Proc. Nat. Acad. Sci. USAVol. 69, No. 2, pp. 353-357, February 1972

A Ribosome Dissociation Factor from Rabbit Reticulocytes(fluoride/E. coli ribosomes/initiation factors)

NICOLETTE H. LUBSEN AND BERNARD D. DAVIS

Bacterial Physiology Unit, Harvard Medical School, Boston, Massachusetts 02115

Contributed by Bernard D. Davis, November 26, 1971

ABSTRACT A ribosome dissociation factor has beendetected in an extract of ribosomal particles from rabbitreticulocytes. This factor dissociates free ribosomes fromreticulocytes and also from Escherichia coli; it does notdissociate ribosomes complexed with peptidyl-tRNA andmRNA. The reaction appears to be stoichiometric ratherthan catalytic; it reaches completion in one minute at370C, but is very slow at 0C, and it is antagonized andreversed by Mg++. Reticulocyte dissociation factor thusclosely resembles that from E. coli. However, the activityhas been found primarily associated with the native largesubunits rather than the small subunits in lysates.

Echerichia coli contains a ribosome dissociation factor (DF)that converts free ribosomes into a stable pair of subunits(1-4); it has been identified with initiation factor F3 (2, 5-7),and in extracts it is found to be attached, like the otherinitiation factors (8), to the native small subunits (1). Thesupply of DF evidently regulates the concentration ofribosomal subunits (1, 9), which are required for the initiationstep in protein synthesis.The ribosome-polysome cycle in eukaryotes closely

resembles that in E. coli. Lysates contain large and smallsubunits, single ribosomes, and polysomes; subunit exchangehas been demonstrated in yeast (10) and slime molds (11);the kinetics of labeling of various ribosomal particles hasprovided evidence for initiation by subunits in reticulocytes(12); there is a special initiating tRNA, Met-tRNAF (13, 14);and three initiation factors have been identified in reticulo-cytes (15-17). Moreover, just as a block in initiation inE. coli leads to the accumulation of free ribosomes (18), sucha block in reticulocytes, by NaF, causes a reversible accumu-lation of 80S ribosomes at the expense of the polysomes (19).This accumulation suggests that in reticulocytes, as inbacteria, subunits are formed, and their concentration isregulated, by complexing of runoff ribosomes with a disso-ciation factor.We now report the identification and some of the properties

of a ribosome dissociation factor from rabbit reticulocytes.While this work was in progress, a similar factor was de-scribed from yeast (20) and, in preliminary reports, fromliver cells* and reticulocytest.

MATERIALS AND METHODS

Preparation of "Fluoride" and "Puromycin" ReticulocyteRibosomes. Rabbit reticulocytes were prepared as described

Abbreviations: DTT, dithiothreitol; DF, dissociation factor.* Lawford, G. R., Kaiser, J. & Hey, W. C. (1971) Fed. Proc. 30,1311 (Abstr.).t Favelukes, G., Sorgentini, D., Bard, E. & Martone, C. (1970)Abstract, Presented at 8th Intern. Congr. Biochem., Montreux.

(17). To promote polysome runoff the cells were incubatedfor 30 min at 370C with 10 mM NaF, in the medium describedby Rabinovitz et al. (21) (except that rabbit transferrin wasomitted). The cells were then chilled, centrifuged, washedtwice in cold saline, and lysed in four volumes of 3 mMMgCl2-2 mM dithiothreitol (DTT). Cell debris was removedby centrifugation for 15 min at 27,000 X g, and the ribosomeswere pelleted from the supernatant by centrifugation at300,000 X g for 2 hr. These ribosomes were suspended (atabout 250 A20) in standard buffer [10 mM Tris (pH 7.4)-100 mM KC1-2 mM DTTJ with 3 mM MgCl2. The prepara-tions were stored either in small portions at -76°C or onice; the ribosomes were stable at 0°C for about 1 month.Free ribosomes were also prepared by incubation of cells

for 30 min with 75 ,ug/ml of puromycin and then treatmentas above, except that the puromycin ribosomes, because oftheir lower stability (see Results), were resuspended inlow-KC! buffer [10 mM Tris (pH 7.4)-10 mM KC1-2 mMDTT] with 1.5mM MgCl2.

Preparation of Polysomes from Reticulocytes. A ribosomalpellet was obtained as described above, except that therunoff step was omitted and the ribosomal particles werepelleted for only 1 hr. The pellet was gently resuspended inthe low-KCl buffer and was layered on a linear 15-30%sucrose gradient (28 ml) in the same buffer. After centrifuga-tion for 2.5 hr at 24,000 rpm in a Spinco SW 25.1 rotor thepolysome fractions, identified by passage through an ISCO(Instrument Specialties Co., Lincoln, Nebraska) gradientanalyzer, were collected, pooled, and pelleted by centrifuga-tion for 10 hr at 300,000 X g. The pellet was resuspended inthe same buffer and stored in small portions at -76°C.

Preparation of Reticulocyte DF. Ribosomes from untreatedreticulocytes (about 20 ml of packed cells) were extracted,as in the preparation of initiation factors (17), with 0.5 MKCl containing 0.25 M sucrose-0.1 mM EDTA-2 mM DTT.The ribosomal wash (5 ml) was chromatographed on DEAE-cellulose (50 ml) essentially as described for initiation factors(17), except that the column was eluted stepwise with 0.1,0.2, 0.3, and 0.4 M KCl. The bulk of the DF activity waseluted by the 0.2-0.3 M step; these fractions were pooled,concentrated by ultrafiltration, and precipitated by theaddition of solid ammonium sulfate to 70% saturation.The precipitate was dissolved in a small volume of modifiedstandard buffer containing 0.1 mM EDTA instead of MgCl2.After dialysis against the same buffer for 3 hr the partlypurified DF was stored in small portions at -76°C.

Protein concentration was measured as described byLowry (22).

353

Dow

nloa

ded

by g

uest

on

Apr

il 2,

202

0

354 Biochemistry: Lubsen and Davis

A B 0.2010%/ 26%

0.10

. 0I100/60

0.10

3 2 3 2ml from bottom

FIG. 1. Dissociation of fluoride ribosomes. Fluoride ribosomesincubated at 0.15 mM Mg++ for 10 min as described in Methods:(A) without DF at 370C, (B) with 25 Aug of DF at 370C, and (C)with DF at 00C. In (D), dissociation, achieved as in (B), wasreversed by addition of MgC12 to 5 mM and incubation for anadditional 10 min. The % dissociation, determined from the areasunder the peaks, is indicated.

DF assays

Fluoride Ribosomes. DF was added as specified to 0.2 A20units of fluoride ribosomes in 100 ul of standard buffer con-taining 0.15 mM MgCl,. In assays with various amounts ofDF in EDTA-buffer, ionic conditions were kept constant byappropriate additions of the same solvent. Unless otherwiseindicated the mixtures were incubated for 10 min at 37CC,chilled on ice, and layered on a 10-30% linear sucrose gradient(10 mM Tris-100 mM KCl-0.5 mM MgCl2). The gradientswere spun for 60 min at 60,000 rpm in the SB405 rotor in anInternational B60 centrifuge. The absorbance profiles, ob-tained with a gradient analyzer, were quantitated by cuttingout peaks and weighing the paper.

Puromycin Ribosomes. Reaction mixtures were the sameas for fluoride ribosomes, except that the Mg++ concentrationwas raised to 0.3 mM to reduce the background dissociation.The sucrose gradients were 15-30% and contained 3 mMinstead of 0.5 mM MgCl2. Centrifugation was for 60 minat 48,000 rpm in the Spinco SW 50.1 rotor. The gradients wereanalyzed as described above.

Results are presented either as percent of total ribosomalparticles sedimenting as subunits or as percent of the initiallypresent 80S riboeomes that were caused to dissociate.E. coli Ribosomes. Ribosomes were prepared from E. coli

strain MRE600 and converted to free ribosomes by washingwith 1 M NH4Cl (6). DF action on these ribosomes wasassayed as described above, except that the incubation mix-ture contained 10 mM Tris-60 mM KC1-3 mM MgCl,-2 mMDTT and the sucrose gradients contained 10 mM Tris-60mM KC1-5 mM MgC12.

Separation and extraction of ribosomal subunits

Ribosomes were pelleted from a reticulocyte lysate for 3 hrat 300,000 X g and resuspended in low-KCl buffer. About

400 A2m0 units were layered on a 28-ml 15-30% sucrosegradient in the same buffer and spun for 7.5 hr at 23,000 rpmin a SW 25.1 Spinco rotor. The "native" 40S, "native" 60S,and 80S fractions were collected. For further purificationeach fraction was pelleted and then resuspended in 100lOof 0.25 M sucrose-0.1 mM EDTA-2 mM DTT (17), whichappeared to minimize aggregation. 1 ml of the low-KCl bufferwas added, the suspension was sedlimented through a sucrosegradient as before, the appropriate fractions were collected,and the ribosomal particles were pelleted by centrifugation.The pellets were each resuspended in 100 ,ul as above and asmall portion was removed for gradient analysis.DF was extracted from each suspension of ribosomal

particles with 0.5 M KC1, as described for initiation factors(17). After centrifugation for 5 hr at 200,000 X g, the super-natant was carefully withdrawn, dialyzed for 4 hr against twochanges of standard buffer with 0.1 mM EDTA, and ana-lyzed immediately for DF activity with 0.25 A260 units ofE. coli ribosomes. The dilution of the extract during dialysiswas estimated by determination of the decrease in proteinconcentration.

RESULTSDissociation factor (DF) from reticulocytes; preparationof responsive ribosomes

Bacterial DF acts only on free ribosomes and not on ribo-somes complexed with mRNA and peptidyl-tRNA (2).Hence, to test for a similar factor from reticulocytes wefirst used ribosomes prepared from reticulocytes incubatedwith NaF (10 mM), which inhibits initiation, perhaps bypreventing the attachment of a 60S subunit to an initiating40S subunit (23). Some ribosomes remain attached to mRNAafter this treatment (24), but most should be free.

Reticulocyte DF was sought in a crude initiation factorpreparation, obtained by extraction of ribosomes with asolution containing 0.5 M KC1 (Methods). Small amounts ofthis extract caused slight but detectable dissociation of thefluoride ribosomes. With larger amounts, however, most ofthe ribosomes aggregated and hence were pelleted on sub-sequent analysis, which revealed only a small peak sedi-menting broadly in the 40S region. The aggregation is likelyto be due to some supernatant component in the initial pellet,

A B C

040

2020

0

50 100 50 100 50 100

DF Odded (mg)

FIG. 2. Response of different kinds of reticulocyte ribosomesto DF. (A) Purified reticulocyte polysomes (0 0) were in-cubated for 5 min at 370C with 100lg/ml of puromycin and 0.5mM GTP in 10 mM Tris-10 mM KCl-1.5 mM MgC1r-2 mMDTT and then diluted 100-fold into the usual DF assay mixture(0.12 Am,. units/assay) for puromycin ribosomes. These ribosomes,and untreated polysomes (X---X) (fragmented during isolation),were incubated with various amounts of DF. (B) 0.15 A,60 unitsof ribosomes from puromycin-treated cells was incubated withDF. (C) Ribosomes from fluoride-treated cells, treated as in (B).

Proc. Nat. Acad. Sci. USA 69 (1972)

Dow

nloa

ded

by g

uest

on

Apr

il 2,

202

0

Reticulocyte Dissociation Factor 355

since ribosomes isolated from a sucrose gradient yielded a non-aggregating extract. The "aggregation factors" were readilyremoved: on DEAE-cellulose chromatography most of theDF activity was eluted in the 0.2-0.3 M KCl fraction, whichcaused no aggregation of ribosomes (i.e., no loss in totalabsorbance) and yielded sharp peaks (see Fig. 1A and B).This fraction was used in all further studies.

Fig. 1B shows a typical dissociation pattern, obtained byincubation of fluoride ribosomes (background dissociation10%) with partially purified DF at 370C. The activity wasdestroyed by heating to 80'C for 5 min (data not shown).Complexed reticulocyte ribosomes (i.e., polysomes isolated

from a sucrose gradient) were not detectably disso-ciated by reticulocyte DF, even in considerable excess; buttreatment with puromycin in the presence of GTP convertedabout 30% of these ribosomes into a DF-responsive form(Fig. 2A). The incomplete conversion was presumably due tothe absence of supernatant factors (whose addition causedaggregation).With fluoride ribosomes the extent of dissociation was

essentially proportional to the amount of DF added (Fig.2C), but the curve leveled off at 35% dissociation of the 80Sribosomes in the test sample. Since this limit might be due tothe presence of complexed ribosomes, we tested the action ofDF on lysates from cells that had been incubated with puro-mycin instead of NaF. As Fig. 2B shows, these "puromycinribosomes" were more completely dissociated by DF thanwere "fluoride ribosomes" (Fig. 2C). However, the puromycinribosomes exhibited considerable background dissociation incontrol incubations without DF, even at a higher Mg++concentration than usual (0.3 mM instead of 0.15 mM).Hence, fluoride ribosomes were used in most subsequentexperiments.

Recently we have prepared free ribosomes that were moresatisfactory by the method of Blobel and Sabatini (25), inwhich polysomes are exposed to puromycin in the presenceof 0.5 M KCl and then the salt is diluted to permit reasso-ciation of the subunits. The background of dissociation waslow and dissociation by DF was almost complete.

Characteristics of the reaction

Dissociation by reticulocyte DF, as by E. coli DF (1), israpid at 370C: with a DF concentration giving submaximaldissociation the reaction was nearly complete within 1 min(Fig. 3). These kinetics, and the essentially linear concen-tration-activity curves of Fig. 2, suggest that dissociation

TABLE 1. Reversibility of the DF reaction byMg++ at 370C and 00C

Incubation conditions DF(,ug) %80S

Unincubated 0 7210 min at 370C (0.3 mM Mg++) 0 44

65 2910 min at 37°C; add Mg++ to 5 mM; 0 67incubate 15 min 65 72

10 min at 37°C; cool to 0°C; add Mg++ 0 65to 5 mM; hold 15 min at 0°C 65 37

Puromycin ribosomes were incubated with and without DF,as indicated, in the appropriate assay mixture. Further treatmentas specified, followed by chilling and analysis.

c0

'5?_O0-e

Minutes0.5 0.9

mM MgCI2

FIG. 3 (left). Rate of dissociation. 700 yl of reaction mixture,containing 0.2 A260 units of fluoride ribosomes per 100lO, wasbrought to 37°C, 100Ml was removed as a zero-time sample, and230,g of DF was added. Samples were taken at intervals andtransferred immediately to chilled gradients.

FIG. 4 (right). Magnesium dependence of the dissociationreaction. Fluoride ribosomes were incubated with 47,g of DF asdescribed in Methods, except that the Mg++ concentration wasadjusted as noted. The % dissociation of a control sample in-cubated without DF under the same ionic conditions was sub-tracted; these values ranged between 10 and 15%.

results from complex formation rather than from damage tothe ribosomes by an enzyme in the preparation.

Also as with E. coli DF (6), the effect of reticulocyte DFwas decreased markedly at increased Mg++ concentrations(Fig. 4). Moreover, dissociation achieved under the standardconditions was completely reversed by raising the Mg++ to5 mM and then incubating for 10 min (Fig. 1D).The reaction is very sensitive to temperature: no response

was detected in 10 min at 0°C (Fig. 1C). Moreover, at 0°Can elevated Mg++ concentration did not reverse dissociationthat had been produced by reticulocyte DF at 370C, thoughit did reverse (as at 37°C) the spontaneous dissociation ob-served with puromycin ribosomes (Table 1). These findingssuggest that reversal of complex formation between DF anda ribosomal particle, like the forward reaction, is very slowat 0°C; and the persistent attachment of DF may preventMg++ from overcoming conformational effects that keep thesubunits separate.

Activity of reticulocyte DF on E. coli rihosomes

The DF preparation used also dissociated E. coli ribosomeswhen tested at a Mg++ concentration of 3 mM. This activitycochromatographed, on DEAE-cellulose and Sephadex G-200,with the reticulocyte DF (unpublished observations). Anotherfraction, chromatographing separately, dissociated onlyE. coli ribosomes; it has not been characterized. Like reticulo-cyte ribosomes, E. coli ribosomes were not dissociated byreticulocyte DF at 0°C, and their dissociation at 37°C wasreversed by an elevated concentration of Mg++ (Fig. 5).Moreover, the reaction is restricted to free ribosomes. Thus,with a preparation containing polysomes as well as singleribosomes an excess of reticulocyte DF dissociated a fractionthat corresponded closely to the free ribosomes (Fig. 6), asdetermined by sedimentation in a gradient containing 50 mMNaCl instead of KCl (26). With a preparation freed of ligandsby washing with 1 M NH4C1 about 80% of the ribosomeswere dissociable (Fig. 6).The crossreaction with E. coli is not symmetrical: an

E. coli DF preparation, in an amount that dissociated 30 ugof E. coli ribosomes, did not dissociate reticulocyte ribosomesdetectably (i.e., <6,ug).

Proc. Nat. Acad. Sci. USA 69 (1972)

,0

0

0

Dow

nloa

ded

by g

uest

on

Apr

il 2,

202

0

356 Biochemistry: Lubsen and Davis

2 3 4 2 3 4ml from bottom

FIG. 5. Dissociation of E. coli ribosomes. E. coli ribosomes(0.25 A2860 units), washed with 1 M NH4Cl, were incubated at3 mM Mg++: (A) without DF at 370C, (B) with 24 /g of reticu-locyte DF at 370C, and (C) with DF at 0°C. (D) Reversibilityby Mg++ was measured as described in Fig. 1, except that thefinal concentration of MgCl2 was 15 mM. The percentage ofA254 in the 70S peak is indicated.

The distribution ofDF among ribosomal particles

For determination of the distribution of DF among theparticles in a reticulocyte lysate the 80S, 60S, and 40Sfratetions were separated by sucrose gradient centrifugationand the DF was extracted from each and assayed. Sur-prisingly, the amount of DF recovered per particle washighest for the 60S fraction. However, this fractioncontained a large number of presumably dimerized 40Ssubunits (27), as shown by resedimentation and by analysisof the RNA in gradients that contained sodium dodecylsulfate. All three fractions were therefore further purified byan additional centrifugation and isolation of the peaks.

After this second centrifugation, the 80S and 40S fractions,and the 40S fraction derived from the initial 608 preparation,were essentially homogeneous on reanalysis. The second-stage60S preparation still contained contaminating 80S and 40Sparticles (Fig. 7). This preparation yielded 4-7 times moreDF activity per mole than did either 40S preparation, and 20times more than the 80S ribosomes (Table 2). These assayswere performed with E. coli ribosomes; assays with reticulo-cyte ribosomes are less accurate, but they gave similarresults.

DISCUSSION

This paper shows that a fraction from a high-salt extract ofrabbit reticulocyte ribosomal particles selectively dissociatesfree but not complexed reticulocyte ribosomes. While this re-sponse could be accounted for by a physiological dissociationfactor (DF), it could also conceivably be due to binding offree Mg++ by the added fraction-especially since the Mg++concentration is low (0.15 and 0.3 mM) in the assays, in whichtest and control samples were adjusted to the same totalionic concentrations. However, the activity of reticulocyteDF was destroyed by heating to 800C. Moreover, the ac-

tivity was not altered by dialysis of the DF solution againstthe same medium in which it was to be tested. Finally, thereticulocyte DF was also highly active with E. coli ribosomes,tested at 3 mM Mg++.

Dissociation could also be caused by damage to the ribo-somes by the action of an RNase or a protease in the addedfraction. However, an enzymatic reaction would be expectedto cause a continuous increase in dissociation with time, butthe DF reaction is complete in about 1 min (Fig. 3). More-over, in all respects tested (kinetics, effects of Mg++ and tem-perature, reversibility, concentration- activity curve) the re-action closely resembles that of E. coli DF, for which pure ma-terial of known molecular weight (5) has provided data thatstrongly support a stoichiometric rather than a catalytic reac-tion (6). Finally, the reaction with reticulocyte DF does notappear to damage the ribosomes. Thus elevation of the Mg++concentration causes the subunits to reassociate at 370C butnot at 0C (Figs. 1 and 6; Table 1), at which dissociation isalso very slow. This finding suggests that the dissociation de-pends on persistent complexing with DF and not on an en-zyme-catalyzed modification of a subunit.The similarity to E. coli DF suggest that reticulocyte DF

should likewise be an initiation factor. Its elution fromDEAE-cellulose between 0.2 and .0.3 M KCl would suggestfactor M2, which elutes at 0.27 M KCl (17), as the mostlikely candidate. Moreover, a preparation of M2, kindlysupplied by Dr. W. F. Anderson, had dissociation activity,while a preparation containing both Ml and M3 did not.However, E. coli F3 (identical with DF) plays a role in therecognition of messenger (29), and it is not certain whether

801

c0

0

0

._

60

40

20

_-0

-o .

/

I L~--A

5 10 15 20DF added (ILI)

2 3 dml from bottom

.20

N(I

.10

FIG. 6 (left). DF dissociability of free and complexed E. coliribosomes. 0.25 A260 units of E. coli ribosomes washed with 1 MNH4Cl, or 0.17 units of an unwashed preparation containing bothfree ribosomes and polysomes, were incubated with differentamounts of reticulocyte DF (6.5 mg/ml) and analyzed. With thepolysomal preparation, 1 ,g/ml of pancreatic RNase was addedafter the incubation and chilling and the mixture was left on icefor 5 min; this treatment converted the polysomes into complexedmonosomes. This preparation, with and without added DF, was

also analyzed in a gradient containing 50 mM NaCl, which selec-tively dissociates free ribosomes. DF and NaCl evidently dis-sociate the same fraction of the ribosomal preparation. 0 *,NaCl gradient; X X, KC1 gradient, unwashed ribosomes;O -O, KC1 gradient, NH4Cl-washed ribosomes.

FIG. 7 (right). Absorbance profile of the 60S fraction after a

second centrifugation. 0.2 A260 units of the 60S preparation (Table2) was analyzed in a 15-30% sucrose gradient with 10 mM Tris-100mM KC1-3mM MgC12.

Proc. Nat. Acad. Sci. TUSA 69 (1972)

Dow

nloa

ded

by g

uest

on

Apr

il 2,

202

0

PReticulocyte Dissociation Factor 357

TABLE 2. DF activtity of extracts from ribosomal fractions

Ribosomal DF activityparticles Total per picomole

Total (pmol) DF of ribosomalFraction A260 40S 60S 80S activity particle

40S 7.7 6.5 0 0 185 2940S from 5.7 4.8 0 0 74 16

initial 60Sfraction

60S 5.4 1.0 1.0 0.2 130 108*80S 70 0 0 16 95 6

Each fraction was analyzed for distribution of ribosomalparticles and for DF content. Moles of ribosomal particles werecalculated on the assumption that one A20 unit equals 90 /g of80S, that the molecular weight of the 80S particle equals 4.0 X101 (28), and that the A260 ratio of the 60S to the 40S subunit is2.7.

* The activity per 60S unit was calculated after correction forthe activity of the contaminating 40S particles, on the assump-tion that the activity per picomole of 40S in this fraction is thesame as the activity per picomole in the 40S fraction itself.

this function is performed in reticulocytes by M2 or M3(16, 30). Until either reticulocyte DF or M2 is purified, theiridentity cannot be considered firmly established.The evolution of parts of the protein-synthesizing ma-

chinery has been remarkably conservative. For example,E. coli factor T and mammalian Transferase I are reportedto be interchangeable on reticulocyte ribosomes, though noton E. coli ribosomes (31). We have found the converse withDF: reticulocyte DF is active (at a permissible Mg++ con-centration) with E. coli ribosomes, but E. coli DF is not activewith reticulocyte ribosomes. Curiously, the dissociation factorof yeast and that of E. coli are reported to show no crossre-action in either direction (20).To distinguish a bound ribosomal factor from a true ribo-

somal protein that is easily extractable, it must be shown to bepresent on the ribosome at one stage in its cycle but absentin another (9). Thus, bacterial DF has been shown to bepresent in the native small subunits but not in the polysomalribosomes (1). This criterion has also been met for reticulo-cyte DF: only a negligible amount was found in the 80Sfraction, compared to the amount in the subunits (Table 2).To our surprise, however, the 60S fraction, even after furtherpurification, yielded several times as much DF as the equiva-lent amount of the 40S fraction; and though the preparationswere not free of crosscontamination an exclusive attachmentto 408 particles is clearly ruled out. Moreover, though 40Ssubunits from reticulocyte ribosomes tend to form dimers,which sediment at 60S (27), these could not account for thefindings: for the 40S particles recoverable from the 60S frac-tion would then be expected to be richer in DF than the second-stage 40S or 60S particles, and they were actually much poorer(Table 2). Nevertheless, since E. coli DF clearly acts by com-plexing with the small subunit, and since both E. coli andreticulocyte DF can act on E. coti ribosomes, we are reluc-tant to conclude that reticulocyte DF acts by binding to the

large subunit. The possibility of redistribution during prepa-ration and fractionation of the reticulocyte lysate cannot beexcluded.

We are deeply indebted to Drs. Edgar Henshaw and CarlHirsch of the Beth Israel Hospital, without whose generousassistance and advice this work could not have been initiated.We are also grateful for grants from the U.S. Public HealthService, the National Science Foundation, and the AmericanCancer Society. This work was submitted by N. H. L. in partialfulfillment of the requirements for the Ph.D. degree in the De-partment of Biological Chemistry.

1. Subramanian, A. R., Ron, E. Z. & Davis, B. D. (1968)Proc. Nat. Acad. Sci. USA 61, 761-767.

2. Subramanian, A. R., Davis, B. D. & Beller, R. J. (1969)Cold Spring Harbor Symp. Quant. Biol. 34, 223-230.

3. Algranati, I. D., Gonzalez, N. S. & Bade, E. G. (1969) Proc.Nat. Acad. Sci. USA 62, 574-580.

4. Albrecht, J., Stap, F., Voorma, H. O., Van Knippenburg,P. H. & Bosch, L. FEBS Lett. 6, 297-301.

5. Sabol, S., Sillero, M. A. G., Iwasaki, K. & Ochoa, S. (1970)Nature 228, 1269-1273.

6. Subramanian, A. R. & Davis, B. D. (1970) Nature 228,1273-1275.

7. Dubnoff, J. S. & Maitra, U. (1971) Proc. Nat. Acad. Sci.USA 68, 318-323.

8. Eisenstadt, J. M. & Brawerman, G. (1967) Proc. Nat. Acad.Sci. USA 58, 1560-1565.

9. Davis, B. D. (1971) Nature 231, 153-157.10. Kaempfer, R. (1969) Nature 222, 950-953.11. Ceccarini, C., Campo, M. S. & Andronico, F. (1970) J. Cell

Biol. 46, 428-430.12. Howard, G. A., Adamson, S. D. & Herbert, E. (1970) J.

Biol. Chem. 245, 6237-6239.13. Smith, A. E. & Marcker, K. A. (1970) Nature 226, 607-610.14. Housman, D., Jacobs-Lorena, M., Rajbhandary, U. L. &

Lodish, H. F. (1970) Nature 227, 913-918.15. Shafritz, D. A., Prichard, P. M., Gilbert, J. M. & Anderson,

W. F. (1970) Biochem. Biophys. Res. Commun. 38, 721-727.16. Prichard, P. M., Gilbert, J. M., Shafritz, D. A. & Anderson,

W. F. (1970) Nature 226, 511-514.17. Shafritz, D. A. & Anderson, W. F. (1970) J. Biol. Chem. 245,

5553-5559.18. Kohler, R. E., Ron, E. Z. & Davis, B. D. (1968) J. Mol.

Biol. 36, 71-82.19. Colombo, B., Vesco, C. & Baglioni, C. (1968) Proc. Nat.

Acad. Sci. USA 61, 651-658.20. Petre, J. (1970) Eur. J. Biochem. 14, 399-405.21. Rabinovitz, M., Freedman, M. L., Fisher, J. M. & Maxwell,

C. R. (1969) Cold Spring Harbor Symp. Quant. Biol. 34,567-578.

22. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall,R. J. (1951) J. Biol. Chem. 193, 265-275.

23. Hoerz, W. & McCarthy, K. S. (1971) Biochim. Biophys.Acta 228, 526-535.

24. Lebleu, B., Huez, G., Burny, A. & Marbaix, G. (1967) Bio-chim. Biophys. Acta 138, 186-188.

25. Blobel, G. & Sabatini, D. (1971) Proc. Nat. Acad. Sci. USA68, 390-394.

26. Beller, R. J. & Davis, B. D. (1971) J. Mol. Biol. 55, 477-485.27. Tashiro, Y. & Morimoto, T. (1966) Biochim. Biophys. Acta

123, 523-533.28. Dibble, W. E. & Dintzis, H. M. (1960) Biochim. Biophys.

Acta 37, 152-153.29. Iwasaki, K., Sabol, S., Wahba, A. J. & Ochoa, S. (1968)

Arch. Biochem. Biophys. 125, 542-547.30. Crystal, R. G., Shafritz, D. A., Prichard, P. M. & Anderson,

W. F. (1971) Proc. Nat. Acad. Sci. USA 68, 1810-1814.31. Krisko, I., Gordon, J. & Lipmann, F. (1969) J. Biol. Chem.

244, 6117-6123.32. Felicetti, L. & DiMatteo, G. F. (1970) J. Mol. Biol. 54,

611-620.

Proc. Nat. Acad. Sci. USA 69 (1972)

Dow

nloa

ded

by g

uest

on

Apr

il 2,

202

0