• No results found

Cellular proteins which can specifically associate with simian virus 40 small t antigen.

N/A
N/A
Protected

Academic year: 2019

Share "Cellular proteins which can specifically associate with simian virus 40 small t antigen."

Copied!
11
0
0

Loading.... (view fulltext now)

Full text

(1)

JOURNALOFVIROLOGY,Sept. 1986, p.692-702 0022-538X/86/090692-11$02.00/0

Copyright ©1986, American Society for Microbiology

Cellular Proteins Which Can

Specifically Associate with Simian

Virus

40 Small

t

Antigen

CHERYL ISAACMURPHY, ILAN BIKEL, ANDDAVID M. LIVINGSTON*

Dana-FarberCancer Institute and Departments of Medicine, Brigham and Women's Hospital and Harvard Medical

School,Boston, Massachusetts 02115

Received 18February1986/Accepted 2 June 1986

When crude, radiolabeled extracts ofvarious cellswere appliedto homogeneous simian virus 40 smallt

antigen-Sepharose adsorbents, three cellproteins (57, 32,and20 kilodaltons[kDa]) bound specifically. Each

also bound toaninsoluble, truncatedtderivative composed oftheCOOH-terminal 123 residues of theprotein.

The binding of theseproteins wasgreatly inhibited afterreduction and alkylationof thetligand. Therefore,

someelement of native conformation, but notall of the primarystructureof t, isnecessaryforthis binding

property, which may constitute a discrete, in vitro biochemical function of this protein. Results of cell

fractionation experiments suggestedthatthe 57- and32-kDa proteinsarenonnuclear cellconstituents,whereas

the 20-kDa protein wascloselyassociatedwithadetergent-washed nuclear fraction. Specific immunoblotting

andcomparative partial proteolyticdigestion analyses indicated that the 57-kDaprotein is tubulin,a major

componentofthecytoskeleton.Inthisregard,tand tubulinwereobservedtocoimmunoprecipitate from crude cell extracts after incubation with monospecific anti-t antibody. Therefore, it is possiblethat t and tubulin

interactinvivo.

Simian virus 40 (SV40) can either lytically infect or

neoplastically transform its host, dependingonthe celltype

involved. In monkey cells, the virus replicates efficiently,

killingthecell in theprocess.Bycontrast, in cells of several

other species, little or noautonomous viral DNAsynthesis

occurs, and a small fraction ofan infected population may

become permanently transformed. In both infected and

transformed cells, two early viral proteins (the largeTand small t antigens) are synthesized. Intact t is not absolutely

required for productive growth of the virus in established

monkey cell lines, althoughitmayfacilitate thisprocess(31,

37).Both Tandtplay important, complementary rolesin the viraltransforming mechanism when cellstobe infected are

cultivated inaspecificmanner(3, 10, 12, 20, 24, 25, 29, 33).

Although anumber ofbiochemical and functional

proper-ties of T have been elucidated, much less is known ofthe

analogousfeatures oft. Itappears tobenonphosphorylated

andcontains twoclustersofcysteinylresidues initsunique

region. At least one of these sequence units also occurs in

the a and 1B subunits ofa group of glycopeptide trophic

hormones, thereby raising the possibility that t and these

proteins share one or more functional properties (11). A

largefractionoftispresentin thecytoplasmof infected cells

(23, 35), but recent immunofluorescence studies revealed

that t can also be readily identified in the nucleus oftwo

different celltypes(9, 21, 40).Whether it functions inone or

the other of these compartments or both is unknown.

Fur-thermore, nothing is known of how t operates in vivo,

although its presence correlates with the subsequent

disso-lution of actin cable structures (2, 14, 15, 25), the new

appearanceofcentriolarstructures(30),amitogeniceffect in

resting, T-containing cells (4, 15), and resistance to the

cellular DNAsynthesisinhibitioneffect oftheophylline (27).

*Correspondingauthor.

Rundellandco-workers(26, 28, 39) showed that

immuno-precipitates ofextractsofSV40-infected monkey cells,

gen-erated with selectedserafromSV40 tumor-bearing animals,

containtwocellproteins (56,000 molecular weight [56K]and

32K), whose presence depends on the simultaneous

pres-enceoft.These resultsraise thepossibility thatthe antien

can form stable and specific complexes with at least two cell-encodedproteins.

Inanotherapproachtosearchingforpotential interactions

oftwith cellularpolypeptidesand for therelevance ofsuch

events to tfunction,weconstructedsolid-phase t-Sepharoe

adsorbentsto search forspecific t-binding proteins incrude

cell extracts. These adsorbents were constructed with t

isolated fromanoverproducingclone ofEscherichia coli(2).

This report contains a description of the use of such

adsorbents to capture t-binding proteins from crude cell

extracts.Italso describes theisolation of three suchproteins andthe identification ofoneof themastubulin and raises the

possibility thattand tubulin caninteract in vivo.

MATERIALS ANDMETHODS

Cells and viruses. Stocks of all cellswere grown onplastic

surfaces(BectonDickinsonLabware)in Dulbeccomodified

Eagle (minimal essential) medium containing 10% calf or

fetal calf serum (Colorado Serum Co.) in a 7%

C02-containing atmosphere. Cos-1 cellswereoriginally obtained

fromY. Gluzman.

The titers of SV402 were determined on Cos-1 cell

monolayers by a VP-1 induction assay (25), and the virus

wasgrownin 100-mmplatesof these cellsatamultiplicityof infection of =1 to 5. dl883 (a T+/t- viral mutant [31]) was growninBSC-1 cellsatamultiplicityof infection of 0.01.All infectionswere performedat 37°C.

Radiolabelingof cells. Cultureswere washed with

methio-nine-free Dulbecco modified Eagle medium and labeled for either4 horovernightwith 0.15 mCi of[35S]methionine(600 692

Vol.59,No.3

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

Ci/mmol) in1.5mlof methionine-free medium

containing

2% serum.

Preparation of cell extracts. Radiolabeled cells were

washed twice at room temperature in phosphate-buffered saline (pH 7.1) and then extracted at0°C for30min with 1.5 mlof buffer(0.01 M Trishydrochloride [pH 8],0.15 MNaCl, 1 mMCaCl2,1 mMMgCl2, 0.5% NonidetP-40[NP-40])per

100-mm plate. The extracts were centrifuged at 10,000rpm

in a Sorvall RC5B centrifuge for 20 min to remove cell debris.

35S-labeled CV-1Pcellextracts, enriched intubulin, were

prepared by first removing soluble

proteins

in a buffer

(PM2G)

which does not disrupt microtubules and then

adding calcium to the buffer to liberate tubulin

specifically

(22, 34).

Purification of tantigen.The

purification

oftfromE. coli

transformed by the high-level-expression

plasmid

pTR865

was described previously (2). The samples ofhomogeneous

t and 14-kilodalton (kDa) tused in this studywere obtained by preparative sodium dodecyl sulfate

(SDS)-polyacryl-amidegelelectrophoresis ofextractA(2).The

proteins

were

eluted from crushed,

unfixed,

and unstained

gel

strips

into

0.2 Mammonium bicarbonate(pH

8.0)-0.4%

SDS for 16 hat

37°C. The eluted

proteins

were

lyophilized

to volumesless

than 1 ml anddialyzedat4°C for48 h. For thefirst24

h, they

weredialyzedagainst1liter of

coupling

buffer(see

below)

or

buffer A (see below), as

indicated,

which contained 1% cholicacid. For thenext24

h, they

were

dialyzed

against

the same buffer containing 0.1% cholic acid.

Dialysis

was then continued for three more

days,

with

frequent changes

of buffer containing 0.1% cholic acid. The

proteins

were then dialyzedagainst

multiple

changes of

detergent-free

buffer for

a final 24-hperiod.

Coupling of proteins to CNBr-activated Sepharose 4B.

Cyanogen bromide-activated

Sepharose

4B was

purchased

fromPharmacia. Thedrymaterial(30 mg)was

suspended

in

1 mMHCI and washedtwice with6 mlof1 mMHClat room

temperature andfiltered. Itwasthenwashed twicein 1 mlof coupling buffer (0.1 M

NaHCO3 [pH 8.3]

containing

0.5 M NaCl).Each batchof

Sepharose

beadswasthen

centrifuged,

thecoupling bufferwas

removed,

and 1 mlof

coupling

buffer containing

'200

,ug of

purified protein(s)

was added. The Sepharose beads wererocked at

4°C

overnight

orfor 2 h at

room temperature. They were then

centrifuged

and

sus-pendedin 0.2Mglycine (pH 8.0).

Rocking

wascontinued for anadditional 2hatroomtemperature. The beads werethen centrifuged and

suspended

in

coupling

buffer. To remove

excessuncoupled ligand, each adsorbentwas washed

alter-natelywith 0.1 M sodium acetate (pH 4.0)-0.5 M NaCland

coupling

bufferthreetimes.

Each portion ofadsorbent to be used in a

given

experi-ment contained 100 ,ul of

packed Sepharose

beads in an

Eppendorf Microfuge

tube. The adsorbent was

extensively

washedbeforeuse withbuffer A(0.01MTris

hydrochloride

[pH 8.4], 0.1 M

NaCl,

1mM

EDTA,

1 mM

2-mercaptoeth-anol, 10%

[vol/vol]

glycerol).

Typically,

a 200-t1l

sample

containing

[35S]methionine-labeled

cell extract was then

added, and thesuspensionwas rockedat roomtemperature for1hor at

4°C

overnight. Theadsorbentwas thenwashed extensively with buffer A containing 1% NP-40. Bound proteinswere eluted in twostages. First, 500 ,ulof buffer A

containing

1 MNaCl

(buffer

B)was

applied,

followed

by

500

,ud

of the same buffer containing 2 M NaCl and S M urea

(buffer C). Each elution step was

performed

at room

tem-perature for 30 min with gentle rocking, after which the

beads were pelleted and the eluate was removed. In some

cases, adsorbents werealso elutedwithbufferA

containing

100 ,ug oft, bovine serum albumin

(BSA),

or reduced and

alkylated

t. In

addition,

some batches of

protein-loaded

adsorbent wereboiled for3 minin SDS

polyacrylamide gel

sample

buffer

(19)

containing

1% SDS. The eluted

proteins

were

precipitated

with 10% trichloroacetic

acid,

and the

precipitate

was

pelleted,

rinsed with90%

ethanol,

dried,

and

dissolved in

sample

buffer before SDS

gel

electrophoresis.

Reductionand

alkylation

oft

antigen.

Samples

of

purified

t

antigen

were

adjusted

to 1%SDS and 0.02 M dithiothreitol

and were then incubated at room temperature for 15 min.

N-ethylmaleimide (100 mM)

was

added,

and theincubation

wascontinuedat

0°C

for1h.

2-Mercaptoethanol

(0.5 M)

was

addedto stopthe

reaction,

and the

protein

was

precipitated

with acetone.

Mammalian cell fractionation. The

technique

used for

mammalian cell fractionation was derived froma

published

procedure (1).

Cells were labeled with

[35S]methionine,

as

described above. After removal ofthe

medium, they

were

washed three times with 10 mM

NaPO4 (pH 7.4)-0.14

M

NaCl and

scraped

with a rubber

policeman

into 1 ml of 10 mM

morpholine

ethanesulfonic acid

(MES; pH 6.2)-0.14

M

NaCl.

The

suspension

was

centrifuged

at 800 rpm in a

Beckman J6B

centrifuge

for 5 min, and the

pellet

was

suspended

in 1.5 ml of

lysis

buffer

(10

mMMES

[pH 6.2],

10

mM

NaCl,

1.5 mM

MgCl2)

and swollen on ice for 30 min.

After25 strokesofDounce

homogenization,

the

suspension

was

centrifuged

at

1,600

rpmfor 10 minat

4°C.

The

super-natant will be referred to as extract A. The

pellet

was

suspended

in 1 ml ofNWB

(10

mM MES

[pH 6.2],

10mM

NaCl,

3 mM

MgCl2,

0.5%

NP-40,

0.1% sodium

deoxycho-late)

and

centrifuged

at

1,600

rpmfor 10 min at

4°C.

The

supernatant will be referredto asextract B. The

pellet

was

suspended

in 200

ulI

of

NWB,

and the

suspension

was

layered

over 1 ml of NWB

containing

1.8 M sucrose.

Centrifugation

was for 10 min at

2,500

rpm. The

pellet

(containing

purified nuclei)

waswashed in NWB and

again

centrifuged

at

1,600

rpm for 10 min. The

supernatant

was

removed,

and the

pellet

was

suspended

in 0.5 ml of buffer

containing

0.01 MTris

hydrochloride

(pH 8.0),

0.15 M

NaCl,

1mM

CaCl2,

1mM

MgCl2,

and1%NP-40.Extractionwasat

0°C

for30min, and the mixturewas

centrifuged

at

1,600

rpm

for10 min. The supernatant willbe referredto asextractC.

Staphylococcus

V8protease

mapping.

Boiled eluatesfrom

t-antigen

columns

containing

the

[35S]methionine-labeled

57K CV-1P

protein

and an

[35S]tubulin-enriched

CV-1P

extract were

electrophoresed

in

parallel through

a 7.5%

SDS-polyacrylamide gel.

After

being

fixed with

ethanol,

washed with water, and

dried,

the

gel

was

autoradiographed.

The radioactive tubulin and

comigrating

57K bands were

excised. Partial V8 protease

mapping

of these two

polypeptides

was

performed by

the methodof Cleveland et

al.

(5).

Immunoprecipitation.

Soluble

protein

fractions

(as

pre-pared above)

were incubated at room temperature for 1 h

with one of the

following

antibodies: rabbit

anti-gel-band-purified-t (2

RI/200 RI

of

extract);

the mouse monoclonal anti-SV40 T/t

common-region

antibody

PAb419

(200

RI

of

tissue culture fluid per 200

RI

of

extract);

or normal rabbit

serum

(2

,ul/200

,ul

of

extract).

Immune

complexes

were

isolated after the addition of a 50%

protein A-Sepharose

mixture

(10

,ul/200

[L of

extract).

Aftera30-minincubationat

room temperature, the immune

complexes

and

Sepharose

beadswereremoved

by

centrifugation

and washedfive times

with 1-ml

portions

ofRIPAbuffer

(1%

sodium

deoxycholate,

0.15 M

NaCl,

0.05 M Tris

hydrochloride [pH

7.2],

0.1%

on November 10, 2019 by guest

http://jvi.asm.org/

(3)

694 MURPHY ET AL.

B

`3SA(;'Oi.- BSACOL tCOL

r F H

- 94K ~vF

443K 4 ; . . . | s ...^~~At

4,1S3ia.

,';i K

C

BSA COL. t+BSA COL tCOL

A C.- F- N

-A B C EF H

57 K 43K

- 32K

2.

7)Kt

..0 %. - 20K

14Kt

a-O

FIG. 1. Identification of specific t-bindingproteins. (A)SV40tantigenwaspurifiedbypreparative SDS gel electrophoresis,asdescribed

inthetext.Samples of purified, intactt(=5,ug)and 14Kt(=2 jig)wereelectrophoresed through15%SDS-polyacrylamidegels, whichwere

thenstained withCoomassie brilliant blue.Standardprotein markersarepresent onthe left.The far right lane isirrelevanttothiswork.(B) CV-1P cellswerelabeledwith[35S]methionineandextracted, andsamples (=50 jig)weremixedwiththeindicated adsorbent,asdescribed

inMaterialsand Methods. The binding reactionwasperformedfor 1 hatroomtemperature.Lanes: A, D,and G, samples(3%) oftheunbound proteins;B, E,and H, alloftheprotein eluted withbufferB;C, F,andI,allofthe protein elutedwith bufferC. Eluateswereprecipitated

with 10% trichloroacetic acidand washed with ethanol before electrophoresis in 15% SDS-polyacrylamide gels. The arrowsindicate the

migration positions ofactin (43K) and the threet-binding proteins(57K, 32K, and 20K).(C)Cos-1 cellswerelabeled with [35S]methionineand extracted,andsamples (=±50 jig)weremixed with the indicated adsorbents. Thebindingreactionwasperformedat4°C overnight. Lanes:A, D,andG, samples(3%)of the unboundproteins; B, E,andH,alloftheproteineluted with bufferB; C, F, andI,allof theproteineluted

withbuffer C. Thearrowsindicatethemigration positions oftwo(32K and20K)t-binding proteins. COL, Columns.

SDS, 1% TritonX-100). The immunecomplexeswereboiled

for 3 mininLaemmli samplebuffer(19) containing1%SDS

and analyzed bySDS-polyacrylamide gel electrophoresis.

Westernblotting.Theprocedureused for Westernblotting

was a minor modification ofthe method of Hohmann and

Faulkner(16). Proteins were analyzed afterelectrophoresis

in polyacrylamide gels and subsequent transferto

nitrocel-lulosepaper strips (Schleicher& Schuell, Inc.). The

trans-ferswereperformed overnight electrophoreticallyat 50 mA

in Tris-glycine-methanol buffer (25 mM Tris, 192 mM

glycine, 20% methanol). Aftertransfer, theimprinted

nitro-cellulose strips were washed in phosphate-buffered saline

and reacted with enzyme-linked immunosorbent assay

(ELISA)C buffer(0.05 M Trishydrochloride [pH 7.8], 10%

BSA)diluted 1:3 inphosphate-bufferedsalinefor 6 hat4°C.

Thestripswereplaced inplasticfood sealerbags,and 50 ml

of ELISA W buffer(1.5mMKH2PO4,8 MNa2HPO4,0.14 M

NaCl, 3mM KCI,0.05% Tween20)wasaddedto eachbag.

Thestripswerethen shakenovernightatroomtemperature.

After the buffer was removed from each bag, antibody

appropriatelydiluted in ELISA A buffer(0.05MTris

hydro-chloride[pH 7.8], 1%BSA, 1mM

MgC92)

wasaddedtoeach

strip andshaken for 2 hatroomtemperature. An

immuno-globulin Mmouse monoclonal antitubulin antibody (61D; a

generous giftof Frank Solomon) was used in these

experi-ments. The stripswerethen washed in ELISA W buffer and

reacted for 2 hat roomtemperature with alkaline phospha-tase-labeled anti-mouseimmunoglobulin (17).Eachstripwas

washed in ELISA W buffer for 30 min and in ELISA W

buffercontaining0.9 M NaCl foranadditional 30min. The

strips werethenwashedbrieflyin 0.1 M Tris hydrochloride

(pH 9.2). The substrate, AS-BI-fast red TR, was prepared

immediately before use as follows. Naphthol-AS-BI

phos-phate (Sigma) was dissolved in 10 mM Tris hydrochloride

(pH9.8)-2mMMgCl2at 1mg/ml. Fast red TR salt(Sigma)

wasaddedat1mg/mltothissolution, whichwasthen mixed for 5 min at room temperature, filtered, and added to the nitrocellulose strips. The enzymatic reaction was then

al-lowedtoproceedfor -10min,andthestripswerewashed in

waterand air dried.

RESULTS

Three cell proteins bind to tantigen in vitro.The SV40 t used in these experiments was purified from an

ampicillin-resistant E.coli clone(865i)which containsaplasmid-joined

SV40 t gene, governed by a strong, hybrid bacterial

pro-moter(2). As much as 15% of the soluble protein ofthese

cells is SV40 t after isopropyl-p-D-thiogalactopyranoside

induction. The antigenwas isolated fromthe washedpellet fraction of an 865i cell lysate by extraction with

urea-containingbuffer. This fractionwaselectrophoresed through

a preparative SDS-polyacrylamide gel; tand a 14K

deriva-tiveof itwereidentified andeluted,and the eluatewasfreed

ofdetergent bydialysis againstcholic acid.Thehigh degree

ofpurityof thesetwoproteinswhen isolatedin thismanner

isshown inFig. 1A.It has been shown thattpurified bythis method retains its actin cable-dissolving activity (2) and is,

therefore, notinert.

A

14Kt t

...

::

_7.

-43K

- 32K

-20K

I

J. VIROL.

lr, _ ....I

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.612.73.552.75.326.2]
(4)

BSA+tCOLL

+1

t

E F G H

m

*-57K

*43K

FIG. 2. Elution of bound protein(s) with t. Extracts of [35S]methionine-labeled CV-1P cells(=50,ug)weremixedwith BSA and BSA + t-Sepharose adsorbents, asdescribed in Materials and

Methods. After extensive washing with buffer A, elution was

attempted with an additional portionofbuffer A (lanes C and G) followedby100,ug of purifiedtin bufferA(lanes B and F). Finally, theadsorbentswere elutedwithbufferC (lanes A and E). Samples (3%)ofthe unbound proteinsareshown in lanesDand H. All of the

protein eluted inthe otherfractions noted above wasappliedtothe

gel.COL,Columns.

Purified t was coupled to cyanogen bromide-activated

Sepharose 4B (see Materials and Methods) either alone to

formt-Sepharose orincombination with BSAto form BSA

+ t-Sepharose. Among the control adsorbents used in the

subsequent experiments were BSA coupled to Sepharose

andSepharose alone. Actin- andmyoglobin-Sepharose were

also used as control adsorbents in some studies. In each

coupling experiment, approximately 200,ul oftotal protein

(50 jug) was incubated with 100 ,ul of activated Sepharose

beads,asdescribed in Materials andMethods, and .90%of

theprotein coupledtothe beads under these conditions. All

subsequentbinding/elution experiments with these

adsorb-ents wereperformed in batch.

The results obtained when the BSA, BSA + t, and t

adsorbents were individually mixed with lysates of

virus-free, [35S]methionine-labeled CV-1P cells are shown in Fig.

1B. All loaded adsorbents were washed extensively with

bufferAtoremovenonspecifically bound proteins (lanes A,

D, and G) and then eluted in series with the same buffer

containing1M NaCl (bufferB)(lanes B, E, and H), followed

bythesamebuffer containingSMureaand 2MNaCl (buffer

C) (lanes C, F, and I). Only actin(43K)waselutedfrom the

BSA adsorbent. In addition to actin, three other proteins

(57K, 32K, and 20K) were eluted from the t-containing

adsorbents. In addition, in some but not all such

experi-ments, a 50K protein appeared in the buffer B and C eluates. Its identity is not known. Moreover, relative to that of the

57K band, the intensities of the 32K and 20K bands varied

from experiment to experiment. For example, the results of

an experiment inwhich an extract of

[35S]methionine-labeled

Cos-1 cells (13) was mixed with each of these three

adsorbents are shown in Fig. 1C. Two of the three bands noted above (32K and 20K) appeared in 1 M NaCl eluates of the t-containing adsorbents (lanes E and H). Notably, the 57K protein did not elute in buffer B or C from a t-containing adsorbent, perhaps because it bound too tightly to this material. Alternatively, it may have been so modified that it

could not bind to the adsorbent. None of the threet-binding

proteins (57K, 32K, 20K) bound significantly to actin- or myoglobin-Sepharose under the same set of conditions which led to their binding to either of the two t-containing adsorbents (see Fig. 6; C. Murphy and D. Livingston, unpublished observations).

57K protein elution with t antigen or SDS. As one test of

whether the binding of the three cell proteins to t-Sepharose

was specifically to the viral protein, BSA- and BSA +

t-Sepharose were mixed in parallel with an extract of

35S-labeled CV-1P cells. The nonbinding (or flowthrough) frac-tion was analyzed (Fig. 2, lanes D and H), and the adsorbent was washed with buffer A (lanes C and G). Then, elution was attempted with purified t (lanes B and F). Finally, both adsorbents were washed with 5 M urea-containing buffer (buffer C) (lanes A and E). No proteins eluted specifically from BSA-Sepharose with t (lane B) or the urea-containing buffer. In contrast, two major proteins, actin (43K) and a 57Kpolypeptide, were eluted from t-Sepharose with purified t (lane F). Previously, one of these polypeptides, the 57K protein, was also observed to elute, specifically, from this absorbentin 2 MNaCl-5 M urea (e.g., Fig. 1B). In addition, although poorly defined in the photograph, very small amountsof a 32K protein and a 20K protein were present in the t eluate in the relevant autoradiogram. Unlike the case for the 57-, 32-, and 20-kDa bands, it is difficult to ascribe actin elution to a specific interaction with t because, unlike any of the other three proteins under consideration, it often bound toall of the control adsorbents used in this work.

When similar amounts of cell extract protein (z50 ,ug)

were applied to two portions of the BSA + t-Sepharose

adsorbent, the amount of the 57K protein appearing after elution with soluble t antigen was greater than when buffer C

was used as an eluant (data not shown). The relevant

evidenceis thatCoomassie blue staining of a gel bearing the t-eluate fraction revealed a 57K band which was regularly

absent from the urea eluates. Becauseincubation of a loaded

t-Sepharose adsorbentwith tresulted in stained 57K protein eluting from the matrix and application of buffer C did not, it waspossiblethat inprevious experimentsthe majority of the

57Kprotein remained bound to the t adsorbent after elution

with buffer C. To test this,identical portions of the BSA and

BSA + t adsorbents were mixed with a

[35S]methionine-labeled CV-1P cell extract, washed, eluted with buffer C,

and boiled in buffer containing 1% SDS. A Coomassie

blue-stained SDS-polyacrylamide gel is shown in Fig. 3, in

which stained amounts of a 57K protein were specifically

present in the SDS eluate of the BSA + tadsorbent (lane I)

but not the BSA adsorbent (lane D). By contrast, actin

eluted from both columns, and some BSA (68K) also

ap-peared in both cases, presumably because of nonspecific

BSA COL.

+t

11

A 8 C D

-.: 4-1

A.C.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.612.67.282.74.404.2]
(5)

696 MURPHY ET AL.

detachment from the matrix. Notably, a comparison of the

pattern ofstained bands in lane I with that of the protein

markers in lane J shows that the 57K protein comigrated

withan authentic tubulinmarker. In otherexperiments, we

detectedsmallamountsofboth the 32K and 20K proteins in

eluatesobtained by boiling tadsorbents in SDS.

Neverthe-less,whenevertested, the majorityofeach of these proteins

eluted in buffer B.

Bindingofvariouscell proteins to amodifiedspecies of t. It

wasasked whether the primary and secondary structureoft

must be intact for it to complex with the three t-binding

proteins under investigation. Yang et al. (39) showed that reductionandalkylationof tantigen withN-ethylmaleimide prevented its coimmunoprecipitation with two cell proteins

(56K and32K). Todeterminewhetherreducedand alkylated

twouldinteract withthe three cell proteinsdescribed above,

samples ofN-ethylmaleimide-treatedandunmodifiedt were

compared aseluants ofvarious protein-loaded adsorbents. Specifically, an extractoflabeled CV-1P cells was mixed withanunmodifiedtadsorbent.Afterwashingwith buffer A,

serial elution was attempted with soluble, alkylated t and

then withanidentical amountofunmodifiedt.Theunbound

protein fraction is shown in Fig. 4, lane A; the effect of

applying reduced andalkylatedt(50 ,ug) isshowninlane B;

A B

C

D

-57K -43K

BSA+tCOL.

r . ...

F~ H I J1

-94 K

-68 K

57 K

-43 K

-29 K

-24 K

-20 K

- 17 K

- 14 K

FIG. 3. Elution ofprotein-loaded t-Sepharose in boiling

SDS-containing buffer. Extracts of CV-1Pcellswere mixed with BSA-andBSA +t-Sepharoseatroomtemperaturefor 1 h.Samples (3%)

oftheunbound proteinsareshown inlanes A and F. Afterwashing with buffer A, bound proteins were serially eluted with buffers B

(lanes B and G) and C (lanes C and H) and by boiling in 1% SDS-containingbuffer,asdescribed in Materials and Methods(lanes

DandI).Samples in lanes B through D and G throughIincludedall of the protein present in the relevant fractions. Proteins were

electrophoresed througha15% SDS-polyacrylamidegel whichwas

stainedwithCoomassie brilliantblue and dried.Proteinmarkersare

showninlanes EandJ. COL, Columns.

FIG. 4. Effect ofreduction and alkylation of t on cell protein binding. t-Sepharosewasusedasanadsorbent foraffinity chroma-tography, and an [35S]methionine-labeled CV-1P extract (=50p.g)

was applied, washed, and then serially eluted with reduced and alkylatedt(50,ug),native t(50 ,ug),andbuffer C. Lanes:A,sample (3%)of theunboundproteins; B, all of the protein eluted with buffer Acontaining reduced and alkylated t; C, all ofthe proteineluted withnative t;D, all of theproteinelutedwithbufferC.

lane C shows whatwaseluted with nativet(50 ,ug). Onlya

faint band comigrating with actin was detected in lane

B,

whereas in lane C both actin and the 57K protein were

present.A final wash oftheadsorbentwith buffer C(lane

D)

led to the further elution of faint actin and 57K

protein

bands. In a related experiment (data not shown), the

57K,

32K,

and 20K proteins were found to bind with a much reducedaffinity toreduced and alkylated t-Sepharose

com-paredwithbindingtounmodifiedt-Sepharose.These results suggestthatsomeelement(s)of the nativestructureoftmay be important for its interaction with the three cellproteins

notedabove.

To testthe possibility that all of the t sequence must be

present for binding of the three proteins, a truncated t

polypeptide (14-kDat;Fig. 5)containing theCOOH-terminal

123 residues of the t antigen (2, 7, 36) was coupled to

CNBr-activated Sepharose. Labeled CV-1P proteins which

boundto andwere subsequently eluted from the adsorbent

werethen electrophoretically analyzed asdescribedabove.

The three cell proteins which bound to the t antigen

adsorbents also bound, to the same degree, to the

14K-t-Sepharose adsorbent (data not shown). Thus, a significant

segment of the primary structure of t is not required for

specific bindingofthe three cell proteins.

Subcellular fractionation of CV-1P cells. The cell extracts

usedin the aboveexperiments were prepared such that most

of the nuclear material wasinitially pelleted and was,

there-fore, not present in the final extract (see Materials and Methods). Thus,itwaspossiblethatcertain nuclear

constit-uents were absent from themixture ofproteins appliedtothe

BSA COL.

!r _ .__

A R C. F

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.612.386.503.78.340.2] [image:5.612.75.291.327.619.2]
(6)

SV40 Small t Antigen

38 f97 103 111 113 116 138 140 143 153 161

82aa 92aa

T/t t

14 Kd t Antigen

7103 11 113 116 138 140 143 153 161

30aa I 92 aa

T/t t

FIG. 5. Relevant features of the primarystructuresof SV40tand 14Ktantigens. The diagramsrepresentthegrossprimarystructures

of nativetand14K t.Thelatter lacks the N-terminal51residuesof intact t. Thus, the 82 N-terminal amino acids of t and the 30 N-terminal amino acids of14K t areshared withlargeT.Bothtand

14Ktcontain the 92-amino-acidunique region.The numberswithin theboxes indicate the locations ofthecysteine residues ofthetwo proteins.

column. To test this possibility and to begin to assess the

intracellular location of the three cellulart-bindingproteins noted thus far, CV-1P cells were fractionated by a

modifi-cation of a previously described procedure (1). Cells were

swollen in a hypotonic buffer and Dounce homogenized.

Nuclei were purified by centrifugation and washing (see

Materials and Methods). Three fractions were then

gener-ated:extractA,the initialsupernatant after Dounce

homog-enization and nuclearsedimentation; extract B, adetergent wash (0.5% NP-40, 0.1% sodium deoxycholate) ofthe nu-clear pellet; and extract C, a 1% NP-40 extract of 0.5% NP-40-0.1% sodium deoxycholate-washed nuclei whichhad been pelleted through a 1.8 M sucrose cushion. SDS gel

A B

A B C

electropherograms of the proteins present in these extracts

areshown in Fig. 6A. Samples of each of theseextracts were

then mixed with t-Sepharose andacontrol adsorbent

(BSA-or actin-Sepharose). In one case (extract B), 14K

t-Sepharose was also tested. Each was then washed with

buffer A and serially eluted with buffers B and C. The t-binding, 57 K protein was detected largely in extract A

(data not shown). By contrast, significant amounts of the

32K and 20Kproteins were not observed in this extract (data

notshown). Extract B contained the 32Kprotein, whichwas

the major protein eluted from both the t-Sepharose column

and the 14K-t-Sepharose column (Fig. 6B). In addition,

when isolated from =:50 ,ug of extract B, this protein was

present as a Coomassie blue-stained band in a 15%

SDS-polyacrylamide gel (data not shown). In some experiments

with extract B, a small amountof the 57K protein was eluted

from t-Sepharose with buffer C. Thus, the 57K protein is

probably not an exclusive component of eitherextract A or

B, but may be present in both.

Itcan be seen (Fig. 6C) that the 20K t-bindingproteinwas

enriched inextract C. When this fraction wasincubated, in

parallel, with t-Sepharose and actin-Sepharose, the 20K

protein bound only to the t adsorbent and was eluted,

primarily, in 1 M NaCl-containing buffer (Fig.6C, lane E),as

previously observed. In another experiment, extract Cwas

also applied to 14K t-Sepharose, and the results show that

the 20K protein also bound specifically to this adsorbent (data not shown). These observations strongly suggest that

the three cell proteins which interact with insoluble t are

differentiallyconcentrated in certain cell fractions.Hence,it ispossible thatthey are concentrated in different cell

com-partments. In particular, the 20K protein may be a nuclear

constituent whoseanchoringinthatenvironment is sensitive

torelatively high concentrationsof nonionicdetergentor to sedimentationinhypertonicmedia orboth.The 20Kprotein

C BSACOL. t COL. t4Kt COL.

A B C D E F GH I

Actin COL. t COL. A B C D E F

a...._

*W..

0--43K

-29K

-20K

I

i

t

- 57K --43K

-43K -32K

on ..20K

FIG. 6. Subcellular fractionation of CV-1P cells. (A) Samples ofeach subcellular fraction prepared as described in the text were electrophoresed througha15%SDS-polyacrylamide gel. Lanes: A, extract A; B, extract B;C,extract C. (B) Samples of extract B (==50

pg)

weremixed witht-, 14K t-, andBSA-Sepharose,as described in the text. Lanes: A, D, and G, sample (2%) of the unboundproteins; B, E, andH,all of theproteineluted withbuffer B;C,F, andI,alloftheprotein elutedwith buffer C. (C) Samples of extract C (==50,ug) were mixed witht-andactin-Sepharose,asdescribedin the text. Lanes: A and D, sample (2%) of the unbound proteins; B and E, all of theprotein eluted with bufferB;Cand F, all of the protein eluted with buffer C. COL, Columns.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:6.612.46.298.81.202.2] [image:6.612.88.516.442.674.2]
(7)

698 MURPHY ET AL.

is not t itself since the CV-1P cells used in these experiments

wereuninfected and lacked SV40 gene products. These data

also suggest that each protein can bind alone to t in the

absence of the others.

Evidence that the 57K protein is tubulin. As noted

previ-ously, the [35S]methionine-labeled57Kproteineluting from t-Sepharose comigrated with an authentic tubulin marker (Fig. 3). In someexperiments, it also had the same charac-teristic autoradiographic appearance as tubulin, migrating through SDS-polyacrylamide gels as a doublet (data not shown), suggesting thepresenceof boththea.and iforms of

tubulin. Thus, three experiments were performed to test whether the 57K protein is tubulin. With a procedure that

was developed by others (22, 34), [35S]methionine-labeled

CV-1P cells were first extracted with a buffer that preserves intact microtubules and then in a calcium-containing buffer which depolymerizes them and solubilizes the tubulin subunits. The first extract contained most of the soluble cytoplasmic proteinsof the target cell. By contrast, a major protein inthecalciumextract was57Ktubulin (Fig. 7A, lane

A). This extract was incubated with both BSA-Sepharose

A

A B C

B

t+BSA COL BSACOL.

A B CDEFGH

--57K -43K

[image:7.612.405.476.76.293.2]

29K

FIG. 7. [35S]tubulin bindingtot-Sepharose. (A) Plates (100 mm) containing confluentmonolayersof[35S]methionine-labeled CV-1P cells were washed with phosphate-buffered saline and incubated

three times in succession with 3 ml of PM2G (0.1 M PIPES [piperazine-N,N'-bis(2-ethanesulfonic acid)], 1 mM MgSO4,2 mM EGTA[ethylene

glycol-bis(O-aminoethyl

ether)-N,N,N',-tetraacetic acid], 20% glycerol, 1 mM 2-mercaptoethanol) containing 0.1% NP-40(lane C),thesamebufferagain (lane B),and PM2Gcontaining

5 mM CaCI2 (lane A). Asample (30 ,ul)ofeach 3-ml fractionwas

electrophoresed througha15%SDS-polyacrylamide gel.Molecular weight markersareindicatedontheright. (B)Samples(=20 ,ug)of

thecalcium-containingfraction(panel A,laneA)of [35S]methionine-labeled CV-1P cells were mixed with t- andBSA-Sepharose, and chromatographywasperformedasdescribed in thetext. Lanes: D

andH, sample (2%)of the unbound proteins; C andG, all of the

proteineluted withbufferB;B andF, all oftheproteineluted with

bufferC;Aand E, all of theprotein elutedin buffer Acontaining 0.05% SDS. Proteinfractionswereprecipitatedwith 10%

trichloro-acetic acid, and the ethanol-washed pellets were redissolved in

sample buffer and electrophoresed through a10% SDS-polyacryl-amidegel. COL,Columns.

A B

^iu

-57K

-

43K

i-

32K

FIG. 8. Western blotting oftubulinand the 57K t-binding pro-tein. Purified authentic tubulin and proteins eluted from a t-Sepharose adsorbent by boiling in SDS-containing buffer were analyzed onapolyacrylamide gel and transferred to nitrocellulose paper. After extensive washing and blocking with BSA, the strip was incubated withmonoclonal antitubulin antibody(61D), followed by incubationwithalkalinephosphatase-labeled anti-mouse immu-noglobulin. The strip was then washed and incubated with a chromogenic substratetoproduce the bands shown. Lane A, Eluate fromboiling t-Sepharose adsorbents in SDS-containing buffer; lane B,authentic tubulin. Protein markers are indicated on the right. and BSA + t-Sepharose. Figure 7B shows the results of elution of these adsorbents with bufferB (lanes C and G), bufferC(lanes B and F), and 0.05% SDS (lanes A and E).

The 57K protein was specifically eluted from BSA +

t-Sepharose, whereas little or none eluted from BSA-Sepharose.

Next, Western blotting was performed on eluates from

t-Sepharosecolumnsbyusingamousemonoclonalantibody directedagainst tubulin. LabeledCV-1Pextracts were incu-bated with t-Sepharose, washed, and eluted by boiling in SDS-containing buffer. The eluates and authentic, purified

tubulin were electrophoresed in parallel through a 15%

SDS-polyacrylamide

gel.The

proteins

werethentransferred

to nitrocellulose paper, and the nitrocellulose strips were

washedandblocked withBSA(seeMaterialsandMethods). The strips were then incubated with a mouse monoclonal antitubulin antibody. After abriefwashing, alkaline phos-phatase-labeled anti-mouse

immunoglobulin

was incubated

with thestrips.Thestripswerethen removedfrom thebags,

washed, and incubated with a chromogenic substrate. The

resultsare showninFig. 8; laneAcontains the eluate from

the t-Sepharose column, and lane B contains

purified

tubulin. The 57K protein eluting from t-Sepharose

clearly

reacted with monoclonal antitubulin antibody and

comi-grated with theauthentic tubulin standard.

In afurthertestof therelationshipbetweentubulin and the

57K t-binding protein, Cleveland-Laemmli staphylococcal V8 partial proteolysis mapping was

performed

on both the

CV-1P 57K

protein

isolated from

t-Sepharose

and on 57K

tubulin isolated from the same cellsby the

Ca2'

extraction

method noted above. Labeled CV-1P extracts were

incu-bated with t-Sepharose,

washed,

and eluted by

boiling

in

J. VIROL.

I

ow

on November 10, 2019 by guest

http://jvi.asm.org/

[image:7.612.63.305.311.540.2]
(8)

57K

Tubulin

57K

r I,I'

r---~~~--.1

01

.1 .01

1

.01

primary structures of the 57K

protein

and tubulin were

indistinguishable.

Coprecipitation

of the 57K

protein

and t from crude cell

extracts. It was asked whether t can form a complex with

tubulin in vivo. Todetermine

this,

several cellextractswere

immunoprecipitated

with rabbit antiserum directed

against

the common N terminus ofthe T and t

antigens

(2). The

results of suchanexperimentperformedwith threedifferent

[35S]methionine-labeled

cell extracts are shown in

Fig.

10.

Specifically,

CV-1P cellswereinfectedwithd1883(T+/t-)or

SV402

(T-/t+)

toobtain extracts containing largeTonlyor

small t

only, respectively.

Cos-1

cells,

which contain both

the

large

Tand smallt

antigens,

werealso used

(Fig.

1OAand

B). Immunoprecipitation

of the d1883-infected cell extract

withtheabove-noted serumyielded largeTonly.However, when twas presentin theextracts (Cos-1 cells and SV402-infectedCV-1P

cells),

smallamountsofa57Kprotein were coprecipitated with t antigen. By comparison with an

[35S]methionine-labeled

standard, this protein, extracted

from multiple gel lanes loaded with anti-t

immunoprecipi-tatesofSV402-infected CV-1P cells, provedtobetubulin

by

Cleveland V8 partial protease mapping (data not

shown).

The 32Kproteinwasnotdetectedinthese experiments, and

itwas notpossibletodetermine whether the 20K proteinwas present, since it comigrates with the small t antigen. The

identity

of the -62-kDa band seen in the SV402/CV-1P

immunoprecipitateis unknown. Conceivably, it is the

-56-kDa bandreportedtobindto tby Yangetal.(39). In the

gel

A

883

COS

I

N

I N

Om

402

I

N

-*-Large T _62K

-57K

FIG. 9. Cleveland staphylococcal V8 protease mapping.

[35S]methionine-labeled 57K protein eluted fromtadsorbents (left

and right lanes) and35S-labeled tubulin from a calciumextractof CV-1P cells (center lanes) were excised from a 7.5% SDS-polyacrylamide gel and subjectedto partial V8 protease mapping. Twoconcentrations ofV8protease (0.01and0.1

Fxg)

wereused for eachprotein asindicated andasdescribed in Materials and Meth-ods. Peptides were electrophoresed through a 15%

SDS-polyacrylamidegel.

B

COS

--LargeT

-of mw.--57K

S

.- Small T

---25K

f

o-Smal f.

SDS-containing buffer to obtain the maximum amount of

35S-labeled

57Kt-bindingprotein. The eluates were

electro-phoresed through an SDS-polyacrylamide gel alongside a calciumextractofthe same cells. The 57Kproteinsineach lanewereidentified byautoradiography, eluted from the gel, and incubatedwith two concentrations of V8 protease, and the digestion products were subjected to SDS gel electro-phoresis (5) (Fig. 9). All labeled peptides derived from the 57K protein comigrated with those from the tubulin

[image:8.612.75.267.77.520.2]

stan-dard. Thus, by the three criteria discussed above, the

FIG. 10. Coprecipitation of the 57K proteinandsmalltantigen. (A) [35S]methionine-labeled extracts from three different sources, dl883-infected CV-1P cells(left lanes), Cos-1 cells (middle lanes), andSV402-infected CV-1P cells(right lanes),were immunoprecipi-tated with rabbit anti-T/t serum (I) or nonimmune normal rabbit serum (N). Immune complexes were washed five times in RIPA bufferandboiledinSDS-Laemmli buffer before analysisona10%

SDS-polyacrylamide gel. The positions of large T, the 62K, 57K,

and 25K proteins, and small t are shown. (B) A separate, but otherwise identical, experiment performedonanother [35S]methio-nine-labeled Cos-1 cell extract.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:8.612.318.544.379.632.2]
(9)

700 MURPHY ET AL.

system used here, we and others (K. Rundell, personal

communication) have found that the 56-kDa t-binding

pro-teinofYang et al. (39) migrates likea =62-kDa band. The

faint 25-kDa band isa truncated largeTderivative, specifi-cally synthesized inthese cellsonthe SV402 template (25).

DISCUSSION

Three cell proteins, present in crude, uninfected cell

extracts, bound specifically and reproducibly to

homoge-neousSV40 t antigen, immobilized by coupling to Sepharose

beads. They could also bind in solution, since free t eluted them from the t-Sepharose adsorbent. Since no more than 123of its174amino acids are required for the antigen to bind to the proteins, this property oft may be governed by a limited portion of its structure. Conceivably, there is a discrete sequence domain which operates in this regard. Given the results with the 14K t protein, one candidate region isthe sequencebetween residues83 and 174, i.e., the

segmentwhich is not sharedwith SV40 largeT.

Two of these proteins (32K and 20K) have not been

identified. Neither is a phosphoprotein, as defined by clas-sical in vivo labeling with 32Pi, and one of them (20K) is closely associated with the nucleus,asdefinedbya biochem-ical extraction method which yields highly purified, intact nucleiwithout significant quantities of attached cytoplasmic tabs. A significant quantity of small t is also present in the

nucleus. Hence, there is a possibility of an interaction in

vivo.

The 32K protein, although associated with the crude

nuclear pellet, could be eluted from this fraction in deter-gent-containing buffer. This raises the possibility that this polypeptide is associated with some form of nonnuclear membranous structure. Moreover, given the fact that it could be isolated as a stained band from t-Sepharose col-umnsloaded with50 ,ugofacrude, subcellularextract, itis likelytobe a major cell protein.

The57K t-binding protein is tubulin. First, it comigrated withauthentic tubulin subunits. Second, radioactive tubulin boundspecificallytot-Sepharose and eluted under the same conditions and with the samespecificityasthe 57Kprotein. Third, the 57K protein and purified

[35S]tubulin

revealed identical partial V8 peptide maps. Finally, the 57K protein andtubulinwereidentifiedascomigratingbandsbyWestern

blotting

in which the probe was a monoclonal antitubulin

antibody.

That the binding ofthe three proteins to purified t is at

least relatively specific is apparent fromthe fact that they failed to bind to Sepharose alone or to Sepharose loaded with threeotherproteins, BSA, myoglobin, andactin.

How-ever, the most

revealing

observation in this

regard

wasthat

all three proteins failed to bind

effectively

to t which had

been reduced andalkylated.Thissuggests thatsomeelement

ofthe native t structure is essential for

binding and,

there-fore, that the observed binding is not the result of nonspe-cific adherence to randomcoil sequences.

Rundell andco-workers (28, 39)previouslydescribed 56K

and 32K cell proteins which appear to

coprecipitate

with

papovaviral t antigens. We obtained a sample containing

theseproteinsfrom K. Rundell andcomparedtheir

electro-phoretic mobilities with the mobilities ofthe 57K and 32K proteins which we observed (data not shown). The 56K protein migrated as a -62-kDa protein in the SDS-polyacrylamide gel electrophoretic system used here; this wasinkeepingwithpriorobservations (K.Rundell, personal

communication).

As such, it and the 57K

protein (tubulin)

did not comigrate (Fig. 10). Furthermore, the 32K protein

observed by Rundell and co-workers (28, 39) also failed to

comigrate with the 32K in vitro binding protein described

here (data not shown). Hence, it also seems unlikely that

they are identical proteins.

It waspossible to demonstrate coimmunoprecipitation of

the 57K protein (tubulin) and t (Fig. 10). A coprecipitating

='62-kDa protein present in the same lanes as the 57K

species may be thet-bindingprotein noted above (28, 39). Notubulinwas observed in an extract of a cell synthesizing large T but not small t, and there was no tubulin precipitated by preimmune rabbit serum or by staphylococcal protein A-Sepharose in the absence of serum (data not shown).

Hence, the coprecipitation phenomenon strongly suggests

the existence of a t-tubulin complex in the extracts tested.

From the intensity of the coprecipitating tubulin bands, it canbe concludedthat a smallfraction of tubulin(<1% ofthe

total tubulin) is present in a complex with t. Indeed, given

the relative intensities of the t and tubulinbands, it seems

likelythatonlyasmallfractionof the ambient t in the extract was bound to tubulin, assuming that the ratio of t/tubulin in such complexes is not very high. Conceivably, this was because the affinity of t synthesized in animal cells for tubulin is lower than thatof bacterial t. Alternatively, onlya

few t or tubulin molecules may be so located within a cell

that complex formation could ensue. Anotherpossibility is thatsuchcomplexes formefficiently but also dissolve readily as a result ofsubsequent events.

Asimilarcoprecipitation experimentwasattemptedon the

above-noted extract with a monospecific antitubulin serum (gift of K. Fujiwara) with negative results. This need not imply that the t-tubulin coprecipitation phenomenon does not signify complex formation between these twoproteins. First, the major, relevant epitopes on the tubulin surface

may have been shielded fromthe antibody as aresult oft

binding. Alternatively,theantibodymay noteffectively bind thetubulin molecules which bind tot, because they

repre-sent asubclassof molecules whicharestructurallydifferent

fromthosewhich do bindto theantibody.Itis also possible

that the available antitubulinserumfacilitated thedisruption ofsuchcomplexes. Hence,

given

the knownability oftand

tubulin toformtight complexesinvitroand thespecificity of

the t-tubulin coprecipitation, as defined with anti-t serum,

one interpretation ofthe finding of t-tubulin complexes in

crude

t-containing

cellextractsisthattand tubulin interact

in vivo. In this regard, we have, thus far, been unable to

detect thedecoration of microtubules byt in fixedcells

by

indirect immunofluorescence methods (B. Wu, C. Murphy,

and D. Livingston,

unpublished observations).

Consistent

with this, the results of biochemical extraction analyses

performed by

the methodof Solomonetal.

(34)

suggest that

iftisboundtomicrotubules in vivoverylittle of the

antigen

is so located. This does not mean that t fails to contact

microtubules in vivo. Rather, it suggests that if it does, it

might do so in the manner ofmicrotubule-associated

pro-teins which

typically

functionin substoichiometricamounts

(fora

review,

seereference38).IfSV40tdoes interact with

microtubulesinvivo, itmightbeworthwhileto

speculate

on the possibility that such an interaction contributes to the

maintenanceofaneoplastic phenotype in somecells,

given

the fact that tubulin is involved in the maintenance ofcell

architecture,

aswellasin

effecting

cell

division,

movement,

and

mitosis,

and is also

possibly

involved in the

regulation

of

cell surface receptor function (8, 18).

We have not seen evidence of the in vitro

t-binding

32K protein in small t immunoprecipitates.

Conceivably,

our J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

(10)

available sera may disrupt or fail to interact with t-32K protein complexes. In preliminary experiments, however, we have detected specific binding of small amounts of complexes containing radioactive 57K, 32K, and t proteins

derived from crude cell extracts to columns of

anti-gel-band-purified-t antibody-Sepharose. If these results are correct,

the reasonfor the apparent discrepancyis not clear.

How-ever, if anti-t antibody binds less efficiently to t once the

latter is bound to one or morecellproteins, bindingof such complexes might be more apparent at relatively high antibody/antigen ratios, such as might be achieved in an

environment provided by an antibody-linked solid-phase

immunoadsorbent.

Finally,evenif tdoesnotformphysiologically meaningful interactions with any of the t-binding proteins described here, thefact that one or more of themcanform specific in

vitro complexes with the antigen may indirectly signify an

importantfunctional propertyof t.If,infact, theybind to a

region of the protein surface which constitutes a discrete domain, theirbindingmay mimic aphysiologically relevant

functionfor which this particular structural region is

essen-tial. Inthat case, thesebinding phenomenacould serveas a

functional marker for theintegrityof such aregion (e.g.,part

or allof the 92-residuesegmentunique to t)and be usedto

probe its relevance to t function in vivo. Such analyses performedwithselectedmutantspeciesof t could be instruc-tive.

ACKNOWLEDGMENTS

We are grateful to Frank Solomon, David Pallas, and Thomas Roberts forhelpful conversations andadvice. We alsothank Ann Desai for herexpert assistance inpreparing the manuscript.

This work was supported by Public Health Service grantsfrom the NationalCancer Institute.

LITERATURE CITED

1. Abrams,H.,L.Rohrschneider,and R. Eisenman.1982. Nuclear location of the putative transforming protein of avian myelocytomatosis virus. Cell 29:427-439.

2. Bikel, I.,T. M.Roberts,M.T.Bladon,R.Green,E.Aman,and D. M.Livingston. 1983.Purification of biologically active simian virus 40 small tumor antigen. Proc. Natl. Acad. Sci. USA 80:906-910.

3. Bouck, N., N. Beales, T. Shenk, P. Berg, and G. diMayorca. 1978. New region of simian virus 40 genome required for efficient viral transformation. Proc. Natl. Acad. Sci. USA 75:2473-2477.

4. Chou, J. Y.,and R. G. Martin. 1975. DNAinfectivity and the induction of host DNA synthesis with temperature-sensitive mutantsof simian virus40.J. Virol. 15:145-150.

5. Cleveland, D. W., S. G. Fischer, M. W. Kirschner,and U. K. Laemmli. 1977. Peptide mapping by limited proteolysis in

so-dium dodecyl sulfate and analysis by gel electrophoresis. J. Biol. Chem. 252:1102-1106.

6. Crawford,L.V., and P. Z. O'Farrell.1979. Effect ofalkylation

onthephysical properties of simian virus40T-antigen species. J. Virol. 29:587-596.

7. Derom, C.,D.Gheysen,and W. Fiers.1982.High-level synthe-sis on Escherichia coli of the SV40 small-T antigen under control of the bacteriophage lambda p2 promoter. Gene 17:45-54.

8. Dustin, P. 1978. Microtubules. Springer-Verlag KG,Berlin. 9. Ellman, M.,I.Bikel, J. Figge,T. Roberts,R.Schlossman, and

D. M.Livingston.1984.Localization ofthe simianvirus40small

t antigen in thenucleusandcytoplasm of monkey and mouse cells. J. Virol. 50:623-628.

10. Fluck, M. M., and T. L. Benjamin. 1979. Comparisons of two early gene functions essential for transformation in polyoma virusand SV40. Virology96:205-228.

11. Friedman, T., R. R. Doolittle, and G. Walter. 1978.Amino acid sequencehomology between polyoma andSV40tumorantigens deduced from nucleotide sequences. Nature (London) 274: 291-293.

12. Frisque, R. J., D. B.Rifkin, and W. C. Topp. 1979.Requirement forthelargeand small Tproteins of SV40 in maintenance ofthe transformed state. Cold Spring Harbor Symp. Quant. Biol. 44:325-331.

13. Gluzman, Y. 1981. SV40-transformedsimian cells support the replication of early SV40mutants. Cell23:175-182.

14. Graessmann, A., M. Graessmann, R. Tjian, and W. C. Topp. 1980.Simian virus40small-tprotein is required for loss of actin cable networksinratcells.J. Virol. 33:1182-1191.

15. Hiscott, J. B., and V. Defendi. 1981. Simian virus 40 gene A regulation ofcellular DNAsynthesis.II.Innonpermissive cells. J. Virol. 37:802-812.

16. Hohmann, A. W., and P. Faulkner.1983.Monoclonal antibodies tobaculovirus structuralproteins: determination of specificities byWestern blot analysis. Virology 125:432 444.

17. Kilton, L. J., M. Bradley, C. Mehta, and D. M. Livingstone. 1981. Rapid and sensitive quantitative immunoassay for the largesimian virus 40 Tantigen.J. Virol. 38:612-620.

18. Kischner, M. 1978. Microtubule assembly and nucleation. Int. Rev.Cytol.54:1-71.

19. Laemmli, U. 1979. Cleavage of structural proteins during the assembly of the head ofbacteriophage T4. Nature (London) 227:680-685.

20. Martin,R.G., V. P.Setlow,C. A. F.Edwards,and D.Vembu. 1979. The roles of the simian virus 40 tumor antigens in transformation of Chinese hamster lung cells. Cell 17:635-643.

21. Montano, X., and D. P. Lane. 1984. Monoclonal antibody to

simian virus 40 smallt.J.Virol. 51:760-767.

22. Osborn, M.,and K. Weber.1977.Thedisplay of microtubulesin transformed cells. Cell 12:561-571.

23. Prives, C., E. Gilboa, M. Revel, and E. Winocour. 1977. Cell-free translation of simianvirus 40earlymessenger RNAcoding for viralTantigen. Proc. Natl. Acad. Sci. USA 74:457-461. 24. Rassoulzadegan, M., B. Perbal, and F. Cuzin. 1978. Growth

control in simian virus 40-transformed rat cells: temperature-independent expression of the transformed phenotype in tsA transformants derivedbyagarselection. J.Virol. 28:1-5. 25. Rubin, H., J. Figge, M. T. Bladon,L. B. Chen,M. Ellman,I.

Bikel,M. Farrell, and D. M. Livingston. 1982. Role of smallt

antigen in the acute transforming activity of SV40. Cell

30:469-480.

26. Rundell, K. 1982. Presence ingrowth-arrested cellsof cellular proteins that interact with simian virus 40 small-t antigen. J. Virol.42:1135-1137.

27. Rundell, K.,andJ.Cox.1979. Simian virus40tantigen affects the sensitivity of cellular DNA synthesis to theophylline. J. Virol.30:394-396.

28. Rundell, K.,E.0. Major,and M.Lampert.1981.Association of cellular56,000-and32,000-molecular-weight proteinswith BK virusandpolyoma virust-antigens. J. Virol. 37:1090-1093. 29. Seif, R.,and R.G.Martin.1979. Growthstateofthecellearly

after infection with simian virus 40 determines whether the maintenance of transformation will be A-gene dependent or

independent. J. Virol. 31:350-359.

30. Seif, R.,and R.G. Martin. 1979. Simianvirus 40 smalltantigen isnot required for the maintenance of transformation but may

act as apromoter(cocarcinogen) during establishment of

trans-formation inrestingratcells. J.Virol. 32:979-988.

31. Shenk,T.E., J. Carbon, and P. Berg. 1976. Construction and analysisof viabledeletionmutantsofsimian virus 40. J. Virol. 18:664-671.

32. Shyamala,M., C. L.Atcheson, and H. Kasamatsu. 1982. Stim-ulation of host centriolarantigeninTC7cellsby simianvirus 40: requirementfor RNA andprotein synthesisand anintact simian virus 40 small-t genefunction.J. Virol. 43:721-729.

33. Sleigh,N.J.,W.C.Topp,R.Hanich,andJ. F.Sambrook.1978. MutantsofSV40withanaltered smalltproteinarereduced in theirabilitytotransformcells. Cell 14:79-88.

on November 10, 2019 by guest

http://jvi.asm.org/

(11)

702 MURPHY ET AL.

34. Solomon, F., M. Magendantz, and A. Salzman. 1979. Identifica-tion with cellular microtubules of one of the co-assembly microtubule-associated proteins.Cell 18:431-438.

35. Spangler, G. J.,J.D.Griffin, H.Rubin, and D. M.Livingston. 1980. Identification and initial characterization ofanew low-molecular-weight virus-encoded T antigen in a line of simian

virus40-transformedcells. J. Virol. 36:488-498.

36. Thummel, C.S.,T. L. Burgess, and R.Tjian. 1981.Properties of

simian virus 40 small t antigen overproduced in bacteria. J. Virol.37:683-697.

37. Topp, W. C.1980. Variable defectivenessforlytic growth of the

dl54/59 mutantsofsimian virus40.J.Virol. 33:1208-1210. 38. Vallee, R. B. 1984. MAP2(microtubule-associated protein 2),p.

289-311. In J. W. Shay (ed.), Cell and muscle motility, vol.5. Plenum Publishing Corp., New York.

39. Yang, Y.-C., P. Hearing, and K. Rundell.1979.Cellular proteins associated with simian virus 40 early gene products in newly infected cells.J.Virol. 32:147-154.

40. Zhu, Z., G. M. Veldman,A. Cowie, A. Carr, B. Schaffhausen, and R. Kamen.1984.Construction and functional characteriza-tion of polyomavirusgenomesthatseparately encode the three early proteins.J. Virol.51:170-180.

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

Figure

FIG.1.ininproteins;thenCV-1Pwithextracted,D,withmigration the Materials Identification of specific t-binding proteins
FIG.2.followedgel.theattemptedandprotein(3%)Methods.[35S]methionine-labeled Elutionof boundprotein(s)witht.Extractsof CV-1P cells (=50 ,ug) were mixed with BSA BSA + t-Sepharose adsorbents, as described in Materials andAfter extensive washing with buffer A
FIG. 4.Aalkylatedwithbinding.was(3%)tography, containing Effect of reduction and alkylation of t on cell protein t-Sepharose was used as an adsorbent for affinity chroma- and an [35S]methionine-labeled CV-1P extract (=50 p.g) applied, washed, and then seri
FIG. 5.theintactofN-terminal14K14K native Relevant features of the primary structures of SV40 t and t antigens
+3

References

Related documents

In a previous work, it was demonstrated that the bacterial transposon TnS is capable of undergoing sequence inversion via recombination between its duplicated IS50 elements

The limited amounts of viral genomic RNA led us to the use of mRNA from infected cells, containing both viral genomic RNA and subgenomic RNAs, as a template for cDNA synthesis.. In

This highly purified preparation of 3DPo1 catalyzed the synthesis of covalently linked dimeric RNA products from poliovirion RNA templates in the presence of an oligo(U)

The promoter region for transcription of the 3.6-kilobase mRNA of hepatitis B virus was identified by the chloramphenicol acetyltransferase assay by using HuH-7 hepatoma cells and

the DHBV core gene promoter is active in differentiated human liver cells and that synthesis and secretion of the processed core proteins are dependent on the expression of the

This adapts the state-feedback distributed control scheme presented in Borrelli and Keviczky (2008), leading to the solution of the stabilization problem for networks with

We conclude that in the estimation of options whose pay- off is determined by statutory accounting rules, which is often the case for European traditional with-profit

Results show that the motion exposure limiting values applied for the assessment of the working environment during maintenance ac- tivities has an impact on the time in which