Nucleotide sequences of the mRNA's encoding the vesicular stomatitis virus G and M proteins determined from cDNA clones containing the complete coding regions.

0022-538X/81/080519-10$02.00/0

Nucleotide

Sequences

of the

mRNA's Encoding the Vesicular

Stomatitis Virus G and M Proteins Determined from cDNA

Clones

Containing the Complete Coding Regions

JOHN K. ROSE* AND CAROLJ. GALLIONE

Tumor VirologyLaboratory, The Salk Institute, San Diego, California 92138

Received 24 March 1981/Accepted 20 April 1981

The complete nucleotide sequences of the vesicular stomatitis virus mRNA's

encodingtheglycoprotein(G) and the matrixprotein (M) have been determined

from cDNA clonesthat contain the complete coding sequences from eachmRNA. The Gprotein mRNA is 1,665 nucleotides long,excluding polyadenylic acid, and

encodesaprotein of511amino acids including a signal peptide of 16 amino acids.

G protein containstwolarge hydrophobic domains, one in the signal peptide and

the other in thetransmembrane segment near the COOH terminus. Two sites of

glycosylation arepredicted at amino acid residues 178 and 335. The close

corre-spondence of the positionsof these sites with the reported timing of the addition

of the two oligosaccharides during synthesis of G suggests that glycosylation occurs as soon as theappropriate asparagine residues traverse the membrane of

therough endoplasmic reticulum. The mRNAencoding the vesicular stomatitis

virus M protein is831 nucleotideslong, excludingpolyadenylicacid, andencodes aprotein of229aminoacids. Thepredicted Mprotein sequence does not contain any long hydrophobic or nonpolar domains that might promote membrane

association.Theprotein is rich in basic aminoacids and contains a highly basic

amino terminal domain. Details of construction of the nearly

full-length

cDNA

clonesare presented.

Vesicular stomatitis virus (VSV) buds from thesurface of the host cellandthereby acquires

amembrane that contains spikes composed of the

single

viral

glycoprotein (G).

Ourprevious studies have shown thatG protein hasa COOH-terminal basic domain of29amino acids which isinternaltothe

lipid

bilayer of the virion and

an

adjacent hydrophobic

segment of20 amino acids which spans the

bilayer (27).

The

NH2-terminal 95% of G isexternaltothe

lipid bilayer

and containstwo

asparagine-linked

complex

ol-igosaccharides (10, 21). G

plays

anessential role

in

binding

of the virus to the host cell and presumably playsarolein directingbudding of virusfrom theplasma membrane of the host cell

(2,5).

The Gprotein is synthesized on membrane-bound

polyribosomes (18)

and inserted into the

rough endoplasmicreticulumas a nascent pro-tein chain (29, 34).A short

hydrophobic signal

sequence of16 amino acids is cleaved

from

the

NH2 terminus after insertion into the

rough

endoplasmicreticulum (12, 15,

22).

Gassumes a transmembrane configurationin the

rough

en-doplasmic reticulum (6, 13) with

only

a small

COOH-terminal segment

exposed

on the

cyto-plasmic face ofthe

rough

endoplasmic

reticu-lum. Transport of Gto the plasma membrane

occursviathe

Golgi

apparatus(la), with inter-organelle transport

probably

occuring

via coated vesicles

(28).

Atalatestageoftransport tothe

plasma

membrane,

one or twomolecules of

fatty

acidare

esterified

toG

(31).

TheVSV matrix

protein (M)

is

thought

tobe

a peripheral membrane

protein

that lines the inner surface of the virion

envelope, perhaps

interacting with the

lipid

bilayer,

the internal portion of

G,

and the

nucleocapsid

core

(re-viewed

by Wagner [35]).

Inadditionto a

struc-tural

function,

M

plays

arolein

directing

bud-ding ofvirusfrom infected cells

(14)

andmaybe involved in

regulating

transcription

of mRNA from the

single

negative

strandof

genomic

RNA

(4, 7, 9,16).

In contrast to

G,

M is

synthesized

on free polyribosomes

(18)

andassociates

rapidly

with

the plasma membrane fraction after

synthesis

(14).Thenatureoftheassociation ofMwiththe

plasmamembranefractionisnotclear.Itcould

beby associationwith the

lipid

bilayer

in

regions

containing

G

protein

or

by

association withother

-proteins such as those in

nucleocapsids

which

are

already

associated with the

plasma

mem-brane

(14).

519

on November 10, 2019 by guest

http://jvi.asm.org/

520

Knowledge of the complete primary amino acid sequences of G and M proteinsis clearly critical to a molecular understanding of their interactions with membranes and other cellular and viralcomponents.Ourpreviousstudieshave employed VSV cDNA clones that containonly fractions of the coding sequences from each

mRNA (24, 27), and only terminalmRNA and proteinsequences werereported. Wereporthere the isolation and the complete nucleotide se-quencesof cDNA clones that containtheentire

codingsequences ofG andM protein mRNA's.

From thesesequenceswepredicttheamino acid sequences of the G and M proteinsanddiscuss features of these predictedsequences.

MATERIALS AND METHODS

Materials. Reverse transcriptasewassupplied by

J.Beard,St.Petersburg,Fla.The Klenowfragmentof DNApolymeraseIandmostrestrictionendonucleases

were purchased from NewEngland Biolabs, Beverly,

Mass. NucleaseS1, PstI,and T4polynucleotidekinase

were fromBoehringer Mannheim, Indianapolis,

Indi-ana, and terminal deoxynucleotidyl transferase was

fromBethesdaResearchLaboratories, Bethesda,Md. Oligo(dT)12-18andoligo(dT)cellulose (T3)werefrom CollaborativeResearch, Waltham,Mass.[a-32P]dCTP and dGTPwerefromAmersham/Searle, Chicago, Ill.,

and unlabeled dNTP'swerefromP-LBiochemicals, Milwaukee,Wis.

RNAsynthesis. VSV mRNAwassynthesized in

vitro ina 10-mlreaction exactly asdescribed

previ-ously (26), with thefollowingmodification.After 4h

the reactionwasstopped byaddition ofsodium dode-cyl sulfateand sodiumacetate tofinalconcentrations of 1% and 0.5M, respectively.The entiremixturewas

then passed through a column containing 0.5 g of

oligo(dT)-cellulose,and thecolumnwaswashed with

10ml of 0.4 Msodiumacetate.The boundmRNAwas

eluted with distilled water and precipitated with ethanol. AllmRNAwasfrom the SanJuan strain of

the IndianaserotypeofVSV.

First-strand DNA synthesis and purification. First-strand DNAcopiesof themRNA'sencodingthe N, NS, M,and Gproteinsweresynthesized ina4-ml

reaction containing300,ug of totalVSV mRNA, 0.5

mM dATP, dGTP, and dTTP, 0.25 mM [a-32P]-dCTP (10 Ci/mmol),30mM /3-mercaptoethanol, 120

mMKCl,10mMMgCl2,4,000U ofreverse

transcrip-tase,300jigofoligo(dT)12 8andacytoplasmicextract

(1 mg of total protein) from baby hamster kidney

(BHK-21) cells prepared as described below. After

incubation for 30min at42°C, the reaction mixture

was extracted with phenol. Unincorporated dNTP's

wereseparated from cDNA by chromatographyof the

aqueousphaseonSephadexG-50.Theexcluded

frac-tionwaslyophilizedinasilicatedglass tube. The yield

oftotal cDNAfrom 300 ,ugof VSV mRNAwas

ap-proximately 40,ugasdeterminedfrom the

incorpora-tion of[a-32P]dCTP. The full-lengthcDNA copiesof

eachmRNAspecieswerepurified by electrophoresis ona1.5%alkaline agarosegel (30 mM NaOH,5mM

EDTA;1.5 mmby18 cmby 20 cm). The cDNA's were located byautoradiography (Fig. 1) or by staining with ethidium bromide and UVillumination. Stained bands wereexcisedfrom the gel,electroeluted, and precipi-tated with ethanol. The yield offull-length cDNA's wasapproximately 0.7

jig

ofG,2,ugof N, and 3ytgof both NS and M. Reaction conditionsgiven were opti-mized with respect to the concentrations of reverse transcriptase, KCl, and cell extracts to give maximal sizes of cDNA's.

Cell extract preparation. BHK cells (8 x 108 cells) growing in a 1-liter suspension werepelleted by centrifugation and suspended in 1.5 times the packed cell volume of 10 mM HEPES (N-2-hydroxyethylpi-perazine-N'-2-ethanesulfonic acid) (pH 7.5), 10 mM KCl, 1.5 mM magnesium acetate, and 7 mM

fl-mer-captoethanol. Cells were broken by 50 strokes of a Dounce homogenizer and centrifuged at 15,000 x g for 20min.The supernatant was dialyzed for 16 h against 1 liter of 10 mM HEPES (pH 7.5), 90 mM KCl, 1.5 mMmagnesium acetate, and 7 mM ,B-mercaptoetha-nol.The total protein concentration in the extract was 15 to20 mg/ml.All steps were carried out at 4°C, and samples of the extract were stored at -80°C.

Second-strand cDNA synthesis. Reactions for second-strand DNA synthesis (200

pl)

included 100 mM HEPES buffer (pH 6.9), 10mMMgCl2, 60 mM KCl,1mMdATP, dCTP, dGTP, and dTTP, 0.5

jig

of apurified cDNA, and 5 U of the Klenow fragment of DNApolymerase I. Reactions were for 15 h at 15°C and were terminatedby extraction with phenol fol-lowed by precipitation with ethanol. Samples were thenresuspended in75pl ofasolution containing3 mM ZnCl2,30mM sodiumacetate(pH.4.5),300 mM NaCl, and 75U ofS1 nuclease. Digestionwasfor 30 min at 370C. S1-treated DNAswere extracted with phenol andprecipitatedwithethanol and thenpurified byelectrophoresis on a 6% polyacrylamide gel.The nearly full-length double-stranded DNAs were de-tectedbyautoradiographyorstainingwith ethidium bromide. Theamountsof double-stranded DNAs

re-covered were 100 ng (G size class), 180 ng (N size class), and450ng(NSand M sizeclass),ascalculated from the cpm of32p in the first-strand DNA. Double-stranded DNAs were electroeluted and precipitated withethanol.

Homopolymer addition andcloning. Homopol-ymer addition wasbya modification ofapublished procedure(30). Addition of dCMP residuesto double-stranded cDNAwasfor10to30minat370C ina

25-pl

reactioncontaining100 to 500ngofDNA,0.2mM dCTP, and an amountof terminaldeoxynucleotidyl transferase (Tdt) that was known to add approxi-mately 10 to 20dC residues to HaeIIIfragments (5 pmol of total ends) of pBR322DNA.Thenumberof dC residues added wascalculated from the incorpo-ration of [a-32P]dCTP (100 Ci/mmol) that was in-cludedin testreactions. Unitsreported bythesupplier were generally 10- to 50-fold greater than whatwe

observed. Addition of 10 to 20dG residues to PstI-cleaved pBR322 DNAwascarried out asabove (0.2 mMdGTP) using 1to 2,ugofDNA.The linearform ofPstI-cleavedpBR322wasalways purified by agarose gel electrophoresis and electroelution before tailing.

on November 10, 2019 by guest

http://jvi.asm.org/

Equimolarquantities of dC-tailed insert DNA and dG-tailed plasmid DNA were annealed for 15 min at4°C in 10

p1

of0.5MNaCI-10mMTris (pH 7.4) and used totransform Escherichia coli strain C600 to tetracy-cline resistance (30). From 10 to 40 colonies were obtained per ng of insert DNA. Approximately 80% of the colonies obtained were ampicillin sensitive and had inserts at the PstI site. Small preparations of plasmid DNA (1 to2 ug) were analyzed for the size of theinsert DNA by PstI digestion. HaeIII and Hinfl digestions gave partialrestriction maps for each insert. Bycomparing these maps with previous data (24, 25, 27) we were able to assign the clonesunambiguously tothe appropriate mRNA's. Although apparently full-lengthdouble-stranded DNAs were used as starting material, alarge fraction (80% for G) of the inserts had deletions at one or both ends.

DNA sequence analysis. All sequence analysis was by the Maxam-Gilbert procedure as described previously (17, 25). DNAfragments (25 to 50 pmol) wereprepared by restriction enzyme digestion, alka-linephosphatase treatment, polyacrylamide gel elec-trophoresis,andelectroelution (25). End labeling was with 500,uCi of[y-32P]ATPand2Uof polynucleotide kinase (24).Labeled DNA strands were separated by electrophoresis onthin 5%polyacrylamide gels (0.35 mmby 16 cm by40cm) rather than the thickergels suggestedby Maxam and Gilbert (17). Thin gels gave betterresolution ofclosely spaced DNA strands. Gels were runat1,200 V with anelectric fan forcooling, and strand separation of fragments up to 800 base pairslongwasaccomplishedbyusing electrophoresis times of less than 8 h. Occasionally fragments for sequencing were obtained by cleavage with a second restriction enzyme followed by gel electrophoresis. Sequencinggels (0.35mmby16 cmby40,85,or152 cm) were electrophoresed at 1,500 to6,000 V, such that thegelremained warm(>40°C) throughout the run. Reaction times were reduced substantially for samples on which the sequence was to be readfor morethan300nucleotides from the labeled end. Gen-erally100,000to500,000dpmof32P-labeledfragment wereloaded per laneon152-cmgelssothatexposure could be carriedout at highresolution without flu-orescentscreens.Long (85or152cm)gelswerecutin sections, transferred to old exposedsheets ofX-ray filmfor backing, covered with plastic wrap, and ex-posedtoKodak XR-5 film for1to 4days.

RESULTS

The basic strategy we employed to obtain

nearly

full-length cDNAclones was: (i) to

opti-mize reverse transcription of

full-length

first-strand DNA

copies

of each

mRNA;

(ii)

to use

isolated, full-lengthfirst strandsastemplates for

second-strandDNAsynthesis;and(iii)topurify

the nearly full-length double-stranded DNAs

(obtained after Si nuclease treatment) by

gel

electrophoresis

before cloning. All cloning was ofdC-tailedinserts into thedG-tailed PstIsite ofpBR322 (3).

Inprevious experimentsweand othersfound thatadditionofa crude cellextract

greatly

en-hanced the synthesis of full-length VSV

mRNA's by theVSV viriontranscriptase(1, 26).

Becauseourinitialreversetranscription experi-ments showed that the yields of full-length cDNA's of VSV G and N mRNA's were low

relative tothe yieldsof cDNA's of the smaller

mRNA's(NS andM), weexaminedthe effect of

suchanextractonreversetranscription. Figure

1showstheproducts of reversetranscriptionof

total VSV mRNA synthesized in the absence andpresence ofthe extract. The yields of full-lengthG and NcDNA's were increased 5- and

2-fold, respectively, whereas the M and NS

cDNA's were increased only marginally

(1.2-fold). Becauseof the dramatic enhancementof

the yield of large cDNA, the cell extract was

included in all syntheses of single-stranded cDNA's for cloning. The mechanism of action of thecellextractisnotknown,but it presumably

acts by inhibiting RNase or bypromoting

un-folding of mRNA secondarystructure.

G and M cDNA clones contain the

com-plete coding sequences. Two recombinant plasmids,

pG1

andpM309,wereobtained which had insert sizes ofca.1,700and850nucleotides, respectively, consistent with theirhavingnearly complete copies of the G and M mRNA se-quences. Todeterminethe exact 5'endpoints of each clonedsequencerelativetothe

mRNA,

we

determined the nucleotidesequenceatthe end of the insertcorrespondingtothe 5' endof the RNA. The

sequencing

gel showing this region from the

pGl

insert is shown in Fig. 2. The sequence preceding the

homopolymer

tails is

12

-Origin

4PG

N

NS,M

FIG. 1. ReversetranscriptsofVSVmRNA

synthe-sized inthe presence(lane 2)orabsence(lane 1)

of

a

cytoplasmiccellextract.Reversetranscriptswere la-beled with[a-32P]dCTPand

analyzed

by

electropho-resis on an alkaline agarose gel. Identities

of

the cDNAbandsin theautoradiogramareindicated.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.503.319.395.431.590.2]

522

complementarytoallbutninenucleotidesatthe

5'end of theGmRNA sequence (24).Similarly,

the sequenceof the pM309 insert(Fig. 2) showed that it containedall but19nucleotides from the

5' M mRNA sequence (24). Because the

trans-lation initiation codons for theG andMmRNA's

are located 30 and 42 nucleotides from the 5' ends ofthe mRNA's(23, 24), these clonesclearly contain thesequencesencoding the NH2 termini ofG andMproteins.Subsequentsequence anal-ysisshowed that these clones contained the 3'-terminal mRNAsequences aswell.

FIG. 2. Sequencinggelsshowing thesequencesin

theregions ofpGlandpM309thatcorrespondtothe

5'-proximal regionsof Gand M mRNA's.Panels: A,

sequencefrom the5'-proximalAluIsite of thepGl

insert through the C-tails. Panel B, sequencefrom

the pBR322 HpaII site proximal to the 5' end of

pM309 readingintothe insert DNA. Numbers indi-catethealignmentof thesequenceswith the mRNA

sequence(see Fig.4and6).

The mechanism of formation of pGl and

pM309 can be explained from the 5'-terminal

sequencesof the mRNAs(Fig.4and6of

refer-ence 24). The priming of second-strand DNA synthesisinpG1 presumably occurred from the

3' end ofthe cDNA looped back to base pair perfectly with positions 13 through 18, leaving

anunpaired loop of six nucleotides. After second

strandsynthesis,digestion of the

loop

with

nu-clease S1 must have left the unpaired

nucleo-tides 10through 12 intact because they appear

in theclone. Inthe formation ofpM309, the 3' end of the cDNA

presumably

looped back to

pair with positions 21

through

24 and began priming. Nuclease S1digestionwascompleteor

nearly complete, generating aclone lacking 19

nucleotides of the 5' mRNAsequence.

Sequencing

strategy for

pGl.

Initial

re-strictionmapping combined with previous

par-tialsequences (24) allowedus tolocate thetwo

HaeIII sites within the

pGl

insert

(Fig. 3).

Initial

sequences were then established from the cen-tralHaeIIIsitereading520nucleotides toward

the 5' end of the mRNA and 350 nucleotides toward the 3' end. Additionalsequence was de-termined from the3' HaeIII site for520 nucleo-tides toward the 5' end of the mRNA. This approach, with sequencing gels 152 cm long, allowedus to establishatentativesequencefor almost the entire Ggeneusing onlytwo restric-tion sites.Arestrictionmap wasthenestablished

from this sequence by computer analysis (32).

Subsequentlywe were able to identify specific restrictionfragments requiredto complete the sequence onboth DNA strands and correct a

few errors that were generated by reading se-quences more than 450 nucleotides from the

labeled end. Figure3shows the locationsof sites

for restriction enzymes that were used in the sequence analysis and the

specific

regions

se-quenced.SeveralEcoRII sites

(CC

T GG) were

encountered in this

analysis.

These sequences

showed gaps in the sequence ladderatthe

sec-ondC residue (20), and thesequences atthese siteswere alwaysconfirmed

by

sequencing

the

complementarystrand.

Thecomplete sequence of theGmRNA pre-dicted from thesequence of the

pGl

clone

(and

previousdata, references24and27)is shown in

Fig. 4. The sequence contains a

single

open

reading frame for translation beginning at the

5'-proximal ATG codon (positions 29

through

31) andending 100nucleotides from the 3' end of the mRNA. That this is the

reading

frame encoding G proteinhas beenconfirmed

by

direct

sequencing

of G

protein

at both ends of the molecule (12, 15,27).

Nucleotide

sequencing

of the M mRNA. A J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.503.52.248.207.569.2]

a~

D v) <i 0 0 HI )I

C I,,H)PH,,H1,,re)H HH H

2 3 4 5 6 7 8 9 10 12 13 14 15 16

genonc . .

sequence

FIG. 3. Restriction map of thepGFI insert DNA. Sites for restrictionenzymnes that were used in sequencing are shown. The left end corresponds to the 5' end of G mRNA. Arrows indicate the regions sequenced andthe direction of sequencing. Numbers are in hundreds of nucleotides starting from the 5' end of the mRNA. Terminal PstI sites are at the junctions ofpBR32 sequences with the dG:dC tails. The 5'-terrninal sequence of G mRNA determnined previously from the genomic sequence is indicated (27). The sequence of the 3'. terminal 470 nucleotides were reported previously from a smaller clone, but also were sequenced on at least one DNA strand inpGI.

sequence from the 5' end of the M

protein

mRNA had been determined from thesequence

ofa DNA

primer

extended from the

adjacent

NSgenealong the genomeintothe M

mRNA-codingregion (24).This sequence

overlaps

and agreeswith thesequencesdetermined from the pM309 insert (Fig. 5) and with thesequenceof

a partial M clone

(pM101;

L.

Iverson,

unpub-lished

results;

12a). The remainder of the M

mRNAsequencewasdetermined

by sequencing

from the restriction sites indicated in

Fig.

5.All

regionswere

sequenced

onbothDNAstrandsor

sequencedatleast twice

(from

different

sites)

to ensureaccuracy.

TheMmRNA is831nucleotides

long

without polyadenylic acid and contains a

single

open

reading framefortranslation

extending

from the 5'-proximal ATG codonat

positions

42

through

44 to aTAGtermination codonat

positions

729

through 831

(Fig. 6).

Other

reading

frames are

blocked

extensively throughout

the mRNA se-quence,

leaving

little doubt that the

predicted

Mprotein sequence iscorrect.

Although

no

di-rect protein sequence is available for

M,

the

sequence of the ribosome

binding

site deter-mined

directly

fromthe mRNA

(23)

is in

com-plete agreement with the sequence

predicted

here from the cDNA clone. The sequence also matches the

partial

RNA sequence of the M

mRNA

determined

by

priming

on

genomic

RNA

fromthe adjacent NSgene

(24).

DISCUSSION

VSV G

protein

structure and modifica-tion. The nucleotide sequence

presented

here for theVSVglycoproteinmRNA

predicts

a pro-teinof511 aminoacids

(57,416

daltons),

includ-ing the NH2-terminal

signal

sequence of 16 amino acids that isnot presenton the mature

protein (12, 15, 27). The

predicted

COOH-ter-minal sequence of G contains a

hydrophobic

domain of20amino acids

(residues

463

through

482)followedby a hydrophilic domain (residues 483 through511). We have presented evidence

previously that the hydrophobic domain spans

thelipid bilayer ofthe viral envelope, and that

the COOH-terminal charged domain residues

insidethe viral envelope leaving more than95% of the protein protruding from the virus (27).

The COOH-terminal basic domain probably in-teractswith intemal virion proteins and

cellular

proteins during budding of the virus from the

plasma membrane.This portion ofGcouldalso

playarole in targeting the proteinto theplasma

membrane.

The VSV Gprotein contains twoapparently

identical Asn-linked complex oligosaccharides (10, 21). The precise locations of the

oligosac-charide attachment sites in G are of special

interest becausethetiming of the glycosylation eventsrelative to the timingofGprotein

syn-thesis has beenexaminedin detail invitro.

Ad-dition of

oligosaccharide

chainsby transfer from

alipid carrieroccurs onthe nascentpolypeptide chain (29, 34). The first addition occurs when about 38% of the chain has been synthesized, and the secondoccurswhen about70% has been

synthesized (29), assuming that therateof pro-teinsynthesisis

uniform

in thein vitrosystem.

N-glycosidic linkage

of

oligosaccharides

to pro-teins occurs at Asn-X-Ser or Asn-X-Thr se-quences(19).Inspection of the predicted G pro-tein sequencereveals 18Asnresidues, but only

twoofthese (amino acid residues 178and 335,

Fig. 4) occur in canonical glycosylation se-quences. We therefore presume that these are theactualglycosylationsites. These sites are at fractional distances of 0.35 and 0.66 from the NH2 terminus. Thenearlyexactcorrespondence ofthepositionsofthesesites with fraction ofG synthesizedwhen

glycosylation

occurs suggests thattransferofoligosaccharides tothenascent

protein chain occurs when theappropriate Asn residues traverse theroughendoplasmic

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.503.63.459.66.162.2]

fl0 :t VA- MK CTS 0LYS PHE THR ILE VAL PHE PRO HIS ASN GLN LYS GLY ASN TRP LYS

TCATTGGGGTGAATTGCAAGTTCACCATAGTTTTTCCACACAACCAAAAAGGAAACTGGA

70 80 90 100 110 120

ASN VAL PRD SER ASN TYR HIS TYR CYS PRO SER SER SER ASP LEU ASN TRP HIS ASN ASP AAAATGTTCCTTCTAATTACCATTATTGCCCGTCAAGCTCAGATTTAAATTGGCATAATG

130 140 150 160 170 180

LEU ILE GLY THR ALA ILE GLN VAL LYS MET PR) LYS 5ER HIS LYS ALA ILE GLN ALA ASP ACTTAATAGG CACAGCCATACAAGTCAAAATGCCCAAGAGTCACAAGG C3TATTCAAGCAG

19 200 210 220 20 240

GLY TRP MET CYS HIS ALA SER LYS TRP VAL THR THR CYS ASP PHE ARG TRP TYR GLY PRD ACGGTTGGATGTGTCATGCTTCCAAATGGG TCACTACTTGTGATTTCCGCTGGTATGGAC

250 260 270 280 290 300

LYS TYR ILE THR 1Q1]. 5ERR ILE ARG SER PHE THR Pr) SER VAL GLV GiLN CYS LYS GLU 5E

CGALASG TTART AITAAACACG TSCECRAAT CGATCTTCACTCCAT TGTA GAACA TGC AAGGAAA

310 320 330 0 0 350 360

ILE GLV GLN THR LYS GLN GLY THR TRP LE ASN PR GLY PHE PR) PR) GLN S CYS GLY

GCATTGAACAAACGAAACAAGGAACTTGGCTGAATCCAGGCTTCCCTCCTCAAAGTTGTG

370 380 390 400 410 420

TYR ALA THR VAL THR ASP ALA GLU ALA VAL ILE VAL GLN VAL THR PR) HIS HIS VAL IL GATATGCAAG M40 TGTGACGGATGCCGAAGCA~4A40TG 45&TGGTGCGTAT~ACTTAT T G TC4i60 GGTGACTC4C70 CAC CA TG Ta C

VAL ASP GLU TYR THR GLY GLU TRP VAL ASP SER GIL PHE ILE ASN GLY LYS CYS SER ASN

TGGTTGATGAATACACAGGAGAATGGGTTGATTCACAGTTCATCAACGGAAAATGCAGCA

490 500 510 520 530 540

TYR ILE CYS PR) THR VAL HIS m. THR TRP HIS SER ASPTYR LYSVAL LYS GLY

ATT ACA TA 5GCC C C ACTGW6CCA TAA C TCW7AC AA CC T G G&AT TCT G ACT G TCAA8 LEW CYS ASP SER ASN LEU ILE SER MET ASP ILE THR PHE PHE SER GLU ASP GLY GLU LW

GGCTATGTGATTCTAACCTCATTTCCATGGACATCACCTTCTTCTCAGAGGACGGAGAGC

610 620 630 640 650 660

SER SER LEW GLY LYS GLU GLY THR GLY PHE ARG SER ASN TYR PHE ALA TYR GLU THR GLY

TATCATCC%TGGGAAAGGAGGGCACAGGgTTCAGAAGTAACTACTTTG5 TATGAAACT2

60 680 690 7007172

GLY LYS ALA CYS LYS MET GLN TYR CYS LYS HIS TRP GLY VAL ARG LW PR SER GLY VAL

GAGGCAAGGCCTGCAAAATGCAATACTGCAAGCATTGGGGAGTCAGACTCCCATCAGGTG

730 740 750 760 770 780

TRP PHE GL] MET ALA ASP LYS ASP LEW PHE ALA ALA ALA ARG PHE PR) GLU CYS PR) GLU

TCTGGTTC9AGATGGCTGATAAGGATCT ,TTGCTGCAi2CAGATTCCJGAATGCCCftg

790 800 1o 38

GLY SER SER ILE SER ALA PR) SER GLN THR SER VAL ASP VAL SER LW ILE GiLN ASP VAL

AAGGGTCAAGTATCTCTGCTCCATCTCAGACCTCAGTGGATGTAAGTCTAATTCAGGACG

850 860 870 880 890 900

GLU ARG ILE LEW ASP TYR SER LEW CYS GLN GL T1R WRP SER LYS ILE ARG ALA GLY LW

T T G A G AGG A T C T T G G A T T AWTCC C TC TGj3gA A G A A ACCC48G A GCAAA A 5EA G AG C G G G

PR) ILE SER PR) VAL ASP LEU SER TYR LWU ALA PR) LYS ASN PR) GLY TM GLY PR) MA

TTCCAATCTCTCCAGTGGATCTCAGCTATCTTGCTCCTAAAAACCCAGGAACCGGTCCTG

970 980 990 1000 1010 1020

PHE

TTILEA ILE LEU LYS TiR PHE GUl THR ARG TiE IIC ARG AGL ASP ILE

CTTTCACCATAATCA

~~A'"TA

'CCTAAAATACTTTGAGACCAGATACATCAGAGTCGATA

1030 1040 1050 1060 1070 1080

ALA ALA PR) ILE LEW SER AR MET VAL GLY MET IE SER GLY THR THR T11 GLU ARG GLU

TTGCTGCTCCAATCCTCTCAAGAATGGTCGGAATGATCAGTGGAACTACCACAGAAAGGG

1090 1100 1110 1120 1130 1140 LEU TRP ASP ASP TRP ALA PR) TYR GLU ASP VAL GLU IIE GLY PR) ASN GLY VAL LW ARG

AACTGTGGGATGACTGGG6tCCATATGAG Al8GTGGACCCAfACGTGGAGGGAGTTCT1G

THR SER SER GLY TYR LYS PHE PRO LEU TYR Mr IIE GY HIS GLY MET LEU ASP SER ASP

GGACCAGTTCAGGATATAAGTTTCCTTTATACATGATTGGACATGGTATGTTGGACTCCG

1210 1220 1230 1240 1250 1260

LEU HIS LEU SER SER LYS ALA GLN V'L PHE GiL HIS PR) HIS ILE GLN ASP ALA MA SE

ATCTTCAT WTAGCTCAAAgGCTCAGGT TTCGAACAT5TC ATC GCCGC2

ATCTTCA1217ATGCCTTACAC128T G 90TC

GLN LEU PR ASP ASP GLU SER LEU PHE PHE GLY ASP THR GLY LEU SER LYS ASN PR) ILE

CGCAACTTCCTGATGATGAGAGTTTATTTTTTGGTGATACTGGGCTATCCAAAAATC CAA 1330 1340 1350 1360 1370 1380 GiL LW ML GiLU GLY TRP pHE SSE SER TRP LYS .. U5

TCGAGCTTGTAGAAGGTTGGTTCAGAT GAA

3125TATTOPT36144

1390 1400 1411 142 A3ItTT

ALU. M: VAL GLY ILE LW CYS ILE

CATAGGGTTAATC A

GG4T

CTCCCA GTTGGTATCCATCTTTTG CATTA

1450 1460 1470 1480 1490 1500 LWIUSLW 1W 115 WI ~GilN IL.8 TYR THR ASP ILE (LU MET ASN S L L I

AATTAAAGCAC ACCAA AAA ACAGATTTATACAGACATAGAGATG1AAC ACTTGGAA

1510 1520 1530 1540 1550 1560

AGTAACTCAAATCCTGCACAACAGATTCTTCATGTTTGGACCAAATCAACTTGTGATACC

1570 1580 1590 1600 1610 1620

ATGCTCAAAGAGGCCTCAATTATATTTGAGTTTTTAATTTTTATG

1630 1640 1650 1660

524

on November 10, 2019 by guest

http://jvi.asm.org/

-4

if

0) 0

e .C

I

t

H 14 M

z m I

i

2 3 4

,<, I H I

I r Il )

5 6 7 8

AA

genomic

At-sequence

FIG. 5. Restriction map of the pM309 insert DNA. Sites for restrictionenzymes used in sequencingare

shown. The left end corresponds to the 5' end of M mRNA. Arrows indicate the regions and directions sequenced; numbersrepresenthundredsof nucleotides. Arrows labeled A and Brepresentsequences

deter-minedfrom cDNA clones pM32andpM37 which contain smaller portions of the MsequencethanpM309.

Terminal PstI sitesareatthejunctions ofpBR322sequenceswiththedG:dC tails. The 5'-terminal mRNA sequence(genomic sequence) reported previously is indicated (24).

lum. Extensivefolding ofthepolypeptide chain

onthe COOH-terminal site of theglycosylation

sites is apparently notrequired to specify gly-cosylation.

Oneto twomolecules offatty acidare

esteri-fiedtoGproteinatalatestageinpassageof G

from the rough endoplasmic reticulum to the plasma membrane (31). The evidence suggests that thefatty acid isesterifiedtoserine residues (31). Ourpreliminary evidence (W. J. Welch, B. M. Sefton, and J. K. Rose, unpublished data) indicates that[3H]palmitate labelcanbe found

associated specifically with the 64-amino acid,

membrane-protectedtailpieceofG isolatedafter proteolysis of intact virions (27). Thereareonly

fiveserine residues in thisportionofG,andthey

are clustered around the NH2-terminal side of

the hydrophobic domain (Fig. 4). Presumably,

one or more of these residues are esterified to

fatty acid. Fatty acid esterified in this region could serve the obvious function ofpromoting stable association of the hydrophobic protein domain with the membrane.

Structure and function of VSV M. The nucleotidesequence of theVSV M mRNA

pre-sented here predicts a proteinsequence of 229 amino acids (26,064 daltons). The predicted amino-terminalsequence of M is highly basic. Eightlysine residuesoccurinthe first 19 posi-tions,and thesearetheonly chargedresiduesin this region (Fig. 6). A triple proline sequence

separates this domain from the remainder of the

molecule. There isevidence from analysisof M protein mutants andfrom in vitro studies that Mmayregulate VSV transcription (4,7, 9, 16). This basic domain might play a role in such

regulation, perhaps by interacting with the

ge-nomic RNA. The predictedsequence indicates

that M is themostbasic of the VSV N, NS, M, and G proteins,having 21 Lys, 10Arg, and8His residues and only 13 Asp and 13Glu residues. Although there isnodirect amino acidsequence

analysis available for M protein, this basic char-acter isconsistent with its isoelectric point de-termined by gel electrophoresis (4). Our at-temptstoobtainanNH2-terminalsequencefrom

M havebeen unsuccessful, suggesting that the NH2 terminus is blocked.

Evidence indicates that M protein is released into the soluble fraction ofinfected cells after

synthesis. It then associates rapidly with the plasma membrane fraction (14). It is notclear, however, whether there isanydirectassociation

of M withthelipid bilayer of the plasma

mem-brane. There is evidence that M will associate with membrane fractionsfromuninfectedcells,

although it is notclear that this interaction is biologically significant (8). Inspectionofthe

pre-dicted M protein sequence (Fig. 6) does not reveal any long hydrophobic or nonpolar

do-mains that might be inserted into the

mem-brane. This situation contrasts with that of the influenza virusmatrixproteinwhich hasa

cen-tralhydrophobicdomainthatmaybe membrane

FIG. 4. Nucleotide sequence ofthe VSV G mRNA and thepredicted protein sequence. The shaded

nucleotidesequenceis the ribosomebindingsite.NH2-terminalandCOOH-terminalhydrophobicdomains

areshaded,as arethetwopotentialglycosylationsitesand basic residues in the COOH-terminalhydrophilic

domain. The bracketed G residue(317) appeared clearlyas aGontwosequencing gels ofthe DNAstrand shown. ThesequenceofthecomplementaryDNA strandusingthesameDNApreparationshowedaclearA residueinsteadoftheexpectedC in thispositionin twoseparateexperiments. Thisresidue and the amino acid encoded shouldthereforebe considered tentative. Nucleotidenumberoneis linked tothe 5'capinthe mRNAsequence.

0.

I

_I

on November 10, 2019 by guest

http://jvi.asm.org/

[image:7.503.106.402.63.181.2]

526 ROSE AND GALLIONE

AACAGATATCACGATCTAAGT

10 20

G

Gtt*$£e:q-bt.C¢.-s-T-C: C.¢...-m..f..

.7...C77CIL

LEW GLY LEU GLY LEU GLY ILE ALA TYR

G GfA AGGGATCG^CACcACfi

GLU G ASPM TH SER MET GW TI ALA PR SM ALA PR ILE ASP LYS SE TE PHE GLY

ATGAAGAGGACACTAGCAT4GGAGTATGCTCCGAGCGCTCCAATTGACAAATCCTATTTTG

13 0 40 150 160 170 180

VAL ASP GLU MEr ASP THR TYR ASP PRO ASN GLN LEU ARG TYR GLU LYS PHE PHE PHE THR

GAGTTGAC GAGGACTGATG TCG ACA TTAAGATATGAGAAATTCTTCTT TA

A0 0 220 230 0

VAL LYS MET T VAL ARG SER ASN ARG PRO PHE ARG TH TYR SER ASP VAL ALA ALA ALA

CAGTGAAAATGACGGTTAG6ATCTAATCGTCCGTTCAGAACATACTCAGATGTGGCAGCCG

250 260 270 280 290 300

VAL SER HIS TRP ASP HIS MET TYR ILE GLY MET ALA GLY LYS ARG PRO PHE TYR LYS ILE

CTGTATCCSATTGG GATCACATGTACATSCGGAATGGCAGGGAAACGTCCCTTCTACAAAA

310 320 303 340 350 360

LEU ALA PHE LEU GLY SER SER ASN LEU LYS ALA T1 PRO ALA VAL LEU ALA ASP GLN GLY

TCTTGGCTTTTTTGGGTTCTTCTAATCTAAAGGCCACTCCAGCGGTATTGGCAGATCAAG

370 380 390 400 410 420

GUL] PR GLU TYR HIS TH HIS CYS GLU GLY ARG ALA TYR LEU PR) HIS ARG MET GLY LYS

GTCAACCAGAGTATCACACTCACTGCGAAGGCAGGGCTTATTTGCCACATAGGATGGGGA

430 440 450 460 470 480

THR PRO PRD MET LEU ASN VAL PRO GLU HIS PHE ARG ARM PRO PHE ASN ILE GLY LW TYR AGACCCCTCC~CATGCTCAATGTACCAGA ,CACTTCAGAAGACCATTCA 3TATAGGTCT

AGACCCCT490CAG 500 10 520

A30

A40

LYS GLY THR ILE GLU LEU TH MET THR ILE TYR ASP ASP GLU SER LW GLU ALA ALA PRO ACAAGGGAACGATTGAGCTCACAATGACC7ATCTACGATGATGAGTCACTG8GAA GCAG T

550 560 7 80 600

MET ILE TRP ASP HIS PHE ASN SER SER LYS PHE SER ASP PHE ARG GLU LYS ALA LW MET CTATGATCTGGGATCATTTCAATTCTTCCAAATTTTCTGATTTCAGAGAGAAGGCCTTAA

610 620 630 640 650 660

PHE GLY LEW ILE VAL GiL LYS LYS ALA SER GLY ALA TRP VAL LW ASP SER ILE SER HIS

TGTTTGGCCTGATTGTCGAGAAAAAGGCATCTGGAGCGTGGGTCCTGGATTCTATCAGCC

670 680 690 700 710 720

PHE LYS ***

ACTTCAAATGAGCTAGTCTAACTTCTAG CTTCTGAACAATCCCCGGTTTACTCAGTCT

730 740 750 760 770

7A0

CCTATT 5,t CCCTCGAACAACTAATAT CCTGTCTTT2TCTATC

C C T A A TTCC798GCCT

CTG800

820

FIG. 6. Nucleotidesequence ofthe VSVMmRNA and thepredictedMprotein sequence. Theshaded nucleotidesequenceis the ribosomebindingsite(23).Basic residuesneartheNH2terminus and thetriplePro

sequence areshaded.

associated (37). The nature of VSV M

protein

association with the viral

envelope clearly

re-quires furtheranalysis.

CG dinucleotide

deficiency

and codon

usage. There is an overall

deficiency

of the

dinucleotideCGin theG andMmRNA's. This deficiencyisalsoseenintheNandNS mRNA's (11). Onarandom basisconsideringbase

com-position, one would predict 74 CGs in the G mRNA-coding region and 34 CGs in the M

mRNA-coding region,whereas only24 and 18,

respectively, are observed. This deficiency is clear frominspection ofthe codon usage in both mRNA's(Fig. 7). ArgininecodonsAGAorAGG

are preferred over CGN arginine codons, and there isinfrequentuse of codons having CG in the second and third positions. The CG

defi-ciencyisalsoobservedbetween adjacent codons. Thatis,if one considers thefrequencyof codons

havingC in thethirdposition and G in the first

position, random assortment would generate 37 CGs in G mRNA and 20 CGs in M mRNA

betweenadjacent codons, whereas only 14 and

9,respectively,occur.Because theCGdeficiency

isobservedbetween

adjacent

codonsaswell as

withincodons, the

deficiency

is

probably

not an

adaptationof the virus to arelativeshortage of cellular tRNA's recognizing codons containing CG. Consistent with thisinterpretation, wefind

that CG is also deficient in the 479noncoding nucleotide ofthe G, M, N, and NS mRNA's where 18CGs areexpected and only 9 are found. Vertebrate DNA is known to be deficient in the CG dinucleotide (33), although the reason for this is not known. Ifthe mechanismof

selec-tionagainst CG is the same for cellular DNA and VSV RNA the selection must be able to operate attheRNAlevelbecauseVSVhas only RNA intermediates in transcription and

repli-cation.

Both the G and M mRNA's aremoderatelyA rich(32 and 30% A,respectively).Thepreference for codonshaving Ain the thirdposition

pre-sumablyreflects this basecomposition (Fig.7).

on November 10, 2019 by guest

http://jvi.asm.org/

[image:8.503.71.445.75.395.2]

Gprotein

U C

UPhe 10 Ser 10 Tyr

16 8

Leu 8 11 Term

4 1

CLeu 9 Pro 10 His

7 3

6 12 Gin

5 2

A lie 14 Thr 10 Asn

15 12

8 7 Lys

Met11 2

G Val 8 Ala 13 Asp

9 6

3 5 Glu

9 1

Mprotein

U C

UPhe 5 Ser 8 Tyr

9 4

Leu 4 2 Term

4 0

C Leu 1 Pro 3 His

3 3

1 7 Gin

4 3

A lle 4 Thr 3 Asn

6 3

1 3 Lys

Met11 2

GVal 2 Ala 6 Asp

2 3

3 6 Glu

2 2 l

FIG. 7. Codon usage in the I Firstpositions ofcodons are in secondpositions are indicated thirdpositionsareindicatedon

ACKNOWLEDGMI We thank Bart Sefton and Linda I mentsonthemanuscriptand otherr

Virology Laboratory for their helpft gratefultoLindaIversonfor providi pMlOl (IversonandRose,inpress)w firmsour5'-terminalsequencefromp: This workwassupported byPubli4

AI-15481 from the National Instituteo

Diseases andCA-14195fromthe Nati4

LrTERATURE CI 1. Ball,L.A.,and C. N. White. 197( ofgenesofvesicular stomatitis

VSV G SEQUENCES 527

Sci.U.S.A. 73:442-446.

la.Bergman,J. E.,K. T. Tokuyasu,and S. J.Singer. A

G

1981. Passage of an integral membrane protein, the 11 Cys 4 U vesicular stomatitis virus glycoprotein, through the 9 11 c Golgi apparatus en route to the plasma membrane. 1 Term O A Proc.Natl.Acad.

Sci.

U.S.A.78:1746-1750.

1 Tr 15 G 2. Bishop, D.H.L.,P.Repik, J. F.Obijeski,N. F. Moore,

0Trp15 G and R. R.Wagner.1975.Restoration ofinfectivityto spikeless vesicular stomatitis virus bysolubilized viral 11 Arg 0 U components. J.Virol.16:75-84.

5 1 C 3. Bolivar, F.,R.L. Rodriguez,M.C.Betlach, and H. 11 3 A W. Boyer. 1977.Construction and characterization of

7 0 G newcloningvehicles. Gene 2:95-112.

--- - - --- 4. Carroll, A.I.,andR. I. Wagner. 1979. Role of the

10 Ser 10 U membrane (M) protein in endogenousinhibitionof in 8 7 A vitro

trnscription

byvesicular stomatitis virus. J.

Virol.

8

5

7rg

AC

29:134-142.

15 Arg 8 C 5. Cartwright, B., C. J. Smale, and F. Brown. 1969.

16 3 G Surface structure of vesicular stomatitis virus. J. Gen.

-- - - --- - Virol.5:1-10.

17 Gly 9 U 6. Chatis, P. A., and T. G. Morrison. 1979. Vesicular 11 4 C stomatitis virusglycoproteinis anchoredtointracellular

16 19 A membranes nearits carboxyl end and isproteolytically

9₉6_{6 G}

G>

_{7. Clinton,}cleaved at its amino terminus. J. Virol. 29:957-963._G.

_M.,

_{S. P.}

_Little,

_{F. S. Hagen, and A. S.} Huang.1978.The matrix(M) protein of vesicular sto-matitis virusregulatestranscription.Cell 15:1455-1462. 8. Cohen, G.,P.Atkinson,and D.Summers.1971.

Inter-A G action of vesicularstomatitis virusstructural proteins 7 Cys 0 U with HeLaplasmamembranes. Nature(London)231:

7 Cys 0 U

~~~121-123.

5 1 C 9. Combard, A., and C.Printz-Ane. 1979. Inhibition of 0 Term 1 A vesicular stomatitis virustranscriptasecomplex by the

0 Trp 3 G virion envelope M protein. Biochem. Biophys. Res.

- - - . Commun. 88:117-123..

3 Arg 2 U 10. Etchison, J.R.,J. S. Robertson and D. F. Summers. 5 0 C 1977. Partial structural analysis of theoligosaccharide 3 O A moieties of the vesicular stomatitis virusglycoprotein O 0 G by sequentialchemical andenzymaticdegradation.

Vi-rology78:375-392.

11.

Gallione,

C.J., J. R.Greene,L E.Iverson,and J. K.

6 Ser 1 U Rose. 1981. Nucleotide sequences of the mRNA's en-0 3 C coding the vesicular stomatitis virus N and NS proteins. 11 Arg 6 A J.Virol.39:529-535.

10 2 G 12. Irving,R.A., F. Toneguzzo, S. H. Rhee, T.Hofmann,

_

- - - and H. P. Ghosh. 1979. Synthesis andassemblyof

9 Gly 5 U membraneglycoproteins:presenceof leaderpeptidein 4 2 C nonglycosylatedprecursorof membraneglycoproteinof vesicular stomatitis virus. Proc.Natl. Acad. Sci. U.S.A.

3 4 A 76:570-574.

10 4 G 12a.Iverson, L., and J. K. Rose. 1981.Localizedattenuation - ---- - - and discontinuoussynthesisduringVSVtranscription.

G and MmRNA's.

1.Cell

23:477-484.

Gdicated

on

theNA's-,

13.Katz,F.N.,andH. F. Lodish. 1979.Transmembrane

[dicated on the left, biogenesis of the vesicular stomatitis virus glycoprotein. acrossthe top, and J.Cell. Biol. 80:416-426.

theright. 14.Knipe, D., D.Baltimore, and H. Lodish.1977.Separate pathwaysof maturation of themajorstructuralproteins

ENTS of vesicular stomatitis virus. J. Virol. 21:1128-1139. 15.Lingappa, V. R.,F. N. Katz,H. F. Lodish,and G. verson forcriticalcom- Blobel. 1978. A signal sequence for the insertion of a members ofthe Tumor transmembrane glycoprotein. J. Biol. Chem. 253:8667-al suggestions. We are 8670.

ngsequencedata from 16.Martinet, C., A. Combard, C. Printz-Ane, and P. vhich overlaps and con- Printz. 1979. Envelope proteins and replication of

ve-M309. sicular stomatitis virus: in vivo effects of RNA' tem-cHealthService grants perature-sensitive mutations on viral RNA synthesis. J. ofAllergyandInfectious Virol.29:123-133.

onal Cancer Institute. 17. Maxam,A.M.,andW.Gilbert.1980.Chemical sequenc-ingof DNA.MethodsEnzymol.65:501-561.

[TED 18. Morrison, T. G., and H. F. Lodish.1975.Site of

synthe-6.Order oftranscription sis of membrane andnonmembrane proteinsof vesicu-virus. Proc.Natl. Acad. lar stomatitis virus. J. Biol. Chem.250:6955-6962.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:9.503.56.435.51.527.2]

19. Neuberger, A., A. Gottschalk, R.D.Marshall,andRI.

G.Spiro.1972.Carbohydrate-peptidelinkages in

gly-coproteins and methods fortheirelucidation, p. 450-490. In A. Gottschalk (ed.) The glycoproteins: their composition,structureand function.Elsevier, Amster-dam.

20. Ohmori, H., J. I. Tomizawa, and A. Maxam. 1978. Detection of5-methylcytosine in DNA sequences. Nu-cleicAcids Res. 5:1479-1485.

21. Reading,C. L.,E.E.Penhoet, andC. E. Ballou. 1978. Carbohydrate structure of vesicular stomatitis virus glycoprotein. J. Biol. Chem. 253:5600-5612.

22. Rose, J. K. 1977.Nucleotide sequences of ribosome rec-ognitionsitesonmessenger RNAsof vesicular stoma-titis virus. Proc. Natl. Acad. Sci. U.S.A.74:3672-3676. 23. Rose, J. K. 1978.Complete sequence of ribosome

recog-nition sitesinVSV mRNAs. Cell14:345-353. 24. Rose, J.K. 1980.Complete intergenic and flanking gene

sequencesfrom the genome of vesicular stomatitis virus. Cell 19:415-421.

25. Rose,J. K.,and L. E. Iverson.1979.Sequences from the 3' ends of VSV mRNA'sasdetermined from cloned DNA.J. Virol.32:404-411.

26. Rose, J. K., H. F.Lodish,and M. L. Brock. 1977. Giant heterogeneouspolyadenylic acid on vesicular stomatitis virus mRNAsynthesized in vitro in the presence of S-adenosyl hemocysteine. J. Virol.21:683-694. 27. Rose, J. K.,W. J.Welch, B. M.Sefton, F. S. Esch,

and N. C. Ling.1980.Vesicular stomatitis glycoprotein isanchored in the viral membrane byahydrophobic domain near the COOH-terminus. Proc. Natl. Acad. Sci.U.S.A. 77:3884-3888.

28. Rothman, J.E.; and R. E. Fine.1980.Coated vesicles transportnewly synthesized membrane glycoproteins from endoplasmic reticulum to plasma membrane in

twosuccessive stages. Proc. Natl. Acad. Sci.U.S.A.77: 780-784.

29. Rothman, J.E., and H. F. Lodish.1977.Synchronized transmembrane insertion andglycosylationofa mas-centmembraneprotein.Nature(London) 269:775-780. 30. Roychoudhury,R., E. Jay, and R. Wu.1976.Terminal labeling and addition ofhomopolymer tracts to duplex DNAfragments by terminal deoxynucleotidyl transfer-ase.Nucleic Acids Res. 3:101-116.

31. Schmidt, M.F., and M. J. Schlesinger. 1979. Fatty acid bindingtovesicular stomatitisvirusglycoprotein: a new typeofpost-translational modification of the viral gly-coprotein. Cell 17:813-819.

32. Staden, R. 1977. Sequence data handling by computer. Nucleic Acids Res. 4:4037-4051.

33. Swartz, M.N., T. A.Trautner,and A.Kornberg. 1962. Enzymaticsynthesis of DNA: further studies on nearest neighbor base sequences in deoxyribonucleic acids. J. Biol. Chem. 237:1961-1967.

34. Toneguzzo,F., andH. P. Ghosh. 1977.Synthesis and glycosylationinvitro ofglycoprotein of vesicular sto-matitis virus. Proc. Natl. Acad. Sci. U.S.A. 74:1516-1520.

35. Wagner, R. R.1975.Reproductionofrhabdovirus, p. 1-93. In H. Fraenkel-Conrat and R. R. Wagner (ed.), Comprehensive virology. Plenum Publishing Corp., New York.

36. Wensink, P.C.,D. J.Finnegan, J. E. Donelson, and D. S. Hogness. 1974. A system for mapping DNA sequences in the chromosomes of Drosophila melano-gaster.Cell 3:315-325.

37.Winter, G., and S. Fields. 1980. Cloningof influenza cDNA into M13: the sequence of the RNA segment encodingtheA/PR/8/34 matrix protein. Nucleic Acids Res.8:1965-1974.

on November 10, 2019 by guest

http://jvi.asm.org/