Functional complementation of nuclear targeting-defective mutants of simian virus 40 structural proteins.

Copyright© 1994, American Society for Microbiology

Functional

Complementation

of Nuclear

Targeting-Defective

Mutants of Simian

Virus

40

Structural

Proteins

NORIO ISHII, AKIRANAKANISHI, MASAYASUYAMADA,t MICHAEL H. MACALALAD,ANDHARUMI KASAMATSU*

Department ofBiology and Molecular BiologyInstitute, University ofCalifomia,

Los

Angeles,

Los

Angeles, California

90024

Received 25 May 1994/Accepted 7 September 1994

Structural proteins of simian virus 40 (SV40), Vp2 and Vp3 (Vp2/3) and Vpl, carry individual nuclear

targeting signals,

Vp3198_2o6

(Vp2316-324)

and

Vpll,,

respectively, which are encoded in different reading

frames of an overlapping region of the genome. How signals coordinate nuclear targeting during virion

morphogenesis was examined by usingSV40variants in which there is only onestructural geneforVplor

Vp2/3,nuclear targeting-defective mutantsthereof,VP2/3202TandVplAN5, ornonoverlappingSV40variants inwhichthegenesfor Vpl and Vp2/3areseparated,and mutant derivatives ofthegenecarryingeitherone or

bothmutations. Nucleartargetingwasassessedimmunocytochemicallyfollowingnuclearmicroinjection ofthe

variant DNAs. WhenVp2/3 and Vpl mutants with defects in the nuclear targeting signals were expressed

individually, the mutant proteins localized mostly to the cytoplasm. However, when mutant VP2/3202T was

coexpressed in the same cell along with wild-type Vpl, the mutant protein was effectively targeted-to the

nucleus. Likewise, theVpIAN5 mutant proteinwas transported into the nucleus when wild-type Vp2/3was

expressedin thesamecells. Theseresultssuggestthat while Vpl and Vp2/3have independent nuclear targeting signals, additional signals,suchasthosedefining protein-protein interactions,playaconcerted rolein nuclear

localization aiongwiththe nuclear targeting signals oftheindividual proteins.

During the late phase of simian virus 40(SV40) infection, cytoplasmically synthesized structural proteins,Vpl,Vp2, and Vp3, are transported into the nucleus within 60 min of their synthesis (23),where they assemble into maturevirions with SV40 minichromosomes (28). The fine structure of SV40

reveals thatthe virion is composed of 72 Vpl pentamers (1, 21), each of which is presumedtointeractwithonemolecule of

eitherVp2orVp3 (1, 21). Since theformation ofSV40 Vpl pentamersoccursinvitro(16),pentamerformationcanalsobe expectedtooccurin thecytoplasmsoonaftersynthesisduring SV40infection. The site atwhichVp2 andVp3interactwith

Vpl following synthesis is not known. They might be trans-portedindividuallytothenucleus,orthey mightinteract in the cytoplasm. On the basis ofgenetic andbiochemical data,we

have proposed that the viral structural proteins destined to form thematurevirion inthenucleus interact in thecytoplasm (19, 22).Asthe structuralproteinsareknowntocarry individ-ual nuclear targeting signals (as discussed below) which are responsible for their nuclear localization following their syn-thesis (7, 15, 17, 29, 30),this hypothesiscanbe tested experi-mentally. The phenotypic rescue of the targeting-defective Vpl, for example, is expected to occurwhen Vp2 or Vp3 is present.

Three structuralproteinsofSV40areencodedinapartially overlappingmanner on the genome. Thegenes of the minor structural proteins, Vp2 andVp3 (herein designated Vp2/3),

aretranslated in thesamereading frame,with theamino acid

sequence ofVp3corresponding to the carboxy-terminal two-thirds ofVp2.Thegenefor themajorcoatprotein, Vpl,begins

38 codons 5' totheVp2/3termination codonand is translated

*Correspondingauthor. Mailingaddress: Department ofBiology,

UniversityofCalifornia,LosAngeles,405Hilgard Ave.,LosAngeles, CA 90024. Phone:(310)825-3048. Fax:(310)206-7286.

tPresent address: Department of Animal Science, College of Agriculture, Kyoto University, Kyoto 606, Japan.

in a differentreading frame. Within the carboxy-terminal 35 residues ofVp2/3,thereexistaDNA-binding domain (6) and a Vpl-interactive determinant (15), in additionto theVp2/3

nucleartransportsignal (NTS)whichmapstoresidues 198 to

206ofVp3 (316to324 ofVp2). TheVp2/3 NTSisnecessary andsufficientto targettheproteinstothe nucleus(7, 9). The

nuclearlocalizationsignal (NLS) ofVpl hasbeen mapped by

deletion analysisofpoliovirus Vpl-SV40 Vplfusionproteins tothe amino-terminaleightresidues of Vpl(29),yetitcannot function as an independent NTS. When these residueswere conjugatedatthe carboxyterminustochickenserumalbumin,

theywereunabletopromote entryof the nonnuclearprotein

into the nucleus (5). How these two differentsignals, Vp2/3-NTS and Vpl-NLS, function in the nuclear targeting of the viralproteinsis notwell understood; hence,werefertothem

asnucleartargeting signals. Bothtargeting signals arewithin the overlapping region. As mutations in theoverlapping region

of Vp2/3 could inevitably introduce alterations in the Vpl

sequence,weconstructedanonoverlappingSV40(NO-SV40),

whoseVp2/3 genes are separatedfrom theVpl gene. In this study,wehave tested for thecomplementationof the nuclear targeting-defective Vpl (or Vp2/3) by Vp2/3 (or Vpl).

MATERIALSAND METHODS

Plasmid construction. The strategy of the construction of the plasmids used in this study is diagrammed inFig. 1. All DNAmanipulationswere performed as described previously (25).

pSV40 containing the wild-type SV40 DNA inserted into

pBR322as ashuttlevectorhas beendescribed earlier (16). pSV-Vpl,in which theVp2andVp3codingsequenceswere deleted from pSV40, was constructed by ligating a 6.9-kb

KpnI-BpullO2I fragment and a 263-bpKpnI-AvaII fragment

frompSV40 with a225-bp AvaII-BpullO2I fragment

synthe-sizedby PCR.The 225-bp fragment,which lacks most of the 8209

on November 9, 2019 by guest

http://jvi.asm.org/

8210 ISHII ET AL.

K 294

Od

small-t| AII 557

L.arge-TVpt

B2533(375) s,By2533PCR

SI/A(651)

_r%%%%

x

E

Vpl

KpnID/bal

pSV-VpIMJ5 fragment

t

~~~/

~~NO-pSV40

VpIAN5 &

K NO-pSV40- Vpl

XV

Vp2t322r

Vp1IN5

0-PSV40

& PI KpnUXbal

NO-pSV40- smallfragment

FH5ZZ A

(1)Accl-.Xbal (2) Xbal/Sacl

(linker) largefragment

K IK

Vp2I3 Vp24V213

AccI-.PBamHl

V4pl-AUJGmutatIon pSV40-AVplMet& Bp Sall(6S1) pSV40-AVp1MS (linker)_ S-p

V-32OKtO22T pS\40.V2rM2T- VP ASall&pSV40- 'Vpl pSV-Vp2/22 Vp3-202Kto202T aVmet P 1 MiUatboii VW2W

AVPI

Removal

p2O2T

v rMev MetASall fBdVp1

S regIon

SI/A(65) (BamHl)

FIG. 1. ConstructionofSV40 andNO-SV40variant plasmidDNAs. Aschematicdiagram forthe construction of

pSV-Vpl,

pSV-VplAN5,

pSV-Vp2/3,PSV-VP2/3202T,NO-pSV40,NO-pSV40-VplAN5,

NO-pSV40-Vp2/3202T,

and

NO-pSV40-Vp2/3202T-VplAN5

ispresented.All of the constructs aswellasthe intermediateplasmidsencodeSV40largeTantigen (Large-T),small tantigen (small-t),andagnoprotein (agno)and harbor all theregulatory elements(as shown forpSV40).Onlythepertinentportionof eachplasmid isshown. Thelocation for the deletion of the thirdtoseventh residues ofVplis markedbyaAN5.The numbersrepresent theSV40 nucleotidenumbers,andthose in

parentheses

arethose

ofpBR322. Abbreviations:Ori,originofreplication; A, AccI;All, Avall;B,BamHI; Bp,Bpu1102I;E,EcoRI; H, HindIII;K,

Kpnl;

Sc, SacI;

SI,

Sall; X, XbaI.

Vp2/3region(SV40 nucleotides [nt]567to

1498),

was gener-ated by using pSV40 as a template and then digested with

AvaIlandBpu1102I.Thesenseprimer,5'-GTCTlTTTATI1lC AGGTCCATGGTCTAGATGAAGATGGCCCCAACAA

AA-3', represents the SV40 sequence upstream ofnt567 and downstream ofnt 1948, and also includesanXbaIsite (under-lined)upstreamofthe startcodon ofVpl(boldfaceletters)to

facilitate subsequent manipulation. The antisenseprimer,

5'-CAAAGGAATTCTAGCCACACTGTAGCA-3', represents the SV40 sequence from nt 1792to 1766.

pSV-VplAN5containsa5-amino-acid deletion of residues 3

through7ofVpl, andwasconstructed byexchanginga288-bp

XbaI-EcoRI fragment of pSV-Vpl for a PCR-generated 273-bp

XbaI-EcoRI

fragment. The 273-bp fragment containing thedesignateddeletionwasgeneratedby usingpSV-Vpl as a

template anddigested byXbaI andEcoRI.The sense primer

was 5'-TGCTCTAGATGAAGATGGCCGGAAGTTGTCC

AGGGGCA-3' (nt 563 to 1543 with deletion ofVp2/3coding sequences), and the antisense primer was as described above.

pSV-Vp2/3, in which only the Vp2/3 structural gene is present, and

PSV-VP2/3202T,

in which lysine 202 of Vp3 is mutated to threonine, were generated from pSV40-AVpl

Met and

pSV40202T-AVplMet,

respectively. Briefly,

KpnI-Bpu

11021,

6.9- and 1.4-kb fragments of pSV40 were separated andthe 1.4-kbfragment was partially digested with HindIII to

yield a 1.2-kb KpnI-HindIII fragment. The 6.9- and 1.2-kb

fragments were ligated together with a 217-bp mutated PCR

fragment togenerate pSV40-AVplMet and

PSV40202T-AVPl

Met, respectively. In these plasmids, two ATG codons at the

Vpl start region were eliminated by substituting two threo-nines (underlinedin the sequence thatfollows)without alter-ing the amino acid sequence ofVp2/3.Thesenseprimers,5'-T

AAAAGCTTACGAAGACGGCCCCAACAAAA(A/C)GA

AAAGGAAG-3' (nt 1490to1530),and the above-mentioned antisenseprimerwereused withpSV40as atemplateforPCR, and theresultingPCR fragmentsweredigestedwith HindlIl-Bpu1102Itoobtain the217-bp fragment.Inthesense primer,

the nucleotide sequence at the Vp3 lysine 202 (AAG) was

degenerate, leading either to the original lysine or to a

threonine substitution.Inordertofacilitate theuseof theAccI

site within the SV40 sequence, the AccIISalI site within

pBR322 (nt 651) of both constructs was destroyed to yield pSV40-AVplMetASalI andpSV40-Vp2/3202T-AVp1MetASalI, respectively, by digestionwithSall,followedby circularization after the repair reaction of the cleaved ends. BamHI linker

(5'-pCGGATCCG-3'; NewEngland Biolabs) was insertedat

the single AccI site of SV40 following the repair reaction.

Finally, pSV-Vp2/3 and

PSV-VP2/32O2T

were constructed by theremoval of the0.9-kb BamHIfragment containing the Vpl

codingsequence.

NO-pSV40 and

NO-pSV40-Vp2/3202T

were constructed from pSV40-AVplMetASalI and pSV40-Vp2/3202T-AVplMet

ASalI, respectively, by inserting an XbaI linker (5'-pCTCTA

GAG-3'; NewEngland Biolabs) at the AccI site. The XbaI-Sacl fragmentfrompSV-Vpl, containing thefull-length Vpl

codingsequence, wasinserted into the XbaI andSacl sitesto

generate NO-pSV40 and

NO-pSV40-Vp2/3202T.

NO-pSV40-VplAN5 and NO-pSV40-Vp2/32o2T-Vp1AN5

J. VIROL.

PBI

on November 9, 2019 by guest

http://jvi.asm.org/

[image:2.612.80.533.75.341.2]

NUCLEAR TARGETING OF SV40 STRUCTURAL PROTEINS 8211 Vpl-NLS p

Vp Vp2t3NTS

3 7

D2TI

I I

3 to 7 Vp2t32

SV40

SV-Vp1la

SV-Vp1ANS 4

S 20

SV-Vp2/3 I.

SV-Vp2/202T

100

so

NO-SV40O

NO-SV40-Vp1AN5 Q

4o0

Uf%-VAirVnI52---- o 20

M%P-v4u-VPifF4202T

NO-SV40-Vp2t3202r

[image:3.612.94.272.69.294.2]

VpIAN5

FIG. 2. Lategenestructureof the variantSV40DNAs. The Vp2/3 (E) and Vpl(U)genes areindicated. Vpl and Vp2/3 nucleartargeting signalsaremarked (arrows)as openand solidsegments,respectively, within each of the coding sequences. The Vp2/3202T mutation is indicatedas awhitevertical line within the nuclear targeting signal, andtheVp1AN5 deletion is indicated byanopenarrowhead belowthe nuclear targeting signal. The numbers adjacent to the coding

se-quencesindicatethe lastamino acid.

were generated by exchanging the XbaI-SacI fragments with

those derived frompSV-VplAN5.

Allconstructionswereverified bydideoxynucleotide

double-stranded plasmid sequencing with Sequenase version 2.0 (United StatesBiochemical). Oligonucleotides forPCRwere

synthesized by the Preparation Laboratory of the UCLA

MolecularBiologyInstitute.

Cells,microin.jection, andplaqueassay. Culture conditions for TC7 cells, a subline of African green monkey kidney epithelialcells, andmicroinjection procedures havebeen

de-scribed previously (7). The derivatives of SV40 DNAs for

microinjection were obtained by digesting the plasmids at BamHIsitestoeliminate thepBR322 regionand

recirculariz-ing the DNAs (3.3 ,ug/ml)with T4 DNA ligase (16.7 U/ml).

After theligation reaction,the DNAswereconcentrated with a Centricon 30 microconcentrator (Amicon),

phenol-chloro-formextracted,ethanolprecipitated,andresuspendedin phos-phate-buffered saline at a concentration of 50 ng/,ul. DNAs

weremicroinjected into the nucleus ofTC7 cells whichwere

grown oncoverslipsmarked foreasylocationofinjectedcells. At this concentration, each cell is expected to receive about 100DNAmolecules. For immunofluorescencestudies, 100 to 200 cells were microinjected with each of the DNAson the

same coverslips and each experiment was repeated at least twice.The procedurefor theplaqueassayhas been described previously(32).

Indirect immunofluorescence. The rabbit and guinea pig

antisera against purified Vp3 and Vpl have been described

previously(8, 18). After the indicated incubationperiod, the microinjected cells were fixed in methanol-acetone (1:1) at

-20°C, and stained first with hamster anti-large T antigen (1:50), guinea piganti-Vpl (1:50), and rabbitanti-Vp3 (1:50)

and then with fluorescein isothiocyanate-,

tetramethylrhoda-A.SV40 B.SV-Vpl

C.SV-Vp2/3 D.NO-SV40

K

.*~

.- K

0 4 8 12 16 20 24

hpmj

IC

41 21 W.

101

101

0 4 8 12 16 20 24

hpmj

FIG. 3. Timecourse forthe appearance of large T antigen, Vpl, andVp2/3 in the variantDNA-injectedcells.TC7 cells were microin-jected with SV40DNA(A),SV-Vpl (B), SV-Vp2/3 (C), or NO-SV40 (D),incubated,and fixed at the indicated times (hpmj). Proteins were visualized by staining first with hamster anti-large T, guinea pig anti-Vpl, and rabbit anti-Vp3 antibodies and then with fluorescein isothiocyanate-goat anti-hamster IgG, tetramethylrhodamine isothio-cyanate-goatanti-guinea pig IgG, and cascadeblue-conjugated goat anti-rabbitIgG.Percentexpressionrepresents eitherthe percentage of T-antigen-positive cellsas afraction ofcellsexhibiting large T-antigen staining (0) in total injected cells or the percentage of Vpl- or Vp3-positive cells as afraction of cellsexhibiting either Vpl (U)or Vp3staining(E)amonglarge-T-antigen-positive cells.

mine isothiocyanate-, and cascade blue-conjugated goat

sec-ondaryimmunoglobulinGs (IgGs) (1:50), respectively. In the triple-antigen detection procedure, we sometimes noted that the Vp3 cascade blue stainingwas difficultto unambiguously

distinguish from the T-antigen fluorescein staining. The sub-cellular localization of theviralproteinswasascertainedbythe

indirect immunofluorescent procedure using anti-Vpl and

anti-Vp3 (see thelegendsofFig.4and5). Washedcoverslips

weremountedontoglassslides and examined byfluorescence

microscopyasdescribedpreviously (7).

RESULTS

Construction ofSV40and its variant plasmids. The struc-turesof the late genes ofSV40and the variantDNAs used in

thisstudyarediagrammedinFig.2. Allconstructsharborboth

largeT-antigenand smallt-antigengenes,

regulatory elements,

and anintact agno gene. SV40variant

plasmids

pSV-Vpl

or

pSV-Vp2/3 contain only one structural gene, Vpl or Vp2/3, respectively, andpSV-VplAN5 and

PSV-VP2/3202T

carry mu-tations in their nuclear

targeting

signals.

In

SV-VplAN5,

the

3rd through 7th amino acids

(Pro-Thr-Lys-Arg-Lys)

of Vpl

which constitute the Vpl NLS weredeleted.

pSV-Vp2/3

and

PSV-VP2/3202T,

which containsathreonine substitution at

Vp3

lysine 202 and has been shown to be critical for nuclear

localization, lack the Vpl coding sequence because ATG

initiating codons in the Vpl

reading

framewere mutated. A

nonoverlapping

SV40 variant

plasmid, NO-pSV40,

was also

made to separate

Vp2/3

and

Vpl

coding

sequences with a

short spacerregionbetween thetwo

coding

segments.Nuclear

targeting

signal

mutations in

Vpl

and/or

Vp2/3

were

intro-II m

-I1

VOL.68, 1994

on November 9, 2019 by guest

http://jvi.asm.org/

[image:3.612.329.549.77.278.2]

8212 ISHII ET AL.

Vpl

SV40

SV-Vpl

SV-Vpl AN5

SV-Vp2/3

[image:4.612.152.465.72.532.2]

SV-VP2/3202T

FIG. 4. Subcellular localization ofsingularly expressed capsid proteins. Cellsweremicroinjectedwithplasmid-derivedSV40DNA(aandb),

SV-Vpl (candd),SV-VplAN5(e andf), SV-Vp2/3(gandh),orSV-Vp2/3202T(iandj)asdescribedinMaterialsandMethods.Cellswerefixed 24hpmj, doubly stained firstwithguineapig anti-Vpl andrabbitanti-Vp3 antibodies andthen withtetramethylrhodamine isothiocyanate-goat

anti-guinea pig IgG andcascadeblue-conjugatedgoatanti-rabbitIgG.Panels a,c,e, g,and i representanti-Vplstaining,andpanelsb, d, f, h,and

jrepresentanti-Vp3stainingofthecorresponding field of cells.

duced to generate

NO-SV40-Vp2/3202T,

NO-SV40-VplAN5,

orNO-SV40-Vp2/32o2T-Vp1AN5.

Time course for the viralgene products encoded by SV40 variantDNAs.Recircularized SV40 DNA as well as its variant DNAs, SV-Vpl, SV-Vp2/3, andNO-SV40, were microinjected into cells, and the appearance of large T antigen and the

corresponding structuralproteins was assessed cytochemically

as afunction ofthe time following injection (Fig. 3). In the SV40 DNA-injected cells, large T antigen was observed in

mostcellsby2hpostmicroinjection (hpmj), and Vpl and Vp3 were observedby4 to 5 hpmj (Fig. 3A). Most proteins were

observedtoaccumulatein the nucleus (see Fig.4a and b and

7). LargeTantigenexpressed from the variant DNAs followed

atimecoursesimilartothat ofSV40,whereas the fraction of cellspositive forVpl orVp2/3wasvariable,dependingonthe

constructs(Fig.3B, C, and D). In cells injected with SV-Vplor

SV-Vp2/3, aproportion of eitherVpl- orVp2/3-positive cells reached a plateau level after 24 hpmj (Fig. 3B and C). Less than half of the T-antigen-positive cells in

SV-Vp2/3-intro-duced cells showed Vp2/3 staining (Fig. 3C). Nonetheless, when the Vpl orVp2/3 staining was detectable at any given

time point, all staining was found unambiguously in the

Vp3

J. VIROL.

on November 9, 2019 by guest

http://jvi.asm.org/

Vp1

NO-SVV40

NO-SV40-Vp1 AN5

NO-SV40-VP2/3202T

[image:5.612.157.471.70.462.2]

NO-SV40-

VP2/3202T-Vp1AN5

FIG. 5. Subcellular localization of capsid proteins expressed from NO-SV40 DNAs. Cells microinjected with NO-SV40 DNA (a and b),

NO-SV40-Vp1AN5 (c and d),NO-SV40-Vp2/3202T(e andf),or

NO-SV40-Vp2/3202T-Vp1AN5

(g and h) were processed at 24 hpmj as described inthelegendtoFig.4. Panels a, c, e, and g representanti-Vpl staining, and panels b, d, f, and h represent anti-Vp3 staining of thecorresponding

fieldof cells.

nucleus (Fig. 4c and h). No Vp2/3 or Vpl was detected in

SV-Vpl-orSV-Vp2/3-injected cells, respectively (Fig. 3B and C and 4d andg).NO-SV40 also showed efficientexpressionof

largeTantigenandthe structuralproteins,withVp2/3 expres-sionlagging behindthatof Vpl(Fig. 3D).On the basis of these

observations, all analyses were performed at 24 hpmj in the

following experiments.

Vp2/3 and Vpl nuclear targeting signals function in an

interdependentmanner.Theimmunostainingpatternsfor the

structuralproteinsinafieldof cellsareshown in

Fig.

4, 5,and

6; and the distribution and subcellular localization of the individualproteinsaresummarized inFig.7. InSV-Vpl (Fig.

4c and d)- and SV-Vp2/3 (Fig. 4g and h)-injected cells, the

corresponding structural proteins

efficiently

localized to the nucleus(Fig.7)asinSV40DNA-injectedcells(Fig.4aand

b),

demonstrating that SV40 Vpl and Vp2/3 contain their own nuclear targeting signals and are able to reach the nucleus

independently. In

SV-VplAN5-injected

cells, the truncated

Vp1AN5

proteinwasfound in thecytoplasm

(Fig.

4eand

7),

as was the

Vp2/3202T

mutant protein in

SV-Vp2/3202T-injected

cells(Fig. 4jand7).The alteredsubcellularstaining patternsin the cellsexpressingmutantproteins confirmedtheimportance

of theN-terminal amino acid stretch ofVpl (29)and thelysine

202residue ofVp3for their nuclear localization (7,

30).

To study how the independent Vp2/3 and Vpl nuclear

targeting signals encodedwithin the

overlapping

gene

region

cooperate with each other in the process of the nuclear

targetingof viralproteins,we nextconstructedan

NO-pSV40,

in which the Vp2/3 and Vpl genes are

designed

to be in tandem. NO-pSV40 DNA, following the removal ofthe

plas-mid DNA, wasjust as capable ofplaque formation as SV40

DNA(datanotshown). UsingNO-SV40 and its variantDNAs,

weexamined the function of eachsignalinthe presenceof the

corresponding counterpart structural

protein. Vpl

and

Vp3

expressedfromNO-SV40localizedtothe nucleusas

expected

(Fig.5aand b and7).While themutant

VP2/3202T

in

SV-Vp2/

3202T-injected

cellswasineffective in nucleartargeting

(Fig.

4j),

itwas effectively localized to the nucleus in the presence of full-length Vpl

(NO-SV40-Vp2/3202T)

(Fig. 5f

and

7).

Like-wise,thetruncatedVplwas

effectively

targeted

tothe nucleus

Vp3

on November 9, 2019 by guest

http://jvi.asm.org/

8214 ISHII ET AL.

Vp1

Vp3

SV-Vpl

SV-Vp2/3

SV-Vpl

AN5

SV-Vp2/3

SV-Vpl

SV-VP2/3202T

SV-Vpl

AN5

Sv-VP2/3202T

FIG. 6. SubcellularlocalizationofVpl andVp2/3 expressedfromacombinationof DNAs.CellsmicroinjectedwithSV-Vpl andSV-Vp2/3

DNAs(a and b),SV-Vp1AN5 andSV-Vp2/3(candd), SV-VplandSV-Vp2/3202T(eandf),orSV-Vp1AN5and

SV-Vp2/3202T

(gandh)were processedat24hpmjasdescribedin thelegendtoFig.4. Panels a, c, e, and g representanti-Vpl staining, andpanelsb, d,f, and h represent anti-Vp3 staining of thecorrespondingfield of cells.

inthe presence ofVp2/3 (Fig.5c and7), althoughthemutant

Vplalonecouldonlypartiallylocalize tothenucleus(Fig.4e and 7). A similar complementation was observed when a

combination of DNAs, SV-VplAN5 and SV-Vp2/3 (Fig. 6c and d and7)orSV-Vpl and

SV-Vp2/3202T

(Fig.6e and f and

7),wasintroduced into the cells, and the subcellular localiza-tion of the mutant viral proteins was indistinguishable from thatseenincells in whichthe wild-type Vpl gene andwild-type

Vp2/3 gene were coexpressed (Fig. 6a through f). Thus, the nucleartargeting defect in one structural protein was comple-mentedby the presence of another structural protein.

In the double mutant (NO-SV40-Vp2/3202T-Vp1AN5)-in-jected cells,the Vp2/3-NTS mutation maintained the SV-Vp2/

3202T phenotype (Fig. 7)

andshowed

cytoplasmic

and

perinu-clear fluorescence (Fig.

5h).

The truncated Vpl was not

observed in any compartment of the microinjected cells (Fig.

5g

and

7).

Whenthetime course for the appearance of largeT

antigen,

Vp2/3,

and Vplwasmonitored in thedouble

mutant-introducedcells, large Tantigen and mutant Vp2/3, but not

Vpl,were clearly detected up to 24 hpmj (data notshown).

When thetwoSV40variantDNAs, SV-VplAN5andSV-Vp2/

3202T'

were

coinjected,

both mutant

proteins

were observed

primarily in the cytoplasm (Fig. 6g and h and 7). Thus, the absence of Vp1AN5 in the double mutant-injected cells ap-pears to reflect a repression ofVpl synthesiswhen the two

genesarecontiguousonthe sameDNAmolecule.

DISCUSSION

Inthis study,the interrelationship of the nucleartargeting signals of SV40 structural proteins Vpl, Vp2, and Vp3 was

tested by using variant SV40s in which only one of the structural proteins was present or by using a NO-SV40 in which the genes forVpl and Vp2/3 are separated. NO-SV40

expressedlargeTantigen,Vpl, andVp2/3, producedaplaque,

and thus could complete a lytic cycle. In the variant SV40 DNAs, the nuclear targeting signal-defective mutant protein

Vp2/3202T

orVp1AN5,whenexpressed alone,remained in the

cytoplasm but when awild-type copyof the other structural

proteinwassuppliedandexpressed,either in cisorin trans, the

mutant protein then localized to the nucleus. These results demonstrate thatwhile the nucleartargeting signaldetermines J. VIROL.

on November 9, 2019 by guest

http://jvi.asm.org/

[image:6.612.147.464.70.457.2]

the subcellular localization ofeachprotein and is responsible for the protein nuclear localization following its synthesis in

the

cytoplasm,

a functional defect of thesignalinonecapsid protein can be complemented by supplying the other capsid

protein

carrying anintactsignal.The presenceof at least one functional

signal

in cis or trans is essential for the observed

complementation.

Theseresults suggestthat additional signals suchasthose

mediating

theprotein-proteininteractionsplaya

concertedrole

along

with the nucleartargeting signal during

SV40

morphogenesis.

In the study presented here, we used

anti-Vp3

which recognizes both Vp2 and Vp3 (18); thus, to

what extent the Vp3 staining represents Vp2 molecules in

which

118 residues unique to Vp2 including myristylated

glycine

(27)

at its aminoterminus cannotbe evaluated.

There are a number of ways in whichspecific proteinscan

accumulate in the nucleus. The nuclear accumulation of

pro-teins

harboring

nuclear targeting signals hasbeen well docu-mented

(11).

Proteinscontainingfunctionallyweaksignalscan

effectively

localizetothe nucleuswhenmorethanafewsignals

are present, as reported for theweakly active SV40

large-T-antigen

NLS-pyruvate kinase fusion proteins (24), defective

signals coupled

to mouse IgG

(20),

orbovine serumalbumin

(12).

Proteins containingnonucleartargeting signalhave also

been shown to localize to the nucleus ifthey interact with

import-proficient

shuttle

proteins

(for

areview,seereference

14).

For

example,

localization of the adenovirus DNA

poly-merase is facilitated

by

another viral

protein,

pTP

(33),

and

NLS-deficient

subfragments

of

hepatitis

delta antigen canbe

transported

into the nucleus

by NLS-containing

full-length

antigens

by

complexing throughleucine zipper sequences(31).

Although

thecontributionofcooperative targeting by multiple

signals

and the

piggyback

transport to

protein

nuclear

accu-mulationhas been inferred

(14),

directproof hasbeenlacking.

Our mutational

analysis

indicated that

only

one wild-type

nuclear

targeting signal

of the two is sufficient for the piggy-back

targeting

ofSV40

proteins.

Reported

evidenceindicatesthatthehost transport

machin-eriesappeartoinfluencethe nuclear

targeting

of viral

proteins

of

polyomaviruses.

Like SV40 structural

proteins,

Vpl and

Vp2/3

ofrelated murine

polyomavirus

possess

defined

target-ing

signals

that functioninmammaliancells

(3, 4).

However,in

insect

cells,

polyomavirus Vp2/3

was

ineffectively

targeted to

the nucleus but became

predominantly

nuclear inthe presence of

polyomavirus Vpl

(10, 13).

Because

Vpl

effectively

accu-mulated in the

nucleus, Vpl-Vp2/3

interaction appears to

mediate

Vp2/3

nuclear localization. Sincewhatwasobserved in host

cells

during

infection

by

polyomaviruses (10, 13, 19,

22,

23,

26)

doesnot reveal the role of individual

protein

compo-nentsfor nuclear

targeting,

these observations ofinsect cells

(10, 13)

as well as of CV-1 cells via vaccinia virus

vector-expressed

viral

proteins

(26)

have been used to argue for a dominant role of

Vpl

in nuclear

targeting

of

Vp2/3

in

poly-omavirus infection

(10,

13,

26).

In this report, we have

con-firmedourearlier observation that SV40

Vp2/3

isabletoreach the nucleus

independently

of

Vpl

(7, 17). Currently,

what

causes this difference other than the difference in hosts in which the recombinant viral

proteins

are

expressed

is not known.

The

experimental

evidence

presented

here is in agreement with the

hypothesis

that the SV40 subvirion

assembly

takes

place

in the

cytoplasm

(19, 22).

A similar

idea

has

recently

been

suggested

for

polyomavirus

on the basis of

protein-protein

interaction studies

(2)

and results obtained

by

coex-pression

of

polyomavirus Vpl

and

Vp2/3

inmouse and insect cells

(10, 13).

Whatremains tobe demonstrated is the identi-ties of

mutually

interactive

protein

determinants and their role

DNA bod=.I

SV40O .----..f. aVp1

SV-Vp1---a-Vp'

SV-VpIAN5.---. a-Vpl

SV-Vp2/.--- aVp3 SV-VP2/3M2 .--- a-Vp3

NO-SV40

aVpl-a-Vp3

NO-SV40-VpI ---

a-Vp{

a-Vp1-p

NO-SV-Vpl } P2 .{

a-Vp3

NO-SV40Vp2l't 3

{-a-Vp3--SV-Vp1M~ f.E a-vp3

sv-ps

'''~~'''''''a-Vpi!

SV-VpIAN5 w p a-Vpl

SV-Vp2M f a-vp3E

SV-Vpl0 --- CEaP3p SV-Vp2f3m0T --- CE-Vp3 SV-Vpl^N5E a-,Vpl

SV-Vp2O3mT2 ... a-Vp3

o 20 40 60 80 100

%PosiiveCells

FIG. 7. Subcellular distribution of Vpl and Vp2/3. Cells were injected with either individual variantDNAs or acombination oftwo variantDNAs asshownatthe left. Thenumber ofcellsexhibiting Vpl or Vp2/3 (100%) was classified as N (-) or C (O), which showed distinctnuclear orcytoplasmicstaining, respectively, and N>C ( )or C>N(u),inwhichstainingwasobservedinboth compartments with thestaining predominantly in the nucleusorinthecytoplasm, respec-tively. Cellswereprocessedasdescribed in thelegendstoFig.4to 6.

in the subvirion

assembly

andprotein nuclear targeting

pro-cesses. Withinthecarboxyl35 residues ofSV40Vp2/3, three

signals, a Vp2/3 NTS, a Vpl-interactive determinant, and a

DNA-binding

domain, are present (6, 7, 15, 17, 30). In

polyomavirus Vp2/3, a Vp2/3 NTS isat thecarboxyl end (4)

anda

Vpl-interactive

determinanthasbeen

recently mapped

withinresidues140to181 of theVp3(2). Polyomavirus Vp2/3

is shorter than SV40 Vp2/3 by 28 residues inwhich both a

Vpl-interactive

determinant and a

DNA-binding

domain of

SV40

Vp2/3

arepresent.Thelocationfor the

Vpl-interactive

determinant of

polyomavirus Vp3

raisesa

possibility

thatSV40

Vp3

has an additional domain that

helps

function in the

protein-protein interaction withSV40Vpl.

Alternatively,

the interactivedeterminantpreviouslyidentified

by

deletion

anal-ysis

(15)

couldhave arisen fromconformational

change

ofthe

proteinduetothe truncationof the

carboxyl

13 residues. The latter

possibility

hasbeen

suggested by

Barouch and Harrison

(2)

onthebasis oftheaminoacid

homology

found in residues

between170 and 181 of

polyomavirus Vp3

and 174 and 185 of

SV40

Vp3.

This,

however,

is ruled out

by

results from our

laboratory

indicating

that the

Vpl-interactive

determinant and the

DNA-binding

domain of

Vp2/3,

both of which occupy the last 13 residues of

Vp2/3,

are

functionally

separable

from each other. Mutantsthat

greatly impair

DNA-binding

activity

can

still interact with Vpl

(8a). Although

the

Vp2/3-interactive

determinant ofVpl hasnotbeen

identified,

itexcludes resi-dues 3 to 7 of Vpl, whose deletion resulted in the loss of nuclear

targeting,

butwas

complemented by

Vp2/3

presumably

through

the

protein-protein

interaction. We can

learn

how additional

signals

defining

the

protein-protein

interaction

con-77.

I I

I

FM

.m

MMM7

on November 9, 2019 by guest

http://jvi.asm.org/

[image:7.612.323.558.74.320.2]

8216 ISHII ET AL.

tribute to the nuclear

targeting

ofvirion

proteins

when the

Vp2/3-interactive

determinantof

Vpl

is identified.

An

intriguing

observationwasthat when theNO-SV40-Vp2/

32o2T-VplAN5

DNAwas

injected,

the truncated

Vpl

wasnot detected in any compartment of the cells

(Fig.

5g

and 7).

Cytoplasmic Vp1AN5

and

Vp2/3202T

were detected in the samecells when the two genes on separate DNA molecules were

coinjected (Fig. 6g). Thus,

the observed

Vpl repression

occurred when the two mutant genes were

arranged

in cis. Since NO-SV40 is

viable,

we assume that the amount ofall

structural

proteins

expressed

in the NO-SV40 DNA-intro-duced cellis

comparable

to that in the normal SV40

DNA-introduced cell. Whether the reduced level of mutant Vpl

observed in the double

mutant-injected

cellsreflectsan exper-imental artifact or a

regulatory

role of

Vp2/3

in

Vpl

produc-tion remains tobe tested.

ACKNOWLEDGMENTS

Weare

grateful

toJohnN.Bradyforsupplyinghamsteranti-SV40

antiserum andtoEric Y.Chen and AndresL.Medinaforassistancein some ofthe stepsofplasmid construction.We also thank ArnoldJ. Berkand DavidA. Deanfor criticalreadingof themanuscript.

This workwassupportedbyPublic HealthService grantCA50574 and inpartbyfundsprovided bythe Committee onResearch of the AcademicSenate of theUniversityofCalifornia,LosAngeles.

REFERENCES

1. Baker,T.S.,J.Drak, and M.Bina. 1988. Reconstructionof the three-dimensionalstructureofsimian virus40and visualizationof the chromatincore.Proc.Natl. Acad. Sci. USA85:422-426. 2. Barouch,D.H.,andS. C.Harrison.1994.Interactionsamongthe

majorand minorcoatproteinsofpolyomavirus.J.Virol. 68:3982-3989.

3. Chang,D.,J.I.Haynes,J.N.Brady,and R. A.Consigli.1992.The

useofadditive and subtractiveapproachestoexaminethe nuclear localization sequenceof the polyomavirusmajor capsid protein

VP1.Virology189:821-827.

4.

Chang,

D.,

J. I.Haynes, J.N. Brady,and R. A. Consigli. 1992. Identificationofanuclearlocalizationsequenceinthe

polyoma-viruscapsid proteinVP2.Virology191:978-983.

5.

Chelsky,

D.,R.Ralph,andG.Jonak.1989.Sequence requirements

forsynthetic peptide-mediatedtranslocationtothe nucleus. Mol. Cell. Biol.9:2487-2492.

6. Clever,J.,D. A.Dean,andH.Kasamatsu. 1993.Identificationof aDNAbindingdomain in simian virus 40capsidproteins Vp2and

Vp3.J.Biol.Chem.268:20877-20883.

7. Clever,J.,and H.Kasamatsu. 1991.Simian virus40 Vp2/3small structural proteins harbor their own nuclear transport signal.

Virology181:78-90.

8. Clever,J.,M.Yamada,andH. Kasamatsu.1991.Importof simian virus 40virionsthroughnuclear porecomplexes.Proc.Natl. Acad. Sci.USA88:7333-7337.

8a.Dean,D.A.,etal.Unpublisheddata.

9. Dean, D. A., and H. Kasamatsu. 1994. Signal- and

energy-dependent nuclear transport of SV40 Vp3 by isolated nuclei: establishmentofa filtrationassay for nuclearprotein import. J. Biol. Chem.269:4910-4916.

10. Delos,S.E.,L.Montross,R. B.Moreland,and R. L.Garcea. 1993.

Expressionofthe polyomavirus VP2and VP3proteinsin insect cells: coexpression with the major capsid protein VP1 alters VP2/VP3 subcellular localization.Virology194:393-398. 11. Dingwall, C., and R. A. Laskey. 1992. The nuclear membrane.

Science 258:942-947.

12.

Dworetzky,

S. I.,R. E. Lanford,and C. M. Feldherr. 1988. The

effects of variations in the number and sequence of targeting signalsonnuclear uptake.J. Cell Biol. 107:1279-1287.

13. Forstova, J., N. Krauzewicz, S. Wallace, A. J. Street, S. M. Dilworth, S. Beard, and B. E. Griffin. 1993. Cooperation of structuralproteins duringlate events in the lifecycleof polyoma-virus. J. Virol. 67:1405-1413.

14. Garcia-Bustos, J., J. Heitman, and M. N. Hall. 1991. Nuclear protein localization. Biochim. Biophys.Acta 1071:83-101. 15. Gharakhanian, E., and H. Kasamatsu. 1990. Two independent

signals,anuclear localizationsignal andaVpl-interaction signal, reside within thecarboxy-35 amino acids of SV40 Vp3. Virology 178:62-71.

16. Gharakhanian, E., J. Takahashi, J. Clever, and H. Kasamatsu. 1988. In vitro assay for protein-protein interaction: carboxyl-terminal 40 residues of simian virus 40 structural protein Vp3 containadeterminant for interactionwithVpl.Proc. Natl. Acad. Sci.USA85:6607-6611.

17. Gharakhanian,E., J. Takahashi, and H. Kasamatsu. 1987. The carboxyl 35 amino acids of SV40 Vp3 areessential for its nuclear accumulation.Virology 157:440-448.

18. Kasamatsu, H.,andA.Nehorayan. 1979. Intracellularlocalization of viral polypeptides during simian virus 40 infection. J. Virol. 32:648-660.

19. Kasamatsu, H.,and A.Nehorayan.1979. Vplaffects intracellular localization ofVp3polypeptide during simian virus40 infection. Proc.Natl. Acad. Sci.USA 76:2808-2812.

20. Lanford,R.E., C. M. Feldherr, R. G. White, R. G. Dunham, and P. Kanda. 1990. Comparison of diverse transport signals in synthetic peptide-induced nucleartransport. Exp. Cell Res. 186: 32-38.

21. Liddington,R.C.,Y.Yan,J. Moulai, R.Sahli,T. L.Benjamin, and S. C. Harrison. 1991. Structure of simian virus 40 at 3.8-A resolution.Nature (London) 354:278-284.

22. Lin, W.,T.Hata,and H. Kasamatsu.1984.Subcellular distribution of viral structural proteins during simian virus 40 infection. J. Virol.50:363-371.

23. Lin, W.,J.L.Shurgot,and H.Kasamatsu. 1986. Thesynthesisand transportofSV40 structuralproteins. Virology154:108-120. 24. Roberts, B. L., W.D. Richardson, and A. E. Smith. 1987. The

effect ofproteincontextonnuclearlocationsignal function. Cell 50:465-475.

25. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning:alaboratory manual, 2nd ed. Cold SpringHarbor Labo-ratory,ColdSpring Harbor, N.Y.

26. Stamatos,N.M., S. Chakrabarti, B. Moss, and J. D. Hare. 1987. Expression of polyomavirus virion proteins by a vaccinia virus vector:associationof VP1andVP2 with thenuclear framework.J. Virol.61:516-525.

27. Streuli, C.H.,andE. Griffin.1987. Myristic acidis coupledto a structural proteinofpolyoma virus andSV40. Nature (London) 329:619-622.

28. Tooze, J. (ed.). 1981. DNA tumor viruses. Cold Spring Harbor Laboratory, Cold Spring Harbor,N.Y.

29. Wychowski,C.,D.Benichou,and M.Girard. 1986. Adomainof SV40capsid polypeptideVP1thatspecifies migrationintothe cell nucleus. EMBOJ.5:2569-2576.

30. Wychowski, C.,D.Benichou,and M.Girard. 1987.The intranu-clear location of simian virus 40 polypeptides VP2 and VP3 dependsonaspecificaminoacid sequence. J. Virol. 61:3862-3869. 31. Xia,Y.-P., C.-T. Yeh, J.-H. Ou, and M. M. C. Lai. 1992. Charac-terization ofnuclear targeting signal ofhepatitis delta antigen: nucleartransportas aprotein complex.J.Virol. 66:914-921. 32. Yamada, M., and H. Kasamatsu. 1993. Role of nuclear pore

complexinsimian virus 40 nucleartargeting.J.Virol.67:119-130. 33. Zhao, L.-J., and R. Padmanabhan. 1988. Nuclear transport of adenovirus DNA polymerase is facilitated by interaction with preterminal protein. Cell 55:1005-1015.

J. VIROL.

on November 9, 2019 by guest

http://jvi.asm.org/