THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 260, No. 23, Issue of October 15, pp. 12803-12809,1985 0 1985 by The American Society of Biological Chemists, Inc. Printed in U.S.A.

Polyoma Virus Major Capsid Protein, VP1 PURIFICATION AFTER HIGH LEVEL EXPRESSION IN ESCHERICHIA COLI*

(Received for publication, March 8, 1985)

Andrew D. Leavittzg, Thomas M. RobertsT, and Robert L. Garceaz From the Divisions of $Pediatric Oncology and 7 Tumor Biology, Dana Farber Cancer Institute, Boston, Massachusetts 021 15

We have expression-cloned in Escherichia coli the major polyoma virus capsid protein, VP1. Under the inducible control of the hybrid tac promoter, VP1 con- stituted between 2 and 3% of the total host cell protein. The expressed VP, was purified to near homogeneity with initial yields to 10%. Optimal expression was temperature-dependent, and significant intracellular degradation could be demonstrated. The final product was obtained as one predominant isoelectric focusing species, without the pattern of post-translational mod- ification seen in virus-infected eukaryotic cells. The purified VP1 from E. coli will be useful as a substrate for the purification of VP1 modification enzymes and in the study of inter-VP1 oligomerization.

Polyoma is an endogenous mouse virus whose icosahedral capsid symmetry has been an important model for the mech- anism of protein subunit packaging. X-ray crystallographic studies of polyoma originally suggested that the capsid con- sists of 12 five-coordinated and 60 six-coordinated morpho- logical subunits (capsomeres) arranged on a T = 7d icosahe- dral surface (Klug, 1965). Recent x-ray diffraction data de- rived from polyoma capsid crystals, however, have shown that the 72 capsomeres are all pentamers, suggesting that the contacts between protein subunits are not all quasi-equivalent (Rayment et al., 1982). The packing of the pentamers in the hexavalent and pentavalent environments has been attributed by Baker et al. (1983) to a “variable bonding potential” of the pentameric capsomeres. Klug (1983) has pointed out that this variability may be a property of a simple multifunctional protein (VP,) which is chemically modified in the different environments or a differential association of identical capso- meres with the minor structural proteins (VP2 and VP3).

The major capsid protein, VP,, is extensively modified by both acetylation and phosphorylation, and these modifica- tions appear to occur prior to virion assembly (Bolen et al., 1981; Ponder et al., 1977; Garcea et al., 1985). In addition, there appears to be a relationship between the modification of VP,, the assembly of intact virus, and the function of the viral early proteins, middle, and small tumor antigens (Garcea and Benjamin, 1983). Mutations in these early proteins lead to a defect in effective virion assembly which is associated with a marked under-modification of VP, (Garcea et al., 1985).

Grants CA30002 (to P. M. R.) and CA37667 (to R. L. G.) awarded by * This work was supported by United States Public Health Service

the National Cancer Institute, Department of Health and Human Services, and by Biomedical Research Grant RR05526 from the Dana Farber Cancer Institute (to R. L. G.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact.

Karen Grunebaum Cancer Foundation predoctoral fellow.

The elucidation of the structural roles of modified VPI mole- cules and the enzymes modifying VP1 therefore connects the divergent areas of virion assembly and the mechanism of action of polyoma’s transforming genes.

In order to study the relationship between VP1 modification and virion structure and to generate a substrate for the purification of VP1 modification enzymes, we have expres- sion-cloned VPI in Escherichia coli. We chose a prokaryotic expression system in an attempt to avoid the post-transla- tional modifications which occur in eukaryotes. The ability to isolate unrestricted quantities of VPl not only makes possible the study of the nature of inter-VP, bonding but also its association with the viral minichromosome in the condensa- tion reaction to form the final virion.

MATERIALS AND METHODS’

RESULTS

Expression of Polyoma Capsid Protein VP1 in E. coli-The general methodology of Guarente et al. (1980) was used to express VP, in E. coli. The first step involved the construction of a plasmid carrying the 5’ terminus of VP1 fused to lac2 of E. coli and is shown in Fig. 1. A 5’ fragment of VP, was obtained from a HinfI digestion of pPY322. The single- stranded ends of all HinfI digest fragments were made blunt with T4 DNA polymerase. Phosphorylated SalI linkers were ligated to the HinfI fragments, and the entire mixture was digested with SalI and HindIII. A 242-bp2 fragment contained the initial (5’ ) 156 bp of VP1 terminating in a HindIII sticky end and 80 bp of 5’ flanking polyoma DNA terminating in a SalI sticky end. Subsequent ligation of this fragment into pLG200 yielded a plasmid PAL100 in which the amino ter- minus of the VP1 gene is fused in frame to a large carboxy! fragment of the lac2 gene.

The second step in constructing the VP, expression vector was to place the necessary transcriptional and translational regulatory elements 5‘ to the VPl-lacZ fusion. Plasmid pGL101, when digested with PuuII and PstI, yields a 851-bp “portable promoter” fragment which contains the reasonably strong lacUV5 promoter and, just 5 bp 5’ to the PuuII site, a ribosome-binding site (Shine-Delgarno sequence). In order to place the lmUV5 promoter and Shine-Delgarno sequence at

“Materials and Methods” are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Document No. 85M-670, cite the authors, and include a check or money order for $1.60 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

The abbreviations used are: bp, base pair; IPTG, isopropyl-8-D- thiogalactopyranoside; SDS-PAGE, sodium dodecyl sulfate-polyac- rylamide gel electrophoresis; PMSF, phenylmethylsulfonyl fluoride.

12803

12804 Cloning and Purification of Polyoma Virus VP1

\ PI Bar

PI.11

FIG. 1. Construction of the VP1 expression plasmid under the control of the lacUV5 promoter. Plasmid pPY322 contains the late region of polyoma from the EcoRI to BarnHI restriction sites cloned into pBR322. A 242-bp fragment from pPY322 was cloned in frame of the @-galactosidase gene in pLG2OO at the Hind11 site to yield pAL100. Ea131 digestion of PAL100 from the SalI site yielded plasmids with variable lengths 5' to the VP, ATG initiation codon which were blunt end-ligated to the lacUV5 promoter and Shine- Delgarno sequence provided from pGL101. Plasmid pAL164 was selected from colonies scoring red on MacConkey/ampicillin selec- tion. Finally, the VPl gene was reconstructed by a trimolecular ligation (see "Results").

the optimal distance 5' to the VP1-lacZ fusion gene of pAL100, with variable amounts of DNA separating the Shine-Delgarno sequence and the initiation codon of VP,, pALlOO was di- gested at the unique SalI site located 85 bp 5' to the VP1-lacZ fusion gene. The linearized PAL100 plasmid was then digested consecutively with nuclease Ba131, S1 nuclease, and PstI. The 851-bp PstI-PuuII fragment of pGLlOl was then ligated to the Bul31-PstI-digested pALlOO fragments and transformants selected on MacConkey/ampicillin agar.

Clones with the expected restriction map were screened for the fusion protein by SDS-PAGE gel analysis of total cell lysates. The clone producing the most fusion protein, pAL164, was saved for further study. Densitometric analysis of total cell lysates (data not shown) demonstrated that the level of fusion protein production was approximately 2-3 times the level of @-galactosidase production from a wild type lac operon ("294). @-Galactosidase assays performed on strains "294 and pAL164/P90C (data not shown) showed the @- galactosidase activity of the fusion protein to be only 1.5 times that found in a fully induced wild type lac operon. The discrepancy may indicate that the &galactosidase-specific activity of the fusion protein is less than the specific activity of an unfused complete @-galactosidase protein.

The third step in generating the VP1 expression vector was to reconstruct the complete VP, gene. Plasmid pAL164 was digested with EcoRI and HindIII, and the small, approxi-

mately 380-bp, fragment was isolated. Plasmid pPY322 was digested with HindIII and PuuII, and the fragment containing the distal 992 bp of the VPl gene as well as the 894 bp of 3' flanking polyoma DNA was isolated. These three fragments were ligated, transformants were screened by restriction en- donuclease digest mapping, and one plasmid, pALVPl, was used for further studies.

To assay for the reconstructed VP1 protein, total cell lysates of E. coli, with or without pALVP1, were electrophoresed on 10% SDS-PAGE gels. Several attempts using strains HBlOl and P9OC failed to show Coomassie Blue-stainable VP1 pro- tein. However, these gels exhibited intensely staining cellular proteins at the expected molecular weight of VP, which would make detection of a foreign protein difficult in this region of the gel. To test for low level production of VP, protein from pALVPI, immunoprecipitation using anti-VP, antibody was performed after labeling with 35S04. The results of this exper- iment (data not shown) demonstrated that a protein from pALVP1/RB791: (a) was specifically precipitated by the anti- VP1 antibody, (b) migrated at the expected molecular weight of VP1 (45,000), ( e ) co-migrated with 35S-labeled VP, from polyoma virions, and (d) was not present in an isogenic strain lacking pALVP1.

Addition of the tac Promoter-VP1 protein expression was improved by changing the h U V 5 promoter to a hybrid pro- moter composed of fragments of the lacUV5 promoter and the E. coli tryptophan promoter (Amann et al., 1983), and Fig. 2 outlines this construction. Plasmid pALVPl was digested with EcoRI and HindIII. The large fragment containing the @-lactamase gene was saved. The small fragment, which con- tains the lacUV5 promoter and the 5' end of the VP, gene, was then digested with HpuII, and the larger fragment con- taining the -10 segment and the 5' end of VP1 was isolated. Plasmid pEA3OO was then digested with EcoRI and CluI, generating a 191-bp fragment containing the -35 segment of the trp operon. The three fragments were then ligated, gen- erating pALVPITAC. Because high levels of VP, may be disadvantageous to cell growth, pALVP1TAC was trans- formed into E. coli strain RB791, which has the ladQ muta- tion, and ampicillin-resistant clones were selected. pALVPITAC mapped correctly and was saved for further study.

To assess the level of VP, protein production under the

PI,....'-SS t-lo" ... ATC.......HimdIIl /

1) Dimst with PI and Hind111

P . h

FIG. 2. Changing the lacW5 promoter to a tac promoter 5' to VPI. Details are described under "Results."

Cloning and Purification of Polyoma Virus VPI 12805

a b c d a b c d

4 9 5

- " "

""

"0 -

*-. . ... Y

-21 - FIG. 3. SDS-PAGE of total cell lysates from E. coli produc-

ing unfused VP,. Cells were grown in TYE media to early logarith- mic phase and split into two equal volume samples. IPTG was added to one sample to a final concentration of 5 mM, and all samples were then incubated for an additional 3 h. Aliquots were then removed from each sample and prepared as described in the legends to Fig. 4. Total cell lysates from equal numbers of cells were then electropho- resed on a 10% SDS-PAGE gel. Lanes a, RB791 with IPTG; lane b, RB791 without IPTG; lane c, pALVPrTAC/RB791 with IPTG; lane d, pALVP1TAC/RB791 without IPTG. The VP, protein in lane c is indicated by the arrowhead a t left. Molecular weight markers are indicated in kilodaltons.

transcriptional control of the tac promoter, RB791 was grown with and without pALVPITAC. As shown in Fig. 3, a protein of approximately 45 kDa accumulated in strain RB791 con- taining pALVP,TAC but not in the isogenic strain lacking this plasmid. Furthermore, the 45-kDa protein was present only after the addition of the inducer IPTG. An additional band of approximately 70 kDa also appears to be induced and may represent another plasmid-encoded protein produced by transcriptional read-through beyond VP,.

To confirm that the 45-kDa protein encoded by pALVPITAC is VP,, strains RB791 and pALVPITAC/RB791 were labeled with "SO, and cell extracts were immunoprecip- itated using anti-VP, antibody. When the "S-labeled immu- noprecipitates were electrophoresed on 10% SDS-PAGE gels, a 45-kDa IPTG-inducible protein was present in strain pALVP,TAC/RB791 but not in strain RB791 (Fig. 4). In addition, multiple bands of less than 45 kDa were present in pALVP,TAC/RB791 ( l a n e b) . These bands were not seen in the absence of IPTG induction (not shown) nor were they present in RB791 alone. Furthermore, the identical banding pattern was reproducible in multiple immunoprecipitation experiments. Thus, since these bands are cross-reactive with anti-VP, antibody, they are not nonspecifically bound E. coli proteins and may represent either degradation products or

FIG. 4. Immunoprecipitation of E. colicell lysates with anti- VPI antibody. Cells were grown in supplemented minimal media, induced with 5 mM IPTG, and labeled with 3sSOc (20 pCi/ml) for 3 h. Total cell lysates and immunoprecipitations were then performed as described under "Materials and Methods." The 101 SDS-PAGE gel shows immunoprecipitates of RE3791 with anti-VP, (lane a ) , pALVP,TAC/RB791 with anti-VP, (lane b), RB791 with preimmune sera (lane c), and pALVP,TAC/RB791 with preimmune sera (lane d). The arrowhead indicates the position of a VP, virion standard.

polypeptides synthesized from internal translation initiation sites.

Fig. 5 shows the results of the two-dimensional gel analysis of VP, synthesized in RB791. As shown in panel B, the VP, protein is a prominent protein in a total cell lysate. The immunocross-reactive proteins of less than 45 kDa, first de- tected on the one-dimensional SDS-PAGE gels (Fig. 4, B ) , are readily seen in panel C. However, they are also present in the total cell lysates (small arrowheads) and therefore are not an artifact of degradation during the immunoprecipitation pro- cedure. Fig. 5C also shows that there are two minor protein species more acidic than the major VP, protein, with the same apparent molecular weight as VP,. The presence of these minor species varies, and it is not clear whether the multiple 45-kDa spots represent VP, degradation products or modified forms of VP,. Two-dimensional gel electrophoresis experi- ments (data not shown) were performed in which VP, protein from sucrose gradient-purified polyoma virions was mixed with anti-VP, immunoprecipitates from pALVPITAC/RB791 before electrophoresis. Such mixing experiments showed that the spot labeled VP, in Fig. 5C co-migrates with unmodified VPl protein from virions.

Thus, the 45-kDa protein encoded by the plasmid pALVPITAC and produced in E. coli is consistent with poly- oma VP, protein by the following criteria. (a ) The molecular weight determined by SDS-PAGE is identical to virion de- rived VP,; (b ) it is immunoprecipitated by anti-VP, antibody; and (c) on two-dimensional gel analysis, it has the same isoelectric point as virus-derived VPl. Furthermore, VP, pro- tein is synthesized in E. coli a t levels detectable by Coomassie Blue staining when total cell lysates are electrophoresed on SDS-PAGE gels (Fig. 3). Densitometric analysis of such a gel shows that the 45-kDa band, less the host bacterial contri-

12806 Cloning and Purification of Polyoma Virus V P ,

C

vpT- "

FIG. 5. Two-dimensional gel electrophoretic analysis of proteins synthesized by pALVPITAC/RB791. Cells were grown in supplemented minimal media into mid-logarithmic phase in the presence of "SO, (20 pCi/ml) and 5 mM IPTG. Total cell lysates and immunoprecipitation were as described under "Materials and Meth- ods." A, RB791 total cell lysate; R, pALVP,TAC/RR791 total cell lysate; C, ALVP,TAC/RR791 immunoprecipitated with anti-VP1 an- tibody. The large arrowheads indicate VP, protein. The small arrow- heads represent presumed VP, degradation products, as discussed in the text. The degradation products are seen both in total cell lysate and in the immunoprecipitation, indicating that the degradation spots in panel C are not simply a result of the technique. The isoelectric focusing dimension is oriented with the left basic and the right acidic.

bution at that molecular weight, constitutes between 2 and 3% of the total host cell Coomassie Blue-stainable proteins.

VP, Protein Stability in E. coli-Because the VP,-P-galac- tosidase fusion protein transcribed from the lacUV5 promoter constitutes 3-4% of the total cell protein, whereas the com- plete VP, protein was not detected on SDS-PAGE gels stained with Coomassie Blue, we tested by pulse-chase experiments the possibility that rapid degradation of the unfused VPI protein contributed to the inability to detect VP, by Coomas- sie Blue staining. Control experiments demonstrated that the

conditions for the pulse-chase experiments were such that: (a) the rate of VP, synthesis was linear, and (b) antibody was present in excess during all immunoprecipitations. Fig. 6B shows the results of pALVP,/RB791 grown at 30 and 37 "C, with one-half of the "S-labeled VP1 protein from a 2-min pulse disappearing in approximately 16 min at 30 "C and approximately 7.5 min at 37 "C. Thus, VP, protein has a

A

I L,,,,,,

11 I I I I I I 0 3 6 9 1 2 1 5 1 8 ~

MINUTES AFTER WLSE LABELING (CHASE 1

FIG. 6. The rate of VPI protein degradation determined by pulse-chase experiments. The pulse-chase experiments and anti- VP, immunoprecipitations were performed as described under "Ma- terials and Methods." Samples were electrophoresed on 10% SDS- PAGE gels. The amount of VP, was determined by densitometry of the gel autoradiogram. The relative amount of VP, present a t each time point was normalized to a value of 10 for the 30 "C pulse sample a t t = 0, and data were plotted using linear regression analysis. A, pALVPITAC/RB791 grown a t 30 "C (0) and 37 "C (A). The corre- lation coefficient is -0.19 for the 30 "C experiment and -0.77 for the 37 "C experiment. B, pALVPI/RB791 grown a t 30 "C (0) and 37 "C (A). The correlation coefficient is -0.96 for the 30 "C experiment and 0.96 for the 37 "C experiment. The normalized value at t = 0 for pALVP,TAC ( A ) is approximately 3-fold greater in absolute amount as that for pALVP, ( R ) .

TABLE I Summary of purification (from 5.5 g of cells)

Stage of purification Vol Total protein" VPIb Yield ml mg mg %

Cell extract 70 371 14.8 100 Polymin P extract 63 289 14.5 98 Ammonium sulfate 11 78 4.7 32

DE52 pooled 6 8.1 2.6 18 Post-DE52 precipitate 1.5 1.8 .7 5 P-11 eluant pooled 2 1.6 1.4 9

* Estimated from Coomassie Blue-stained gel profile by densitom-

post-dialysis

Determined by Bio-Rad protein assay (Bradford, 1976).

etry.

Cloning and Purification of Polyoma Virus VP, 12807

S A B C D E F

* VP1

FIG. 7. SDS-PAGE of VP, purification fractions. The follow- ing fractions were electrophoresed in 10% SDS-PAGE gels prepared as described (see “Materials and Methods”): molecular weight stand- ards as indicated in kDa kilodaltons (lane s), total cell lysate (lane A), Polymin P extract (lane R ) , soluble materials after dialysis from ammonium sulfate-precipitated Polymin P extract ( l a n e C) flow- through from DE52 column (lane D), precipitate after dialysis of lane D (lane E ) , and 1 M NaCl eluant of P-11 column (lane F).

A

FIG. 8. Two-dimensional gel electrophoretic analysis of pu- rified VP,. Shown are portions of two-dimensional gels electropho- resed as described under “Materials and Methods.” Panel A shows the pattern of “3-labeled polyoma virion VP,, prepared as previously described (Garcea et al., 1985). Panel R shows the pattern of the 1 M NaCl eluant from the P-11 column (Fig. 7, lane F). The predominant VP, isoelectric subspecies are indicated by arrowheads. The major VP, spot in both samples has an identical isoelectric point.

temperature-dependent half-life when synthesized in E. coli. Unexpectedly, the results in Fig. 6A also show slower degra- dation of VP,, both a t 30 and 37 “C, when the tac promoter is present uersus the lacUV5 promoter. Densitometric compar- ison between pALVP, and pALVPITAC of the pulse-labeled VP, ( t = 0) reveals a 3-fold increase in the relative synthesis of VP, between the tac and lac constructions.

Purification of VP, from Cell Extracts-We reasoned that VP, is a structural virion protein tightly associated with the DNA genome and would thus adsorb to anionic supports. Because of the lack of an enzymatic assay, we followed the

purification by SDS-gel electrophoresis. The cell extraction procedure (see “Materials and Meth-

ods’’) efficiently solubilized VP, until the dialysis step follow- ing ammonium sulfate precipitation of the Polymin P extract. At this stage, a precipitate formed which contained up to 60% of the total VP, extracted. The VP, could be recovered from this precipitate by solubilization in 1 M NaCl, dialysis to 250 mM NaCl, and then proceeding to the phosphocellulose (P- 11) column step described below. The soluble extract after dialysis to 100 mM NaCl was first applied to a DE52 cellulose column equilibrated with buffer A, pH 7.2,lOO mM NaCI, and the flow-through plus 4 column volumes collected. These fractions contained all the VP, applied. The eluate was pre- cipitated with ammonium sulfate to 35% saturation (see “Ma- terials and Methods”), and the precipitate was dissolved in buffer A, pH 7.2, 100 mM NaCl and diaiyzed against this buffer. The precipitate which formed during this dialysis was removed by centrifugation a t 8000 x g for 15 min and con- tained substantially pure VP, which could be solubilized again in 1 M NaCI. The soluble fraction was then applied to a phosphocellulose (P-11) column equilibrated with buffer A, pH 7.4,lOO mM NaCl and washed with buffer A, pH 7.4,250 mM NaCl, and VPI was subsequently eluted with 1 M NaCl.

Table I follows the yield of this purification scheme, not accounting for recovery of the VP, precipitated during the first dialysis. Fig. 7 follows the purity of VP, at each step. Densitometric determination of VP, yield in Table I reflects the amount of protein in the major VP, band, and all other bands are considered contaminants. The lower bands in Fig. 7, lanes E and F, correspond to anti-VP, antibody immuno- reactive peptides (see Fig. 5C) and are most likely VP, deg- radation products. With recycling of the initially precipitated VP,, up to 50% final yields have been obtained. Fig. 8B is a Coomassie Blue-stained two-dimensional gel of the P-11- eluted material shown in Fig. 7, lane F. The most intense spot, labeled VP,, has the same isoelectric point as unmodified polyoma virion VP, (Fig. 8A). Two or more spots at the same molecular weight of the major spot in panel B correspond to those seen in Fig. 5C. The isoelectric focusing pattern of E. coli-produced VP1 (panel B ) is distinct from that seen in virus-infected cells (panel A ) where greater than 50% of the subspecies are modified (Hunter and Gibson, 1978; Hewick et al., 1977; Bolen et al., 1981).

DISCUSSION

The purpose of the present study was to generate large quantities of VPI so that the chemistry of inter-VP, bonding and VP, association with the viral minichromosome may be studied in further detail. We found that VP, produced in E. coli under control of the lacUV5 promoter was detected only by immunoprecipitation with anti-VP, antibody after [35S] methionine labeling. By changing to the hybrid tac promoter, VPI could be detected by Coomassie Blue staining of total E. coli extracts and constituted 2-3% of the host cell protein. This result is consistent with previous reports concerning the tac promoter (Bikel et al., 1983) and may reflect not only enhanced transcription but also a saturation of the degrada- tive pathways for VP,. Alternatively, it is possible that in- creased levels of VP, result in a certain amount of self- aggregation which is protective against degradation. However, it should be noted that VP, does not form the type of insoluble complexes found for other eukaryotic proteins made in E. coli (Bikel et al., 1983). The 2-fold difference in protein half-life between induction a t 32 “C versus 37 “C which we observed may reflect a partial activation of the heat shock proteolytic

12808 Cloning and Purification of Polyoma Virus VP1

system of E. coli (Goff et al., 1984), which may represent one degradative pathway for VP,.

We observed multiple anti-VP1 immunoreactive peptides smaller than VP, in both whole ceIl lysates and during VP, purification. Although the majority of these presumed proteo- lytic fragments may result from E. coli processing, VP, may have an intrinsic autoproteolytic activity. Bowen et al. (1984) have detected a serine protease activity associated with VP,, identified by radiolabeling the protease with diisopropyl fluo- rophosphonate, which yields two well-defined proteolytic sub- species of 43.5 and 40 kDa. Autodigestion of VP, may aid virion uncoating in specific intracellular compartments, and future studies with the E. coli-purified VP, hopefully will clarify this activity.

The distinctive chemical properties of VPI, deduced from the amino acid sequence and its function as a probable DNA- binding protein, greatly facilitated its purification. The prin- cipal difficulty we anticipated was the tendency of VP, to spontaneously aggregate. Indeed, although most VP1 was sol- ubilized from the bacterial cell debris, dialysis to low ionic strengths did result in the precipitation of a fraction of the VP1. Further studies of the chemical properties of VP1 will require a soluble preparation. Brady and Consigli (1978) have previously shown that VP, from virions, when completely denatured by guanidine HCl and 2-mercaptoethanol, can be gel-filtered through Sepharose CL-GB to obtain a monomer species of about 46 kDa. Subsequent step dialysis to nonde- naturing conditions resulted in capsomere-like structures as defined by velocity sedimentation and electron microscopic appearance. This protocol provides a methodology for the initial studies of inter-VP1 interactions using the cloned pro- tein.

We chose expression in an E. coli vector in order to avoid the post-translational modifications of VPl which occur in a eukaryotic system. Nonetheless, our two-dimensional gel pro- files of purified VP, show two to three distinct isoelectric forms at the same molecular weight as our presumed unmod- ified subspecies. Although these subspecies may also represent proteolytic heterogeneity, we have not definitely ruled out that E. coli enzymes are nonspecifically modifying VPl to a small extent. Our interest in VP1 modification results from the observation that the host-range nontransforming mutants of polyoma underphosphorylate VPI and that this defect is contributory to the failure of these mutants to produce wild type levels of virus in certain cell types (Garcea and Benjamin, 1983). Unmodified VP1 will be useful as a substrate for iden-

tifying and purifying the enzymes involved in modification. The definition of these enzymes may aid in establishing a link between the viral transforming genes and pathways of cellular metabolism.

Acknowledgments-We thank Ireta Ashby for typing the manu- script and Lynne Montross for expert technical assistance.

REFERENCES

Amann, E., Brosius, J., and Ptashne, M. (1983) Gene (Amst.) 25 ,

Baker, T. S., Caspar, D. L. D., and Murakami, W. T. (1983) Nature

Bradford, M. M. (1976) Anal. Biochem. 7 2 , 248-254 Bikel, I., Roberts, T. M., Bladon, M. T., Green, R., Amann, E., and

Livingston, D. M. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, 906- 910

Bolen, J. B., Anders, D. G., Trempy, J., and Consigli, R. (1981) J. Virol. 37,80-91

Bolivar, F., Rodriguez, R. L., Greene, P. J., Betlach, M. C., Heynecker, H. L., Boyer, H. W., Crosa, J. H., and Falkow, S. (1977) Gene (Amst.) 2,95-113

Bowen, J. H., Chlumecky, V., D'Obrenan, P., and Colter, J. S. (1984) Virology 135,551-554

Brady, J. N., and Consigli, R. A. (1978) J. Virol. 27, 436-442 Garcea, R. L., and Benjamin, T. L. (1983) Proc. Natl. Acad. Sci. U.

Garcea, R. L., Ballmer-Hofer, K., and Benjamin, T. L. (1985) J. Virol.

Goff, S. A., Casson, L. P., and Goldberg, A. L. (1984) Proc. Natl.

Guarente, L., Lauer, G., Roberts, T. M., and Ptashne, M. (1980) Cell

Hewick, R. M., Waterfield, M. D., Miller, L. K., and Fried, M. (1977)

Hunter, T., and Gibson, W. (1978) J. Virol. 28, 240-253 Klug, A. J. (1965) J. Mol. Bid. 11 , 424-431 Klug, A. (1983) Nature 303 , 378-379 Laemmli, U. K. (1970) Nature 227 , 680-685 Lauer, G., Pastrana, R., Sherley, J., and Ptashne, M. (1981) J. Mol.

Appl. Genet. 1,139-147 Legerski, R. J., Hodnett, J. L., and Gray, H. B. (1978) Nucleic Acids

Res. 5,1445-1464 Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular

Cloning, A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

Miller, J. H. (1972) Experiments in Molecular Genetics, pp. 431-436, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

O'Farrell, P. H. (1975) J. Bioi. C h m . 250,4007-4021 Ponder, B. A. J., Robbins, A. K., and Crawford, L. V. (1977) J. Gen.

Rayment, I., Baker, T. S., Caspar, D. L. D., and Murakami, W. T.

167-178

303,446-448

S. A. 80,3613-3617

54,311-316

Acad. Sci. U. S. A. 81,6647-6651

20,543-553

Cell 11,331-338

Virol. 3 7 , 75-83

(1982) Nature 2 9 5 , 110-115

Cloning and Purification of Polyoma Virus VP1 12809 SUPPLEMENTAL MATERIAL TO

POLYOMA VIRUS MAJOR CAPS10 PROTEIN VP1: PURIFICATION

AFTER HIGH LEVEL EXPRESSION I N ESCHERICHIA COLI

ANOREY 0. LEAVITT, THOMAS M. ROBERTS, and ROBERT L. GARCEA

MATERIALS

Tryptone, yeast extract, and MacConkey agar were from Difco. Polymin P was obtained from Bethesda Research Laboratories. Staph-A Sepharose was from Phamacia. All r e s t r i c t i o n enzymes. nuclease Bal-31, and T4-DNA l i g a s e were from New England Biolabs. T4-DNA polymerase was from P.L. Biochemicals. Inc.. and

"klenow-fragment" was from Boehringer Mannheim. XAR-5 X - h a t X-ray film for autoradiography was frm Kodak. [35S]methionine (800 Cifmnol) and Na235S04 (lO-lOOhnCi/mnal) were frm New England Nuclear. Anti-VP1 antibody was t h e g i f t of B i l l Murakami. and was ra i sed i n rabb i t s aga ins t SOS-gel p u r i f i e d v i r i o n VP1.

It i s s p e c i f i c t o a l l m o d i f i e d forms o f VPl (Garcea etc, 1985).

METHODS

- Media: Complete media (TVE) contained 10 g tryptone, 5 g yeast extract, and 8 g

NaCl p e r l i t e r . Minimal media was M9 as described by M i l l e r (1972). L-aminO acids were used at 20 ygfml. B i o t i n was used a t 1 pglml, and glucose at 0.5% w/v. MacConkey agar was made according to the manufac turer w i th lac tose as t h e

sugar source. Ant ib io t i cs (bo th p la tes and broth) were used as fol lows unless o therw ise spec i f ied : ampic i l l i n a t 35 gfml, t e t r a c y c l i n e a t 12.5 pgfml.

Bacter ia l St ra ins and Plasmids: Plasmids used are as pBR322 (Bol ivar gal, 1977), pGLlOl [Lauer et a, 1981), pLG200 (Guarente g dl, 1980), pPV322 (Roberts, unpublished). Bacterial strains used are HBlOl (F-, hsdS2O (r -B m-B).

recA 13, ara-14, proA2, l a c 41, gal K2, rps l20 (sm'), xyl-5, mti-1, supE44.A-), P9OC (F-, V l a c p r o x l l l - ) , RB791 ( l a c IQ, l a c P8). "294 (Endo I , rk-, mK+B-,

pro-).

P

Plasmid Constructions: Restr ict ion enzymes were used w i t h s l i g h t m o d i f i c a t i o n o f t h e "low", "medium; or "h igh" sa l t buf fers accord ing to Maniat is et%. (1982).

All other enzymes were used according to t he supp l i e r . L iga t i ons were incubated overn ight a t 16%. DNA transformation and a lka l ine lys is for p lasmid min i -preps were p e r f o m d as described (Maniat is etc, 1982). C e s i m c h l o r i d e DNA

preparat ion was performed by using the large scale alkal ine lysis technique with t h e a d d i t i o n o f 2 mgfml u r i d i n e t o i n c r e a s e p l a s m i d y i e l d . I s o l a t i o n of DNA

fragnents from acrylamide was performed by s ta in ing t he DNA with methylene blue, o r s t a i n i n g an adjacent reference lane with ethidium brmide, t o l o c a l i z e t h e band o f i n t e r e s t . The DNA was subsequently cut fran the gel and mashed w i t h a tef lon p lunger i n a p l a s t i c tube. Two mls o f ext ract ion buf fer (500 M NaCl. 10

mM Tr is , pH 8.0, 1 IM EDTA, pH 8.0) was added, the mix mashed well , 1 ml of phenol: chloroform: isoamyl alcohol (25:24:1) (P:C:IA) added. and the acrylamide

was mashed again. This mix was vortexed well and allowed t o stand at 4OC overnight, centrifuged, and t h e aqueous l a y e r removed and spun through

s i l i con ized g lass wool. The DM4 was then prec ip i ta ted wi th e thanol . Ext ract ion

of ON4 frm agarose gels was done by the method o f Man ia t i s et. (1982).

Nuclease Bal-31 Digestion: Cesium c h l o r i d e p u r i f i e d DNA was d iges ted a t the

u n i q u e r e s t r i c t i o n s i t e i m d i a t e l y 5 ' t o t h e f u s i o n gene. The buffer was changed by add i t ion o f an appropriate volume of lox Bal -31 buf fer (Legersk i - al, 1978). The mix was incubated at 3OoC, nuclease Bal-31 was added, and a l i q u o t s were removed every 30 s from 0 t o 6 mins. A l iquots were vortexed wi th

phenol t o stop the react ion. The pooled al iquots were extracted once w i t h P:C:IA, once w i t h C:IA, and prec ip i ta ted w i th e thano l . The amounts of DNA and nuclease Bal-31 used were determined by control experiments described by Legerski " et a l . (1978) t o e s t a b l i s h an appropriate range of digestion. DNA was subsequently digested with S1 nuclease t o assure b lunt ends.

/3-galactosidase assay: The 19-galactosidase assay was performed according t o the procedure of M i l l e r (1972).

Protein Electrophoresis and Radiolabel ing of Proteins: Protein gel electrophoresis was performed according t o t h e method of Laemnli (1970). Gels were e i ther : a ) s ta ined w i th Coomassie B lue s ta in (.OS% wfv, 10% w/v methanol, 10% wfv ace t i c ac id ) , o r b ) f i xed w i th 10% v fv g lac ia l ace t ic ac id and 10% methanol, rinsed with water, dried, and autoradiography performed against XAR-OMAT x - r a y f i l m a t room temperature. Molecular weight markers used i n protein gel e lectrophoresis were / j-galactosidase 116 koa, phosphorylase b 95 koa bovine serum albumin 66 kDa, ovalbumin 45 kDa, carbonic anhydrase 31 koa, and

soybean t r y p s i n i n h i b i t o r 21.5 kDa. Bacter ia l prote ins were labeled by e i t h e r of

two protocols : a) ce l ls were grown i n M9 minimal media w i t h Na235S04 (20 yCifm1, o r b) c e l l s were grown i n M9 minimal media u n t i l e a r l y l o g phase, and then Pulse

label led wi th [35Slmethionine (18 pCifml) as described below.

Pulse-Chase Experiments: Cells were grown i n M9 minimal media u n t i l e a r l y l o g phase. An a l i quo t equ iva len t t o X volume [ml) = 1.500" was removed, spun a t lO,OOOxg, and resuspended i n 7 0 0 ~ 1 o f M9 media. The c u l t u r e was then placed

at the desired temperature for 1 min, 15 p l of 0.5M IPTG was then added, t h e ce l ls incubated for an addi t ional 10 mins. [35Slmethionine was then added, and

t h e c e l l s were re- incubated for 2 mins. The c e l l mix was imned ia te ly d i lu ted 1:l w i t h M9 minimal media con ta in ing amp ic i l l i n and 4% wfvol unlabeled methionine,

spun a t lO,OOOxg, washed w i t h 5 0 0 ~ 1 H9 con ta in ing amp ic i l l i n and 4% w/vol

methionine, and resuspended i n 7 0 0 ~ 1 of M9 minimal media conta in ing ampic i l l i n and 4% w/v methionine (no IPTG). Seventyp l were removed as a pu lse a l iquot , the cel ls concentrated by c e n t r i f u g a t i o n and f r o z e n r a p i d l y i n a dry ice-ethanol bath. The remainder was a l i quo ted i n to 70 p l samples and incubated with ag i ta t ion. A t designated t imes af ter rmoval of the pulse sample. chase samples were spun, aspirated, and placed i n a dry ice-ethanol bath. Subsequently,

samples were imnunoprecipitated as described below.

Imunoprec ip i ta t ion: The c e l l p e l l e t s frm the pulse-chase experiments were thawed, 2 O p l o f 1% SDS, 10 IM T r i s pH 8.0, lnf4 EDTA was added, and the sample

was placed at 100°C for 4 mins. Six pl of t he bo i l ed sample were then added t o 650 pl o f 2% T r i t o n X-100, 50 M T r i s pH 8.0, 150 mM NaC1, Id EDTA pH 8.0 and

vortexed well , and antibody (1-8)r l) was added t o t h e mix. The mix was vortexed and then incubated at O°C f o r 2 h. 50 pl of Staph-A-sepharose beads were then added and the mix was again incubated at O°C f o r 30 minutes wi th gent le ag i ta t ion

every 5 mins. The mixture was then washed twice wi th 500 pl 1% T r i t o n X-100, 50 M T r i s pH 7.5, and 1 M NaCl, and once w i t h 500 p l 10 IM T r i s pH 8.0. 70 p l of SDS-PAGE sample b u f f e r (10% glycero l , 5% 2-mercaptoethanol, 2.3% SDS, 0.0625 M

T r i s pH6.8, 0.005% bromophenol blue) were then added t o t h e mix, the samples

placed at 100°C f o r 5 minutes, centrifuged at 10,DDDxg f o r 8 mins. and then a p p l i e d t o an SDS-PAGE gel.

Densi tanetry: Quant i tat ion of p r o t e i n bands from autoradiograms was performed by

d e n s i t m t r y u s i n g a Quick-Scan scanner (Helena Laboratories). Scan r e s u l t s (graphic) were photocopied, and the weight of peaks were used t o compare the

r e l a t i v e amount o f p r o t e i n i n v a r i o u s bands.

Two-dimensional gel electrophoresis: Two-dimensional gel electrophoresis was performed according t o t h e methods of O'Farre l l (1975) . In the f i rs t d imension ( i soe lec t r i c focus ing) , ampholines were pH 3.5-10fpH 5-71pH 7-9, 1:2:2 ( v o l f v o l f v o l ) . The second dimension gel contained 10% acrylamide.

Growth Curves: Growth curves *ere performed in the s tandard fash ion (Mi l le r ,

1972), w i t h absorbance readings taken at a wavelength o f 550nm.

Preparat ion o f Prote in Ext ract for VPI Pur i f icat ion: 2 l i t e r s of s t r a i n RB791 containing the p lasmid pALVPITAC were grown a t 37' and a t e a r l y l o g a r i t h i c phase

IPTG was added for a t o t a l of 6 h. A f t e r t h i s p e r i o d t h e c e l l s were p e l l e t e d and frozen at -2OOC. The c e l l s (wet weight 5.69) were thawed i n 4 volumes o f b u f f e r

A/pH7.9/250 mM NaCl(5O nf4 T r i s pH7.9. 5% glycero l , 2M EDTA, 1 5 M 2-mercaptoethanol. 250mM NaCl), lysozyme added t o 200 pgfml and the suspension

incubated at 4OC for 20 min. Sodium deoxycholate (10% w/v) was then added t o a f inal concentrat ion of .05% wfv and the suspension sonicated at 200 watt-seconds i n a Bron-sonic 1510 for three 30 s bursts. The sonicated suspension was then incubated at 10°C for 20 min, a t which time an equal volume of buffer

AIpH7.9f250mn NaCl was added a long wi th PMSF t o 1mM. A l l subsequent steps were car r ied ou t a t 4OC. The mixture was sonicated as before, and then centr i fuged at

10,OOOxg for 20 min. Polymin P (10% w/v, pH7.9) was added s low ly t o t he supernatant t o a f ina l concent ra t ion o f .35% wfv and the mixture incubated for 15 min. The tu rb id so lu t i on was then centr i fuged at 6,OODxg f o r 15 min. The p e l l e t was resuspended i n 5 c e l l volumes of buf fer BfpH7.9llM NaCl ( 1 M T r i s pH7.9, 5%

glycero l , ,ld EDTA, 1M NaCl, 15mM 2-mercaptoethanol) using a dounce homogenizer. Af ter cent r i fugat ion a t 6,OOOxg for 10 min the pe l le t was re-extracted wi th 2

volumes of the same buffer and the supernatants combined. Amonium sulphate was slowly added t o 35% satura t ion , the so lu t ion s t i r red fo r an addi t ional 3 h, and

then centr i fuged at 8,OOOxg for 15 min. The p e l l e t was resuspended i n 1 c e l l v o l of buffer B/pH7.9/200mM NaCl, and dialyzed against buf fer BfpH7.2f100mM NaCl.

The dialyzed extract was centr i fuged at 8,OOOxg f o r 15 min, and the supernatant used as the p ro te in ex t rac t f o r f u r the r pu r i f i ca t i on .