Dodecyl Sulfate-Derived Polypeptides of Polyoma Virionst JOSEPH B

JOURNAL OF VIROLOGY, Apr. 1980, p. 119-129 Vol. 34, No. 1 0022-538X/80/04-0119/11$02.00/0

Separation of Neutralizing and Hemagglutination-Inhibiting Antibody Activities and Specificity of Antisera to Sodium Dodecyl Sulfate-Derived Polypeptides of Polyoma Virionst JOSEPH B. BOLEN AND RICHARD A. CONSIGLI* Section of Virology and Oncology, Division ofBiology, Kansas State University, Manhattan, Kansas 66506 Antisera to the sodium dodecyl sulfate (SDS)-polyacrylamide gel-derived polyoma virion polypeptides were used in immunoprecipitation experiments with ethylene glycol-bis-N,N'-tetraacetic acid (EGTA)-dissociated polyoma virions and capsids to determine the specificity of the antipolyoma polypeptide sera. Additionally, a technique for applying "2I-labeled immunoglobulins to SDS- polyacrylamide gels was used to explore the antigenic specificities of the antisera. The results demonstrated that antisera directed against the SDS-gel-derived VP1, VP2, and VP3 did not react with native polyoma proteins, but would react with the appropriate antigens on denatured polyoma proteins. Antisera against the histone region of such gels reacted with native and denatured polyoma VP1. Separation of neutralizing antibodies from hemagglutination inhibition (HAI) antibodies to polyoma in antisera directed against the histone region of polyacrylamide gels was done by using a polyoma capsid affinity column. The antibodies eluted from this column which did not react with capsids possessed only neutralizing activity, whereas antibodies which bound to capsids possessed only HAI activity. These isolated immunoglobulin G fractions were then used in immunoprecipitation experiments to demonstrate that the antigenic determinants responsible for the HAI activity of the serum were contained on a 16,000-dalton polypeptide, whereas those antigenic determinants responsible for neutralizing activity were contained on a 14,000-dalton polypeptide. Both ofthese polypeptides present in the histone region of the SDS-gels appeared to be derived from the major virion protein VP1. In a previous report from this laboratory (11), with polyoma virions for specific binding sites antisera prepared from sodium dodecyl sulfate on the surface of mouse kidney cells (MKC) and (SDS)-polyacrylamide gel electrophoresis were unable to inhibit virion infection of MKC. (PAGE) of polyoma virion polypeptides was Capsids could, however, block virion adsorption used to demonstrate the cytoplasmic synthesis to guinea pig erythrocytes (GPRBC). Virion ad- of polyoma VP1, VP2, and VP3 and subsequent sorption to and infection of MKC was found to transport of these proteins to the nucleus of be independent of the ability of virions to agglu- infected cells. Additionally, anti-VP1, anti-VP2, tinate GPRBC (1). Since capsids lack the recep- and anti-VP3 sera did not form precipitin bands tor(s) necessary for infection but possess the with virions during immunodiffusion assays, nor moieties necessary for hemagglutination (HA), did these antisera demonstrate the capacity to we have been able to use capsids as an affinity cause hemagglutination inhibition (HAI). These matrix to separate the HAI antibodies from the antisera possessed little virus-neutralizing activ- neutralizing antibodies present in the anti- ity. Antiserum produced against the histone re- histone sera. gion of the SDS-gels, however, demonstrated This investigation was therefore directed to cytoplasmic and nuclear immunofluorescence in provide a more detailed study of the antibody polyoma-infected cells, was capable of neutral- specificities to the SDS-derived polyoma virion izing virus infectivity and inhibiting hemagglu- polypeptides and to gain insight into the anti- tination, and formed precipitin bands with viri- genic relationship between the proteins of poly- ons in immunodiffusion assays. oma virions and capsids. The data presented in A recent report from our laboratory demon- this report indicate that little, if any, antibody is strated that polyoma capsids did not compete directed against the polyoma histones. Rather, the major antigen(s) in the histone region of the t Contribution no. 80-123-J from the Kansas Agricultural gels appears to be products derived from the Experiment Station, Kansas State University. major virion protein VP1. Furthermore, this re- 119 120 BOLEN AND CONSIGLI J. VIROL. port demonstrates that HAI antibodies can be CsCl gradients and run on SDS-polyacrylamide gels separated from neutralizing antibodies in the were used as the antigen source as described previ- antihistone serum. This ability to separate these ously (11). Antisera were produced in New Zealand activities allows, for the first time, the identifi- white rabbits, and inmmunoglobulin G (IgG) was puri- of the of fied from whole serum as described previously (11). cation polypeptides polyoma virions Preparation of .25I-labeled IgG. The IgG frac- responsible for hemagglutinin and receptor functions of the various antisera was labeled with 125I by tions. the Enzymobead method (product information bulletin 1060, Bio-Rad Laboratories). Unincorporated io- MATERIALS AND METHODS dine was removed on Sephadex G-25 column chro- Cell and virus propagation. Preparation of pri- matography in 0.01 M phosphate buffer, pH 7.2. The mary cultures of MKC has been described (12). Wild- specific activities of '25I-labeled IgG fractions were 1 type small-plaque polyoma virus was used to infect x 105 to 4 x 105 cpm per,g of protein. cells at a multiplicity of infection of 10. Infected cul- Preparation of EGTA-DTT- and iodoaceta- tures were maintained in serum-free Dulbecco-modi- mide-treated "WI-labeled virion and capsid poly- fied Eagle medium (10). peptides. "nI-labeled virions and capsids adjusted to Virus purification. Virus was purified from the the same protein concentration and HA titer were infected-cell lysate as previously described (10). CsCl dissociated as described by Brady et al. (2, 3), using gradients used to purify the virus were prepared as ethylene glycol-bis-N,N'-tetraacetic acid (EGTA) and described by Brunck and Leick (4) and described in dithiothreitol (DTT), into DNA-protein complex and greater detail previously (2, 3). Empty capsids were capsomere subunits. Iodoacetamide-treated 125I-la- purified from infected-cell lysates by isolating the light beled virions and capsids were prepared by modifying virus band from cesium chloride gradients (buoyant the procedure of Kasamatsu and Flory (9). Briefly, density, 1.28 g/cm3). The capsid band was collected, "26I-labeled virions and capsids were treated with 5% diluted with Tris buffer (0.01 M, pH 7.4) to a density SDS, 10 mM Tris (pH 7.4), 5 mM EGTA, and 5 mM of 1.20 g/cm3, layered onto a four-step CsCl velocity DTT for 45 min at room temperature. The dissociated gradient, and centrifuged at 35,000 rpm for 3 h as products were sedimented through a 10 to 30% linear described previously (2). The isolated capsids were sucrose gradient (10 mM Tris, pH 7.4-5 mM EGTA- dialyzed against Tris buffer (0.01 M, pH 7.4) for 18 to 0.15 M NaCl) for 40 min at 40,000 rpm (Beckman SW- 24 h at 4°C, and then the capsids were layered onto a 50.1). The polypeptides were pooled and dialyzed seven-step CsCl velocity gradient consisting of 0.6 ml against 0.5% SDS, 10 mM Tris (pH 8.0), and 1 mM each of 1.294, 1.283, 1.272, 1.266, 1.261, 1.256, and 1.252 EGTA. DTT was added to 10 mM, the proteins were of CsCl per cm3. The gradients were centrifuged at incubated at 450C for 30 min, and iodoacetamide 35,000 rpm for 3 h. The capsid band was collected, (Sigma Chemical Co.) was added to a final concentra- diluted with Tris buffer (0.01 M, pH 7.4) to a density tion of 14 mM. The reaction was allowed to proceed in of 1.20 g/cm3, and layered onto a five-step CsCl veloc- the dark for 45 min at room temperature. The treated ity gradient consisting of0.5 ml each of 1.294 and 1.266 polypeptides were dialyzed against 0.5% Triton X-100, g of CsCl per cm3 and 0.8 ml each of 1.261, 1.254, and 10 mM Tris (pH 7.8), and 5 mM EGTA overnight. 1.252 of CsCl per cm3. The gradients were centrifuged Immunoprecipitation. EGTA-DTT- or iodoacet- at 35,000 rpm for 2.5 h. The capsid band was collected amide-treated 125I-labeled virions and capsids (50 pi) and centrifuged on a second five-step CsCl gradient as were incubated with 10 pl of the various antipolyoma described above. polypeptide IgG for 90 min at room temperature. Goat The capsid band was isolated and dialyzed at 4°C (IgG fraction) anti-rabbit IgG (Miles Laboratories, for 18 to 24 h against Tris buffer (0.01 M, pH 7.4) to Inc.) was added (50 pl), and the preparation was al- remove residual CsCl. The above scheme for purifying lowed to react for 90 min at room temperature. The empty capsids was determined to be necessary to precipitate was pelleted by centrifugation, washed ensure the removal of residual pseudovirions. Capsid three times with buffer (10 mM Tris, pH 7.8-0.25% preparations which did not undergo the extensive pu- Triton X-100-5 mM EGTA-0.15 M NaCl), and elec- rification steps above were found to be free of histones trophoresed on 15% SDS-polyacrylamide tube gels as when SDS-polyacrylamide gels were stained with Coo- described previously (11). The gels were cut into 2- or massie brilliant blue. However, when such prepara- 4-mm slices, and radioactivity was determined by liq- tions were subsequently labeled with "I and rerun on uid scintillation spectroscopy. similar gels, the developed autoradiograms demon- 1251-labeled IgG staining of polyoma polypep- strated the presence of histones. All capsid prepara- tides in SDS-polyacrylamide slab gels. Unlabeled tions were therefore labeled with "I and analyzed on polyoma virions (500 pl) were disrupted in 2% SDS, SDS-polyacrylamide gels before being used for exper- 5% 2-mercaptoethanol, 10% glycerol, 10 mM Tris (pH iments. Virions and capsids were also analyzed by 6.8), and 0.001% bromophenol blue by boiling for 5 electron microscopy and for plaque production to en- min. The sample was applied to an SDS-15% poly- sure purity and biological activity. acrylamide gel (100 by 140 by 0.75 mm) prepared as Preparation of radioactively labeled polyoma previously described (3). The samples were electro- virions. Purified polyoma virions or capsids were phoresed at 20 mA until the tracking dye began to labeled in vitro with "I by the method of Frost and elute from the gel. After electrophoresis, the gels were Bourgaux (7). prepared by the method of Kasamatsu and Flory (9) Antiserum production and virion antigen for antibody staining. Before addition of the labeled preparation. Polyoma virions isolated from shallow antibodies, however, the gels were sliced vertically into VOL. 34, 1980 POLYPEPTIDES OF POLYOMA VIRIONS 121 replicate 1-cm strips. Duplicate gel strips were then say, using bovine serum albumin as a standard. Radio- incubated with 1 mg of l25I-labeled antibodies in 15 ml activity was quantitated in a toluene-Triton (3:1) scin- of BBS buffer (0.1 M H3BO3-0.025 M Na2B40r-0.15 tillation fluid with a Beckman LS-233 liquid scintilla- M NaCl, pH 8.0) containing 0.01% NaN3 and 1 mg of tion counter. bovine serum albumin per ml for 24 h at room temperature. The unbound antibodies were removed by RESULTS repeated washes with 10% BBS, 0.15 M NaCl, 0.01% Immunoprecipitation of EGTA-DTT-dis- NaN3, and 0.5% Triton X-100 over a 4- to 5-day period, with changes every 12 h. The gels were then cut into sociated virions and capsids. Antisera di- 2-mm slices, and the counts per slice was determined rected against the three structural proteins VP1, by liquid scintillation spectroscopy. VP2, and VP3 did not react with intact virions FA. The indirect method of fluorescent antibody in immunodiffusion assays (11). However, we (FA) assay was employed, using fluorescein-conju- were interested in determining whether these gated goat anti-rabbit globulin (Microbiological Asso- antisera would react with EGTA-DTT-dissoci- ciates). ated polyoma proteins. Under these nondena- HAI. Dilutions of IgG were mixed with either 2 or turing conditions, polyoma virions are dissoci- 10 hemagglutinating units (HAU) of polyoma virions ated into capsomere subunits and DNA-protein or capsids and incubated at room temperature for 1 h. thereby exposing additional regions GPRBC (0.75%) were added, and the HA pattern was complex (3), allowed to develop at 40C. on the virus proteins which might be recognized Virion neutralization assays. (i) Plaque reduc- by the anti-structural protein sera. Such an ap- tion. Polyoma virions were diluted to contain 25 PFU proach did not prove to be fruitful since anti- and mixed with an equal volume of IgG dilutions. The VP1, anti-VP2, and anti-VP3 failed to recognize reaction was incubated for 60 min at 37°C, and the antigenic sites on nondenatured polyoma virion virion-IgG mixture was plaqued on primary mouse or capsid proteins (data not shown). However, embryo cultures (5). when the same experiment was performed with (ii) FA. Polyoma virions were mixed with an equal the antibodies prepared against the histone re- volume of IgG dilutions. The reaction was incubated gion of the gel, two major protein species were for 60 min at 370C, and the virion-IgG mixture was added to MKC. After a 2-h absorption period at 370C, precipitated (Fig. 1A). The main protein peak the cells were washed, medium was added, and the migrated in the region where histones would be infection was allowed to proceed for 48 h. The extent expected, and the smaller peak migrated at the of infection was then determined by an immunoflu- site where VP1 was found on parallel gels. The orescent assay as described above. This assay was addition of antivirion antibodies to the EGTA- quantitated by counting multiple fields of cells and DTT-dissociated virions yielded a similar pro- scoring for positive fluorescent nuclei. tein pattern except that more VP1 and fewer Affinity chromatography with polyoma cap- histone-size proteins were precipitated when sids. Polyoma capsids were purified as described compared with antihistone sera (Fig. 1A). Dur- above except that all buffers were made with morpho- was that the line-propanesulfonic acid (MOPS), pH 7.0 (Sigma), ing these experiments it found rather than Tris-hydrochloride. The capsids (6 x 107 amount of VP1 precipitated by the antihistone to 8 x 107 HAU/ml) were coupled to Affi-Gel 10 antibodies decreased as a function of the age of affinity chromatography gel (Bio-Rad Laboratories) the purified virus preparation and the length of as described in Affi-Gel 10 product bulletin. After the time the virus was allowed to incubate in the capsid-gel coupling, the gel was transferred to a plastic EGTA-DTT dissociation mix (data not shown). column (1 by 4 cm) and washed extensively with 0.1 However, the amount ofantihistone precipitated M MOPS (pH 7.0). The packed column volume was material migrating in the histone region of the 2.0 ml. To the equilibrated column was added 1.0 ml gel increased when virions were incubated under of the anti-histone IgG (100 ug/ml). The IgG was similar conditions. The VP1 did not appear to allowed to react for 1 h at room temperature with the flow valves closed. Ten-drop fractions (0.6 ml/10 be a contaminant since repeated series of wash- drops) were collected until no additional protein was ings of the immunoprecipitant failed to decrease detected (usually about 20 fractions). The flow valves the amount of VP1 present. Also, experiments were again closed, 5 M MgCl2 in 0.1 M MOPS (pH in which a large excess of normal rabbit IgG was 7.0) was added, and the contents were allowed to added to the dissociated virions or capsids failed incubate at room temperature for 1 h. The flow was to trap or coprecipitate any VP1 in the pellet then resumed, and an additional 20 fractions were after the addition of the goat anti-rabbit serum. collected. The fractions which contained MgCl2 were When polyoma virions were dissociated with dialyzed against 0.1 M MOPS (pH 7.0) before biolog- EGTA-DTT for 30 min at room temperature ical activities and protein were determined. a small amount of Ca2" (10-3 to 10-4 was Quantitative assays. Cesium chloride densities and M) were determined for the refractive index with a Bausch added to the dissociation mix, the addition of & Lomb refractometer and calculated by using the anti-histone or antivirion antibodies precipitated equation of Vinograd and Hearst (13). Protein concen- mainly virion VP1 (Fig. 1B). Antivirion IgG also trations were determined by the Bio-Rad protein as- immunoprecipitated some VP2 and VP3. Cal- 122 BOLEN AND CONSIGLI J. VIROL. VPI VP2 VP3 aration or the length of incubation in the EGTA- 40 - A J 4 ; 4ISTONESJ DTT dissociation mix did not have an apprecia- ble effect on the amount VP1 precipitated. Ad- ditionally, capsids are devoid of histones and yet 30 - antibodies directed against the virion histone region on polyacrylamide gels can react with capsids to cause HAI and precipitate VP1 from 20 - EGTA-DTT-dissociated capsids. Thus, it seemed probably that additional polypeptides present in the histone region of SDS-gels were N l0- responsible for the observed reactivity of the 0 antihistone antisera and that the appearance of x these polypeptides was related to the protease activity found in polyoma virions but not capsids _ I I _ (Bolen and Consigli, manuscript in preparation). 0 10 20 30 Immunoprecipitation of iodoacetamide- H SDS-treated virions and capsids. The anti- In 40 - a sera to the denatured (SDS-PAGE) structural proteins failed to immuno-precipitate nondenatured virion or capsid proteins. However, these 30- antisera demonstrated strongly positive cytoplasmic and nuclear fluorescence in polyoma- infected cells (11). To determine the specificities of these antisera directed against the structural 20 protein regions on SDS-gels, we sought to dis- rupt the virions and capsids under mild dena- turing conditions which would favor protein rec- 10 ognition by the antisera. When capsids were treated in this manner, anti-VPl immunoprecipitated only VP1, anti-VP2 precipitated VP2, and , anti-VP3 precipitated VP3 (Fig. 2A). When io- 10 20 30 doacetamide-SDS-treated "25I-labeled capsid polypeptides were reacted with antivirion IgG, SLICE NUM BER all three capsid proteins were immunoprecipi- FIG. 1. JImmunoprecipitation of EGTA-DTT-dis- tated (Fig. 2B). Addition of anti-histone IgG sociated U"5-labeled virions (A) "2'1-labeled virions immunoprecipitated only VP1 (Fig. 2B). When were dissociated for 1 h at room temperature with 10 the same experiment was repeated with labeled mM EGTW 4 and 3 mM DTT (pH 7.5), and 0.1 mM virions, somewhat different results were ob- bovine sernum albumin was added to react with any tamed. The reactivities of anti-VP2 and anti- excess DTT; polypeptides were immunoprecipitated t ame a s of drted apils with antih'istone (0) or antivirion (0) IgG and elec- VP3 were the same as when dirupted capsids trophoresed as described in Materials and Methods. were used (Fig. 20). The addition of anti-VP1 (B) 125I-lal5eled virions were dissociated for 30 min resulted in the immunoprecipitation of three under the (conditions described above, l0o- M CaC,2- polypeptides species (Fig. 20). The major peak 0.1 mM bovine serum albumin was added, and sam- (VP1) had a calculated size of 44,000 daltons pies were treated as above. The positions of VP1, (44K) when compared with parallel gels contain- VP2, VP3, and histones in a parallel gel are indi- ing standard molecular weight markers. The cated. Symtbols are as for (A). second peak had an apparent size of 16K, whereas the third peak's size was approximnately cium was added to the dissociation products in 14K. Antihistone IgG also immunoprecipitated order to Iprevent the protease activity found in the 44K, 16K, and 14K proteins, as well as an associatioon with purified polyoma virions (see intermediate species with an approximate size of below). 29K (Fig. 2D). Addition of antivirion IgG im- Repeatiing the entire sequence of experiments munoprecipitated all three structural proteins, with pol)roma capsids revealed that, as with as well as a broad peak of material migrating in virions, alntiserum to the three structural virus the histone region, and a 29K species. proteins failed to react with dissociated capsids, Effect of calcium chelation on the SDS- and antihistone and antivirion sera immunopre- PAGE of polyoma polypeptides. In the im- cipitated only capsid VP1 (data not shown). munoprecipitation experiments which used Furthermlore, the age ofthe purified capsid prep- EGTA-DTT-derived viral proteins, we observed VOL. 34, 1980 POLYPEPTIDES OF POLYOMA VIRIONS 123

VM VPIVP that the amount of both large and small poly- * 4 + peptide species precipitated depended upon the conditions of virion disruption. In the experiments which used SDS-iodoacetamide-derived viral proteins, an additional polypeptide species migrating between VP2 and VP3 was observed. Since Ca2" has been shown to play an important role in the stability of polyoma virions (2) and 4 capsids (Bolen, unpublished data), experiments were designed to determine whether Ca2" chelation of virus particles before SDS-PAGE 2 would affect the number ofvirus polypeptides in I -IM-- the SDS-gels. Indeed, alterations in the poly- 40 50 peptide electrophoretic pattern were detected if labeled virions were allowed to incubate in the presence of 5 mM EGTA for 18 h at 370C (Fig. 3). Ofparticular interest (Fig. 3B) was the broad- ening of the VP1 region of the gel and the appearance of an additional species of polypeptide migrating between VP2 and VP3. Such alterations were not observed if the virions were incubated in the presence of 10-3 M Ca2" (Fig. 30). Some degree ofalteration was also observed in virions which were incubated in Tris buffer alone under the same conditions (Fig. 3A). Sim- 40 with failed to the 10 20 30 -5 ilar experiments capsids show SLICE NUMBER appearance of any additional polypeptide species (data not shown). Staining of polyoma polypeptides with "2'I-labeled IgG in SDS-polyacrylamide gels. The preceding experiments suggested that the histone region of SDS-polyacrylamide gels VPI VPpVP3 contained polypeptide fragments probably of 10o-c I.uSiT VP1 origin which resulted in the production of rabbit antibodies against the VP1 fragments. Furthermore, the generation of such fragments and others (e.g., 29K species) seemed to depend upon the ability of certain compounds such as EGTA and SDS to chelate virion-associated calcium. a To determine the ability of the various antisera to recognize the polyoma polypeptides under conditions which would favor the generation of any additional polypeptide subspecies, poly- QE1 q atoo io 40 0 oma virions were allowed to incubate in the D presence of EGTA overnight. The unlabeled H 4 - FIG. 2. Immunoprecipitaton ofSDS-mercaptoeth- 6- anol-iodoacetamide-dissociated l251-labeled capsids 12-2 and virions. (A) "LI-labeled capsids were dissociated, immunoprecipitated, and subjected to SDS-PAGE as

described in Materials and Methods. Symbols: @, Anti- VP1 IgG; 0, anti- VP2 IgG; X anti- VP3 IgG. (B) As described in (A). Symbols: *, antihistone IgG; 0, t antivirion IgG. (C) 1251-labeled virions were dissociated, immunoprecipitated, and subjected to SDS- PAGE as described in Materials and Methods. Sym- bols are as in (A). (D) As described in (A). Symbols SUC2 30 40 so are as in (B). The positions of VP1, VP2, VP3, and SLICE NtJMIER histones in a parallel gel are indicated. 124 BOLEN AND CONSIGLI J. VIROL. VP' ily with the VP3 region of the gel and to lesser DIMER VP2 VP3 extent with the VP2 and lOOK 50 - regions (Fig. 40). fA 4 4 I HISTONES Addition of 1251-labeled antivirion IgG reacted 30- with the lOOK region, dimer, and VP1 (Fig. 4D). 10 .S i These antibodies also showed broad reactivity 8- I9| to the VP2-29K-VP3 region and the histone 6- 4 4- \ VP, 2- . ~~A A DIMER HISTOffSI 1a. tE UNumd 8- || 2 K 10 20 30 40 50 16- cq 30 A '~~~~~~~~4- Ail - A21 X 10, 12- 2 8- 29K 10 lb 20 30 40 50 g- 6 - 18 H 4 - 4 0 - cm 2 -sl 20 10 0 30 40 0 12 - 10 10 20 30 40 50 <~~~~~~~~CiN20: c 29K 0 20- 29K

rAA,Amr% ~~0 16-14- H 12- N 10- ]0 20 0 40 5 lb 20 30 SLICE NUMBER 40 so FIG. 3. Effect of calcium chelation on the migra- tion ofpolyoma polypeptides during SDS-PAGE. (A) '251-labeled virions were incubated in 0.01 M Tris (pH 7.5)-0.15 M NaCl at 37°C for 18 h and then subjected to SDS-PAGE. (B) '251-labeled virions were incubated in 0.01 M Tris (pH 7.5)-0.15 M NaCl-5 mM EGTA at 37°C for 18 h and then subjected to SDS-PAGE. (C) 1251-labeled virions incubated in 0.01 M Tris (pH 7.5)- 0.15 M NaCI-0.001 M CaCl2 at 37°C for 18 h and then electrophoresed as above. The positions of dimer, VP1, VP2, 29K region, VP3, and histones in aparallel tube gel are indicated. virions were then disrupted by boiling in SDS and 2-mercaptoethanol and electrophoresed on SDS-polyacrylamide slab gels. When these gels were stained with '25I-labeled IgG directed SLICE NUMBER against VP1, six reactive polypeptides were ob- FIG. 4. Staining ofpolyomapolypeptides with 12I- served (Fig. 4A). The largest reactive species labeled IgG after SDS-PAGE. Unlabeled polyoma had a size of about lOOK. The remaining reactive virions were incubated in 0.01 M Tris (pH 7.5)-0.15 polypeptides corresponded to VP1 dimer (88K), M NaCl-5 mM EGTA at 37C for 18 h and then VP1 (44K), the 29K and the two subjected to SDS-PAGE. The gels were prepared for species, species staining with '251-labeled IgG as described in Mate- which migrated in the histone region. 125I-labeled rials and Methods. (A) Anti- VPI IgG. (B) Anti-VP2 anti-VP2 IgG reacted as expected with the VP2 IgG. (C) Anti- VP3 IgG. (D) Antivirion IgG. (E) Anti- region of the gel and to a lesser degree the VP3 histone IgG. The positions of dimer, VPI, VP2, 29K region of the gel (Fig. 4B), as well as the lOOK region, VP3, and histones in a parallel slab gel are region. '251-labeled anti-VP3 IgG reacted primar- indicated. VOL. 34, 1980 POLYPEPTIDES OF POLYOMA VIRIONS 125

- region of the gel. When labeled antihistone IgG 20 A DIMER VPI VP2 VP3 was added to the gels, the same pattern of reac- 18- 4 4 IHISTNESI 4~29K tivity obtained with anti-VP1 was observed (Fig. 16- 4E). However, the main reactive species were in the histone region of the gel rather than in the 14 - VP1 region. 12- AA To determine the degree of cross-reactivity A~ between the anti-VP1 and antihistone antisera, 2018 10 20 30 40 50 experiments were designed to determine CM0 whether unlabeled anti-VP1 IgG could block the 18- 29K reactivity of labeled anti-histone IgG and vice 16 - versa. In the control experiments, we found that 14 unlabeled anti-VP1 IgG would inhibit all the H 12 _ reactivity of labeled anti-VP1 IgG. Similar re- N Io - RNMW"MU _ sults were obtained with unlabeled and labeled 10 20 30 40 50 antihistone IgG (data not shown). When excess 20 - 29K unlabeled antihistone IgG was added to the gels 18- I before "25I-labeled anti-VP1, the labeled regions 16 - of the gel corresponded to only the 100K, dimer, - VP1, and 29K regions (Fig. 5A). No reactivity 14 was observed in the histone region. The addition 12 - of unlabeled anti-VP1 IgG before the labeled antihistone IgG (Fig. 5B) resulted in the block- 20 30 40 ing of all the antihistone reactivity in the first 20 SLICE NUMBER fractions of the gel. Only a small amount of FIG. 5. Inhibition of staining ofpolyoma polypep- reactivity was observed in the 29K and histone tides with '2I-labeled IgG by unlabeled IgG. Unla- regions. As an additional control, unlabeled anti- beled polyoma virions were treated and electropho- VP2 IgG or anti-VP3 IgG was added to the gels resed as described in the legend of Fig. 4. (A) Unla- before either labeled anti-VP1 IgG or anti- beled antihistone IgG was then added and allowed such (in this exam- to incubate for 24 h, excess antibodies were removed histone IgG. In experiments by extensive washing, and the bound antibodies were ple, unlabeled anti-VP2-l'I-labeled anti-VP1), fixed with 0.1% glutaraldehyde. 25I-labeled anti-VP1 no decrease in the immunological activity of the IgG was then added, and the experiment was carried labeled antibodies was observed (Fig. 50). These out as described in Materials and Methods. (B) Same results showed that no VP1- or VP1 fragment- experiment as in (A) except unlabeled anti-VP1 IgG specific antibodies were present in the anti-VP2 was added before 125I-labeled anti-histone IgG. (C) or anti-VP3 serum. Same experiment as in (A) except that unlabeled anti- These experiments demonstrated that the an- VP2 IgG was added before 1251-labeled anti-VP1 IgG. tibodies present in the antihistone IgG could The positions of dimer, VP1, VP2, 29K region, VP3, block all the antigenic sites that anti-VP1 could and histones in a parallel slab gel are indicated. recognize on the fragments of VP1 which migrated in the histone region. However, anti-VP1 temperature for lh. GPRBC were added, and reacted with additional antigenic determinants the amount of HAI was determined. It appeared present in the 100K, 88K, 44K, and 29K regions. that the antihistone IgG had an equal ability to On the other hand, the fact that unlabeled anti- inhibit the HA activity of capsids and virions VP1 could block almost all the reactivity of the (Fig. 6A, B). antihistone antibodies suggests either that the In addition, the various dilutions of the IgG number of unique antigenic determinants pres- were then incubated with capsids (10 HAU) for ent on the polypeptides in the histone region is 1 h before the addition of virions (10 HAU). The small or that the anti-VPl antibodies stearically virions were also allowed to incubate for 1 h inhibit the antihistone antibodies from stably before the GPRBC were added. By preincubat- binding to their antigen-combining sites. ing the antihistone IgG with capsids, HAI anti- HAI and neutralization of virions and bodies were blocked so that a higher concentra- capsids by antihistone serum. We were in- tion of IgG was required to produce an equiva- terested in determining whether the antihistone lent degree of HAI (Fig. 6C). When the same IgG would have an equivalent inhibitor effect on experiment was repeated except that virions HA by virions and capsids. Virions and capsids rather than capsids were used to block the HAI were adjusted to the same concentration with antibodies, the same results were obtained (data respect to HA (10 HAU) and incubated with not shown). These serum-blocking experiments various dilutions of antihistone IgG at room were then repeated, except that after the addi- 126 BOLEN AND CONSIGLI J. VIROL.

4X100 - A 4.100 B peak (fractions 27 through 31) which had HAI F80- }80- activity, but no apparent neutralizing activity. a. H60-z / 60- These results were in agreement with the serum- U. blocking experiments and demonstrated that, 0 40 - 0 40- indeed, HAI antibodies were distinct from neu- z tj 20- 0 tralizing antibodies. Similar results were ob- w w tained when antivirion IgG was passed through 820-10X0-2 o0S' o61,0IX&4121 1cId this column (data not shown). IgG DILUTION Immunoprecipitation of EGTA-DTT-dis- 4100- *. ,OO D sociated "2'I-labeled virions and capsids. C ~~~~~~~~~~~~U. 0L Previously it was shown that the antihistone o80- '80- IgG could immunoprecipitate two z IL polypeptides H 606 I 60a which migrated in the histone region of SDS- 0 U. ~40- 40- gels if polyoma virions were dissociated under z~~~~~~~ conditions which chelated the virion-associated 0 20- w20- calcium. Under similar dissociation conditions with polyoma capsids, the antihistone antibodies 00 I04 I06 1O 10 la I I0 O* could precipitate only 44K VP1. It was also of IgG DILUTION interest to determine what polypeptides would FIG. 6. HAI and plaque reduction by antihistone be immunoprecipitated from dissociated capsids IgG. (A) Polyoma virions were adjusted to 10 HA U and virions by the capsid affinity column-sepa- and incubated with various dilutions of antihistone rated antibodies. '251-labeled capsids were dis- IgG at room temperature for 1 h. GPRBC were added, sociated with EGTA and DTT and immunopre- and the amount of HAI was determined. (B) As in cipitated. Antihistone IgG and anti-HA IgG pre- (A), except that 10 HAU of capsids was used. (C) 10 HA U ofpolyoma capsids was incubated with various cipitated only capsid VP1 (Fig. 7A). Antirecep- dilutions of antihistone IgG for 1 h at room temper- tor IgG did not react with the capsid proteins. ature; 10 HAU ofpolyoma virions was then added, In similar experiments with "2'1-labeled virions, and the incubation continued for an additional hour. antihistone IgG precipitated virion dimer, VP1, GPRBC were added, and the amount of HAI was and 29K polypeptide, as well as the two low- determined. (D) As described in C, except that after molecular-weight polypeptides (16K and 14K, the 2nd hour ofincubation the virion-antibody-capsid respectively) (Fig. 7B). Anti-HA IgG immuno- mixture was diluted 1:10,000 in Dulbecco-modified precipitated virion dimer, VP1, the 29K poly- Eagle minimal medium, and a plaque assay was peptide, and the 16K polypeptide. Antireceptor performed. Symbols: *, Virionsplus antihistone IgG; IgG immunoprecipitated some VP1 and the 14K 0, capsids plus antihistone IgG plus virions. All undiluted IgG (100) was adjusted to 10 mg ofprotein per ml. z(D 20 - -10 Un z tion of the 10 HAU of virions to the antibody- Fcz 0 zo capsid reaction mixture, the virion-antibody- Ha: 15- 4 capsid solution was diluted and a plaque assay - zE&-- was performed. This experiment (Fig. 6D) dem- H4 -6 i '. onstrated that the capsids were unable to block >o Z ,, J 10- ° *> the neutralizing antibodies present in the anti- z0 a -4 histone serum. Co0 mz Affinity chromatography with polyoma ZD 5- o capsids. The experiments described above dem- 04 -2 cr oj 5 onstrated that capsids could block HAI antibod- -D_ ies present in the antihistone serum but were unable to block polyoma-neutralizing antibodies FRACTION present in the same serum. To separate these two activities, a capsid-Affi-Gel 10 affinity col- FIG. 7. Fractionation of antihistone IgG by cap- umn was prepared. When antihistone IgG was sid-Affi-Gel 10 affinity chromatography. Antihistone passed through this column (Fig. 7), the anti- IgG (100 Mtg) was fractionated on a 2-ml capsid-Affi- bodies which did not bind to the capsids yielded Gel-10 column as described in Materials and Meth- ods. Protein concentration (A), HAI (0), and virion fractions which showed neutralizing ability neutralization assay using the FA technique (0) were (fractions 12 through 15) without the ability to performed on each fraction as described in Materials cause HAI. After the addition of 5 M MgCl2 to and Methods. The arrow denotes the point at which the column, the bound antibodies eluted in a 5 M MgCl2 in 0.1 M MOPS (pH 7.0) was added. VOL. 34, 1980 POLYPEPTIDES OF POLYOMA VIRIONS 127 polypeptide. These results clearly demonstrate TABLE 1. Plaque reduction with affinity that the antigenic determinants responsible for chromatography-resolved IgG the HAI antibodies present in the antihistone IgG % serum reside on the 16K polypeptide (hemagglu- Antibody sample tein No. of cof Re- tinin), whereas those responsible for neutraliza- (IgG) tein plaques col duction reside on the 14K polypeptide (receptor). coc tro tion Plaque reduction with capsid affinity (gJ chromatography-resolved IgG. The two None 25 100 0 fractionated sera were also assayed to determine Normal rabbit 100 25 100 0 Antihistone 100 1 4 96 their respective abilities to neutralize polyoma Antihistone 10 12 48 52 virions in a plaque reduction assay. Antireceptor Antihistone 1 24 96 4 IgG at a concentration of 4 ,ug inhibited the Antireceptor 4 10 40 60 number of virion plaques formed by 60% (Table Antihemagglutinin 5 24 96 4 1). Anti-HA IgG at about the same concentration (5 fig) produced only a 4% reduction of plaques. The plaque reduction ability of the inhibit virion infectivity, and form precipitin antihistone IgG is also shown. These findings bands with virions in immunodiffusion assays clearly demonstrate quantitatively the separa- (11). The data presented in this report confirm tion of the two activities present in the anti- these results. However, it now appears the major histone IgG. antigens in this region of the gels resulted from the comigration of polypeptide fragments of the DISCUSSION major virion protein VP1 with the histones The data presented here demonstrate that rather than solely histone proteins. This concept antiserum produced against the SDS-derived is supported by several lines of evidence. Anti- polypeptides of polyoma virions results in the histone IgG immunoprecipitated only virion formation ofspecific antibodies which react with VP1 under conditions which stabilized the pro- polypeptides used as the antigen source and that teins (i.e., in the presence of 1o-3 to 10-' M Ca2") the antigenic determinants responsible for the (Fig. 1B). Under conditions which tended to neutralizing and HAI antibodies in the histone destabilize the virion proteins (e.g., 5 mM EGTA region of polyoma virion SDS-gels are cleavage or SDS), antihistone sera immunoprecipitated products ofpolyoma structural protein VP1. The and reacted with polypeptides which appeared, data also demonstrate that HAI activity can be in size at least, to be histones (Fig. 1A, 2D, 4E). separated from the neutralizing activity in this However, when virions were placed under con- antiserum. ditions of Ca2" chelation and subjected to SDS- The antisera directed against the three struc- PAGE, additional polypeptide species were ob- tural proteins VP1, VP2, and VP3 did not react served, particularly at 29K and in the histone with these proteins in their native configuration, region (Fig. 3). but did recognize the proteins under conditions The ability to separate receptor from hemag- which mimicked their denatured state after glutinin activities on virus particles was de- SDS-PAGE. The reactivity of these three anti- scribed in an earlier report (1) in which it was sera to SDS-iodoacetamide-treated capsids shown that polyoma capsids lack the receptor clearly demonstrates this point (Fig. 2A). It was activity present in virions, whereas both possess expected that the anti-VP2 and anti-VP3 sera equivalent HA activity. In the same report it would cross-react to some degree since these was demonstrated that the ability of polyoma polypeptides share a number of tryptic peptides virions to successfully infect mouse cells was (8). These data suggest that many of the anti- independent of the capacity of virions to agglu- genic determinants reside in the more hydro- tinate GPRBC. This concept is further sup- phobic regions of these three proteins and that ported in this report, since capsids were able to the native proteins express a different set of block HAI antibodies present in antihistone se- antigenic determinants. This was further veri- rum without blocking neutralizing antibodies fied by the fact that unlabeled antivirion IgG from the same serum (Fig. 6). When capsids could only partially inhibit '25I-labeled anti-VP1, were coupled to an agarose gel resin and used as anti-VP2, and anti-VP3 from reacting with vir- an affinity matrix, the neutralizing antibodies ion proteins after SDS-PAGE (unpublished did not react and were eluted from the column data). (Fig. 7). Addition of high salt to remove the Previous experiments suggested that antisera antibodies which bound to the capsids produced to the polyoma histone region of SDS-gels had antibody fractions which had HAI activity but the ability to inhibit virion hemagglutination, lacked neutralizing ability (Fig. 7). In plaque 128 BOLEN AND CONSIGLI J. VIROL. reduction experiments (Table 1), the amtirecep- The activation of the protease apparently re- tor antibodies demonstrated high leve,ls of neu- quires the chelation of virion-associated calcium tralizing ability, whereas anti-HA aintibodies and results in the specific cleavage of the 44K demonstrated little or no ability to Xneutralize VP1 into smaller polypeptides. virus infectivity. The immunoprecipittation ex- Although we do not fully understand why periments demonstrated that the HAI activity virions but not capsids react in such a manner, was directed to antigenic determinantts present this phenomenon is perhaps related to the pol- on the 16K (hemagglutinin) pollypeptide, yoma virion-associated protease originally de- whereas the neutralizing activity was dlirected to scribed by Friedmann (6). The protease activity antigenic determinants present on the 14K (re- reported by Friedmann, under conditions which ceptor) polypeptide (Fig. 8B). we now know chelate virion-associated Ca2", also The dual antibody activities ofthe arntihistone led to the appearance of a 29K protein which serum appear to be dependent upon a protease shared tryptic peptides with virion VP1. How- activity which is found in association with pol- ever, VP1 proteolytic fragments smaller than yoma virions but lacking in polyomal capsids. 29K were not identified. Thus, it is not unrea- sonable to assume that the chelation of virion- VP' associated Ca2" activates a virion-associated protease which results in the cleavage of virion VP1 10 - A I VP2 VP3 into a number of smaller polypeptides. Some of FHIS,TONESI these VP1 cleavage products might then migrate N 8- in the histone region of SDS-gels and act as an 0 x DIMER antigen source when injected into rabbits. Thus, 2 6- it appears that during the preparation of the a. 0 virions for use as an antigen source in the pro- H to 4- duction of antisera to the SDS-derived virion CN polypeptides (11), chelation of the virion-asso- 2- ciated calcium occurred. Although known cal- I~~[ cium chelators were not present in the stock A A virion preparations used as the antigen source, Oa 10 20 30 40 so prolonged storage of virion stocks in Tris buffer SLICE NUMBER or phosphate-buffered saline results in consid- erable degradation of virion VP1 (unpublished VP; data). Therefore, it appears that short incubation periods under conditions which favor calcium chelation (e.g., presence of EGTA or SDS) iTONESA or storage for prolonged periods results in the activation of the virion-associated protease. The surprising aspect of this phenomenon is that the VP1 polypeptides present in the histone region represent the antigenic determinants (amino acid sequences) which are responsible for virion jv hemagglutination and attachment to host cells (receptor). From this standpoint it is interesting that capsids, which lack the protein or protein modification present on virions which is respon- so so sible for receptor activity to host cells (1), also SLICE NUMBSER lack the ability to generate cleavage products. FIG. 8. Immunoprecipitation of EGTAL-DTT-dis- Furthermore, experiments which used two-di- sociated '25I-labeledpolyoma capsids and Xvirions. (A) mensional gel electrophoresis (isoelectric focus- 1251-labeled capsids were dissociated wiith 10 mM ing in the first dimension followed by SDS- EGTA and 3 mM DTT (pH 7.5) for 1 h at room PAGE in the second dimension) to compare temperature. Bovine serum albumin (0.1 mM) was polyoma virion and capsid proteins have shown added to reach with any excess DTT; thern the poly- that capsids lack several of the VP1 peptides were immunoprecipitated with a subspecies IgG (A), anti-HA IgG (0), or antireceptor IgG (0) found in virions (Bolen and Consigli, manuscript and electrophoresed as previously describe d. (B) 1251 in preparation). The specificity of the protease labeled virions were dissociated and immiunoprecip- to generate the 29K, 16K, and 14K polypeptides itated as described in (A). The positions;of dimer, suggests that this activity is specific and may be VP1, VP2, the 29K region, VP3 and hi,stones are important in the infection process. Although the indicated. Symbols are as in (A). function of the protease in the life cycle of pol- VOL. 34, 1980 POLYPEPTIDES OF POLYOMA VIRIONS 129 yoma virus is currently unknown, studies to cium ion found associated with purified virions. J. Virol. determine the true origin and role of this pro- 23:717-724. 3. Brady, J. N., V. D. Winston, and R. A. Consigli. 1978. tease during infection are now underway in this Characterization of the DNA-protein complex and cap- laboratory. somere subunits derived from polyoma virus by treat- The ability to isolate the polypeptides which ment with ethyleneglycol-bis-N,N'-tetraacetic acid and contain the antigenic determinants responsible dithiothreitol. J. Virol. 27:193-204. 4. Brunck, C. F., and V. Leick. 1969. Rapid equilibrium for the HA and receptor activity of polyoma isopycnic CsCl gradients. Biochim. Biophys. Acta 179: virus could represent a new beginning of our 136-144. understanding of this virus. These antisera can 5. Consigli, R., H. Minocha, and H. Abo-Ahmed. 1966. now be used to probe the virion surface in order Multiplication of polyoma virus. II. Source of constitu- ents for viral deoxyribonucleic acid and protein synthe- to gain insight into virus structure and functions. sis. J. Bacteriol. 92:789-791. Additionally, biochemical and biophysical char- 6. Friedmann, T. 1976. Structural proteins of polyoma vi- acterization of the isolated hemagglutinin and rus: proteolytic degradation of virion proteins by exog- receptor polypeptides will lead to a clearer un- enous and by virion-associated proteases. J. Virol. 20: 520-526. derstanding of their roles in early polyoma in- 7. Frost, E., and P. Bourgaux. 1975. Decapsidation of fection. polyoma virus: identification of subviral species. Virol- ogy 68:245-255. ACKNOWLEDGMENTS 8. Hewick, R. M., M. Fried, and M. D. Waterfield. 1975. This investigation was supported by Public Health Service Nonhistone virion proteins ofpolyoma: characterization grant CA-07139 from the National Cancer Institute. J.B.B. of the particle proteins by tryptic peptide analysis by was supported by a Public Health Service postdoctoral fellow- use of ion exchange columns. Virology 66:408-419. ship through grant HD-00422 from the National Institutes of 9. Kasamatsu, H., and P. J. Flory. 1978. Synthesis of the Health. R.A.C. is a Senior Scientist with the Mid American SV40 viral polypeptide VP1 during infection. Virology Cancer Center Program. 86:344-353. We would like to express our appreciation to Diane Potts 10. McMillen, J., M. S. Center, and R. A. Consigli. 1976. and Viola Hill for their excellent technical assistance and Sue Origin of polyoma virus-associated endonuclease. J. Vi- Stefanski for help in preparation of this manuscript. We also rol. 17:127-131. thank T. C. Johnson, J. V. Hughes, and V. D. Winston for 11. McMUllen, J., and R. A. Consigli. 1977. Immunological their helpful discussions. reactivity of antisera to sodium dodecyl sulfate-derived polypeptides of polyoma virions. J. Virol. 21:1113-1120. LITERATURE CITED 12. Smith, G. L, and R. A. Consigli. 1972. Transient inhibition of polyoma virus synthesis by Sendai virus (par- 1. Bolen, J. B., and R. A. Consigli. 1979. Differential ainfluenza I). I. Demonstration and nature of the inhi- adsorption of polyoma virions and capsids to mouse bition by inactivated virus. J. Virol. 10:1091-1097. kidney cells and guinea pig red blood cells. J. Virol. 32: 13. Vinograd, J., and J. E. Hearst. 1962. Equilibrium sed- 679-683. imentation of macromolecules and viruses in a density 2. Brady, J. N., V. D. Winston, and R. A. Consigli. 1977. gradient. Prog. Chem. Org. Nat. Prod. 20:372422. Dissociation of polyoma virus by the chelation of cal-