of trypsin. (C) Trypsin and soybean inhibitor; incubated for 30 minat O° with 10 Mg of trypsin and 30 pig of soybean inhibitor.

Dow

nloa

ded

by g

uest

on

June

11,

202

1

1358 BIOCHEMISTRY: RABINOWITZ ET AL. PROC. N. A. S.

Discussion.-Mild proteolytic digestion leads to partial disruption of the largeribonucleoprotein complexes isolated from heart muscle. Several observationssupport the conclusion that it is the proteolytic activity of chymotrypsin and tryp-sin that is causally responsible for the phenomenon: the purity of the proteolyticenzymes, especially the absence of detectable contamination with ribonucleaseactivity; the failure of the enzymes to activate latent ribosomal ribonuclease;the absence of an influence of chymotrypsin after it is treated with diisopropyl-fluorophosphate, and of trypsin in the presence of soybean trypsin inhibitor.The effect of chymotrypsin on the state of aggregation of ribonucleoprotein

particles from heart muscle is not readily explicable on the basis of the propertiesof the ribosomes, for the ribosomes from heart can be shown to be free of significantcontamination with mitochondrial fragments and accord closely in character withribosomes from other mammalian tissues and from microorganisms.7' 9 The ribo-nucleoprotein particles from heart are in fact equally as effective as those fromliver in their capacity to catalyze protein synthesis from sRNA-phenylalanine-C'4.It would seem reasonable to conclude then that the rapidly sedimenting materialfrom heart muscle is ribosomal aggregates. The difference in the sedimentationcharacteristics of ribonucleoprotein assemblies from heart muscle, and especiallythe influence of chymotrypsin on those aggregates, seems not to be due to a dif-ference inherent in the individual ribosomes but rather to the nature of the proteinthat is being synthesized.There are several possible explanations for the influence of proteolytic digestion

on the state of aggregation of the ribosomes from muscle: first, that theaggregation is due to nonspecific association of ribonucleoprotein particles withmuscle proteins, perhaps the myofibrils which seem to have an affinity for ribo-somes.9 However, the relative purity of the ribosomal preparation, especiallythe high RNA content, is against significant contamination with extraneous protein.Further, the ribosomes isolated from microsomes are less likely to be contaminatedwith structural protein than are the ribosomes prepared from the whole cell homog-enate, yet they display an identical sedimentation pattern and the same responseto chymotrypsin treatment. While the possibility of nonspecific association ofribonucleoprotein particles with contaminating protein in the preparation cannotbe excluded with certainty, it does not seem likely.

If the large ribonucleoprotein aggregates are held together by nascent polypeptidechains, as the results suggest, then one would predict that puromycin, which causesrelease of the growing peptide, might, like chymotrypsin, cause the breakdown ofthe aggregates. While the possibility has not yet been tested with ribonucleo-protein assemblies from heart muscle, it has with similar preparations from othertissues. Puromycin causes polysomes from liver,'7 and from chick embryos,10 tobreak down at an accelerated rate due to the release of single ribosomes, and theantibiotic decreases the concentration of polysomes in reticulocytes. 18 The resultswith puromycin accord then with the postulation that ribonucleoprotein aggre-gates from a number of tissues are held together in some measure by nascent poly-peptide chains.The report by Kretsinger et al. 10 of the synthesis of collagen on polyribosomes from

chick embryos records a number of features that bear a striking resemblance tothose we have observed. They find that the polyribosomes from chick embryos

Dow

nloa

ded

by g

uest

on

June

11,

202

1

VOL. 52, 1964 BIOCHEMISTRY: RABINOWITZ ET AL. 1359

concerned with the synthesis of collagen (a triple-stranded fibrous protein) have anextremely high sedimentation velocity, and that ribonuclease treatment of thepolysomes liberates only a portion of the 80S ribosomes, leaving behind a ribonu-clease-resistant fraction. From their observations they suggest that collagen issynthesized on very large polysomes and that the aggregates are held together bysomething other than RNA, a conclusion in accord with our results with ribonu-cleoprotein complexes from heart muscle.The association of the disaggregation of the ribonucleoprotein complex and the

release of the leucine radioactivity has led us to focus our attention on the nascentpeptide as being important in holding the aggregates together. It does not, ofcourse, follow that the association is causal. For example, to mention only oneother possibility, Noll et al.'7 have proposed that the transfer enzyme(s) play animportant role in maintaining the structure of the polysomes. They imagine thatthe enzyme forms an interlocking structure with the ribosomal surface, the end ofthe nascent peptide chain, transfer RNA, and messenger RNA. Chymotrypsinmight then cause the breakdown of the ribonucleoprotein complex by proteolysisof the transfer enzyme "lock," and the release of leucine radioactivity might merelybe fortuitous; alternatively, both the enzyme and the nascent peptide chain maybe important in the maintenance of the structure of the aggregate, but cleavage ateither or both sites sufficient for the release of a portion of the 70-80S ribosomes.For the moment we favor the interpretation that the structure of the ribonucleo-

protein complex from heart muscle is the result of a functional interaction betweenthe ribosome, messenger RNA, and the nascent peptide. This is consonant withthe association of the disaggregation of the complex and the release of the leucineradioactivity after proteolytic digestion. Moreover, muscle is known to containproteins composed of subunits with a strong tendency to interact.'9 20 Since thestructural proteins of muscle have a strong tendency for spontaneous association,it is possible that interactions occur between ribosomal bound polypeptide chainscontaining the several subunits of the protein being synthesized; the associationmight occur either during the isolation procedure or as an integral part of the proc-ess of protein synthesis. The model accounts for the high sedimentation velocityof the ribonucleoprotein material from heart muscle. It also accounts for the releaseof 70-80S particles by both ribonuclease (particles with no association of their pep-tide chain with another subunit, i.e., held to the aggregate only by messenger RNA)and chymotrypsin (single 70-80S particles bound only by an association of its com-pleted polypeptide chain with a nascent peptide subunit still attached to the poly-some). It is, of course, important to distinguish between the several possible explana-tions of the effect of proteolytic digestion on the state of aggregation of ribonucleo-protein particles from heart muscle and to determine the biological significance.

* The work was supported by grants from the National Institutes of Health, the Chicago HeartAssociation, the Life Insurance Medical Research Fund, and the John A. Hartford Foundation.

t Recipients of United States Public Health Service Career Development Awards.I Postdoctoral trainees of the National Heart Institute.§ The Argonne Cancer Research Hospital is operated for the Atomic Energy Commission by

the University of Chicago.1 Risebrough, R. W., A. Tissieres, and J. D. Watson, these PROCEEDINGS, 48, 430 (1962).2Warner, J. R., P. M. Knopf, and A. Rich, these PROCEEDINGS, 49, 122 (1963).8 Gierer, A., J. Mol. Biol., 6, 148 (1963).

Dow

nloa

ded

by g

uest

on

June

11,

202

1

1360 BIOCHEMISTRY: SAUER ET AL. PROC. N. A. S.

4Wettstein, F. O., T. Staehelin, and H. Noll, Nature, 197, 430 (1963).6 Gilbert, W., J. Mol. Biol., 6, 374 (1963).6Penman, S., K. Scherrer, Y. Becker, and J. Darnell, these PROCEEDINGS, 49, 654 (1963).7Rampersad, 0. R., and I. G. Wool, Federation Proc., 23, 316 (1964).8 Zak, R., M. Rabinowitz, B. Beller, and L. DeSalle, Federation Proc., 23, 163 (1964).9 Rampersad, 0. R., R. Zak, M. Rabinowitz, I. G. Wool, and L. DeSalle, to be published.10 Kretsinger, R. H., G. Manner, B. S. Gould, and A. Rich, Nature, 202, 438 (1964).11 Eagle, H., Science, 130, 432 (1959).12 Korner, A., Biochem. J., 81, 168 (1961).13 Munro, A. J., R. J. Jackson, and A. Korner, Biochem. J., 92, 289 (1964).14 Haselkorn, R., V. A. Fried, and J. E. Dahlberg, these PROCEEDINGS, 49, 511 (1963).16 Marcus, L., R. K. Bretthauer, R. M. Bock, and H. 0. Halvorson, these PROCEEDINGS, 50,

782 (1963).16Elson, D., Biochim. Biophys. Acta, 36, 372 (1959).17 Noll, H., T. Staehelin, and F. 0. Wettstein, Nature, 198, 632 (1963).18Williamson, A. R., and R. Schweet, Nature, 202, 435 (1964).19 Young, D. M., W. F. Harrington, and W. W. Kielly, J. Biol. Chem., 237, 3116 (1962).20 Huxley, H. E., J. Mol. Biol., 7, 281 (1963).

2-MERCAPTOETHYLAMINE AND 3-ALANINE AS COMPONENTS OFACYL CARRIER PROTEIN*

BY F. SAUER,t E. L. PUGH,4 SALIH J. WAKIL, ROBERT DELANEY,§AND ROBERT L. HILL

DEPARTMENT OF BIOCHEMISTRY, DUKE UNIVERSITY MEDICAL CENTER

Communicated by Philip Handler, October 21, 1964

The synthesis of long-chain fatty acids from acetyl CoA and malonyl CoA bysoluble extracts from E. coli has been studied by Vagelos and co-workers1' 2 andby Bloch and his group.3 These workers found that the responsible enzyme systemwas separable into a heat-stable and a heat-labile fraction, both of which were re-quired for the synthesis of palmitic and cis-vaccenic acids. Recently, Majeruset al.4 and Wakil et al., independently, found that the activity of the heat-stablefraction resided in a protein with a molecular weight of about 9,000. Duringoperation of the system, the heat-stable protein accepts the acyl groups fromacetyl or malonyl CoA so as to form covalently linked acetyl and malonyl deriva-tives which have been found to be the immediate substrates for the fatty acidsynthesizing enzymes. Thus, the heat-stable protein serves as a coenzyme ratherthan as an enzyme. The free form acts as an acyl acceptor while the acylated formcan serve as acyl donor; all of the reactions in which the fatty acyl chain is elon-gated, reduced, and dehydrated appear to occur while the chain is in an acyl linkageto this protein. For this reason the heat-stable protein has been designated anacyl carrier protein, abbreviated herein as "ACP."

Vagelos and his group4 as well as we5 have presented evidence which suggeststhat fatty acid biosynthesis proceeds according to the following sequential steps:

CH3COSCoA + ACPSH = CH3COSACP + CoASH (1)HOOCCH2COSCoA + ACPSH = HOOCCH2COSACP + CoASH (2)

CH3COSACP + HOOCCH2COSACP oCH3COCH2COSACP + ACPSH + CO2 (3)

Dow

nloa

ded

by g

uest

on

June

11,

202

1