THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 258 No. 1, Issue of .January 10, pp. 571-578,1983 Printed'in U. S. A.

Cell-free Translation of Frog Virus 3 Messenger RNAs INITIATION FACTORS FROM INFECTED CELLS DISCRIMINATE BETWEEN EARLY AND LATE VIRAL mRNAs*

(Received for publication, July 21, 1982)

Rajendra Raghow and Allan Granoff From the Division of Virology and Molecular Biology, St. Jude Children's Research Hospital, Memphis, Tennessee 38101

Cell-free protein-synthesizing extracts prepared from rabbit reticulocytes, wheat germ, or cultured baby hamster kidney cells efficiently translated frog virus 3 early mRNAs; in contrast, late mRNAs were translated poorly under similar conditions. However, the transla- tional efficiency of the late viral mRNAs was markedly enhanced in cell-free extracts prepared from frog virus 3 (FV 3)-infected baby hamster kidney cells and in nuclease-treated rabbit reticulocyte extracts by the ad- dition of a 0.5 M KC1 wash from FV 3-infected cell ribosomes; the 0.5 M KC1 wash (initiation factors) from uninfected cells had no such effect. Total cytoplasmic RNA from infected cells was fractionated according to size on sucrose gradients and fractions containing dif- ferent concentrations, and relative proportions of early and late mRNAs were translated in either native or initiation factor-supplemented extracts. Under these conditions, the translation efficiency of early mRNAs was unchanged, while the translation of late mRNAs increased 2-7-fold. Thus, the in vitro discriminatory activity of the 0.5 M wash was not dependent on the complexity of the mRNAs present in the translation mixture. We show also that in native extracts, under conditions of blocked polypeptide chain elongation, early mRNAs are initiated preferentially. However, late as well as early mRNAs are initiated equally well in reticulocyte extracts under similar experimental con- ditions when supplemented with crude initiation fac- tors from infected cells. These data support the conclu- sion that the translational enhancement of FV 3 mRNAs in vitro is mediated by a virus-specified or virus-modified initiation factor(s) and likely represents a regulatory mechanism of protein synthesis operative in vivo (Willis, D. B., Goorha, R., Miles, M., and Granoff, A. (1977) J. Virol. 24,326-342).

FV 3' is a member of the family Iridouiridae; its genome is a linear double-stranded DNA of 100 megadaltons. The expression of FV 3 genes and their regulation have been studied in considerable detail (1). Viral transcription occurs sequentially in at least three distinct phases; immediate early and early, which take place within the first 2 h after infection,

* This work was supported by Research Project Grant CA07055 and Cancer Center Support (Core) Grant CA21765 from the National Cancer Institute and by American Lebanese Syrian Associated Char- ities. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' The abbreviations used are: FV 3, frog virus 3; FHM, fathead minnow; BHK, baby hamster kidney; IF, initiation factor; Icp, in- fected cell proteins; ICR, infected cell mRNAs; KoAc, potassium acetate; Mg(oAc)?, magnesium acetate.

and late, which occurs thereafter ( 2 ) . Protein synthesis is regulated at both the transcriptional and post-transcriptional levels, with translational control the major regulatory mech- anism for both immediate early and early transcripts (2-4). For example, both classes of early RNAs are not only present late in infection, they are still being synthesized despite an inability to be translated ( 2 ) . Utilizing metabolic inhibitors and temperature-sensitive mutants, evidence has been ob- tained for the role several viral proteins play in this regulation (4-6). To gain an understanding of the mechanisms involved in post-transcriptional regulation of FV 3 genes, we undertook translation of viral messenger RNAs in a variety of initiation- competent in vitro protein-synthesizing systems. In this re- port, we present evidence for a modulator protein(s), probably virus-specified, that selectively enhances the translation of late viral mRNAs.

EXPERIMENTAL PROCEDURES

Cells and Virus-FHM cell monolayers were grown at 33 "C in Eagle's minimal essential medium containing 10% fetal calf serum. Suspension spinner cultures of BHK cells were propagated in Joklik's- modified Eagle's minimal essential medium containing 10% fetal calf serum. Procedures for the propagation, purification, and assay of FV 3 have been described previously (7,8).

Preparation of Cell-free Translation Extracts-Conditions for the preparation of rabbit reticulocyte lysates have been described (9). In uitro translation extracts from cultured BHK cells, either uninfected or FV 3-infected 6-h postinfection, were also prepared according to the method described by Villa-Komaroff et al. (9) with two modifi- cations. (i) Cell extracts were made in 4-(2-hydroxyethyl)-l-pipera- zineethanesulfonic acid buffer instead of Tris-HCI. (ii) Cell lysates were supplemented with 50 p~ hemin as recommended by Weber et al. (IO). Micrococcal nuclease treatment of cell-free translation ex- tracts from cultured cells, as well as from reticulocytes, was essentially as described (11). Translation extracts from wheat germ (Bethesda Research Laboratories) were optimized for K' and Mg" ion concen- trations as recommended by the manufacturer. All cell-free extracts were stored at -80 "C in small aliquots.

Protein Synthesis Assay-Each 25 p1 of in vitro translation assay mixture contained 10 pl of lysate and the following components: 184 pM spermidine, 1 pM ["'S]methionine (1m Ci/mmol), 12 mM creatine phosphate, 50 pg/ml of creatine phosphokinase, 5 mM dithiothreitol, 90 pM concentration each of the I9 amino acids, and I mM EDTA. The concentration of K' and Mi2+ ions was optimized using KoAc and Mg(oAc),. Messenger RNA concentration (generally not exceed- ing 5.0 p1 in volume) in the translation extracts was varied in different experiments (see legends to figures). Reaction mixtures were incu- bated at 30 "C for 90 min, after which 1-5-pl aliquots were withdrawn for quantitation of radioactivity incorporated into polypeptides or for gel electrophoretic analyses.

Sodium Dodecyl Sulfate-Polyacrylamide Gel Eleetrophoresis- Samples (1-5 pl) of cell-free protein synthesis extracts were analyzed on 5-1570 gradient polyacrylamide slab gels using the gel system of Laemmli (12). Gels were subjected to 2,5-diphenyloxazole impregna- tion (13), dried, and fluorographed using Kodak XR-5 x-ray fiim.

Messenger RNA Preparation-Cytoplasmic RNA from FV %in- fected FHM or BHK cells (6 h postinfection) was extracted as described previously (4). Early FV 3 mRNAs were isolated from

571

572 Translational Discrimination of FV 3 m R N A s

infected cells treated with fluorophenylalanine (2). Since FV 3 mRNAs lack poly(A) (14), conventional procedures for mRNA en- richment by oligo(dT)-cellulose chromatography (15) could not be used. For some experiments, viral mRNAs were, therefore, selected by hybridization to FV 3 DNA immobilized on cellulose filters ac- cording to the method described by Raskas and Green (16). Purified mRNAs were ethanol-precipitated a t least twice before a final ethanol precipitation in 0.2 M KoAc, pH 5.4, at -20 "C. Messenger RNA precipitates were then washed once in 70% ethanol (-20 "C), dried under vacuum, and taken up in sterile distilled H,O. Before transla- tion, the mRNA solution in H,O was heated to 100 "C for 2 min, cooled quickly, and added to an in vitro translation extract.

Size Fractionation of Viral mRNA-FV 3-infected FHM cells were labeled with ["Hluridine (50 pCi/ml) between 2 and 6 h postin- fection. Radiolabeled cytoplasmic RNA (200 pg) was layered over a gradient of 15-30% sucrose in TENS buffer (0.005 M Tris-HCI, pH 7.4.0.001 M EDTA, 0.1 M NaCI, and 0.5% sodium dodecyl sulfate) and centrifuged in an SW 27 rotor a t 27,000 rpm for 18 h. T o minimize aggregation, RNA samples were taken up in distilled H20, heated to 100 "C for 2 min, and diluted with equal volumes of 2X TENS buffer prior to sedimentation. We collected 0.25-ml fractions while monitor- ing their absorbance a t 260 nm. Calf liver tRNA (5 pg), 50 pI of 1 M KoAc, and 3 volumes of absolute ethanol were added to each fraction and RNA was precipitated overnight a t -20 "C. RNA from alternate gradient fractions was divided into three equal aliquots. Two aliquots from each fraction were translated in either native or initiation factor- supplemented reticulocyte extracts, while the RNA in the third aliquot was precipitated and electrophoresed on denaturing formam- ide-acrylamide gels as previously described (2).

Preparation of Crude Initiation Factor Fraction-BHK cells grown in spinner cultures and infected with FV 3 (4) were used to isolate crude initiation factors. The methodology was essentially that of Schreier and Staehelin (17) with minor modifications as described below. Uninfected or FV 3-infected (6 h postinfection) cells (2.5 X IO9) were suspended in extraction buffer (Buffer E: 5 mM 4-(2-hy- droxyethy1)-I-piperazineethanesulfonic acid, pH 7.6, 2 mM Mg(oAc),, 0.1 mM EDTA, and 10% Triton X-100). Cell slurry was kept a t 0 "C with frequent agitation on a Vortex mixer and centrifuged a t 10,000 rpm for 10 min. The supernatant fraction was layered on top of two layers a t 7.0 ml each of 1.8 and 1 M sucrose in Buffer E and centrifuged a t 27,000 rpm for 18 h in an SW 27.1 rotor to pellet polyribosomes. After rinsing with sterilized double-distilled water, polysomes were resuspended in Buffer E and l/x volume of 4 M KC1 was added. The mixture was stirred a t 0 "C for 15 min and centrifuged a t 60,000 rpm for 2.5 h in an SW 65 rotor. Solid (NH.AS0, (final concentration 361 mg/ml) was added to this high salt wash fraction; the resultant precipitate was sedimented (10.000 rpm, 10 min), solubilized in Buffer E containing 10R glycerol and 100 mM KoAc, and dialyzed against 2 liters of the same buffer for 24 h at 0-4 "C. Small aliquots of crude initiation factor fractions were stored in liquid N2 until further use.

RESULTS

Cell-free Protein-synthesizing Extracts Efficiently Trans- late FV3 Early mRNAs But Not Late Ones-Fig. 1 illustrates and compares the polypeptides synthesized in reticulocyte and wheat germ extracts programmed either with purified globin mRNA (to compare the relative efficiency of the two extracts) or with cytoplasmic mRNA extracted from FV 3- infected FHM cells a t 6 h postinfection. A comparison of the polypeptides synthesized in response to FV 3 mRNAs in the two translation extracts in vitro and a corresponding spectrum of proteins made a t 6 h postinfection showed noticeable dif- ferences. The wheat germ system translated mRNAs greater than 5.5 X lo5 very poorly (Fig. 1, lane g). Furthermore, translation of viral mRNAs in the wheat germ extract also resulted in enhanced background radioactivity in the small molecular weight region of the gel, presumably due to a greater frequency of prematurely terminated translation prod- ucts (la). We also found that the translation of globin mRNA occurred less efficiently in wheat germ extracts (Fig. 1, lanes b and f ; 81,500 cpm versus 156,000 cpm in reticulocyte ex- tract).

Regardless of the differences in the overall efficiencies of translation in the two cell-free extracts, a more striking result

was obtained when we compared the putative in vitro FV 3 gene products with the in vivo ones. An inspection of in vitro products made in the reticulocyte or wheat germ extracts programed with total FV 3 mRNAs indicated that some of the putative viral gene products were either translated poorly (e.g. ICP 55) or not a t all (ICP 72 and 68; Fig. 1, lanes c and g). A comparison of the relative molar abundances of several viral polypeptides made in vivo and in vitro (Table I ) corrob- orated the conclusion drawn from the gel patterns; i.e. trans- lation of late FV 3 mRNAs is relatively inefficient in cell-free systems. Co-electrophoresis of viral polypeptides made in wheat germ and reticulocyte extracts programed with total FV 3 mRNAs revealed a similar qualitative pattern (Fig. 2). However, there were significant quantitative differences. An apparent preponderance of low molecular weight polypeptides in the wheat germ extract, depicted graphically by plotting

RETICULOCYTE WHEAT GERM

" - -73 A

tr

J -68 A'

-55 A

-36

- 18.

a b c d e f g h FIG. 1. A comparison of cell-free protein-synthesizing activ-

ities of rabbit reticulocyte or wheat germ extracts programed with either globin mRNA or with total cytoplasmic RNA ex- tracted from FV 3-infected FHM cells (6 h postinfection). Con- ditions of incubation and analysis of the translation products on 5- 15% gradient Laemmli gels are described in the text. Lanes a-c are reticulocyte extracts programed with no mRNA (a ) , 10 pg/ml of globin mRNA (b ) , or 100 pg/ml of total FV 3 mRNA (c). Lanes e-g are wheat germ extracts programed without mRNA (e) , with 10 pg/ ml of globin mRNA (f ), or with 100 pg/ml of FV 3 mRNA (g). Lanes d and h represent proteins synthesized in FV 3-infected BHK cells between 5 and 6 h postinfection. Some of the early (0) or late (A) gene products have been identified. The numbers indicate molecular weight X 10' of infected cell proteins (ZCP).

TABLE I Relative molar abundances of viral polypeptides synthesized in

vivo and in vitro Molecular weights of polypeptides were determined as described

previously (2). The individual bands representing polypeptides syn- thesized either in infected cells a t 6 h postinfection or made in reticulocyte extracts programmed with total FV 3 mRNAs were excised and radioactivity was measured. Relative molar ratios were calculated by dividing radioactive counts in a given polypeptide by its estimated molecular weight and normalizing all values so that early ICP 36 equals 1 unit. Other early ICPs are 115 and 18; ICPs 72, 6 8 , and 55 are late ones.

Polvneotide Molar ratio in uivo Molar ratio in vitro

ICP 115 ICP 72 0.15 tO.OO1 ICP 68 0.19 0.003 ICP 55 2.07 0.61

ICP 18 4.55 5.24

~~

0.15 0.14

Icr' 36 1 .o 1.0

Tra.nslationa1 Discrimination of FV 3 mRNAs 573

. . 4 . . . .

0 10 20 30 40 50 60 SLICE NUh'O,ER

FIG. 2. Co-electrophoresis of viral polypeptides made in wheat germ and reticulocyte cell-free extracts. Conditions of in vitro translation are the same as described under Fig. 1 except that translation products were radiolabeled with "C-aminoacids in wheat germ and with "H-aminoacids in reticulocyte extract. Following in- cubation, 5 pl from each extract were co-electrophoresed on a 5-158 polyacrylamide gradient gel. Radioactivity in the individual 1.5-mm slices of the gel strip was determined (top). The positions of several early (0) and late (A) viral proteins have been identified; the numbers indicate molecular we;:ht X IO'. The relative distribution of homol- ogous polypeptides made in the two translation systems is depicted by plotting "C/'H ratios throughout the gel strip (bottom).

I4C/"H ratios along the gradient gel (Fig. 2), corroborates the visual pattern shown in Fig. 1, lane g, suggesting a greater frequency of premature termination in this sytem. Thus, wheat germ was clearly a less suitable system, and all subse- quent experiments utilized nuclease-treated reticulocyte ex- tract.

Based on these experiments, we hypothesized that (i) the absence of some of the late viral gene products in the in vitro extracts was due to a preferential breakdown of the late mRNAs in these cell-free systems, or (ii) the cell-free extracts could lack a factor(s) present in the infected cells that was required for the translation of late FV 3 mRNAs. We tested both these hypotheses in a series of experiments described below.

Late FV 3 mRNAs Are Stable in the Cell-free Reticulocyte Extracts-One conceivable mechanism to account for the comparatively scant amounts of late viral polypeptides syn- thesized in reticulocyte extracts would be a preferential deg- radation of late FV 3 mRNA in the extracts. If this were the case, the addition of a high salt wash from infected cells should enhance the stability of the late mRNAs, thus mimick- ing the type of translational control found in the preceding experiments. We, therefore, compared the relative stabilities of early and late viral mRNAs in reticulocyte extracts with or without the high salt wash fraction. Following a 30-min incu- bation of equal aliquots of radiolabeled FV 3 mRNAs in cell- free extracts, we analyzed the phenol-extracted RNAs in denaturing polyacrylamide gels. The patterns of mRNAs in- cubated in reticulocyte extracts, either native or supplemented with high salt wash from infected cell polysomes (Fig. 3, lunes 6 and c) , resemble closely the pattern of the input RNA (Fig. 3, lane a ) ; ICR 597 and 534, the putative mRNAs coding for the two major late proteins ICP 68 and 55, respectively, appear equally stable in these cell-free translation extracts (Fig. 3).

We conclude, therefore, that there is no preferential break- down of late mRNAs in the reticulocyte extracts and that the translational control exhibited could not be explained by a

mechanism by which an infected cell protein(s) preferentially stabilizes the late mRNAs. It has been suggested that 5'- terminal caps play an important role in the stability of mRNAs; decapped messages are degraded preferentially (19).

-345 j22 -290

-. - 74. a b c

FIG. 3. Fluorogram illustrating the relative stability of FV 3 mRNAs in cell-free systems. Radiolabeled ["Hluridine total cytoplasmic mRNA from FV 3-infected FHM cells was incubated with either native reticulocyte extract ( 6 ) or with reticulocyte extracts supplemented with 0.5 M KC1 wash fraction from infected cell poly- somes (c ) . The prqfile of the input mHNA is depicted in lane a. Various infected cell mRNAs (ICH) are identified by their apparent molecular weight (X IO'). Some of the prominent early (0) and late (A) ICKs are also indicated.

BH K RETICULOCYTE ICP

"

" - -1"

a b c d e f g h FIG. 4. Fluorogram of sodium dodecyl sulfate-polyacryl-

amide gel electrophoresis of [3'S]methionine-labeled cell-free translation products. Nuclease-treated extracts from virus-infected BHK cells ( a and h) and reticulocyte (f-h) were programed with total infected cell mRNA (6 h postinfection), and translational products were analyzed as described in the text. Lane a, FV 3-infected BHK cell extract without mRNA; lane 6, same as a but containing 100 pg/ ml of total cytoplasmic FV 3 RNA; lanes c and e, polypeptides labeled in FV 3-infected BHK cells in LiLlo; lane d, structural polypeptides in the purified FV 3 as molecular weight marker; lane f. native reticu- locyte extracts programed with 100 pg/ml of total infected RNA; lane g, same as f but supplemented with 0.5 M KC1 ( 5 pl) wash fraction from uninfected BHK cells; lane h, same as f but containing 5 pl of 0.5 M KC1 (2 mg/ml) wash fraction from FV 3-infected BHK cell (6 h postinfection). 0, early IC%; A, late ICKs.

574 Translational Discrimination of FV 3 mRNAs

4s I

I 2 3 4 5 6 7 8 9 IO II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 FIG. 5. Size fractionation of FV 3-infected cell cytoplasmic RNA. FV 3-infected cells were labeled with

["Hluridine (50 pCi/ml), between 2 and 6 h postinfection, and 200 pg of cytoplasmic RNA were centrifuged over a 15-30% sucrose gradient (27,000 rpm, 18 h). Fractions (0.25 ml) were collected while monitoring them a t Awl. RNA from alternate fractions was divided into three equal aliquots; an aliquot from each fraction was analyzed by formamide-acrylamide gel electrophoresis and fluorography (lanes 2 through 24). Three samples from each end of the gradient profile have been cropped out since there were no detectable mHNA species in these fractions. The positions of cellular RNAs corresponding to 4 S, 18 S, and 28 S were identified from Am, profiles. Lane 25 represents total infected cell RNA (ZCR); various early (0) and late (A) mHNA species are identified by their apparent molecular weight (X 10').

4s I

18s I

28s I

ICP

-36

I 2 3 4 5 6 7 8 9 IO II 12 13 14 15 16 17 18 19 20 21 22 232425 26 27 FIG. 6. Translation of size-fractionated viral RNA in native reticulocyte extracts. FV 3 mRNA was size

fractionated on a sucrose gradient. RNA in one-third aliquots from alternate fractions of the gradient was translated in reticulocyte extracts and translation products were analyzed by 5-15'% gradient polyacrylamide gel electrophoresis and autoradiography. Lanes 2-24 correspond to translation products of mHNAs represented in lanes 2-24 in Fig. 5. Lane 25 represents the endogenous translation products. Infected cell polypeptides (ZCP) representing either total (both early and late, lane 26) or only early (lane 27) polypeptides are shown for comparison. Molecular weights (X 10') of ICPs are denoted on the side. The translation efficiency of two early (ICP 115 and 18) and three late (ICP 72,68, and 55) polypeptides was quantitated (see text). 0, early IC&, A, late IC&.

Thus, the lack of differential stabilities of early and late FV 3 mRNAs is consistent with our earlier findings demonstrating similarly modified ?-termini of FV 3 mRNAs belonging to these two temporal classes (20). In the next section, we de- scribe experiments to test our second hypothesis.

Initiation Factors from FV3-infected BHK Cells Translate Both Early and Late mRNAs Efficiently-FV 3 grows in BHK cells to titers similar to its growth in F H M cells, and the overall regulatory patterns of FV 3 macromolecular syntheses in F H M and BHK cells are also similar (2-4). Since large numbers of cells were required for the preparation of cell-free extracts, we carried out subsequent in uitro experiments in extracts prepared from a clone of BHK cells adapted to grow

in suspension culture; the source of FV 3 mRNAs was either BHK or F H M cells.

The spectrum of polypeptides synthesized, as well as their relative molar abundances in cell-free extracts from FV 3- infected BHK cells (Fig. 4, lane b) , corresponded with their in uiuo counterparts (Fig. 4, lanes c and e) . Extracts from unin- fected BHK cells failed to efficiently translate late mRNAs (data not shown). Since preferential utilization of initiation factors (21, 22) or tRNAs (23) has been shown to exert a discriminatory influence on translation of various mRNAs, we tested for their effect.

Reticulocyte extracts supplemented with tRNAs from yeast, rat liver, or from FV 3-infected BHK cells did not

Translational Discrimination of FV 3 mRNAs 575

FIG. 7. Translation of size-frac- 4s tionated viral RNA in initiation fac- t tor-supplemented reticulocyte ex- tracts. The third aliquot of RNA from alternate fractions of the gradient (see Fig. 5) was translated in reticulocyte ex- tracts supplemented with crude initia- tion factors from infected cells. Lanes 1- 24 show the cell-free translation prod- ucts of RNAs corresponding to fractions 1-24 in Fig. 5. Endogenous translational products are shown in lane 25. Markers depicting total infected cell polypeptides (lane 26) and early ICPs (lane 27) are also shown. The numbers on the side indicate molecular weights (X IO’) of ICI’s. The translation efficiency of two early (ICP 115 and 18) and three late (ICP 72, 68, and 55) PolvDeDtides was

- 18s I

compared (see Tables i l and k ) . 1 2 5 4 3 b ( 8 Y IU II I Z I 5 1 4 1 5 1 6

significantly alter the overall molar abundance of mRNA translation products (data not shown). Similarly, native retic- ulocyte extract supplemented with crude initiation factor(s) from uninfected cells (Fig. 4, laneg) also failed to substantially improve the translation of late mRNAs. In contrast, the polypeptides made in reticulocyte extracts supplemented with a 0.5 M KC1 wash (initiation factors) from ribosomes of FV 3- infected BHK cells (Fig. 4, lane h ) restored translation of FV 3 late mRNA and the pattern resembled that obtained in vivo (Fig. 4, lane e ) . As with the infected cell extract (Fig. 4, lane b) , the relative molar abundances of corresponding polypep- tides were similar to their infected cell counterparts (as judged by the relative intensity of the corresponding bands; Fig. 4, lanes e and h) .

These data clearly implicate a virus-specified or virus-mod- ified initiation factor($ present in FV 3-infected cells, but not in uninfected ones, which appears to selectively enhance trans- lation of FV 3 late mRNAs. The close agreement in the relative molar abundances of polypeptides in vitro and in vivo suggests that the factor(s) from FV 3-infected BHK cells exerts the same regulatory control in vitro as is operative in vivo.

Infected Cell Initiation Factors Selectively Enhance Translation of FV 3 Late mRNAs-To obtain rigorous ex- perimental evidence for translational discrimination, one should compare translation of mixtures of RNAs in which the concentration of the individual species is accurately deter- mined. A comparison of the molar abundance of individual mRNAs and their putative protein products synthesized in extracts with or without the putative factor(s) should reveal if specific factor-mediated translational discrimination occurs (24). Such an analysis is difficult with FV 3 mRNAs because accurate estimates of the concentrations of individual mRNA species in a complex array of transcripts synthesized in the infected cell cannot be made. For example, the purification of individual mRNA species by formamide gel electrophoresis and DNA:RNA hybridization techniques results in poor yields (20). The problem of quantitation is further complicated be- cause FV 3 mRNAs lack poly(A) (14); hence, cDNA probes cannot be easily made by reverse transcription. Therefore, we adopted an alternative approach involving size fractionation of viral mRNAs by centrifugation on a linear 1.530% sucrose gradient. One of the predictable consequences of such a frac- tionation would be a random variation in the concentration and relative proportions of early and late viral mRNA species in different fractions across the gradient. The data presented in Fig. 5 clearly support this prediction. Analysis of an aliquot of mRNA from different fractions by formamide-acrylamide

TABLE I1 Relative efficiencies of translation of mRNAs coding for ICP 115

and 18 (early polypeplides) Cytoplasmic RNA from FV 3-infected cells was fractionated on a

sucrose gradient, and RNA in each fraction was translated in either native or initiation factor-supplemented reticulocyte extracts, and the products were analyzed as described in the legend to Figs. 5-7 and in the text. Bands representing ICI’ 115 (fractions 12 through 17) and ICP 18 (6 through 11) were excised and processed for quantitation of radioactivity as described in the text. Background radioactivity (ap- proximately 70 cpm) was subtracted from the values given.

ICP 18 ICP 115

Fraction No. Initiation Fraction No, factor-

hitiation factor-

Native supple- Native supple. mented

6 1762 1899 12 7 1678 1388 8

1 3 3288 2816 14

9 2328 1829 15 10 1446 1413 I6 11 515 678 17

rnented

142 178 523 473 462 579

1202 928 750 619 625 631

TABLE 111 Relative efficiencies of translation of mRNAs coding for ICP 72,

68, and 55 (late polypeptides) Detailed conditions of RNA fractionation and cell-free translation

are described under the legends to Figs. 5-7 as well as in the text. Protein bands representing ICP 55,68, and 72 (in fractions 12 through 19) were excised and radioactivity was determined. Background ra- dioactivity (approximately 70 cpm) was subtracted from these values.

ICP ss ICP 68 lcr’ 72

Fraction Initiation Initiation No.

Initiation factor- factor- factor-

Native supple- mented mented rnented

Native supple- Native supple-

12 385 86 1 103 271 94 319 13 852 2652 98 269 134 430 14 888 3544 137 462 103 650 15 610 3906 162 583 155 F24 I 6 882 4261 262 676 297 715 17 645 3103 247 655 103 509 18 526 1923 155 308 190 579 19 201 893 105 241 287 433

gel electrophoresis and fluorography clearly revealed the ex- pected variation in mRNA composition (early or late) as well as concentration. Since these mRNAs were labeled under steady state conditions, the intensity of a particular band reflects its amount. Therefore, the translation of size-fraction- ated mRNA is a suitable method for testing the specificity of

576 Translational Discrimination of FV 3 mRNAs

15

10

5

10

k X

15 0

10

5

A 80s 4

BOTTOM FRACTION NUMBER TOP

FIG. 8. Binding of FV 3 mRNAs to native or infected cell IF- supplemented reticulocyte ribosomes. "H-labeled viral mRNA was incubated in 200 pI of native reticulocyte extract ( A ) or an extract containing the high salt fraction from infected cells ( B ) . Both reaction mixtures contained anisomycin ( 1 mg/ml) to block polypeptide chain elongation. After a 5-min incubation at 30 "C, 200pl of high salt buffer (1 M NaCl and 60 mM Mg(oAc)?) were added. The entire contents of the reaction were centrifuged through a 15-30% linear sucrose gra- dient at 40,000 rpm for 90 min in a Beckman SW 41 rotor. Fractions (0.5 ml) were collected and monitored at Ale,. Twenty-five microliter aliquots from each fraction were taken to determine the distribution of "H radioactivity. The arrow indicates the A?,,, peak of 80 S ribosomes.

the viral factor(s) under varied mRNA concentration, as well as varied mRNA composition, in the translation mixture.

A comparison of the in vitro translation products of mRNAs from individual gradient fractions in native (Fig. 6) and IF- supplemented reticulocyte (Fig. 7) extracts establishes three points. (i) There is good correlation between the sizes of the polypeptides and their putative mRNAs in the fractions across the gradient. (ii) All viral polypeptides detected in the infected cell in vivo (both early and late) are translated in the native reticulocyte system. This contrasts with the results obtained when we translated unfractionated total infected cell RNA when predominantly early polypeptides were made (see Figs.

1 and 4). We can account for this by the nonsaturating concentrations of mRNAs present in the various fractions. However, if the concentration of mRNAs in individual frac- tions was 2100 pg/ml (as in the experiments of Figs. 1 and 4), the late mRNAs were no longer effectively translated (data not shown). (iii) There is a marked enhancement in the intensity of late gene products in the initiation factor-supple- mented extracts.

When we compared the incorporation of ["'S]methionine into two early (ICP 115 and 18) and three late (ICP 72, 68, and 55) polypeptide species synthesized in the two cell-free extracts, there was no significant difference in the relative incorporation of radioactivity in ICP 115 and 18 (two early polypeptides) in the native or initiation factor-supplemented reticulocyte extracts (Table 11). In contrast, the efficiency of translation of the three late polypeptides (ICP 72, 68, and 55) improved from 2-7-fold in the initiation factor-supplemented extracts (Table 111). We conclude, therefore, that a virus- specific or -modified initiation factor(s) specifically enhances translation of late mRNAs, and this enhancement occurs regardless of the relative molar abundance of early and late FV 3 mRNA species in the translation mixture.

Early FV 3 mRNAs Are More Efficiently Bound in 80 S Initiation Complexes-To substantiate the conclusion drawn in the preceding experiment and to gain some insight into the underlying mechanisms, we analyzed directly incorporation of viral mRNAs into 80 S initiation complexes. The underlying rationale of these experiments has been described by Rose and Lodish (25). Reticulocyte extracts either with or without initiation factors from infected cells were programed with viral mRNAs under conditions of blocked polypeptide chain elon- gation; i.e. the reaction mixture contains 1 mg/ml of aniso- mycin, a specific inhibitor of chain elongation (26). Stable 80 s initiation complexes (in 0.5 M NaCl and 30 mM Mg(oAc)s) are resolved by centrifugation. A quantitative comparison of the fraction of radiolabeled mRNA found in 80 S initiation complexes in native or infected cell IF-supplemented reticu- locyte extracts revealed no significant differences when ana-

-2401235 A

- 7 4 A

o b c FIG. 9. Fluorograph depicting the FV 3 mRNAs found in 80

S initiation complexes in the native and IF-supplemented re- ticulocyte extracts. Radiolabeled RNA in fractions representing 80 S initiation complexes was extracted and electrophoresed on form- amide-acrylamide gels and fluorographed as described before (2). Pattern of input mHNA ( a ) , mRNA from native 80 3 complexes ( b ) , or IF-supplemented complexes (c) is illustrated. Several infected cell RNAs (ICR) belonging to early (0) or late (A) temporal classes are shown with their molecular weight (X IO').

Translational Discrimination of FV 3 mRNAs 577

lyzed by centrifugation (66.5 versus 70.3%, respectively, Fig. 8, A and B). However, when we compared the mRNA compo- sition of 80 S initiation complexes between native and IF- supplemented extracts by formamide-acrylamide gel electro- phoresis (2), there were significant differences. The initiation complexes of native reticulocyte extracts, while containing abundant early mRNA species, are noticeably deficient in several late mRNAs (Fig. 9, lane b ) . In contrast, both early and late mRNAs are well represented in the initiation com- plexes from IF-supplemented reticulocyte extracts (Fig. 9, lane c ) ; the pattern of mRNAs from IF-supplemented initia- tion complexes resembles more closely the pattern of mRNA found in vivo (Fig. 9, lane a). The addition of a 0.5 M KC1 wash fraction from uninfected cells does not significantly alter the pattern of mRNA found in the reticulocyte initiation complexes (data not shown). We conclude, therefore, that the late FV 3 mRNAs are not efficiently initiated in “native” reticulocyte extracts, and this inefficiency can be overcome by a soluble factor(s) isolated from infected cell polysomes.

DISCUSSION

Under a variety of physiological conditions, protein synthe- sis in eukaryotic organisms is regulated at the translational level. For example, the dramatic change in the pattern of protein synthesis in the oocytes of the surf clam, Spisula solidissima, after fertilization could be explained by develop- ment of stage-specific recruitment of subsets of mRNAs from a common pool of messages present at both these stages (27). Recently, convincing evidence for translational control was also presented in a study of protein synthesis in the heat- shock response in Drosophila melanogaster (28). Similarly, regulation of ovalbumin mRNA translation in the oviducts of chickens after a combined administration of estrogen and progesterone (29) and control of proinsulin mRNA translation in the rat pancreatic islets following glucose administration (30) were demonstrated to be at the translational level. The induction of translational control in response to such diverse stimuli indicates that there may be multiple mechanisms leading to the same final outcome, preferential translation. Since initiation of translation in eukaryotes is a multistep, multicomponent process (31), such a notion of regulation is not unreasonable.

However, the literature of eukaryotic protein synthesis also contains several conflicting claims purporting to show specific translational controls (21). The cause of this discrepancy may partly be inherent in the use of cell-free systems (21, 24). Lodish (21) has argued that different mRNAs may vary in their rate constants for initiation, elongation, and termination. Assuming this to be the case, when translation is performed in vitro under conditions of mRNA saturation, it is possible that one of the components (e.g. initiation factors, tRNAs, or ribosomes, etc.) required for translation of all mRNAs may be rate-limiting. There would be two consequences of such a limitation. First, the mRNAs with the highest affinity for this rate-limiting component would be at an advantage in being translated. Second, the addition of the rate-limiting compo- nent to the cell-free extract would mimic an apparent prefer- ential translational control. This is not the case with FV 3 because addition of tRNAs or initiation factors from unin- fected cells did not alter translational efficiency.

The experimental evidence presented in this report indi- cates that the relative efficiencies of translation of the late FV 3 mRNAs can be modulated in vitro by a factor(s) present in infected cells only. It has also been shown that the putative factor(s) is ribosome-bound and is likely to be virus-coded or virus-modified. Unequivocal proof of it being a viral gene product must, however, await genetic experiments. We d-

ready know that the factor(s) is heat-denaturable and de- stroyed by protease, but it is not affected by nucleases.z

Absolute specificity of the effect of a putative factor(s) on the translation of late FV 3 mRNAs is difficult to prove. Available techniques cannot rule out the requirement of large amounts of a factor for the translation of one set of mRNAs which is needed in small quantities for the translation of another set.

Although nonspecific augmentation of translation of early as well as late mRNAs was observed when size-fractionated mRNAs were translated in native reticulocyte extracts under nonsaturating mRNA concentrations, it cannot adequately explain the increase in translation of late mRNAs in the presence of added infected cell initiation factor(s). The data clearly show that, superimposed onto the nonspecific increase in the translational efficiency of all mRNAs, there is a specific translational modulation of the late mRNAs mediated by a factor(s) in the infected cell because (i) late FV 3 mRNAs are poorly translated in cell-free extracts from uninfected BHK cells, reticulocytes, or wheat germ; (ii) the translation of late mRNAs is significantly increased (2-7-fold) in infected BHK cell extracts as well as in reticulocyte extracts supplemented with the 0.5 M KC1 wash from infected cell polysomes; in contrast, the addition of the infected cell factor(s) does not alter the translational efficiency of early mRNAs; and finally (E) it is evident from the translation of the size-fractionated mRNAs that the enhancement in the translatability of late mRNAs is quite independent of their molar abundance as well as independent of the types of other messages present in the mixture.

The precise mechanism underlying the type of translational regulation we have found remains to be elucidated. The translational discriminatory mechanisms (based on the struc- tural differences between two mRNA species) analogous to the one proposed for picornavirus-infected cells (32-34) cannot be directly applied to the present situation. The structure of 5’-termini of early and late FV 3 mRNAs is similar (20). However, whether differences in the internal methylation between the two sets of mRNAs (20) play a role in transla- tional regulation is not known.

In addition to the translational control of FV 3 genes mentioned in the introduction (2-4), there is translational control at the level of switch-off of host cell protein synthesis by an FV 3 structural protein (35). In cells infected with FV 3, initiation of host cell mRNAs is inhibited, while viral mRNAs are selectively translated (35, 36). It is worth noting that the in vitro protein-synthesizing extracts from FV 3- infected cells also retain a selective inhibitory activity on cellular mRNAs; the translation of globin mRNAs is inhibited by more than 70% in FV 3-infected BHK cell extracts as compared to globin mRNA translation in extracts of unin- fected BHK cells.’

The molecular dissection of translational control requires accurate determination of various kinetic parameters of pro- tein synthesis as have been outlined in a theoretical treatment of this subject by Bergmann and Lodish (37). In this regard, we should be able not only to isolate and characterize the regulatory factor(s) responsible for translational discrimina- tion of late FV 3 mRNA, but also to further analyze transla- tional discrimination between cellular and viral mRNAs.

technical assistance. Acknowledgments-Susan Carr and Susan Wood provided skillful

REFERENCES 1. Goorha, R., and Granoff, A. (1979) in Comprehensive Virology

(Fraenkel-Conrat, H., and Wagner, R. R., eds) Vol. 14, pp. 347-

R. Raghow and A. Granoff, unpublished data.

578 Translational Discrimination of FV 3 mRNAs

399, Plenum, New York

Virol. 24,326-342

27

495-497

448

335

2. Wiuis, D. B., Goorha, R., Miles, M., and Granoff, A. (1977) J.

3. Elliott, R. M., Bravo, R., and Kelly, D. C. (1980) J. Virol. 33, 18-

4. Raghow, R., Willis, D. B., and Granoff, A. (1980) virology 100,

5. Goorha, R., Willis, D. B., and Granoff, A. (1979) J. Virol. 32,442-

6. Willis, D. B., Goorha, R., and Granoff, A. (1979) Virology 98,328-

7. Naegele, R. F., and Granoff, A. (1971) Virology 44,286-295 8. Tan, K. B., and MacAuslan, B. R. (1971) Virology 45,200-207 9. Villa-Komaroff, K., McDowell, M., Baltimore, D., and Lodish, H.

10. Weber, L. A., Feman, E. R., and Baglioni, C . (1975) Biochemistry

11. Pelham, H. R. B., and Jackson, R. J. (1976) Eur. J. Biochem. 67,

12. Laemmli, U. K. (1970) Nature (Lond.) 227,680-685 13. Lasky, R. A., and Mills, A. D. (1975) Eur. J. Biochem. 56, 335-

14. Willis, D. B., and Granoff, A. (1976) Virology 73,543-547 15. Aviv, H., and Leder, P. (1972) Proc. Natl. Acad. Sci. U. S. A . 69,

16. Raskas, H. J., and Green, M. (1971) Methods Virol. 5,247-269 17. Schreier, M. H., and Staehelin, T. (1973) J. Mol. Biol. 73, 320-

18. Brawerman, G. (1974) Annu. Rev. Biochem. 43,621-642 19. Furuichi, Y., LaFiandra, A., and Shatkin, A. J. (1977) Nature

F. (1974) Methods Enzymol. 30, 709-723

14, 5315-5321

247-256

341

1408-1412

349

(Lond.) 266, 235-239 20. Raghow, R., and Granoff, A. (1980) Virology 107,287-294 21. Lodish, H. F. (1976) Annu. Reu. Biochem. 45, 39-72 22. Golini, F., Thach, S. S., Birge, C. H., Safer, B., Merrick, W. C.,

and Thach, R. E. (1976) Proc. Natl. Acad. Sci. U. S. A . 73, 3040-3044

23. Smith, D. W. E. (1975) Science (Wash. D. C.) 190, 529-535 24. Birge, C. H., Golini, F., and Thach, R. E. (1979) Methods En-

25. Rose, J. K., and Lodish, H. F. (1976) Nature (Lond.) 262,32-37 26. Grollman, A. P. (1967) J. Biol. Chem. 242, 3226-3233 27. Rosenthal, E. T., Hunt, T., and Ruderman, J. V. (1980) Cell 20,

28. Lindquist, S. (1980) J. Mol. Biol. 137, 151-158 29. Robins, D. M., and Schimke, R. T. (1978) J. Biol. Chem. 253,

30. Itoh, N., and Okamoto, H. (1980) Nature (Lond.) 283, 100-102 31. Hunt, T. (1980) Trends Biochem. Sci. 5, 178-181 32. Padilla, M., Canaani, D., Groner, Y., Weinstein, J. A., Bar-Joseph,

M., Merrick, W., and Shafritz, D. A. (1978) J. Biol. Chem. 253, 5939-5945

33. Rose, J. K., Trachsel, H., Leong, K., and Baltimore, D. (1978) Proc. Natl. Acad. Sci. U. S. A. 75, 2732-2736

34. Helentjaris, T., Ehrenfeld, E., Brown-Luedi, M. L., and Hershey, J. W. B. (1979) J. Biol. Chem. 254, 10973-10978

35. Raghow, R., and Granoff, A. (1979) Virology 48, 319-327 36. Cordier, O., Aubertin, A. M., Lopez, C., and Tondre, L. (1981)

37. Bergmann, J. E., and Lodish, H. F. (1979) J. Biol. Chem. 254,

zymol. 60, pp. 375-380

487-494

8925-8934

Ann. Virol. (Paris) 132,25-39

11927-11937