0022-538X/91/073421-08$02.00/0
CopyrightC) 1991,American Society for Microbiology
Characterization of V Protein
in
Measles
Virus-Infected Cells
ELIZABETH A. WARDROPt ANDDALIUS J. BRIEDIS*
Department ofMicrobiology and Immunology, McGill University, 3775 University Street, Montreal, QuebecH3A2B4, Canada
Received 28 January1991/Accepted 22 March 1991
AneditedmRNAtranscribed fromthephosphoprotein (P)geneof measlesvirus (MV) has beenpredictedto encodeacysteine-rich proteindesignated V. ThismRNA containsasingleadditional nontemplated G residue whichpermitsaccesstoanadditional protein-coding reading frame.Suchanedited Pgene-specificmRNA has beendetectedinMV-infectedcells, butnocorresponding proteinhasyetbeenidentifiedinvivo.Wereportthe useof antiseradirected against synthetic peptides correspondingtofive different regions ofthepredictedMV
V proteinamino acidsequence toanalyse MV-specificproteins synthesized in vivoand in vitro. The MV V
protein (40 kDa) was detected in MV-infected cells in a diffuse cytoplasmic distribution, a predominant
subcellularlocalization distinct from that ofvirusnucleocapsids. The proteinwasfoundtobephosphorylated andtobemaximally synthesizedat16hpostinfection,whenMV-specific structural protein synthesiswasalso
maximal.Antiserumdirected againstapeptide(PV2) correspondingtoamino acids 65to87 ofthe Vprotein
aminoacidrecognized thePproteinbutnotthe Vprotein, indicatingthatthePand Vproteinsmaybefolded
differentlyat ornearthis region sothatthe PV2 sequence is inanexposed position at thesurface oftheP proteinbutnot atthesurface ofthe Vprotein.
Measles, ahighly contagious acute viral disease of child-hood, is caused by measles virus (MV), a member of the morbillivirus subgroup of theparamyxovirus family of neg-ative-stranded RNA viruses. Inrare cases,persistent
infec-tionwithMVresults in subacute sclerosingpanencephalitis,
a uniformly fatal central nervous systemdisease (25). For-tunately, infection with MV can be prevented by
adminis-tration ofalive attenuated virus vaccine(7, 8).
TheMV genome is a nonsegmented single-stranded neg-ative-sense RNA withanestimatedMrof4,500,000 (3, 14). The complete genomic sequence of MV has now been determined. It contains approximately 15,900 nucleotides,
the exact number depending on the virus strain (17). Six
structural proteins have their genes sequentially arranged
along the MV genome: the nucleoprotein (NP), the
phos-phoprotein (P), the matrix protein (M), the fusion protein (F), the hemagglutinin (HA), and the large or polymerase
protein (L). In addition, a small nonstructural polypeptide
(C)hasbeen identified inMV-infected cells(4).
It hasrecently been demonstrated that multiple different speciesofmRNAtranscriptscanbederived fromPgenesof a number of paramyxoviruses (6, 12, 19, 23, 24, 26, 27).
Although the process has been termed a form of RNA
editing,these different mRNAspeciesappeartoresult from
thecotranscriptional additionofnontemplatedGresiduesat atleastonespecificlocation(27).Thiscan causechangesin theprotein-coding readingframe insomemRNAs, leadingto thesynthesis ofatleasttwodifferentproteinscoterminalat theiraminotermini: the Pproteinandacysteine-rich protein termed V. Two slightly different coding strategies exist
among paramyxoviruses for the expression of P and V
proteins. TheMV and Sendai virus P proteins areencoded
by mRNAs faithful to the genomic sequence, while the
synthesisof Vprotein requiresanadditionalnontemplatedG residue within its mRNA sequence (6, 27). On the other
*Corresponding author.
tPresent address: Bio-Mega Inc., 2100, rue Cunard, Laval, QuebecH7S 2G5, Canada.
hand, during infection with simian virus 5 (SV5), mumps
virus, or parainfluenza virus type 2 or type 4, mRNAs faithful to the original genomic sequence encode V while mRNAs containing nontemplated G residues encodeP (12, 18, 19, 26). The relative frequency of mRNAs containing suchG insertions hasbeenestimatedtobe 50% forMV(6),
45% forSV5 (26), 31%for Sendai virus (27),and 40to50% formumpsvirus (19, 24).
An edited mRNAderived from the MV P gene has been identified in vivo (6) by cDNA cloning, cDNA and RNA
sequencing, and invitrotranslation. TheVprotein-specific
mRNAwasfoundtocontainanextraGnucleotideinserted at a site comprising three genomically encoded G residues
corresponding to positions 752 to 754 of the originally published P gene sequence (4). The edited mRNA was
predictedtoencodeaVprotein containing 299 amino acids, andacorresponding proteinbandcould beseenafterinvitro
transcription and translation of the corresponding cDNA
clone. The authors did not, however, provide any direct
evidence thataprotein correspondingtothe Vcoding region
was presentinMV-infected cells.
Inthis articlewe report the generationand
characteriza-tionof antisera directedagainstaseriesofsynthetic peptides corresponding to different regions of the predicted amino acid sequence of the MV V protein. We have used these antiseraas probesto definitely identify and tocharacterize the in vivo pattern of V protein expression in MV-infected cells.
MATERIALS ANDMETHODS
Cells and viruses. Vero cells were grown in Dulbecco's modification of Eagle's medium supplemented with 10% fetal bovineserum.The Edmonston strain ofMV(American
Tissue Type Collection) was propagated as previously de-scribed (2).
Generationofantiseraagainst synthetic peptides. The syn-thesis of five peptides correspondingtodifferentregions of
the MV Vprotein (Table 1) wascommissionedfrom
Immu-nodynamics (La Jolla, Calif.). Peptides were coupled to
3421
on November 10, 2019 by guest
http://jvi.asm.org/
TABLE 1. Amino acid sequences of synthetic peptides usedto
generatemonospecific and polyspecific rabbit antisera against the MV P and Vproteins
Location within
Peptide Amino acid sequence Vproteina
PV1 IRALKAEPIGSLAIEEAMAA 16-35
PV2 LSAIGSTEGGAPRIRGQGPGESD 65-87
PV3 DIGEPDTEGYAITDRGSAPIS 154-174
Vi SKVTLGTIRAR 256-266
V2 RTDTGVDTRIWYHNLPEIPE 280-299
a Numberingreflectsthe sequenceofMV Vproteinassuming addition ofa single nontemplated Gresiduetothethreetemplated G residuesatpositions 252 to 254of thepublishedMV P genesequence(4)assubsequently corrected
(6).
keyhole limpet hemocyanin (KLH) or bovine serum albu-men(BSA) by addition ofan amino-terminal cysteine
resi-due before treatment with mmaleimidobenzoyl-N-hydroxy-succinimide ester. Peptide V2 (corresponding to the exact
carboxylterminusof V)was leftwithafree carboxylgroup
at its carboxyl terminus, while all other peptides were
synthesizedwithcarboxyl-terminal carbaminogroups.
DuplicateNewZealandWhite rabbitswereinitially
immu-nized with peptide-KLH conjugate in Freund's complete adjuvant. Approximately 0.2 mg ofpeptide conjugate was
injected subcutaneouslyateach of five sitesalongtheback. Rabbitsweresubsequently reimmunizedatotalof fourtimes
at 4- to6-week intervals in a similar manner with
peptide-conjugate in Freund's incomplete adjuvant. All
resulting
antisera were tested for their ability to react with the
peptides used in immunization by enzyme-linked immuno-sorbent assays (ELISAs) with peptide-BSA conjugates as
theantigen. Preimmunesera wereusedasnegativecontrols. Construction of plasmids for in vitro transcription of
mRNAsspecific tothe PandVproteins. The MV Vtranscript differs from the MV P transcript by the addition of a
nontemplated Gresidueatthe stretch of threetemplated G
residues at positions 752 to 754. We had previously
de-scribedcloning ofacDNA ofthe MV P coding region into thePstI site ofpBR322, creating pMV-P(2).We have now
usedasite-specific mutagenesis protocolto covertthis intoa
copy ofthe V protein-coding region. The P coding region consistedof1,565nucleotides bounded bya5'XhoIsite and
a3' HincII site. The plasmid containing this coding region
was digested with Sacl to release a DNA fragment 765
nucleotidesinlength thatcontainedtheGinsertion site.The
fragment was then ligated into the SacI site of M13mpl9 (Bio-Rad), and oligonucleotide-directed site-specific
muta-genesis was performed according to the instructions ofthe
supplier. The primer used for mutagenesis (Regional DNA
SynthesisLaboratory, Universityof Calgary, Calgary,
Can-ada)had the sequence
5'-dCCATTAAAAAGGG_jCACAG
AG-3',where the underlined G was the additional
nontem-plated nucleotide. Confirmation of successful mutagenesis was obtained by dideoxy chain termination sequencing of
alkali-denatured plasmid DNA performed with
deoxyade-nosine-5'-[35S]thiotriphosphate
(500 Cilmmol; AmershamCorp.)and T7 DNA polymerase (Sequenase; United States
Biochemical Corp.) accordingtothe protocol of the
suppli-ers. The primer used for sequencing was 5'-dCTTTCCGA AGCTTGG-3' (Regional DNA Synthesis Laboratory,
Uni-versityofCalgary). The mutated Sacl fragment was inserted
into the Sacl site of pMV-P to create plasmid pMV-V,
containing the V coding sequence. Plasmids pMV-P and
pMV-V
weredigested
withXhoIandHinclltoreleasethe P and Vcoding
sequences, which were thenseparately
in-serted into an SP6/T7 vector
(Amersham)
to createpAM18MV-P
andpAM19MV-V,
respectively.
Invitrotranscriptionandtranslation. Invitro
transcription
in the presence ofthe capanalog
diguanosine
triphosphate
[m7G(5')ppp(5')Gm;
Pharmacia] wasperformed
with SP6polymerase
according
to the instructions of thesupplier
(Pharmacia).
Theresulting
RNAtranscripts
weretranslated in micrococcal nuclease-treated rabbitreticulocyte
lysate
according
to the instructions of thesupplier (Promega
Corp.).
In vivo radiolabeling of infected-cell proteins. Confluent
monolayers
of Vero cells were infected with MV at amultiplicity
of infection of10PFUpercell. Cellswerepulse-labeledasdescribed
previously (2)
with eitherTran[35S]label
(ICN
Radiochemicals),
[35S]cysteine, [35S]methionine,
or32p;
(AmershamCorp.)
in minimum essential medium defi-cient inmethionine, cysteine,
orphosphate, depending
on thelabeled substance added.Analysis of
MV-specific proteins.
In vivo- and invitro-synthesized
proteins
wereincubatedin RIPAbuffer contain-ing 0.1% BSAand theappropriate
antiserum for2 hat4°C.
Forimmunoprecipitation
of invivo-synthesized
proteins,
100 ,ul of cell
lysate
was mixed with 5 ,ul of antiserum. Forimmunoprecipitation
ofinvitro-synthesized
proteins,
5,ul
ofthe 50-,ul total translation reaction was mixed with 5
,ul
of antiserum. ProteinA-Sepharose
beads(Pharmacia)
weresubsequently
addedtoeachmixture,
andthe mixtureswere then incubated for 1 h at4°C.
Immunecomplexes
werewashed three times with RIPA buffer
containing
1% BSA and twice with RIPA buffer without BSA. Then 35p.l
ofprotein lysis
bufferwasadded,
incubatedat37°C
for10min,
heatedto
95°C
for3min,andanalyzed
by
10%discontinuous sodiumdodecyl
sulfate-polyacrylamide gel
electrophoresis
(SDS-PAGE)
according
to the methoddescribedby
Laem-mli(13).Indirect immunofluorescence. Vero cells were grown to
40%
confluence onmicroscope
coverslips
before infectionwith MV.UninfectedVerocellswereusedascontrols.After 16
h,
cells were washed withphosphate-buffered
saline(PBS)
and fixedby
treatment with 2.5%paraformaldehyde
for20min. Cellswere
permeabilized by
treatmentwith0.5% TritonX-100 in PBSfor10min.Following
three washes withPBS,
cellsweretreatedwithnormal goatserum(Cedarlane)
for30minand thenwith
primary
antiserum for1h. Acontrolwas done without addition of
primary
antibody
for both infectedanduninfected cells. Cells werethenwashed threetimes with PBS and treated with goat anti-rabbit
immuno-globulin
G(IgG) coupled
to fluoresceinisothiocyanate
(FITC) (Cedarlane)
for30min. Cellswerewashedagain
withPBS before inversion over a
drop
ofglycerol
containing
Citofluor
(Amersham)
onglass
slides forphotography.
RESULTSPreparation of antibodies directed against synthetic pep-tides. In order to study the
expression
of Vprotein
in MV-infectedcells,
fivepeptide
sequences werechosen from thepredicted
amino acid sequence ofthe Vprotein
for theproduction
ofantipeptide
antibodies(Table 1).
Peptides
were chosen to maximize theirhydrophilicity
while alsooptimizing
location andsecondary-structure
predictions.
Threeof thepeptides synthesized
(PV1, PV2,andPV3)maptotheamino-coterminal
region
ofthePandVproteins,
andantibodies directed
against
thesepeptides
wereexpected
toon November 10, 2019 by guest
http://jvi.asm.org/
u I u I u I
> it
it
u I u I u I u I
200000 97 400
68000
2 0- V V > >> > > > _ - c N
a.. a.CL4 a. a. a. > > > >
M C M C M C M C M C M C MC
200000
97 400
68 000
43 000
43 000
[image:3.612.327.553.81.269.2]29 000
FIG. 1. Analysis of [35S]methionine-labeled polypeptides from uninfected (U) Vero cells as well as from Vero cells 16 h after
infectionwith MV (I). Cell lysates were immunoprecipitated with
preimmuneserum(Pre)orpeptideantiserum PC20,PV1,PV2,PV3, Vi,orV2. Sampleswere analyzedon a10% SDS-polyacrylamide
gel. Positions ofmolecular weightmarkers (not shown)areindicated ontheright.
be able to recognize both P and V. The remaining two peptides (Viand V2)maptotheuniquecarboxyl terminus of
the Vprotein,andantibodies directed against these peptides
wereexpectedtobe able torecognize theVproteinalone. Aftersynthesis, the peptideswereconjugated with KLH, and each was used to immunize two New Zealand White rabbits. The titers of the resulting antibody response were
determinedbyELISA with preimmuneserumas anegative
control. Afteraseriesoffiveimmunizations foreachrabbit, all antisera had titers ofat least 105 except that directed against Vi. This last antiserumreached atiter of only 5 x
104. Nevertheless, the antibody titer of the antiserum
di-rectedagainst theVi peptide would normally beconsidered adequate for use in immunoprecipitation. Preimmune sera
from all 10 rabbits werepooled and then usedas anegative control. Additionally, peptide antiserum PC20 (generously provided by C. D. Richardson), directedagainsta carboxy-terminal region unique to the P protein (20) and thus not expressedinthe Vprotein, wasalso usedas acontrol.
Detection ofVproteininMV-infectedcells. To determine
whether theMVV protein was indeed synthesized invivo,
[35S]methionine-labeled proteins from MV-infected Vero
cells were immunoprecipitated with our different antisera
and analyzed by SDS-PAGE. The MV P protein has a
predicted size ofapproximately 72 kDa (17), while the V
protein has beenpredicted tobeapproximately 40 kDa (6).
Noproteinsweredetectedwhenuninfected-celllysateswere
immunoprecipitated with any of our peptide antisera or when infected-cell lysates were immunoprecipitated with
preimmune serum(Fig. 1).However,twopolypeptideswith
migrations correspondingto70and40 kDaweredetected in MV-infected cell lysates in patterns corresponding to the
predicted specificity of the different peptide antisera and wereconsideredlikelytobethe PandVproteins. Evidence in support of such identification is that (i) the 70-kDa
polypeptide migratedatthesamepositionasthe P
immuno-precipitated by thePC20 antiserum (Fig. 1,lane 2), (ii) the
29 000
FIG. 2. Analysis of polypeptides from Vero cells 16 h after infection withMVfollowingdifferentialradiolabeling with
[355]me-thionine (M) or [35S]cysteine (C) and immunoprecipitation with preimmune serum (Pre) or peptide antiserum PC20, PV1,PV2,PV3,
Vi,
orV2. Samples were analyzed on a10%SDS-polyacrylamide gel. Positions of molecularweight markers (not shown)areindicated ontheright.70-kDapolypeptidewasrecognized by antisera predictedto
bespecifictoboth the PandVproteins (PV1,PV2,andPV3) butnotby antiserapredictedtobespecifictothe Vprotein
alone
(Vi
and V2), and (iii) the 40-kDa polypeptide wasrecognizedby theV-only specific antiserumV2 aswellasby
antiseraspecifictoboth Pand V (PV1 and PV3) but notby
the P-onlyspecific antiserum PC20.
Despite multiple attempts at immunoprecipitation and
autoradiographs ofmuch longer exposure and greater
inten-sitythan those shown, neitherthe
Vi
northe PV2 antisera could detecta band correspondingto the V protein. How-ever, theVi
peptide was predicted tobe the mosthydro-phobic ofoursynthetic peptides. The region of amino acid sequence unique to V encoded after the G insertion site contains only 62 amino acids. The
Vi
peptide sequenceseemed, after that of V2, to be the most favorable for antibody recognition based on secondary-structure
predic-tions. Suchpredictionsdo not, however, guaranteeantibody recognition ofthe native protein. The PV2 antiserum was able to immunoprecipitate the 70-kDa polypeptide but not
the 40-kDa
polypeptide,
although the sequence of the pep-tidewasderivedfromaportion oftheP/V amino-coterminal region. This may indicate that the P and V proteins arefolded differently at or near this region so that the PV2 sequence is in an exposed position at the surface of theP
proteinbutnot atthe surface ofthe Vprotein.
Immunoprecipitation of differentially radiolabeledproteins
from MV-infected cells. In order toprovide more evidence that the 70- and 40-kDa polypeptides indeed
represented,
respectively, the P and V proteins,MV-infected cellswere
separately and differentially radiolabeled with either
[35S]methionine
or[35S]cysteine,
immunoprecipitated
withour different
antisera,
and thenanalyzed by
SDS-PAGE (Fig. 2). The MV P protein contains 11 methionine and 6cysteine residues, whereas the V
protein
ispredicted
tocontain5methionine and 11
cysteine
residues.Itcanbeseen in thecontrol lanes, where the PC20antiserum wasused,
on November 10, 2019 by guest
http://jvi.asm.org/
[image:3.612.73.290.83.272.2]PV1 V2
UN 12 14 16 18 20 22 UN 12 14 16 18 20 22
200000
InVitro InVivo
NORNA P V
PV1V2 PV1V2 PV1 V2 PV1V2 200 000
97400
97400
68000
43000
*_ 68000
43000
4,. 29000
FIG. 3. P and V protein synthesis late in infection. Vero cells were labeled with [35S]methionine at the times noted (in hours) following infection with MVandimmunoprecipitated with peptide antiserum PV1 or V2. Samples were analyzed on a 10% SDS-polyacrylamide gel. Positions of molecular weight markers (not shown)areindicated on theright. UninfectedVerocells(UN)were used as acontrol.
that the efficiency ofcysteine labeling was notas great as
thatof methioninelabeling. WhileabandrepresentingtheP
protein labeled with methionine was clearly visible, no
equivalentbandrepresentingPproteinlabeled withcysteine
could beclearly identified.Thiscould be duetoanumber of
factors, includingdifferentialstabilityof the radiochemicals
or their differential transport into cells, as well asdifferent
cellularamino acid pool sizes. The Pproteinwas alsomore
readily labeled with methionine than with cysteine when
immunoprecipitated with the PV1, PV2, or PV3 antiserum
(Fig. 2). Radiolabeling ofthe 40-kDapolypeptide with
cys-teine, onthe otherhand,wasrelativelyeasiertoaccomplish (Fig. 2, lanes immunoprecipitated with PV1, PV3, and V2
antisera). The data therefore support the identification of these twopolypeptides asthe MV P andVproteins.
Time course of V protein synthesisduringMVinfection. To determine the time course ofV protein synthesis in vivo,
cells were labeled with
[35S]methionine
at different timesfollowingMVinfection. Labeledproteins were immunopre-cipitated with PV1 or V2 antiserum and analyzed by SDS-PAGE. Neither the P nor the V protein was detectable before 8 to 10 h postinfection (data not shown). Figure 3
shows levels of synthesis ofthe P and V proteins at later times after infection. Maximal synthesis occurred at 16 h afterinfection,atwhichtimeapproximately80to90% of the cells exhibited cell fusion. The time of maximal in vivo synthesis of MV structural proteins for the virus stock used had also previously been determined to be 14 to 18 h after
infection (data not shown).
Analysis of polypeptides synthesized in vitro. To confirm the specificity of the peptide antisera for the P and V
proteins, theantisera weretested for their ability to
immu-noprecipitate in vitro-synthesized P and V. Two plasmids, pAM18MV-P and pAM19MV-V, containing, respectively, the P and Vcoding regions,wereconstructed so as tobe able
todirectthe invitrotranscriptionof RNAs specific to the P
and V
proteins.
Rabbitreticulocytelysate
wassubsequently29000
FIG. 4. Comparison ofPand Vproteins synthesized in vivoand
in vitro. [35S]methionine-labeled polypeptides from Vero cells 16 h after infection with MV and [35S]methionine-labeled polypeptides resulting from in vitro translation in rabbit reticulocyte lysate of RNA transcribed in vitro from linearized DNA from plasmids pAM18MV-P (P) and pAM19MV-V (V) wereimmunoprecipitated withpeptide antiserum PV1 orV2 and analyzed on a 10% SDS-polyacrylamide gel. Positions of molecular weight markers (not shown) are indicated ontheright.
used for in vitro translation reactions with the in
vitro-transcribedRNAs.
Polypeptides of 70 and 40 kDawere immunoprecipitated from in vitro translation reactions by peptide antisera and
were not presentin control invitro translationreactions to
which RNAwas not added (Fig. 4). In vitro-synthesized P protein was consistently found tobe less abundant than V protein. This is probablythe result of the inherently lower
efficiencyof thereticulocyte lysate systemin thetranslation
of higher-molecular-weight species but may also reflect differential in vitro stability of the proteins. Proteins were also immunoprecipitated from MV-infected cell lysates to compare the migration of in vivo- and invitro-synthesized
proteins.Polypeptides synthesizedinvitrocorrespondingto
both Pand Vmigrated slightlyfaster thanthecorresponding proteins synthesized in vivo. This probably indicates that these proteins exhibit different degrees of posttranslational modificationin vivo and in vitro.
Proteins synthesized in vitro were differentially labeled with either [35S]methionine or [35S]cysteine of equivalent total andspecificactivityandanalyzedby SDS-PAGE (Fig. 5). Under these conditions, the intensity of labeling ofthe 70-kDapolypeptidewas atleastasgreatwithmethionineas
withcysteine. The intensity of labeling of the 40-kDa
poly-peptide was, however, significantly greater with- cysteine thanwith methionine. This pattern of labelingis consistent with the different expected ratios of methionine to cysteine in the P and V proteins. Since stability, cell transport, cellularamino acidpool size, etc., are not factorsin vitro,
these results stronglyconfirm theidentityof theseproteins.
Analysis of 32P-labeled P and V proteins. In order to
on November 10, 2019 by guest
http://jvi.asm.org/
[image:4.612.60.299.78.276.2] [image:4.612.367.507.80.314.2]canclearlybe seen that the V protein isalsophosphorylated
in vivo.
Subcellular localization of the MV P and V proteins. In ordertodetermine the exact cellular localization of the P and V proteins in MV-infected cells, indirect
immunofluores-cenceanalysiswasperformed.Uninfected cells exhibitedno
fluorescence with any of theantisera (Fig. 7). Similarly, no
fluorescence was observed in uninfected or MV-infected cells that were not treated with primary antibody (data not
shown).The PV2antiserum(specificonlytoPproteinwhen used for immunoprecipitation) resulted in specific fluores-cencein small discrete cytoplasmic inclusions. This result is similar to that described previously (4) for a peptide antise-rum against a carboxy-terminal region of P which is not presentin V. Hence, thespecificity of immunofluorescence resulting from use of the PV2 antiserum does seem to represent P alone. The V2 antiserum(specificonly to the V protein when used for immunoprecipitation) resulted in a different pattern, a diffuse cytoplasmic fluorescence. The PV1 antiserum (specific to both the P and V proteins when used for immunoprecipitation) gave a result appearing to combine thepatterns obtained with the PV2 and V2 antisera.
FIG. 5. Analysis of invitro-synthesized P and V proteins differ-entially radiolabeled with either[35S]methionine(M)or[5S]cysteine (C)atequivalent total and specific activities. Linearized DNAfrom plasmids pAM18MV-P (P) and pAM19MV-V (V)wastranscribed in vitro, and the resulting mRNAsweretranslated in vitro inarabbit
reticulocyte lysate. Polypeptides were analyzed on a 10%
SDS-polyacrylamide gel. Positions of molecular weight markers (not shown)areindicatedonthe right.
determine whether the V proteinwas phosphorylated,
MV-infected cells were pulse-labeled with
32pi
at 16 h after infection,immunoprecipitated, and analyzed bySDS-PAGE (Fig. 6). The P protein has previously been shown to be highly phosphorylated (10), and ourresults confirmthis. ItS@£ S >
>>
> >>>
>CL 0_ .0.>
u I u I u I u I u u I u
200 000
97400
'YX'- - ~~~~~~~~68000
43 000
[image:5.612.133.225.83.307.2]29000
FIG. 6. Analysis of polypeptides from uninfected Vero cells (U)
andfrom Vero cells 16 hafterinfection with MV (I). Cellswere pulse-labeled for 1 h with32Pi.Celllysateswereimmunoprecipitated withpreimmuneserum(Pre)orpeptide antiserum PC20, PV1,PV2, PV3,Vl,orV2.Sampleswereanalyzedon a10%o SDS-polyacryl-amidegel. Positions of molecular weight markers (not shown)are indicatedontheright.
DISCUSSION
Since Thomasetal. (26) reported that the SV5 P protein
wasexpressed fromanedited mRNA transcript,anumber of groups have reported similar RNA editing in transcripts arising from the Pgenes ofanumber of other paramyxovi-ruses. The discovery of this type of RNA editing came
together with the discovery of a previously unknown
paramyxovirus-encoded protein which has cometo be des-ignated the V protein. Analogous V proteins havenowbeen
detected in cells infected withanumberof paramyxoviruses.
Pprotein-specific monoclonal antibodies were used for the
detection of V protein in SV5- and parainfluenza virus type 2-infected cells (23, 26). Antisera directed against synthetic peptides were similarly used to detect V in mumps virus-infected cells(19, 24). An edited P gene-specific mRNA has been detected in MV- and Sendai virus-infected cells, but therehas beennodirectevidence for the in vivo existence of
either a Sendai virus- or MV-specific V protein. Synthetic
peptide-specific antisera directed against the MV Vprotein
wereproduced in the present study for this purpose. With
these antisera, acysteine-rich polypeptide was detected in
MV-infectedcellsbyimmunoprecipitation and
immunofluo-rescence analysis. The V proteinwasfound tomigrate as a
40-kDapolypeptide. Rimaet al. (21) reported the detection ofan MV-specific 40-kDa protein possibly related tothe P protein that comigrated with the MV M protein and used partialproteasedigestion tosuggestthat itwasnotsimplya
breakdownproduct ofP. Wenowreportthe definite identi-fication of the MV V protein, andwe have also found itto comigrate withthe Mprotein (datanotshown). This proba-bly explains previous difficulties in recognizingand identify-ing the MV V protein. No strong signal at 40 kDa had
previously been detected in MV-infected cells after
immu-noprecipitation with antibodies directed against the exact amino terminus of the Pprotein(4).Thismaybe becausethe
carboxyl terminus ofparamyxovirus P proteins is, in
gen-eral, moreimmunogenic than the restof the molecule(28).
Thegelmigration ofP and Vproteinsimmunoprecipitated
from MV-infected cells was compared with that of those
immunoprecipitatedfrom in vitro translation reactions. The
invitro-synthesized P and Vproteinswerefoundtomigrate slightly faster than those immunoprecipitated from
MV-P v
M C M C
200000
97400
68000
43 000
a. a
29000
on November 10, 2019 by guest
http://jvi.asm.org/
[image:5.612.66.293.483.668.2]FIG. 7. Indirectimmunofluorescenceanalysis of theMV PandVproteinsinuninfected Verocells and Vero cells 16 h afterinfection with MV. Cellswerefixed in2.5% paraformaldehyde in PBS,permeabilized with0.5% Triton X-100 inPBS,and then incubatedsequentiallywith normal goat serum,peptide antiserum, and FITC-conjugated goat anti-rabbit IgG. (A)Uninfected, PV2antiserum;(B) MVinfected, PV2 antiserum; (C) uninfected, V2antiserum; (D)MVinfected,V2antiserum; (E) uninfected,PV1antiserum;(F) MVinfected,PV1antiserum. UVphotomicrographs takenat x400magnification. Photomicrograph exposure times forpanels A,C,and Ewereadjustedtobe thesame asthosefor panels B, D, and F.
infected cells (Fig. 4). This probably indicates that different posttranslational modifications occur in vivo and in vitro.
Such differences would not be surprising since the rabbit reticulocyte lysate system has been shown to contain a
variety of posttranslational processing activities whichmay leadtogel migration differences. These include proteolysis (15), phosphorylation (9), acetylation (22), and isoprenyla-tion (29). The Pprotein is knowntobe phosphorylated, and sincethe Vproteinshares theamino-terminal 231 of its 299 amino acids with the amino terminus of the P protein, the possibility ofphosphorylation of the V protein was
exam-ined. Immunoprecipitation of32P-labeledproteinsfrom MV-infected cells indicated that the Vproteinwasindeedhighly phosphorylated(Fig. 6). Differential phosphorylationinvitro
and in vivo of the Vprotein may explain differences ingel migration. This does not exclude the possibility that other
modifications may distinguishthe V protein synthesized in
vivo fromthat synthesizedinvitro.
Patersonand Lamb (19)have reportedthat anadditional
mumpsvirus-specific protein,designatedI,of
approximately
19kDa is encodedbyanadditionalspecies
ofeditedmRNAtranscribed from the virus P gene. In this transcript, four
on November 10, 2019 by guest
http://jvi.asm.org/
[image:6.612.158.461.61.546.2]nontemplated G residues are added instead of thetwowhich give rise to the P mRNA. This results in a shift into yet another reading frame distal to the G insertion point which is distinct from that of either P or V. Takeuchi etal. (24) have reported immunoprecipitation of a mumps virus-specific
protein of approximately 24 kDa with antiserum directed against a synthetic peptide corresponding to a part of the
amino-coterminal region of P and V. This protein may well be the product of an mRNA to which four nontemplated G residues are added. An analogous protein could potentially beencoded by MV if G residues were added to the insertion
point in the P gene transcript to cause a shift into the +1
reading frame relative to that of P (i.e. 2, 5, 8, etc.). Translation of such a message would terminate only five codons afterthe insertion site. Cattaneo et al. (6) reported that 2 of 19 cDNA clones specific to MV P gene-encoded mRNA species had insertions of three G's, but no clones wereobtained which could encode a putative MV I protein. However,thismayonlyreflect examinationof aninsufficient number orunrepresentative sampling of cDNA clones. We wereunabletodetect any suchputativeMV-specificprotein with our antisera forimmunoprecipitation analysis of
MV-infected cell lysates. We were, in some instances, able to detect proteins of low molecular weight, but these did not appear to be additional species of amino-coterminal yet
carboxy-dissimilar proteins since they were also
immuno-precipitated with the PC20 antiserum, which is directed
against acarboxyl-terminal region of the P protein (Fig. 2). The exact nature ofthese proteins has not yet been fully determined.
The MV Vprotein isdiffuselypresentin the cytoplasmof MV-infected cells (Fig. 7). The P protein localizes in a
distinctly differentpattern, in small brightcytoplasmic inclu-sions. Earlier studies have shown that such cytoplasmic inclusionsalsocontainviralnucleocapsids(11, 16). Since the P protein thus appears to be a marker for nucleocapsids,
most MV V proteindoes not therefore appear to be associ-atedwith nucleocapsids duringthe period of maximal virus
protein synthesis. In fact, when MV-infected cells were
analyzedwith an antiserum able torecognize both P and V, amixture of both of thesedistinctfluorescencepatterns was seen.Thisdoesnot,however, exclude the possibilitythat a very small proportion of V protein may nonetheless be
associated with nucleocapsids. A number ofinvestigators
have reported that anti-P antibodies are also able to
immu-noprecipitate the V, NP, and possibly also L proteins of some paramyxoviruses and have taken this to imply an
association ofVwithtranscriptional complexes(24, 26). We
also found that asmall amount oftheNP protein could be
immunoprecipitated when either P- orV-specific antiserum
was used. However, we have previously found that small amounts ofthe NPcould be immunoprecipitated with
anti-serum directed against the MV hemagglutinin (1). The
sig-nificance ofthis phenomenon remains unclear.
The function of the paramyxovirus V protein remains
obscure. An open reading frame for V protein is highly
conservedamongparamyxoviruses, and itthus seemslikely that it has an important biologicalrole. Since the cysteine-rich domain in theunique region ofparamyxovirus V
pro-teins resembleszincfinger motifsthat have beenidentifiedas
participatinginbinding withnucleicacidsorproteins,it has beensuggested that the Vproteinmay be aregulatory factor involved in virus RNA orproteinsynthesis(6, 26). Delinea-tion of the exact nature of the funcDelinea-tion of the Vproteinwill
involve additional research, likely requiring the ability to
produce large amounts of purified protein for functional
assays or elsethe developmentof virusmutantsdeficient in
theexpression ofVprotein.
ACKNOWLEDGMENTS
Wethank Talia Pakos for excellent technicalassistance andPeter Liston and Francoise Blain forhelpful discussion.Wethank Chris-topher Richardson for the generousgift ofpeptide antiserum PC20 andforstimulating scientific discussion.
This workwassupported byaresearchgrant from the Multiple SclerosisSociety of Canada.E.A.W. was supportedbyfellowships
from theRoyal VictoriaHospital Research Institute and theRoyal Victoria Hospital Department of Medicine. D.J.B. is a Chercheur-Boursier-Clinicien ofleFonds de laRechercheenSanteduQuebec.
REFERENCES
1. Alkhatib, G., and D. J. Briedis. 1986. The predicted primary structureof the measles virushemagglutinin. Virology 150:479-490.
2. Alkhatib, G.,B.Massie,and D.J.Briedis. 1988. Expressionof bicistronic measles virus P/C mRNA by usinghybrid adenovi-ruses: levelsof Cprotein synthesized in vivoareunaffectedby the presenceor absenceof the upstream P initiatorcodon. J. Virol. 62:4059-4069.
3. Baczko, K., M. Billeter, and V. ter Meulen. 1983. Purification and molecular weight determination of measles virusgenomic RNA. J.Gen. Virol. 64:1409-1413.
4.
Beilini,
W. J., G. Englund, S. Rozenblatt, H. Arnheiter, and C. D. Richardson. 1985. Measles virus P gene codes fortwo proteins.J. Virol. 53:908-919.5. Billeter, M.A., K. Baczko, A. Schmidt, and V.terMeulen.1984. CloningofDNAcorrespondingtofour different measlesvirus genomicregions. Virology 132:147-159.
6. Cattaneo, R., K. Kaelin, K. Baczko, and M. A. Billeter. 1989. Measles virus editingprovides anadditionalcysteine-rich pro-tein. Cell 56:759-764.
7. Enders, J. F. 1962. Measles virus: historicalreview, isolation and behaviorin various systems. Am. J. Dis. Child. 103:282-287.
8. Enders, J. F.,S. L.Katz,and A.Holloway. 1962.Development
of attenuated measles virus vaccines. A summary of recent investigation. Am.J.Dis. Child. 103:335-340.
9. Gibbs, P. E. M., D. C. Zouzias, and I. M. Freedberg. 1985. Differentialposttranslational modification of human typeI ker-atins synthesized in a rabbit reticulocyte cell-free system. Biochim. Biophys. Acta824:247-255.
10. Graves, M. C. 1981. Measles virus polypeptides in infected cells: studies by immune precipitation and one-dimensional
peptide mapping.J. Virol.38:224-230.
11. Hogan, R. N., F. Rickaert, W. J. Bellini, C. Richardson, M. Dubois-Dalcq,and D. E. McFarlin. 1983. Early appearance and co-localization of individualmeaslevirusproteins using double-labelfluorescentantibody techniques, p. 421-426. In R. Com-pans and D. Bishop (ed.),The molecularbiologyof
negative-strandviruses. AcademicPress, Inc.,NewYork.
12. Kondo, K.,H. Bando, M.Tsurudome,M. Kawano, M.Nishio, and Y. Ito. 1990. Sequence analysisof the phosphoprotein (P)
genes ofhumanparainfluenzatype4Aand4Bviruses and RNA editing at transcript of the Pgenes: thenumberof G residues added isimprecise. Virology 178:321-326.
13. Laemmli,U. K.1970.Cleavageof structuralproteinsduringthe assembly of the head ofbacteriophage T4. Nature (London)
227:680-685.
14. Lund,G.A.,D. L.J. Tyrell,R. D.Bradley,andD.G. Scrabba. 1984. Themolecularlengthof the measles virus RNA and the structuralorganizationof measlesnucleocapsids.J.Gen. Virol. 65:1535-1542.
15. Mumford, R. A., C. B. Pickett, M. Zimmerman, and A. W. Strauss. 1981. Protease activities present in wheat germ and rabbitreticulocyte lysates. Biochem. Biophys. Res. Commun. 103:565-572.
16. Nakai, T.,F.L.Shand,and A. F. Howatson.1969.Development
of measles virusin vitro. Virology38:50-67.
on November 10, 2019 by guest
http://jvi.asm.org/
17. Norrby, E., and M. M. Oxman. 1990. Measlesvirus, p.
1013-1044. In B. N. Fields andD. M. Knipe (ed.), Fields virology, 2nded. Raven Press,Ltd.,NewYork.
18. Ohgimoto,S., H. Bando, M. Kawano, K. Okamoto, K. Kondo,
M. Tsurudome,M. Nishio, andY.Ito. 1990.Sequenceanalysis of Pgeneof humanpara-influenzatype2virus: P and cysteine-richproteins are translatedby two mRNAs thatdiffer by two
nontemplated G residues.Virology177:116-123.
19. Paterson, R. G., and R. A. Lamb. 1990. RNA editing by G nucleotide insertioninmumpsvirusPgenemRNAtranscripts. J. Virol.64:4137-4145.
20. Richardson, C. D., A. Berkovitch, S. Rozenblatt, and W. J. Bellini.1985. Use ofantibodies directed againstsynthetic pep-tidesforidentifyingcDNAclones,establishing readingframes, anddeducingthegeneorderofmeasles virus. J. Virol. 54:186-193.
21. Rima, B. K., S. A. Lappin, M. W.Roberts, andS. J. Martin. 1981. A study of phosphorylation ofthe measles membrane protein. J.Gen.Virol. 56:447-450.
22. Rubinstein, P., and J. Deuchler. 1979.Acetylatedand
nonacety-lated actins inDictyostelium discoideum. J. Biol. Chem. 254: 11142-11147.
23. Southern, J. A., B. Precious, and R. E. Randall. 1990. Two
nontemplated nucleotide additionsarerequiredtogeneratethe PmRNA ofparainfluenzavirus type2since the RNAgenome encodesproteinV.Virology 177:338-390.
24. Takeuchi,K., K.Tanabayashi,M.Hishiyama,Y. K.Yamada, A.
Yamada, and A. Sugiura. 1990. Detectionandcharacterization
ofmumpsvirus Vprotein.Virology178:247-253.
25. ter Meulen, V., J. R. Stephenson, and H. W. Kreth. 1983.
Subacute sclerosingpanencephalitis.Compr. Virol.18:105-185.
26. Thomas, S. M., R. A. Lamb, and R. G. Paterson. 1988. Two
mRNAsthatdifferbytwonontemplatednucleotides encode the aminocoterminal proteinsPand Voftheparamyxovirus SV5. Cell54:891-902.
27. Vidal, S., J. Curran, and D. Kolakofsky. 1990. Editingof the SendaivirusP/C mRNAbyG insertionsoccursduring mRNA synthesisviaavirus-encoded activity. J. Virol.64:239-246. 28. Vidal, S., J. Curran, C. Orvell, and D. Kolakofsky. 1988.
Mappingofmonoclonal antibodiestotheSendaivirus Pprotein and the location of itsphosphates. J. Virol.62:2200-2203. 29. Vorburger, K., G. T. Kitten, and E. A. Nigg.1989.Modification
ofnuclear laminproteins byamevalonic acidderivativeoccurs inreticulocytelysatesand requires thecysteine residueof the C-terminalCXXM motif. EMBO J. 8:4007-4013.