0022-538X/82/030792-09$02.00/0
Poliovirus Empty Capsid Morphogenesis: Evidence for
Conformational
Differences
Between
Self-
and
Extract-Assembled Empty Capsids
J. ROBERT PUTNAK AND BRUCE A. PHILLIPS*
Departmentof Microbiology, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania 15261
Received 14 September1981/Accepted5November 1981
In this paper we describe the use of specific proteinases, surface-specific radioiodination, and antigenic reactivity in conjunction with isoelectric focusing for probing the conformations of different polioviral empty capsid species. Naturally occurring empty capsids (called procapsids) with an isoelectric point of 6.8 were resistant to proteolytic digestion by trypsin or chymotrypsin, as were empty capsids assembled in vitro in the presence of a cytoplasmic extract prepared from poliovirus-infected HeLa cells. In contrast, self-assembled empty
capsids (isoelectric point,5.0) weresensitivetobothproteinases. Capsid proteins
VPO and VP1 were attacked predominantly, whereas VP3 was resistant to
cleavage. Unpolymerized14Sparticles possessedatrypsin sensitivitywhich was
qualitatively similar to that of self-assembled empty shells. Surface-specific
iodination of virions and procapsids labeled VP1 exclusively. In contrast,
radioiodination of self-assembled empty capsids labeled predominantly VPO.
After radioiodinationthe sedimentationcoefficientcorrectedto water at20°C, the
isoelectric point, and the trypsin resistance of the procapsids remained
un-changed. Procapsids and extract-assembled empty capsids were N antigenic,
whereas self-assembled empty capsids were H antigenic. Self-assembled empty
capsids were notconverted to pH6.8 trypsin-resistant structuresby incubation
with avirus-infected cytoplasmic extract. However, 14S particles assembled in the presence ofamock-infected extract formed empty capsids, 20% of which resembled extract-assembled emptyshells as determinedbythe above-described criteria. These and related findings are discussed in terms of empty capsid structure andmorphogenesis.
The assembly of poliovirus empty capsids from precursor 14S particles was first demon-strated to occur in vitro in the presence of
cytoplasmic extracts prepared from
poliovirus-infectedHeLacells(13, 16). Later,itwasshown
that 14S particles can self-assemble, forming emptycapsid-like structures visiblebyelectron
microscopy (14). Unlike extract-mediated
as-sembly, this latter reaction demonstrated a marked dependence upon the concentration of 14Sparticles. The nature of the assembly-facili-tating activity in extracts is not known, but this
activitydoes notappear to be due to free
endog-enous14Sparticles (12).
We recently reported that self- and extract-assembledemptycapsids could bedistinguished from one another by (i) isoelectric focusing under nondenaturing conditions and (ii) rate-zonal centrifugation on sucrose density
gradi-ents (17). However, the polypeptide composi-tions of thesecapsids, as well as the apparent sizes and isoelectricpoints (pI)of their individ-ual polypeptides, appeared to be identical.
Therefore, we concluded that the two species of invitro empty capsids differed only in conforma-tion.
The topography of virus particles, including thetopography ofpicornaviruses,hasbeen stud-ied extensively by usingsurface-specific labeling
techniques, suchasradioiodination(5, 8, 9, 19).
When these techniques were performed with chloramine T, H202, or other oxidizing agents, tyrosine residues were labeled almost exclusive-ly. In an effort to obtain surface specificity, a glucose oxidase-lactoperoxidase enzyme system has been used with carrier-free 1251 (5). More recently, a solid-state oxidizing reagent,
chlor-oglycoluril, wasdeveloped (6). Based upon the
criterion ofpreservationof virusinfectivity,this reagent appeared to be more gentle than either chloramine Torlactoperoxidase. Studies have alsoindicated thatchloroglycolurilisasefficient aschloramineTandmore surface-specific than lactoperoxidase (10).
Another technique for studying topography hasbeenthe useofdegradative enzymes. With
792
on November 10, 2019 by guest
http://jvi.asm.org/
POLIOVIRUS EMPTY CAPSID MORPHOGENESIS
the exception of foot-and-mouth disease virus, picornavirusesareveryresistanttothe actionof proteolytic enzymes. Treatment of foot-and-mouth disease virus with trypsin caused a de-creasein its infectivity and cell attachment and altered its antigenicity (3). An analysis of tryp-sin-treated virusby polyacrylamide gel electro-phoresis (PAGE) in the presence of sodium dodecyl sulfate (SDS) showed that VP1 was cleaved, suggestingthat thispolypeptidewas at the surface of the virus particle. Both VP1 and VP2 were cleaved when the 12S particles pro-ducedasby-products of virus degradationwere exposedtotrypsin (7).
Inthisworkwestudied the surface configura-tions of the various polioviral capsid speciesby using proteolytic enzymes, surface-specific ra-dioiodination, and specific immune antisera. We presentevidence concerning the conformational differences between self- and extract-assembled empty capsids, as well as heat-denatured pro-capsids. We found thatvirus-infected cytoplas-mic extracts possess a morphopoietic factor(s) which directs empty capsid assembly and deter-minescapsid conformation.
MATERIALSANDMETHODS
Virus.Type 1poliovirus (Mahoney strain)wasused
for this work. The purification methods used have
been describedpreviously (16).
Cells.HeLacells(S3)werecultivated insuspension
culturebyusingEagle minimal essential medium
sup-plemented with Spinner salts, 2 mM glutamine, and
5%calfserum.Noantibioticswereused in routine cell
propagationcultures.
Preparation of virus-related particles, purification,
radio-labeling,andassembly reactions. Welabeled14S
particles,emptycapsids,andpoliovirionswith14C-or
3H-aminoacid mixtures andpurifiedthemas
previous-ly described (17). Likewise, assembly reactions were
carriedout asdescribedpreviously (13, 14).
Protein determinations.Proteinwasmeasuredbythe
Bradfordmethod(2).
Agarosegelisoelectricfocusing. Isoelectricfocusing
ofpoliovirus-related particles and polypeptides was
carried out in gels at pH 3.5 to 9.5, as described
previously (17).
Acrylamide gel isoelectric focusing. Focusing was
carriedoutingels containing4.8%acrylamide
mono-mer,0.27%bisacrylamide,8 M urea, and2%Nonidet
P-40in double-deionized water.These reagentswere
passedover amixed-bed ion-exchangeresin(Bio-Rex
RG501-X8) before use. Then2%ampholine (pH 3.5 to
9.5;LKB) wasadded, and the gel was polymerized in
the presenceofasolution containing 0.01%
ammoni-umpersulfate, 0.05% TEMED
(N,N,N'N'-tetrameth-ylethylenediamine), and 0.005% riboflavin phosphate
for1hunderafluorescentlight. Samples were
solubi-lized in1%SDS-5% 3-mercaptoethanolat 100°C for 3
min andcooled, and then 9 volumes of 10 M urea-4%
Nonidet P-40 was added. The gels were prefocused at
200 Vfor1h,and then the samples were focused from
thecathodeat200Vfor1h, at 400Vfor1 h, at 600V
for5h, andat800Vfor 30 min. The catholyte was 0.5
N NaOH, and the anolyte was 0.5 N H3PO4. After
focusing, the gels were fixed and fluorographed as
previously described (17).
Radioiodinations. Radioiodination of the purified
particles was carried out in the presence of
1,3,4,6-tetrachloro-3a,6a-diphenylglycoluril
(chloroglyco-luril). The choroglycoluril (dissolved in 0.1 ml of
chloroform) was added to a borosilicate glass test tube
(12 by 75 mm), and the solvent wasevaporated under
an N2 gas stream. The protein to be iodinated was
added together with 500 ,uCi of carrier-free Na
"25I
buffered with 20 mM Na2HPO4 (pH 7.4) in a total
volumeof 0.1 ml. The reaction wascarried outfor 10
min at0°C, and then the mixture was transferred to a
polypropylene tube containing 10 ,ulof 250 pM
KI-0.01% 3-mercaptoethanol. The iodinated proteins
wereseparated from thefree "25Ibycentrifugationon
15 to30%o sucrosegradients inphosphate buffer (pH
7.2) containing 25p.MKI and0.001%
P-mercaptoeth-anol. Centrifugationwas in an SW41 rotor at35,000
rpmfor 4 h at18°C.Fractionswere collected from the
bottom of the tube and assayed for radioactivity.
Immunoprecipitation of virus-related particles.
Spe-cific anti-N and anti-Hsera weregenerouslyprovided
byR.Rueckert and A. Mosser(Universityof
Wiscon-sin).Anti-N serumwaspreparedagainst purified po-liovirus in rabbits and was adsorbed with heated virus. Anti-H serum to heated virus was made and adsorbed with native virions. Dilutions of antisera were
pre-paredin NET (0.1 MNaCl,1mMEDTA, 0.01 MTris,
pH 7.2) or, more satisfactorily, in IPbuffer (20mM
P04, pH 7, 0.1 M NaCl, 1 mMEDTA, 0.1% bovine
serum albumin, 0.01% Triton X-100, 0.01% SB-14
[Zittergent; Calbiochem]) in 1.5-ml polypropylene
tubes. Thevolumes used were 15 to 30
RIp.
Aknownamountof labeled antigen (1,000 to 5,000 dpm) in 5
RI
orlesswasadded to eachdilution tube. After1 hof
incubation atroomtemperature (20°C)with continu-ousvigorousshaking,sufficientIPbufferwasaddedto
bringthevolume to 85p.1,and then 15,u1ofprotein
A-Sepharosebeads(50% suspensioninNET)wasadded.
Incubation was continued for 60 to 90 min, as
de-scribedabove. The tubes werecentrifugedat15,600x
g in an Eppendorf model 5412 centrifuge, and the
resulting supernatant fluids were carefully collected
and counted. Thebeads were washed twice with 100
RI of cold IP buffer and then suspended in 50p.lof
buffer and counted. Preimmune antisera at several dilutions (usually 1:10 and 1:100) were run as
con-trols.Specificimmune precipitation was calculated by
thefollowingtwomethods: (i) percentageof
disinte-grations immunoprecipitated = [(disintegrations per
minute in immunepellet- disintegrations per minute
incontrolpellet)/(disintegrationsper minute recovered
-disintegrationsperminute in control pellet)] x 100;
and(ii) percentage ofdisintegrations
immunoprecipi-tated= 100-[(disintegrations per minute in immune
supermatant)/(disintegrations perminuteincontrol
su-pernatant)]. Recoveries were 80 to 100%o, and the
values obtained by the two methods were nearly
identical. Controlpelletscontained approximately5%
orlessof the recoveredradioactivity.
Treatment of particles with proteolytic enzymes.
Preparations of particles or proteins were treated with trypsin (tolylsulfonyl phenylalanyl chloromethyl
ke-tone treated) or chymotrypsin in sodium phosphate
buffer(pH 7.4)at20°C. A ratio of enzyme to protein of
793
VOL. 41,1982
on November 10, 2019 by guest
http://jvi.asm.org/
794 PUTNAK AND PHILLIPS
1:10(wt/wt) was used. The reaction was stopped by
adding a 2- to 10-fold excess of soybean trypsin
inhibitor or 0.1 mMphenylmethylsulfonylfluoride (for
chymotrypsin).
Materials. 14C-amino acids (100 .Ci/ml) and
3H-amino acids (1 mCi/ml) were obtained from New
England Nuclear Corp., Boston, Mass. Na125I (100
mCiIml;carrier-free; in NaOH, pH 8 to 10) was also
obtained from New England Nuclear Corp.
Chlorogly-coluril(IODOGEN)wasobtained from Pierce
Chemi-calCo., Rockford, Ill. Bio-Rex mixed-bed resin was
obtained fromBio-Rad Laboratories, Richmond,
Cal-if.Trypsin (209 U/mg;N-tosyl-L-argininemethylester)
and soybean trypsin inhibitor were obtained from
Worthington Diagnostics, Freehold, N.J.
Chymotryp-sin A4 was obtained from Boehringer Mannheim,
Indianapolis, Ind. Phenylmethylsulfonyl fluoride was
obtained fromCalbiochem,LaJolla, Calif. Protein
A-Sepharosebeads werepurchased from Pharmacia Fine Chemicals, Inc., Piscataway, N.J.
RESULTS
Todescribe moreprecisely the kinds of empty
capsids used in these experiments, we used a
previously definedterminology, in which the pl and density (in grams per cubic centimeter) in CsCl are designated by a superscript and a subscript, respectively(17). Inthis terminology
procapsids derivedfrom infected cellsare
abbre-viated PRO, and empty capsids assembled in vitro are abbreviated E.C.
Trypsin sensitivity of poliovirus-related parti-cles. Polioviral procapsids
(PROT:83),
extract-assembled empty capsids(E.C. 9), self-assem-bled emptycapsids(E.C.5:
%),and 14Sparticles weretested forsusceptibilitytotrypsin. Purifiedor partially purified particles were treated at
20°Cwithtrypsin at a ratio of enzyme to protein of 1:10 (wt/wt). The reaction mixtures were sampled at 0, 10, 20, 40, and 60 min, and the samples wereanalyzedby SDS-PAGE. Figure 1 shows ananalysis ofthe'4C-aminoacid-labeled
particles after trypsin treatment. As this figure
shows,
PRO6
81 andE.C.6-9
were very resistant totryptic cleavage, as judged by the intactness of their polypeptides. However,E.C.5:
%, the product of the self-assembly of 14S particles, and the14S precursor particles themselves were sensitive to trypsin;VP0andVP1 were attacked selectively.Densitometric scanning of a fluorogram such asthatshown in Fig. 1permitted kinetic analyses
ofthe reactions(Fig. 2). The reactions of PRO
1.31 (Fig. 2A)and
E.C1
19
(Fig. 2B)
withtrypsin
showed that both of these empty capsids were relatively insensitive. A limited amount of VPO appeared to be vulnerable. From these data, it wasdifficult to assess whether VP1 or VP3 was attacked.After60 min,75 to80%of the
structur-alpolypeptidesremained intact. Whether a
siz-R6.8
E.C.
PRO
6 o E 6.80 0 0 0 0 0 0
y- N1 v co 0w- N V (
14
S
03 %- CM qt (O
E o o o 0
o cs- q XD
a b C d e f g h i k I m n 0 p q r s t
FIG. 1. Effect oftrypsinonpoliovirus-relatedparticles. 14C-labeled procapsids
(PRO`
1;lanesathrough e),extract-assembled empty capsids (E.C.8 ; lanes fthrough j), 14S particles (lanes k through o), and
self-assembledemptycapsids
(E.C.1-09;
lanespthrough t)werepurified aspreviouslydescribed(17). They werereacted withtrypsin(tolylsulfonyl phenylalanylchloromethylketonetreated)ataratiooftrypsintoproteinof
1:10(wt/wt) for 0, 10, 20, 40,or60minat20°C.Afterthisincubation,a2-to10-foldexcessofsoybeantrypsin
in-hibitorwasaddedtothemixtures,whichwerethensolubilizedby adding2%SDSand100 mMdithiothreitolat
100°C for 3 min. Electrophoresis was on 15% polyacrylamide gels as previously described (17). After
electrophoresisthegelswerefixed,andVPO, VP1,andVP3 werevisualizedbyfluorography.
0
Vp~~~~~~~~~~~~~~~~~~
w-X4..i 4&V. ;: ijjjSoZ!S:d}X^.
VP3 S. ..
J. VIROL.
46b
tm.
on November 10, 2019 by guest
http://jvi.asm.org/
[image:3.491.54.447.392.603.2]POLIOVIRUS EMPTY CAPSID MORPHOGENESIS 795
FIG. 2. Kineticsofcleavageofpolioviralstructural
proteins by trypsin. (A) Purified "4C-labeled
procap-sidsweretreated withtrypsin(tolylsulfonyl
phenyla-lanyl chloromethyl ketone treated) at20°Cfor0, 10,
20, 40, or 60 min and then subjectedtoPAGE in the
presenceofSDS,asdescribed in thelegendtoFig.1.
The percentagesofVP0(0), VP1 (A),and VP3 (0)
that remained uncleaved were determined from a
densitometric scan ofthe developed fluorogram and areexpressed aspercentages of the radioactivity
re-maining.(B) Purified extract-assembled empty capsids (E.C.68). (C) Purified self-assembled empty capsids (E.C.5). (D)Purified 14S particles.
able fraction of these particles were completely trypsin resistant or whether approximately 20%
ofthe polypeptides in any given particle were
accessible to attack was not determined. In contrast,
E.C.1%29
(Fig. 2C) and 14S particles(Fig. 2D) were attacked at much higher initial
rates, and a substantial percentage ofVPO and VP1 molecules were cleaved to small peptides after 10 minof incubation. Cleavage of theVPO in both types ofparticles was essentially com-plete by40 min. There appeared to be a small fraction of VP1 molecules associated with
E.C.5
29which
resisted attack. However, VP3 wasunique in its trypsin resistance in both typesofparticles. These results also suggested that
theVPO in
E.C.5:0
particles was more accessi-ble to trypsin initially than the VPO in 14S particles.Treatment ofall four types ofparticles with chymotrypsin yielded similar results in terms of relative susceptibilities (data not shown).
Sensitivity
of heatedprocapsidstotrypsin treat-ment. Previously, we reported that whenPRO6
81particles were heated at 40 to 46°C, they assumed apI of 5.0 and a density of 1.29g/cm3 (17). Wealso reportedpreliminary results which indicatedthat heatedprocapsids lost a substan-tial amount of their VP0 molecules, yet sedi-mented at 70S orgreater.The change in the pI values of theprocapsids caused by heating is shown in Fig. 3. The reaction was complete within 20 min, and the heated particles focused at pH 5.0. A similar reaction was demonstrable at 37°C (data not shown). The sedimentation coefficient (correct-ed to water at 20°C) of heated procapsids was
determinedby sucroserate-zonalsedimentation
together with differentially labeled, unheated procapsids. Figure 4 shows thatheated procap-sids had asedimentation coefficient whichwas5 to10%greater than that of unheatedprocapsids.
Purified procapsids labeled with 14C were
heated at 43°C for up to 20 min and then ana-lyzed for trypsin sensitivity. The trypsin
sensi-tivity of these procapsids, which was
deter-mined as described in the legend to Fig. 1, was increased by heating. After heating, predomi-nantly VPO was cleaved to smallpeptides,which migrated at the gel front (data not shown; see Fig.2C). VP1 was alsoattackedbut wascleaved only incompletely, generating a highly stable fragment that was only slightly smaller than
CO
O
*-~
£E
0 C 0E
Nv
anode
pH 5.0_
pH
6.8-,
rIFt 4
I
cathode
a
b
c
d
FIG. 3. Effect of heating on the pI of purified
poliovirus procapsids.Procapsids were heated at43°C
for 2, 5, 10, or 20 min. Isoelectric focusing was
performed inaprefocused (200 V, 1 h) agarose gel.
The pH was determined by using an Ingold
micro-probe surface electrode. VOL. 41,1982
on November 10, 2019 by guest
http://jvi.asm.org/
[image:4.491.266.435.406.617.2]796 PUTNAK AND PHILLIPS
70a3I
.2-I
S10 1 ba 1o 155 30
fractions
FIG. 4. Effect of heat on the sedimentation
coeffi-cientofpurifiedpoliovirus procapsids. (A) Unheated
14C-labeledprocapsids (0)weremixed with unheated
3H-labeled procapsids (A)andcentrifuged on a 15 to
30%osucrosegradient in an SW41 rotor at35,000rpm
for 4h at 18°C. (B) 14C-labeledprocapsids (0)were
heated at 43°C for 20 min before being mixed with
unheated3H-labeledprocapsids(A).Themixture was
centrifuged as described above. The sedimentation
coefficients wereestimated by integration.
VP1. The identification of this fragment as a
cleavageproduct of VP1 is described below.
Heated procapsids treated with trypsin were analyzed by rate-zonal centrifugation. Most of the radioactivity sedimented at 80 to 85S, with only a small amount at the top of the gradient
(datanotshown). Therefore, most of the
cleav-age products produced by trypsin treatment must have remained associated with the parti-cles under theseconditions.
Surface-specific radioiodination as aprobe of particle conformation. Since
PRO1.81
and129 were resistant to trypsin, we had no
information aboutthe surfaceconfigurations of
theseparticles. To study the procapsids further, we used a recently introduced surface-specific radioiodination technique (10).
Virus particles,
PRO681,
andE.C.59
were purified, and 1 ,ug of purified material was reactedwith0.1, 1.0, or 10.0 ,g of chloroglyco-luril in the presence of 500 ,uCi of carrier-freeNa125I, as described above. lodinated particles
were separated from free 125I by sucrose rate-zonalcentrifugation. Wefoundthatprocapsids wereiodinated to specific radioactivities of2 x
106,
4 x 106 and5 x 106dpm/,ug with 0.1, 1.0, and 10.0 ,ug of chloroglycoluril, respectively. When theiodinated procapsidpreparationsweresubjectedtoSDS-PAGE andautoradiographed,
labelwasfound almost exclusively inVP1(Fig. 5, lane e).
Wedetermined thataratio of chloroglycoluril to protein of 1:1 was optimal for radioiodina-tion, and this ratiowas used in theiodination of poliovirus and
E.C.':
%. A specific radioactivity of 5.5 x 106dpm/,Lgwasobtained for virus,and a value of 8 x105
dpm/nLg
was obtained forE.C.5:
%.LikePRO681,
poliovirusparticles were also iodinated primarilyin VP1 (Fig. 5, lane d). In contrast,E.C.129
particles were iodinated primarily in VPO, although some impuritieswere also labeled (Fig. 5, lane f).0
0.o
'ttt Ns N N
_
cc_.-VP0
-VP1
._ _ _ 0-VP3
[image:5.491.50.243.58.276.2]a
b c
d
e f
FIG. 5. Analysisof thepolypeptidesof
poliovirus-relatedparticles aftersurface-specificradioiodination.
Afterpurification,theparticles were radioiodinated in
the presence ofchloroglycoluril (1 ,ug/,ug ofprotein)
and 500 ,Ci ofcarrier-free 125I. The reaction was
carried out at 0°C for 10 min. The particles were
reisolated by sucrose rate-zonal centrifugation and
thensubjectedtoPAGEasdescribed in thelegendto
Fig.1.Lanesa,b, andccontained"C-labeledextract
(ext.), poliovirus particles (P.V.), and procapsids
(PRO) forcomparison. Lane d contained 125I-labeled
poliovirus particles (1251P.V.), laneecontained
125i-labeled procapsids (1251 PRO), and lane f contained
125I-labeled self-assembled empty capsids (125I
E.C.5-°). The structural polypeptides VPO, VP1, and
VP3 areindicated byarrows.
J. VIROL.
on November 10, 2019 by guest
http://jvi.asm.org/
[image:5.491.277.420.231.529.2]POLIOVIRUS EMPTY CAPSID MORPHOGENESIS 797
Theradioiodinatedprocapsidsweretested for susceptibilitytotrypsin,aswellaspl.Wefound thatthese procapsidsweretrypsinresistantand focusedatpH6.8(datanotshown). Thus,based on these criteria, the particles had not been denatured. The iodinated procapsidswerethen heated at 43°C for 15 min, after which they focusedatpH 5.0 (datanotshown). The heated procapsids were also treated with trypsin for 20 min at 20°C and then analyzed by SDS-PAGE (Fig. 6). A 32-kilodalton cleavage product of VP1, designated VPla, was observed. In addi-tion, the smaller cleavage products VPlb (17 kilodaltons) and
VPlc
(11 kilodaltons) werealso detected.Antigenic determinants of E.C.1.9 and
E.C.I50
. Based onprevious reportby Mosseretal. (A. S. Mosser,S. W. Hong, J. Icenogle, and R. R. Rueckert, Abstr. Annu. Meet. Am. Soc. Microbiol.1981, T125,p.258),weexaminedthe
antigenic determinants ofprocapsids, aswellas
extract-mediated and self-assembled empty cap-sids. Poliovirions (datanotshown)and procap-sids isolatedfrom virus-infectedHeLacells and
purified in CsCl gradients were N antigenic.
E.C.1 29
formedby incubating14Sparticleswithw 0L 10X 0.
i ^
~~~~~Pro6-1. pr6.
01
a T- EC6.
10,
ANTI-H s
0
5
0 c;
satedPro
1 2 3 4 5 -O
AbConc.(-LogDil'n.)
FIG. 7. Reactivities of procapsids (Pro681),
ex-tract-assembled emptycapsids(E.C.t ),
self-assem-bledempty capsids
(E.C."
°:),and heatedprocapsids(broken line) to specific anti-N and anti-H antisera.
Notethat heatedprocapsids didnot reactwithanti-N
serum at a10-2 dilution (A). Particleswere-obtained
as described in the text. Heated procapsids were
obtained by treatment ofpurifiedprocapsids at43°C
for15min.AbConc.,Antibody concentration;Dil'n, dilution; IMMUNE-PPTE,immuneprecipitate.
ib4_
p.-VP1 *--VP1a-.00h omn
WYalioqgl
.4laop
4-VPlb
-VPlc
a
b
cd
ef
FIG. 6. Effect of trypsin on 1"MI-lat
procapsids. Poliovirus procapsids were
radioiodinated as described inthe legei
Lane a, "M5I-labeled procapsids; lane b
procapsids treated with trypsin; lanec
procapsids heatedat43°C for 20 min and
withtrypsinat20°Cfor 20 min; lane d,r
labeled procapsids; lane e, 14C-labele
treatedwithtrypsin for 20 minat20°Cas
the legend to Fig. 1; lane f, 14C-labele
heated at 430C for 20 min and then
trypsin.Electrophoresiswascarriedout
acrylamide gels asdescribedinthelege
The major trypsin cleavage products
VPla, VPlb,andVPlc.
infected cell extracts were also N antigenic. In contrast, self-assembled empty capsids i
h
-n
VP 0(E.C.1
%) and heated(43°C,
15min)
procapsids-.-vYP were H
antigenic
(Fig.
7). These and other r-VPla? findings are summarizedinTable1.P+ VP3 Evidence for a
conformational
difference
be-tween
self-assembled
empty
capsids(E.C.50
°)and-VP1b mock-infected extract-assembledPreviously(17), wereported that a small number
empty
capsids.of procapsid-like, pl 6.8 empty shells were
VP-
c formed when concentrated 14Sparticles
wereallowed to assemble in the presence of cytoplas-micextract frommock-infectedcells. To charac-terize these empty capsids, we studied their
susceptibility totrypsin.
Self-assembled empty capsids
(E.C.51%29)
andbeled
VPiin
empty
capsids
assembled in the presence ofapurified and mock-infectedextractwereisolated
by
sucrosend to Fig. 5. rate-zonal
centrifugation.
In thisexperiment,
2,521I-labeled
E.C.1os
sedimented at 71S, whereasmock-in-125I-labeled fected extract-assembled empty capsids
sedi-Ithen treated mented at 75S (data not shown). Both types of
*eference
14C- empty capsids were treated with trypsin asde-d
procapsids
scribed in the legend to Fig. 1 (Fig. 8). The sdescribed in resulting gelpatterns
were distinctly different,treated
with
indicating
that a different setofcleavage
sitesin 20%poly- was
accessible
ineachtypeofparticle.
Notable :ndto Fig. 1. were the 60%o conservation of VPO and the of VP1 were almost totalconservation of VP1 inmock-infect-edextract-assembled empty capsids.
VOL.41, 1982
i
on November 10, 2019 by guest
http://jvi.asm.org/
[image:6.491.251.448.76.273.2] [image:6.491.48.242.352.540.2]TABLE 1. Characteristics of empty capsids formed in vitrobyassembly of14Sparticles in the absence (i.e.,
self-assembly) or presence of poliovirus-infected cell extracts
Type ofempty I Density inCsCl a
Ant.g.i.t
Surface Trysincapsids (g/cm3) s ntigemcity polypeptide susceptibility
Self-assembled 5.0 1.29 71 Hb VPO Yes
Extract-assembled 6.8 1.29c 75 N VP1 No
a
s20,W,
Sedimentation coefficient corrected to water at200C.
bSome N determinants may exist.
cProcapsids differed fromextract-assembledempty capsids only in density in
CsCl
(i.e., 1.31g/cm3).
The trypsin-treated particles were also ana-lyzed by sucrose rate-zonal centrifugation. The sedimentation coefficient (corrected to water at
20°C) of trypsin-treated
E.C.5:0
was shiftedfrom 71S to 60S, whereas the sedimentation
coefficient of the mock-infected
extract-assem-bled empty capsid species remained unchanged
aftertrypsinization (data not shown). The ability
of mock-infected or virus-infected extracts to convert E.C.1 particles into trypsin-resistant
structures could not be demonstrated, which was consistent with the inability of extracts to convert E.C. ° into E.C. 9 or PRO 81, as determined by pl analysis (17). These findings strongly suggested that mock-infected extracts contain afactor thatdirectsalimitednumberof
14S particlesinto a similar kind of empty shell
(E.C689) as the putative morphopoietic factor in poliovirus-infected cells.
DISCUSSION
Previously, we reported that the
self-assem-bled emptycapsidsof poliovirustype 1 hada I of5.0 andadensityof 1.29 g/cm3(i.e.,E.C.129) and that the empty capsids assembled in the presenceof infected-cell extracthadapIof6.8
(i.e.
E.C.619)
(17).Procapsids,whichapparentlyexist in virus-infected cells, alsofocused atpH
A
E
c 4
vpo - . _ _
vp1 -- -: -.oI
VP3 -is_ ma_ _ _
6.8, but they had a density of 1.31 g/cm3 (i.e., PRO
13).
At that time we ascribed the differ-ences in pl values todifferences in capsidcon-formation because the two species possessed
polypeptides which were identical in apparent
molecular weight and pI. In thispaper we pre-sent more direct evidence for differences in capsid conformation.
Wefoundthat
PRO68
1 andE.C.69
were bothvery resistant to the proteinases trypsin and
chymotrypsin. Incontrast,
E.C.1:29
wasreadilyattacked by both enzymes, as were the precur-sor 14S particles. The polypeptides of E.C.l12
and 14S particles that were most readily at-tacked were VPO and VP1. We conclude from these results that the configurations of VPO and VP1 are such that extensiveregionsof these
molecules lie exposed at the surfaces ofthese
particles. The configurations of these
polypep-tides must bedifferent in PRO6
31
andE.C.1
29,so astobe resistant toproteolysis.This does not eliminate the possibility thatsignificant regions of thesepolypeptidesare onthe exteriorsofthese
shells, but only means that their secondary,
tertiary, and quaternary conformations make
them inaccessible to the active site of the
en-zyme. VP3 was the only polypeptide that was resistant toproteolyticcleavage in all four
parti-c
0
B
0 0 0 0
v N t CD
-woom
FIG. 8. Comparison of the trypsin sensitivity of self-assembled empty capsids (E.C.1: (B) with the
sensitivity ofcapsidsassembled in the presence ofcytoplasmic extractfrom mock-infected cells (A). After assemblythe emptycapsidswerepurified bysucroserate-zonalcentrifugation,treatedwithtrypsin,and then analyzed byPAGEasdescribed in thelegendtoFig.1.
on November 10, 2019 by guest
http://jvi.asm.org/
POLIOVIRUS EMPTY CAPSID MORPHOGENESIS 799
cles. This finding appears to be moot because
under nocircumstances was VP3 degraded
ex-tensively by trypsin or chymotrypsin (see be-low).
Each structural polypeptide appeared to have
a different reactivity with trypsin, as
demon-strated by the differences in the initialcleavage rates (Fig. 2). Cleavage at one site may make previously hidden sites accessible. Forexample, VP3 in 14Sparticles wascompletely resistantto
cleavage until most of the VP0 and VP1
mole-cules had been digested (Fig. 2D); however, even after 60 min most of this polypeptide remained intact.
Itis not evident whether the limiteddegrees of trypsin sensitivity demonstrated by
PRO68
1and E.C.1.29were
due to asmall number of sensitive sites in every particle or to alarge number of sites in a smallpopulationofparticles. Previous-ly, we reported thatPRO6:81
underwent a con-version to pI 5.0 particles when it was heated (17)(Fig. 3). The heatedprocapsidsalsobecametrypsinsensitive. Although it isconceivablethat
asmall percentage ofprocapsidsare denatured during purification, much of which is carried out at 20°C, 20 to 25% appears to be too large an amountand is inconsistent with thehighenergy
ofactivationobserved (about 450kJ/mol;
previ-ously, weerroneously reported thatthis energy
of activation was 119 kJ/mol) (17). Also, we
never found more than a few percent of the
particles in procapsid preparations with a pI of
5.0.Therefore,wetentatively favor the idea that
there are afewtrypsin-sensitive sites on every
particle and that these sites reside on certain
VPO molecules or VP1 molecules or both. We have reported the results ofpreliminary
experimentswhich indicated that heated
procap-sids lose most of their VPO molecules (17). It now appears that contaminating proteinases could be responsible for this observation.
In-deed, VPOwaspredominatelycleavedin heated
procapsids by trypsin. Heated procapsids also showed an increased sedimentation rate of ap-proximately 80S (Fig. 4). The exact nature of
this heat-induced change is unclear, although
again it most likely reflects a change in capsid
conformation. It is noteworthy that although
E.C.5%0
and heated procapsids were both H antigenic (Fig. 7), they were not identical in susceptibility to trypsin (Fig. 1 and 6). Procap-sids in extracts appeared to be more resistant toheating than purified
PRO6.81
(unpublisheddata), but this may have been due to a protective effect exerted by other proteins or differences in
theionicenvironment.
Aftertrypsin treatment the sedimentation co-efficient of heated procapsids was conserved, indicating that most of the cleavage products remained particle associated, probably held in
place by electrostatic or hydrophobic interac-tions. These bonds should be susceptible tohigh concentrations of salts or to urea and detergents. On theother hand, trypsin treatment of
E.C.29
yielded particles with a sedimentation coeffi-cient of about 60S and a largeamountof radioac-tivity at the top of thegradient.
Since virus particles,
PRO61
, andE.C.69
were resistant to proteolytic cleavage,we were unable to obtain topographical information by this method. Therefore, chloroglycoluril-cata-lyzed radioiodination of exposed tyrosine resi-dues was used to probe the surface configura-tionsof these particles. Both virus particles andPRO68
1 were labeled primarily in VP1. Thehypothesis that VP1 is the predominant surface
polypeptide in most picomavirions has been
supported by thefollowing lines of evidence: (i) the radioiodination of VP1 by the lactoperoxi-dase and chloramine-T method (4, 5, 9), (ii) surfacelabeling ofVP1withaceticanhydride (8) andN-succinimyl proprionate (19), (iii) the tryp-sinsensitivity of VP1 in foot-and-mouth disease virus (3), and (iv) virion neutralization by anti-body to VP1 (9). Coxsackievirus appears to be the only exception to date (1). On the other hand, published data dealing with the topogra-phyofprocapsids is scanty (5). Itwascommonly
believedthat theconformation of polioviral
pro-capsids was different from that of virions since
procapsids reportedly did not attach to
virus-susceptible cells or cross-reactimmunologically with virions (18). Recently however, data
con-flicting with this notion have been obtained
(Mosser et al., Abstr. Annu. Meet. Am. Soc. Microbiol. 1981). Also, procapsids and virus
particleshavealmostidenticalisoelectricpoints
(17), and both types of particles were
radioiodin-ated primarily in VP1 (Fig. 5). Evidently, the
cleavageofVP0 that occurs upon RNA
encapsi-dation may not have the profound effect upon
capsidconformation thatwas oncebelieved.
In contrast tothe otherpolioviral capsid
spe-cies,
E.C.5l0
shells wereradioiodinatedprimari-ly in VPO, a finding thatcomplements the
sus-ceptibility ofthispolypeptidetotrypsin (Fig.1).
The factthat VP1 was also cleaved by trypsin butwas notiodinated may indicate that cleavage of VPO isaprerequisiteto VP1 susceptibilityto trypsin (Fig. 2) or that the exposed portions of VP1 lack tyrosine residues. In any case, the exposed regions of VPO andVP1 aredifferent in self-assembledemptycapsids than in procapsids or
E.C.61
9 particles, and this difference may accountfor their different pI values. Finally, ourexperiments showed that
PRO6-8
andextract-assembledemptycapsids(E.C.1:29)were N anti-genic, whereas self-assembled empty capsids
(E.C.5
:) and heated PRO681
were Hantigenic(Fig.7). Thus, there appears to be a correlation
VOL. 41,1982
on November 10, 2019 by guest
http://jvi.asm.org/
between the pl and the predominant antigenic determinantsof poliovirus empty shells;i.e.,pl 5.0 empty shells are H antigenic, and pl 6.8 emptycapsids areNantigenic. Note that polio-virions have a pI of 6.7 to 6.8 and are N antigenic.
How is the conformation of empty capsids determined? The conformation of E.C.l% is determined by information contained within its constitutent14S particles. These capsids appear to require no accessory factors for assembly. However,
E.C.5:0
particles are not found in vivo, and those formed in vitro by the self-assemblyreaction could not be converted into pl 6.8 procapsid-like structures by incubation inthe presenceofvirus-infected cellextracts (17).
Evidently, additional information in the formof an assembly factor or factors is required for the assembly of PRO681 in vivo and
E.C.689
in vitro.Wefoundthatincubation ofpartially purified 14S particles with mock-infected cell extracts resulted in the assembly of empty capsids, some (about 20%) of which focused at pH 6.8 (17) and were more resistant to trypsin than self-assem-bled empty capsids (Fig. 8). The significance of an activity in uninfected cells which affects empty capsid conformation, at least in vitro, remains obscure. Perhaps this activity fortu-itously mimics avirus-coded morphogenetic fac-tor. Another possibility is that it interacts with the viral proteins which function in capsid as-sembly, analogous to the interaction between the bacterial protein groE and phage p31 in T4 headmorphogenesis (11). However, the fact that this activity was not detectable in HeLa cells infected with poliovirus defective-interfering particles (15) or a temperature-sensitive mutant (C. K. Drescher and B. A. Phillips, Abstr. Annu. Meet. Am. Soc. Microbiol. 1979, S89, p. 254; C. K. Drescher, J. R. Putnak, and B. A. Phillips, submitted for publication) strongly sug-gests that the putative morphopoietic factor is not aderepressed or induced host protein.
ACKNOWLEDGMENTS
Wegratefully acknowledge the excellent technical assist-ance of Mary-LouWong-Chong.
This work was supported by Public Health Service grant Al-08368 to B.A.P. from the National Institutes of Health.
ADDENDUM IN PROOF
Recently (Marongiu et al., J. Virol. 39:341-347,
1981), it has been reported that emptycapsids isolated
from poliovirus-infected cells were dissociated into
14S subunits under mild alkalineconditions (pH 8.5)at
4°C. Empty capsids made in vitro by self-assembly of
isolated 14S particles could not be similarly
dissociat-ed. Performing the same experiment on our own types
of empty shells, we found that PRO`1 and E.C.49
were alkali labile, whereas E.C.51% was stable under
the same conditions. The14S particles produced from
the pH 8.5 dissociation of PRO0 31 at 4°C did
self-assemble when the pH was lowered to 7.4 and the
temperaturewasraisedto37°C;atypical self-assem-bled shell, E.C.5%, wasthe only product of such a reaction.
LITERATURECITED
1. Beatrice, S. T., M. G. Katze, B. A. Zajac, and R. L. Crowell. 1980. Induction ofneutralizing antibodies by the coxsackievirus B3 virion polypeptide, VP2. Virology 104:426-438.
2. Bradford, M. 1976. A rapid and sensitive method for the quantitation ofmicrogramquantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254.
3. Brown, F. 1978. Structure-functionrelationships in the picornaviruses, p. 49-72. In R. Perez-Bercoff (ed.), The molecular biologyofpicornaviruses. NATO Advanced StudyInstituteSeries,vol. 23. Plenum Press, New York. 4. Carthew, P. 1976. The surface nature ofproteins ofa bovine enterovirus, before and after neutralization. J. Gen. Virol. 32:17-23.
5. Carthew,P., and S. J. Martin. 1974. The iodination of bovineenterovirus particles.J.Gen. Virol. 24:525-534. 6. Fraker,P.J., and J. C.Speck, Jr. 1978.Protein and cell
membrane iodination with a sparingly soluble chloroa-mide, 1,3,4,6-tetrachloro-3a,6a-diphenylglycoluril. Bio-chem.Biophys. Res. Commun. 80:849-857.
7. King, A. M. Q., and J. W. F. Newman. 1980. Tempera-ture-sensitive mutants of foot-and-mouth disease virus withaltered structuralpolypeptides. I. Identification by electrofocusing. J. Virol. 34:59-66.
8. Lonberg-Holm,K., and B. E. Butterworth. 1976. Investi-gationof the structureof polio-andhuman-rhinovirions throughtheuseofselective chemicalreactivity.Virology 71:207-216.
9. Lund, G.A., B. R. Ziola, A. Salmi,andD. G. Scraba. 1977. Structure of themengovirion.V.Distribution of the capsidpolypeptides with respect tothe surface ofthe virusparticle.Virology78:35-44.
10. Markwell, M. K., and C. F. Fox. 1978. Surface-specific iodination of viruses and eucaryoticcellsusing 1,3,4,6-tetrachloro-3a,6a-diphenyl-glycoluril. Biochemistry 17:4807-4817.
11. Murialdo,H., andA.Becker.1978. Headmorphogenesis ofcomplexdouble-stranded deoxyribonucleic acid bac-teriophages. Microbiol.Rev.42:529-576.
12. Perlin,M., and B.A.Phillips. 1975. In vitroassemblyof polioviruses. IV. Evidence for the existence oftwo as-semblysteps in theformation of emptycapsidsfrom 14S particles.Virology 63:505-511.
13. Phillips,B.A.1969. In vitroassembly of polioviruses.I. Kinetics of theassemblyofemptycapsidsand the role of extractsfrom infectedcells.Virology39:811-821. 14. PhUUlips,B. A.1971. In vitroassemblyofpolioviruses.II.
Evidencefor the selfassemblyof 14Sparticlesinto empty capsids.Virology44:307-316.
15. Phillips,B.A., R. E. Lundquist,and J. V. Maizel, Jr. 1980.Absenceof subviralparticlesandassembly activity in HeLacells infected withdefective-interfering particles ofpoliovirus. Virology100:116-124.
16. Phillips,B.A.,D. F.Summers,andJ.V.Maizel,Jr. 1968. Invitroassemblyofpoliovirus-relatedparticles.Virology 35:216-226.
17. Putnak, J. R., and B. A. Phillips. 1981. Differences between poliovirusempty capsids formed in vivo and thoseformed in vitro:arole for themorphopoieticfactor. J.Virol. 40:173-183.
18. Rueckert,R. R.1976.On thestructureandmorphogenesis ofpicornaviruses,p. 131-200. In H. Fraenkel-Conrat and R. R. Wagner (ed.), Comprehensive virology, vol. 6. PlenumPress,New York.
19. Wetz, K., and K. 0. Habermeli. 1979. Topographical studiesonpolioviruscapsidproteins bychemical modifi-cation andcross-linking with bi-functional reagents. J. Gen.Virol. 44:525-534.
on November 10, 2019 by guest
http://jvi.asm.org/