JOURNAL OF VIROLOGY, May 1979, p. 481-4880022-538X/79/05-0481/08$02.00/0

Vol. 30, No. 2

Capsid Protein Precursor Is One of Two Initiated Products ofTranslation of Poliovirus RNA In Vitro

STEVE HUMPHRIES,`* FRED KNAUERT,' AND ELLIE EHRENFELD'2Departments of Cellular, Viral and Molecular Biology' and Biochemistry,' University of Utah Medical

Center, Salt Lake City, Utah 84132

Received for publication 28 November 1978

Previous studies in our laboratory have demonstrated that cell-free systemstranslating the Mahoney strain of poliovirus type I RNA utilize two uniqueinitiation sites. In this study, defective-interfering particles of poliovirus, whichcontain deletions in the region encoding the capsid proteins, are shown to initiatetranslation of proteins in vitro at these same two sites. Both the standard virusand the defective-interfering virus RNA direct the synthesis of two polypeptideslabeled with n-formyl-methionine (fmet) at their amino termini. The size of thesmaller fmet polypeptide synthesized in vitro by the defective virus appearsidentical in size to that of the standard virus. However, the larger-molecular-weight fmet polypeptide is reduced in size from 115,000 to 69,000 daltons. Thiscorrelates exactly with the reduced size of the precursor to the capsid proteinssynthesized by the defective virus in vivo and with the size of the deletion in thedefective virus RNA (1,200 bases). This provides genetic evidence that the115,000-dalton fmet polypeptide synthesized in vitro by the standard virus isNCVPla, the precursor to the coat proteins. Although the identity of the small(5,000 to 10,000 daltons) fmet polypeptide is not clear, several lines of evidenceenable us to exclude the possibility that it is VP4, the smallest viral capsid protein.

The generally accepted model for picornavirusRNA translation is that initiation occurs at asingle site at or near the 5' end of the RNA.However, using n-formyl-[3S]methionine([3S]fmet) donated from [35S]fmet-tRNAfmet tospecifically label the amino termini of initiatedpolypeptides, Celma and Ehrenfeld (4) demon-strated that in a cell-free system made frompoliovirus-infected cells, two different amino-terminal tryptic peptides are synthesized. Therelative proportions of the two peptides synthe-sized varied as a function of the Mg2" concentra-tion in the reaction.Recent work in our laboratory (13) has re-

solved two major [3S]fmet-labeled polypeptidesby sodium dodecyl sulfate-polyacrylamide gelelectrophoresis which result from in vitro trans-lation of viral RNA. The relative proportions ofthese polypeptides show an Mg2" concentrationdependence which correlates with that shownby the two amino-terminal tryptic peptides iden-tified by Celma and Ehrenfeld. Synthesis of a115,000-dalton polypeptide predominates atMg2" concentrations between 1.0 and 2.5 mM,whereas synthesis of a 5,000 to 10,000-daltonpolypeptide predominates at higher Mg2" con-centrations (2.5 to 4.0 mM). Knauert and Ehren-feld further showed that fmet tryptic peptide II(4) is derived from the high-molecular-weight

polypeptide, and fmet tryptic peptide I is derivedfrom the low-molecular-weight polypeptide.The present study uses a genetic approach to

obtain information on the identity of these twofmet polypeptides and uses a mutant of polio-virus, the defective-interfering (DI) particle firstisolated by Cole et al. (6). DI particles infectcells, shut off host cell protein synthesis, andreplicate their RNA as efficiently as the stan-dard virus (5), but no progeny particles are pro-duced because the defective genome RNA lacksthe information coding for some of the capsidproteins (5). In DI particle-infected cells,NCVPla, the precursor to the capsid proteins,is reduced in size by between 30,000 and 40,000daltons. Since the sequences coding for NCVPlahave been mapped at the 5' end of the poliovirusgenome RNA (21, 22), NCVPla is likely to bethe amino-terminal portion of the translationproduct and, therefore, to be labeled with fmetin vitro. If the high-molecular-weight, fmet-la-beled polypeptide synthesized in vitro isNCVPla, we would expect the size of the fmetpolypeptide in DI-infected cell extracts to bereduced by 30,000 to 40,000 daltons.

MATERIALS AND METHODSCells and virus. The growth and maintenance of

HeLa S3 cells was as described by Celma and Ehren-481

482 HUMPHRIES, KNAUERT, AND EHRENFELD

feld (4). Virus stocks enriched for DI-1, -2, and -3 (akind gift from E. Wimmer) were used to infect HeLacells, and the progeny virus was purified from themixed infection as previously described (5, 6). Thedefective virus particles were separated from standardvirus particles by two cycles of centrifugation on CsCldensity gradients in NTE buffer (10 mM NaCl, 10mMTris [pH 7.4], 2 mM EDTA) containing 1% Brij 58 (6).Centrifugation was carried out in cellulose nitratetubes (0.625 by 3 inches [ca. 1.6 by 7.6 cm], spun inthe angle 65 rotor at 30,000 rpm for 15 h at 4°C. Thevisible virus bands were collected and rebanded inCsCl density gradients. Purified virus was diluted withNTE buffer and pelleted by centrifugation in the angle65 rotor at 60,000 rpm for 1 h. Virus particles weresuspended in NTE buffer and stored at -70°C. Analiquot was diluted into water, and the absorbance at260 nm was measured using a Beckman model 25spectrophotometer. The concentration of virus parti-cles was calculated assuming an absorbance of 1 at 260nm equals 10'3 standard virus particles and 0.85 x 1013DI-1 defective virus particles (5).

In vivo labeling of viral RNA. To prepare viralRNA, HeLa cells infected with 1,000 particles of eitherDI or standard virus per cell were labeled in thepresence of 5 ,ug of actinomycin D per ml, by theaddition of 5 ,uCi of [14C]uridine (50 mCi/mmol; NewEngland Nuclear) per ml from 2.5 to 4.5 h postinfec-tion. Cells were washed once in Earle saline and lysedin RSB (10 mM Tris [pH 7.4], 10 mM NaCl, 1.5 mMMgCl2) made 0.5% in Nonidet P-40. After removal ofcellular debris by centrifugation, the supernatant wasmade 0.5% sodium dodecyl sulfate and the RNA wasdeproteinated with phenol and precipitated withethanol.

In vivo labeling of viral proteins. Cells infectedwith either standard or DI virus at 1,000 particles percell were labeled by the addition of 20 ACi of [35S]_methionine (>500 Ci/mmol; New England Nuclear)per ml 2.5 to 4.0 h postinfection. Cells were lysed asdescribed above, and postmitochondrial supernatantwas made 1% sodium dodecyl sulfate and stored at-20°C until required.

In vitro protein synthesis. Cell-free extracts(S10) of either standard or DI virus-infected cells wereprepared as previously (4) 3.75 to 4 h after infection.FormyI-["S]methionine-tRNAf"' prepared from totalrabbit liver tRNA (Grand Island Biological Co.) wascharged and purified as previously described. Condi-tions for the incorporation of [35S]fmet by cell-freeextracts were as described by Celma and Ehrenfeld(4). The incubation also contained 15 Ml per 100 [lIreaction mixture of ribosomal salt wash containinginitiation factors, prepared from uninfected HeLa cellsby the method of Kaufman et al. (12). [35S]fmet-la-beled tryptic peptides were prepared and analyzed aspreviously described (4).

Preparation of VPg. Preparation of VPg was es-sentially as described by Lee et al. (16). At 2.5 hpostinfection with either standard or defective virus,cells were washed in saline and resuspended in me-dium minus lysine. A 100-,uCi/ml amount of [3H]lysine(60 Ci/mmol; New England Nuclear) was added, andthe incubation was continued for a further 2.5 h. Thecells were lysed as described above, and viral RNA

was extracted using a 25:24:1 mixture of phenol, chlo-roform, and isoamyl alcohol. After ethanol precipita-tion, the RNA was dissolved in 50 mM ammoniumacetate, pH 5.0, and digested with 100 U of T2 RNaseper ml, 200 U of TI RNase per ml, and 150 U ofpancreatic RNase A per ml (all enzymes from Calbi-ochem) at 37'C for 90 min. The RNA nucleotides werethen adsorbed to DEAE-cellulose as described by Leeet al. (16), and the material that was not retained onthe column was analyzed further. Proteinase K diges-tion was performed in 10 mM Tris (pH 7.4)-i mMEDTA-1% sodium dodecyl sulfate with 1 mg of pro-teinase K (E. Merck, Darmstadt, Germany) per ml at37°C for 2 h.Acrylamide gel electrophoresis. Viral proteins

were analyzed on 6 to 24% sodium dodecyl sulfate-polyacrylamide slab gels as described by Laemmli (14).Electrophoresis was run at a constant current of 35mA and a maximum voltage of 150 V until the brom-ophenol blue dye reached the bottom of the gel (ap-proximately 16 h). If necessary, gels were prepared forfluorography according to Bonner and Laskey (2).Autoradiography was on Kodak SB 5 film for 2 to 5days.Methylmercury agarose gel electrophoresis.

Methylmercury agarose gel electrophoresis was per-formed as described by Batt-Humphries et al. (S.Humphries, C. Simonsen, and E. Ehrenfeld, Virology,in press), using a modified method of Bailey andDavidson (1). Agarose gels, 0.8%, were run at 80 V (40mA) for 13 h. Fluorography of gels was as describedby Batt-Humphries et al. (Virology, in press).

RESULTS

Defective poliovirus particles were grown andpurified from mixed stocks of standard virus andDI-1, -2, or -3. Virus particles banding at adensity of 1.32 g/cm3 (compared with 1.34 g/cm3for standard virus) were rebanded in CsCl andused to infect HeLa cells in the absence of stan-dard helper virus. ['4C]uridine-labeled viralRNA was prepared from standard and DI virus-infected cells and analyzed on a methylmercuryagarose gel. Figure 1 shows that the RNA fromall three defective virus stocks is similar in sizeand that none are significantly contaminatedwith standard virus RNA. The band width ofthe RNA from DI-3 is greater than that for DI-1 or DI-2 and may reflect heterogeneity in thevirus stock. The sizes of the defective virusRNAs range from 84.2% (DI-1) to 81.9% (DI-3)of the standard genome. This represents a dele-tion of 1,200 to 1,400 bases, in agreement withpreviously reported values (5).The proteins synthesized in vivo during infec-

tion by defective viruses were labeled by theaddition of [35S]methionine to the medium, andthe proteins were analyzed by polyacrylamidegel electrophoresis (Fig. 2a). With one exception,the proteins synthesized by the defective virus

J. VIROL.

IN VITRO TRANSLATION OF POLIO DI RNA

v STD STO42+S POLIO DI( 1(2) DI(3)POLIO 28S 18S31 S VIRUS VIRUS rRNA rRNA

rasp

U. :. i;

4.0-

0

ao

x

0CU

O 3.0-

1-J3a: JcJI

2IoL)

I. 1.5-

f 127 8 9 10 I 12b D ISTANCE MisG RATE D(cm}

13 14

FIG. 1. (a) Size analysis of RNAs from purified poliovirus DI particles on methylmercury gels. Sampleswere run on a 0.8% agarose gel in 5mM methylmercury hydroxide as described. The molecular weights of theDI RNAs were estimated from a semilog plot of the molecular weight against distance migrated, as shown in(b). Standards used were 42S vesicular stomatitis virus (VSV) virion RNA (molecular weight [MW] = 3.6 x106; 24); 31S VSVL mRNA (MW = 2.1 x 106; 19) (both a kind gift from S. Batt-Humphries); standard poliovirion RNA (MW = 2.6 x 106; Y. F. Lee, A. Nomoto, and E. Wimmer, submitted for publication); 28S HeLacell rRNA (MW = 1.65 x 10P). The estirnated molecular weights for the DIRNAs were: DI-1, 2.19 x 106 (84.2%of standard); DI-2 and -3, 2.12 x 106 (81.9% of standard).

between 2.5 and 4 h after infection are synthe-sized by the standard virus. However, in the DI-1-infected cells, NCVPla, the precursor to thecapsid proteins, is absent, and a new proteinmigrating between NCVP3 and NCVP4 is ap-parent. Furthermore, as expected, the viral cap-sid proteins VP2, VP3, and VP4 could not bedetected, although polypeptides with mobilitiessimilar to those of VP1 and VPO, the directprecursor to VP2 and VP4, can be seen. Fromthe known molecular weights of the poliovirusproteins, and using a T4-infected E. coli proteinmarker (a gift from G. Stetler), the molecularweight of the new protein synthesized by thedefective virus was estimated to be 69,000. Thisreduction of approximately 45,000 daltons isroughly equivalent to 375 amino acids and cor-relates well with the observed 1,200 bases of

RNA deleted from the defective genome of DI-1.We next compared the proteins made in vitro

by cell extracts (S10) prepared from DI-1 andstandard virus-infected cells, using [3S]methio-nine to label the proteins synthesized. As can beseen in Fig. 2b, the proteins made in vitro cor-

respond well with those made in vivo; the115,000-dalton capsid precursor is not made bythe defective S10, and a new polypeptide ofabout 69,000 daltons is present. VP3, which issynthesized by the standard S10, is not observedin the DI-S10; as in the in vivo samples, thereare bands of reduced intensity that migrate inthe region of VP1 and VPO. VPO is only finallycleaved to VP2 and VP4 after incorporation intovirus particles (10), and since only the earlystages of virus assembly have been observed in

up

a

VOL. 30, 1979 483

484 HUMPHRIES, KNAUERT, AND EHRENFELD

VIRUS -<T> 3I em

VP.MCV

w I rPRO7f....<............- .__i _=

.1..

4-MOAsaw E_

g...., -....

FIG. 2. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis ofproteins synthesized in vivoand in vitro by standard and DI virus. [35S]methionine-labeled proteins made (a) in vivo between 2.5 and 4h postinfection and (b) in infected cell S10 extracts were run on 6 to 24% polyacrylamide gels as described.

vitro (for review, see 20), the absence of VP2and VP4 in these extracts is not unexpected.We next examined the polypeptides synthe-

sized in the S10, which could be labeled with[35S]fmet donated from [35S]fmet-tRNAf`et Fig-ure 3 shows that two major fmet-labeled poly-peptides can be detected in the standard virus-infected S10. At low Mg2" levels (2.5 mM), syn-

thesis of the high-molecular-weight polypeptidepredominates, and at high Mg2e levels (4.0 mM)the low-molecular-weight polypeptide is prefer-entially made. In the DI-S10, the size of thesmall fmet protein appears to be unchanged,although since resolution is poor in this regionof the gel, small differences in molecular weightwould not be detected. However, the size of the

high-molecular-weight fmet polypeptide is re-

duced from 115,000 to about 69,000 daltons, cor-

responding in size to the protein synthesized bythe defective virus in vivo. The high-molecular-weight fmet polypeptide is similarly reduced insize in DI-3 virus-infected S10 lysates, appearingas a diffuse band of about 69,000 daltons (notshown). This is presumably a consequence ofthe heterogeneity in this virus stock noted ear-

lier. These data demonstrate genetically that the115,000-dalton polypeptide labeled with formyl-[35S]methionine in standard virus-infected S10lysates in NCVPla, the precursor to the capsidproteins.The two initiation sites utilized during polio-

virus translation in vitro were originally detected

N. -'7

J. VIROL.

-, ",;% :- I y'-'.-,' -L L.' " A- -.

....

j

ID 0.kl-)

.1 .-. .-j'.

i11i,P "" e.

I': -

.-.o-i f, -,-

WoolV..., .:..

:8MM

IN VITRO TRANSLATION OF POLIO DI RNA 485

DI (1)

j3Sj fmet4.OmM 2.5mMMg++ Mg++

STANDARD

35Sj me.

4.0mM 2.5mMMg++ Mg++

NCVP l

vp --VPI

VP2 -

VP3 -

VP4_h..

FIG. 3. Sodium dodecyl sulfate-polyacrylamidegelanalysis ofpolypeptides made in infected cell extractslabeled by [35S]finet-tRNA 't Samples were run on

6 to 24% polyacrylamide gels, and the gel was fluo-rographed as described.

by tryptic peptide analysis. The polypeptideslabeled in vitro by the incorporationof [ S]fmnetwere thus digested with trypsin, and the trypticpeptides were analyzed by high-voltage paperelectrophoresis at pH 1.9 (Fig. 4). As expected,the DI-infected S10 directs the synthesis of two[3S]fmet-labeled tryptic peptides, which coelec-trophorese with the two tryptic peptides fromthe standrd virus-infected S10. There appears tobe no significant difference in the relative pro-portions of the two tryptic peptides synthesizedby the defective or standard virus-infected S10at the different Mg2+ concentrations used.At this time, we have no direct evidence for

the identity of the low-molecular-weight [3S]-fmet-labeled polypeptide. Two known low-mo-lecular-weight proteins synthesized in virus-in-fected cells are VP4 and VPg, the protein foundcovalently attached to the 5' end of virion RNA

(9, 15, 16). VP4 is not synthesized in cells in-fected with DI particles alone (Fig. 2a). It thusbecame of interest to determine if the defectivevirus makes VPg. To test this, cells infected witheither standard or defective virus DI-1 werelabeled with [3H]lysine for 2.5 to 4 h postinfec-tion, and RNA was prepared from a cytoplasmicextract as described by Lee et al. (15). Theresulting RNA was digested to completion withT2, Ti, and pancreatic RNases, and the nucleo-tides were adsorbed to a DEAE-cellulose col-umn. From both the standard- and DI-infectedcytoplasmic RNA, >96% of the acid-precipitableradioactivity was not retained on the columnand eluted in the void volume. Figure 5 showsthat this eluted material migrated on a poly-acrylamide gel as a single low-molecular-weightprotein of mobility similar to that of the [35S]-fmet-labeled low-molecular-weight protein. Fur-thernore, the eluted material from both thestandard- and the DI-containing RNA migratestoward the cathode on high-voltage paper elec-trophoresis at pH 3.5 and is completely sensitiveto proteinase K digestion (data not shown).From this we conclude that the defective virusboth codes for and synthesizes the small protein,VPg, that has been found attached to the 5' endof poliovirus RNA.

DISCUSSION

Previous work from our laboratory has dem-onstrated that in poliovirus-infected cell ex-tracts, two different sites are utilized for theinitiation of protein synthesis (4, 11, 13). Wehave determined the identity of the polypeptidesynthesized from one of these sites, using thedefective virus particles first described by Coleet al. (6). In agreement with their observations,we find that the defective virus does not directthe synthesis of NCVPla, the 115,000-daltonprecursor to the viral capsid proteins, but in-stead a protein of about 69,000 daltons is made(5).The reduced size of the protein is in good

correlation with the observed reduced size of thegenome. The extent and location of the deletionhave recently been determined by Ti oligonu-cleotide fingerprinting and by heteroduplexmapping in the electron microscope (E. Wim-mer, personal communication). The deletion ap-pears to leave intact the first 10% (or about 700bases) of the 5' end of the genome and extendsfor about 1,000 to 1,300 bases to a point 25 to30% from the 5' end of the RNA.The DI protein is large enough to code for

some of the viral capsid proteins, although it isnot known whether correct cleavage of the al-tered precursor can occur. It is possible that VP1(molecular weight, 32,000) or VPO (molecular

VOL. 30, 1979

486 HUMPHRIES, KNAUERT, AND EHRENFELD

-5mM 4 fr2 'K; .-' ;+-*

9q+M* Mgt0 YV

Fi¢. 5. Sodium dodecyl sulfate-polyacrylamidegelelectrophoresis analysis of standard and DI-VPg.Swaples prepared as described in the text were run

on a 10 to 24%polyacrylamide gel and fluorographedas described.

FIG. 4. High-voltage paper electrophsis of [35SJfnet-labeled tryptic peptide,different Mg2" concentrations. Trypticproducts synthesized by standard virusextracts and by DI-1-infected cell extrapared and electrophoresed as describe,fmet was applied as a marker andelectrophoresis by staining with chloroThe dried electropherogram was expos

SB-5 film for 7 days. The exposed filmlocate the radioactive regions on thegram, which were cut out and countecMg2+, the relative amounts of tryptic p,II synthesized in the standard virus-infe25 and 75% and, in the DI-1 S10, 23 anc

weight, 39,000) may be cleaved from the 69,000-dalton DI protein, since bands reduced in inten-sity are seen in the appropriate regions of thegel. We have observed in some gels, however,that the putative VPO band synthesized by theDI virus in vivo does not comigrate with VPOfrom standard virus (data not shown), and theidentity of this polypeptide band in the defectivevirus-infected cells is not clear. Cole and Balti-

oresis analy-more (5) could find no evidence for any capsid

s initiated at proteins and concluded that the DI protein is

c peptides of highly unstable and is rapidly degraded to amino¢-infected cell acids.Icts were pre- The proteins synthesized by infected cell ex-

!d. Unlabeled tracts in vitro mirror those made in vivo. Aslocated after expected, the standard virus -extract directs the

platinate (7). synthesis of NCVPla, which is absent from DI-sed to Kodak infected extracts. Probably because it is not sot was used to rapidly degraded, the new DI protein can beoZettrnnherm-d. At 2.5 mMeptides I and!ctedSlOwered 77%, respec-

tiuely. At 4.0 mM Mg2+ the relative amounts synthe-sized were 86 and 14% (standard) and 84 and 16%(DI-1).

J. VIROL.

- rm

5mM 40 m .

i; F. `.y '..:

7-1.

'Y

".". p 2" 'I.v F -

..'r .: .r

.... 1, :'

W. 'V'

IN VITRO TRANSLATION OF POLIO DI RNA 487

more clearly seen in vitro than in vivo samples.Two major polypeptide species can be labeled invitro with [35S]fmet, from both standard- andDI-infected cell extracts. The size of the smallfmet polypeptide appears identical in both cellextracts, but the size of the large fmet poly-peptide is reduced in the DI virus-infected ex-tract and corresponds in size to the DI proteinof 69,000 daltons synthesized in vivo, which rep-resents the shortened NVCPla. This demon-strates conclusively that the 115,000-dalton largefmet polypeptide observed in standard virus-in-fected celis extracts is NCVPla, the precursor tothe viral capsid proteins.The fmet-labeled tryptic peptides from the

DI-infected S10 coelectrophorese with the stan-dard tryptic peptides from both high- and low-molecular-weight polypeptides. This is not un-expected since the peptides are small, probably13 to 15 amino acids (11), and thus apparentlydo not extend from the initiation site forNCVPla into the deleted region of the genome.The location of the initiation site for the smallfmet polypeptide is unknown. Villa-Komaroff etal. (23) have also reported that both standardand DI-1 virus RNA direct the synthesis of thesame fmet-labeled tryptic peptide in a HeLa cellextract, although they only detected a singleinitiation peptide for both RNAs.The identity of the low-molecular-weight fmet

polypeptide is not clear. In this gel system res-olution of low-molecular-weight polypeptides ispoor, and comigration is not a good criterion foridentification. Although VP4, the small capsidprotein, does migrate in this region of the gel,we feel that it is unlikely that the small fmetpolypeptide is VP4. First, although VP4 is de-rived from the amino-terminal region ofNCVPla (18), the amino acid-terminal trypticpeptide of the small-molecular-weight poly-peptide initiated in vitro is different in its pe-nultimate amino acid from that of the high-molecular-weight polypeptide (4, 13). This indi-cates that the small polypeptide is not simply adirect cleavage product of NCVPla, as would beexpected for VP4. Second, the small fmet-la-beled polypeptide is detected in a standard virus-infected extract under conditions where VPO isnot cleaved (10, 20), and no VP4 can be detectedby gel analysis. Third, the defective virus, whichdoes not make VP4, synthesizes the small fmnetpolypeptide.An alternative candidate for the small fmet-

labeled polypeptide is VPg, the protein that isfound covalently linked to the 5' end of viralRNA. The function of VPg and the location ofits coding sequences are currently unknown. Onecharacteristic ofVPg is that it cannot be labeled

in vivo with methionine (8), and we have thusexamined the polypeptides made in a standardvirus-infected S10 that can be labeled with[3H]lysine and [3S]methionine. Using polyacryl-amide gel analysis, we have observed that a low-molecular-weight polypeptide that comigrateswith the small [3S]fmet-labeled protein can bespecifically labeled with [3H]lysine and not with[3S]methionine (data not shown). Furthernore,synthesis of this polypeptide is enhanced at highMg2e concentrations, which correlates with theobserved increased level of synthesis of the smallfmet polypeptide. This evidence is only sugges-tive that the small fmet polypeptide might beVPg, and the identity of this polypeptide iscurrently under investigation in our laboratory.We have demonstrated that the DI virus syn-

thesizes a protein in vivo that appears to beidentical to the VPg found on the 5' end of thestandard virus RNA. The position of the codingsequence for this polypeptide must therefore bein the first 10% or the last 70 to 75% of thepoliovirus RNA genome.

ACKNOWLEDGMENTISThis work was supported by grants from the National

Science Foundation and from the National Institutes ofHealth (AI 12387). F.K. is the recipient of a National Institutesof Health postdoctoral fellowship.We also thank Oliver C. Richards for advice in the prepa-

ration of VPg.

LMRATURE CITED1. Bailey, J. M., and N. Davidson. 1976. Methylmercury

as a reversible denaturing agent for agarose gel electro-phoresis. Anal. Biochem. 70:75-85.

2. Bonner, W. M., and R. A. Laskey. 1974. A film detectionmethod for tritium-labeled proteins and nucleic acids inpolyacrylamide gels. Eur. J. Biochem. 46:83-88.

3. Celma, M. L, and E. Ehrenfeld. 1974. Effect of polio-virus double-stranded RNA on viral and host-cell pro-tein synthesis. Proc. Natl. Acad. Sci. U.S.A. 71:2440-2444.

4. Celma, M. L., and E. Ehrenfeld. 1975. Translation ofpoliovirus RNA in vitro: detection of two initiationsites. J. Mol. Biol. 98:761-780.

5. Cole, C. N., and D. Baltimore. 1973. Defective interfer-ing particles of poliovirus. II. Nature of the defect. J.Mol. Biol. 76:325-343.

6. Cole, C., D. Smoler, E. Wimmer, and D. Baltimore.1971. Defective interfering particles of poliovirus. I.Isolation and physical properties. J. Virol. 7:478-485.

7. Easley, C. W. 1955. Combination of specific color reac-tions useful in the peptide mapping technique. Biochim.Biophys. Acta 107:386-388.

8. Fernandez-Munoz, R., and U. Lavi. 1977.5' Termini ofpoliovirus RNA: difference between virion and nonen-capsidated 35S RNA. J. Virol. 21:820-824.

9. Flanegan, J. B., R. F. Petterson, V. Ambros, M. Hew-lett, and D. Baltimore. 1977. Covalent linkage ofprotein to a defined nucleotide sequence at the 5' ter-minus of virion and replicative intermediate RNA ofpoliovirus. Proc. Natl. Acad. Sci. U.S.A. 74:961-965.

10. Jacobson, M. F., and D. Baltimore. 1968. Morphogen-esis of poliovirus. I. Association of viral RNA with coatprotein. J. Mol. Biol. 33:369-378.

VOL. 30, 1979

488 HUMPHRIES, KNAUERT, AND EHRENFELD

11. Jense, HI, F. Knauert, and E. Ebrenfeld. 1978. Twoinitiation sites for translation of poliovirus in vitro:comparison of LSc and Mahoney strains. J. Virol. 28:387-394.

12. Kaufman, Y., E. Goldstein, and S. Penman. 1976.Poliovirus-induced inhibition of polypeptide initiationin vitro on native polysomes. Proc. Natl. Acad. Sci.U.S.A. 73:1834-1838.

13. Knauert, F., and E. Ehrenfeld. 1979. Translation ofpolio virus RNA in vitro: studies on [nS]fmet-labeledpolypeptides initiated in cell-free extracts preparedfrom poliovirus infected HeLa cells. Virology, in press.

14. Laemmli, U. K. 1970. Cleavage of structural proteinsduring the assembly of the head of bacteriophage T4.Nature (London) 227:680-685.

15. Lee, Y. F., A. Nomoto, B. M. Detjen, and E. Wimmer.1977. A protein covalently linked to poliovirus genomeRNA. Proc. Natl. Acad. Sci. U.S.A. 74:59-63.

16. Lee, Y. F., A. Nomoto, and E. Wimmer. 1976. Thegenome of poliovirus is an exceptional eukaryoticmRNA. Prog. Nucleic Acid Res. Mol. Biol. 19:89-96.

17. Phillips, B. A. 1972. The morphogenesis of poliovirus.Curr. Top. Microbiol. Immunol. 58:157-174.

18. Rekosh, D. 1972. Gene order of the poliovirus capsid

proteins. J. Virol. 9:479-487.19. Rose, J. K., and D. Knipe. 1975. Nucleotide sequence

complexities, molecular weights, and poly(A) content ofthe vesicular stomatitis virus mRNA species J. Virol.15:994-1003.

20. Rueckert, R. R. 1976. On the structure and morphogen-esis ofpicornaviruses, p. 131-213. In H. Fraenkel-Conratand R. Wagner (ed.), Comprehensive virology, vol. 6,Plenum Press, New York.

21. Summers, D. F., and J. V. Maizel. 1971. Determinationof the gene sequence of poliovirus with pactamycin.Proc. Natl. Acad. Sci. U.S.A. 68:2852-2856.

22. Taber, R., D. Rekosh, and D. Baltimore. 1971. Effectof pactamycin on synthesis of poliovirus proteins: a

method for genetic mapping. J. Virol. 8:395-401.23. Villa-Komaroff, L, N. Guttman, D. Baltimore, and

H. F. Lodish. 1975. Complete translation of poliovirusRNA in a eukaryotic cell-free system. Proc. Natl. Sci.U.S.A. 72:4157-4161.

24. Wagner, R. R. 1975. Reproduction of rhabdoviruses, p.

1-93 In H. Fraenkel-Conrat and R. R. Wagner (ed.),Comprehensive virology, vol. 4. Plenum Press, NewYork.

J. VIROL.