Genetically defined nuclear localization signal sequence of bovine ...

Journal of General Virology (1994), 75, 2463-2467.

2463

Printed in Great Britain

Genetically defined nuclear localization signal sequence of bovine papillomavirus E1 protein is necessary and sufficient for the nuclear localization of El-fl-galactosidase fusion proteins X i a o L e n g X i a o and Van G. W i l s o n * Department of Medical Microbiology and Immunology, Texas A & M University Health Science Center, College Station, Texas 77843-1114, U.S.A.

The 605 amino acid E1 protein of bovine papillomavirus type 1 (BPV-1) is a multifunctional nuclear protein required for viral DNA replication. A nuclear localization signal (NLS) sequence was previously defined by point mutations in three short adjacent clusters of basic amino acids located in the amino-terminal region of the E1 protein. In this study, we used a fusion protein approach to evaluate the contribution of other regions of the E1 protein to nuclear transport. The nearly fulllength E1 gene and six non-overlapping subfragments were each fused in-frame with the lacZ gene in a eukaryotic expression vector. Each clone was electroporated into COS-1 cells, and the intraceUular location of the El-fl-galactosidase fusion proteins was deter-

mined by immunofluorescence. Only the constructs containing the full-length E1 or a single subregion (E 1-259; amino acids 84 to 166) produced fusion proteins that entered the nucleus. Point mutations in the NLS sequences of the E1-259-1acZ construct prevented nuclear translocation of the corresponding fusion protein. This confirms the previous result that the cluster of basic amino acids is critical for nuclear transport. Furthermore, the data obtained in this investigation indicated that the region of El containing the NLS sequence was not only necessary, but was also sufficient for nuclear localization. No other region of E 1 contained independent nuclear localization activity.

The E1 protein of bovine papillomavirus type 1 (BPV-1) is a regulatory protein expressed during the early stage of virus infection. The E1 open reading frame (ORF) is the largest of BPV-1 and encodes a 605 amino acid protein with an apparent M r of 68K (Blitz & Laimins, 1991 ; Sun et al., 1990). Early studies showed that frameshift and termination mutations throughout the E 1 0 R F eliminated BPV-1 replication and led to integration of the viral DNA into the host genome (Berg et al., 1986; Lusky & Botchan, 1985, 1986; Rabson et al., 1986; Stenlund & Ustav, 1991), suggesting a role for E1 in viral DNA replication. Additional studies have now confirmed that the El protein, along with the E2 full-length gene product, are essential trans-acting elements for viral DNA replication (Ustav & Stenlund, 1991). Consistent with the role in viral DNA replication is the observation that E1 is a nuclear phosphoprotein (Santucci et al., 1990; Sun et al., t990). In eukaryotic cells, the nucleus is separated from the cytoplasm by a nuclear membrane. The nuclear membrane is punctuated by the presence of membrane structures termed nuclear pore complexes (Miller et al., t991). These have an average diameter of 9"0 nm and

function as the gateway for the nuclear-cytoplasmic exchange of macromolecules (Newport & Forbes, 1987; Schekman, 1985). The nuclear uptake of karyophilic proteins has been demonstrated to be a highly selective, signal-mediated process (De Robertis et al., 1981; Dingwall & Laskey, 1986; Dingwall et al., 1982; Richardson et al., 1988). Proteins with an M~ larger than 60K must be actively translocated into the nucleus, and the proteins that are selectively transported into the nucleus have been shown to possess a specific signal sequence termed a nuclear localization signal (NLS) (Garcia-Bustos et al., 1991; Kalderon et al., 1984a, b; Newmeyer et at., 1986). A putative NLS for BPV-1 E1 was first mapped to the amino-terminal 223 amino acids, because a construct expressing a truncated E1 lacking this region accumulated exclusively in the cytoplasm (Sun et al., 1990). Within the amino-terminal region of E1 there are three clusters of three or four consecutive basic residues located between amino acids 84 and 114. A point mutational analysis of the basic amino acid sequences found that these sequences are necessary for nuclear translocation of E1 (Lentz et al., 1993). However, while the genetic approach indicated that these basic

0001-2389 © 1994 SGM

2464

Short communication

amino acid sequences were necessary, it could not determine whether they were sufficient for E1 nuclear transport. In this study, an alternative approach using El-fl-galactosidase fusion proteins was employed to define the E1 sequences required for nuclear transport. The nearly full-length E 1 0 R F and six non-overlapping subfragments (shown in Fig. 2) fused in-frame with the lacZ gene W~re originally constructed in a prokaryotic vector, pORF1. The E1 constructs in pORF1 were termed the pC1 series as reported previously (Wilson & Ludes-Meyers, 1992). To express the El-flgalactosidase fusion proteins in mammalian cells, the E l - l a c Z fusion genes were transferred to a eukaryotic expression vector, pCHll0. Plasmid pCH110 contains the origin of replication and early promoter sequences from simian virus 40 (SV40) for replication and expression of any foreign genes inserted downstream of the SV40 early promoter sequence (Hall et al., 1983). Each pE1 series plasmid was linearized by SalI digestion and an adapter (oligonucleotide sequence 5' AAGCTTGGATGT 3'/3' TTCGAACCTACAGCT 5') which contains a HindIII site, an ATG initiation codon and a SalI overhang, was added on both ends. The modified DNA was digested with HindIII plus ClaI, and fragments (1312 to 2800 bpy containing E1 and part of the lacZ gene were gel-purified. Each fragment was ligated to the 6117 bp fragment purified from the HindIII plus ClaI digestion of plasmid pCH 110 to create the pCHE1 series in which the E1 gene or its subfragments were fused inframe with the lacZ gene. The fusion constructs were confirmed by restriction digestion pattern and DNA sequencing using the dsDNA Cycle Sequencing System (BRL) as recommended by the manufacturer. The COS-1 cell line (Gluzman, 1981) was used for the pCHE1 series gene expression since it constitutively expresses the SV40 large T antigen which is required for amplification of the pCHE l series plasmids. The plasmid DNA was introduced into COS- 1 cells by electroporation transfection according to Gearing et al. (1989) with minor modifications. Briefly, 5 x 106 cells in 180 pl of PBS pH7.3 were mixed with 10 to 20lag caesium chloride gradient-purified fusion plasmid or p C H l l 0 plasmid DNA. The mixture, in a total volume of 200 pl, was chilled on ice' for 5 min and electroporated in a 0.4cm gap cuvette at 300V and 125pF ( t = 8 . 5 to 12.5 ms) on a Gene Pulser (Bio-Rad). The electroporated cells were returned to ice for 5 min and cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% (v/v) fetal bovine serum (FBS). The transfection efficiency was monitored by an X-Gal in situ staining assay (Lira & Chae, 1989), and was typically 15 to 40% in surviving COS-1 cells. The expression of E1 fl-galactosidase fusion proteins was monitored by SDS-PAGE and Western blotting (Fig. 1). Seventy-two

1

4

2

5

6

7

I

10 11

12

I91

160K

l16K

Fig. 1. E1 fl-galactosidase fusion proteins were expressed from the SV40 early promoter of pCHE1 series plasmids in COS-1 cells. SDS-PAGE and Western blotting were performed as described in the text. fl-Galactosidase-specific polyclonal antibody (Amersham) was used for lanes 1 to 9 and monoclonal antibody (Promega) was applied for lanes 10 to 12. The position of two Coomassie blue-stained M r markers, fl-galactosidase (116K) and Escherichia coli RNA polymerase subunit (160K), are shown to the left. Lane l, purified fl-galactosidase (3 ng); lane 1, mock E1 control [sheared salmon sperm DNA (SSS DNA)] ; lane 3, pCH 110; lane 4, pCHE 1-212; lane 5, pCHE 1-249; lane 6, pCHE1-259; lane 7, pCHE1-259M; lane 8, pCHE1-312; lane 9, pCHE1-333; lane 10, pCHE1-413; lane 11, pCHEI-1700; lane 12, purified fl-galactosidase (3 ng).

j lac Z I I i

E l ORF

I 840

2663

849

Fusion

Clones

El Amino Acids

pCHEI-1700

2 586

pCHEI-249

2-75

]ntracellular Oistrlbutlon

NUCleUS

2617 1096

Cytoplasm

DCHEl-259

84 166

pCHEI-333

175-278

Cytoplasm

Nucleus

pCHEI-4~3

291-398

Cytoplasm

OCHEI-312

421-517

Cytoplasm

DCHEI-212

527 586

Cy[oplasm

;369

1777 2107

2425

Fig. 2. Summary of the intracellular distribution of the El-flgalactosidase fusion proteins. Shown are the E1 DNA fragments used for cloning, the names of the resultant constructs, the predicted El amino acids expressed in each fusion, and the subcellular location of each fusion protein. The intracellular distribution was determined by immunofluorescence as shown in Fig. 3. For fusions designated 'Cytoplasm', greater than or equal to 95 % of positive cells showed fluorescence only in the cytoplasm. Fusions designated 'Nucleus' had approximately 90 % of positive cells with solely nuclear fluorescence and 10% of cells with both nuclear and cytoplasmic signals, with the signal ratio of nuclear > cytoplasmic. For all fusions, 300 to 500 positive cells were examined.

hours after electroporation, cells were harvested from culture dishes on ice with a rubber policeman, washed once with ice-cold PBS pH 7.5, and lysed by boiling for 5rain in SDS gel sample buffer (50mM-Tris-HC1 pH 6"8, 10 % glycerol, 2 % SDS, 2% 2-mercaptoethanol, 0.01% bromophenol blue). Each sample was electrophoresed on 8 % SDS-polyacrylamide gels according to Laemmli (1970), and the proteins were electrophoretically transferred from the SDS gels to Hybond-ECL Western nitrocellulose membrane (Amersham) at 2-5mA/cm 2 of gel in a Polyblot unit (American Bionetics). Immunological detection of the expressed El-fl-galactosidase fusion proteins on the membrane

Short communication

2465

ia)

Fig. 3. Intracellular distribution of El-fl-galactosidasefusion proteins. The intracellular locations of fl-galactosidase(expressed by pCH110) and El-fl-galactosidasefusion proteins (expressed by pCHEI series plasmids) in transfectedCOS-1 cells were detected by immunofluorescenceas detailed in the text. (a) SSS DNA-transfectedcells (mock); (b to i) cells transfectedby pCH110 (b), pCHE11700 (c), pCHE1-249 (d), pCHE1-259 (e), pCHE1-333 (f), pCHE1-413 (g), pCHE1-312 (h) and pCHE1-212 (/). was performed according to an enhanced chemiluminescence procedure from Amersham as described by the manufacturer. The primary antibodies were as indicated in the figure legend and were used at 1000 to 1500-fold dilutions. As shown in Fig. 1, each construct expressed a protein of the appropriate M r for the predicted El-flgalactosidase fusion proteins. In addition, each construct appeared to express comparable levels of fusion protein. The intracellular distribution of El-fl-galactosidase fusion proteins in transfected COS-1 cells was determined by an indirect immunofluorescence assay according to the protocol previously described (Stein et al., 1991). Electroporated COS-1 cells were cultured in 10% F B S - D M E M in glass slide-based Lab-Tek culture chambers (Nunc) for 72 h. The slides were fixed in 2 % paraformaldehyde (10 rain) and permeabilized in 1% Triton X-100 (10 min). After three washes with 0"3% Tween in 0.02 M-PBS pH 7.3 the slides were incubated overnight in a humid chamber at 4 °C with mouse anti@ galactosidase monoclonal antibody (Amersham) at a 300-fold dilution in diluent (1% BSA-0.3% T w e e ~

PBS). The slides were then incubated for 1 h with a biotinylated anti-mouse secondary antibody (Amersham) at a 200-fold dilution in the above diluent followed by 1 h incubation with streptavidin-Texas Red (Amersham) at a 200-fold dilution in the diluent. Slides were mounted in glycerol with p-phenylenediamine (9.25 mM-p-phenylenediamine, 90% glycerol in PBS pH 8-0), visualized under a Zeiss photomicroscope III equipped for epifluorescence illumination and photographed on Kodak TMX-3200 film. Fig. 2 summarizes the intracellular distribution of the various El-flgalactosidase fusion proteins determined by indirect immunofluorescence assay, and representative pictures of the fluorescent patterns are shown in Fig. 3. These results indicate that the unfused fl-galactosidase protein expressed by pCH 110 was localized in the cytoplasm in transfected COS-1 cells. In contrast, fl-galactosidase fused to the nearly full-length E1 sequence was transported into the nucleus. However, of the six subregion El-fl-galactosidase fusion proteins, only the fusion protein expressed by the pCHE1-259 clone (amino acids

2466

Short communication

(a) Wild-type Mutant

84

105

I

I

112

I

. . . . . . I~ R I( . . . . . . . . . . . .

I( R R I( ... I( It R . . . . . .

ECHEI-259

......

l? I; I . . . . . .

pCHEl-259M

6 l? . . . . . . . . . . . . . .

N G .........

(b)

Fig. 4. Point mutations in the clusters of basic amino acids destroyed the nuclear transport of the El-JLgalactosidase fusion protein. (a) Wild-type sequences of three closely spaced clusters of basic amino acids in the E1-259 fragment and their corresponding point mutations in E1-259M. (b) Subcellular localization of the corresponding fusion proteins as determined by immunofluorescence: left, pCHE1-259; right, pCHE1-259M. 84 to 166 of El), which includes the genetically defined N L S sequence, was localized in the nucleus. These results confirmed the importance of this region in nuclear localization as previously reported by Lentz et al. (1993). As shown in Fig. 4(a), there are three closely spaced clusters of basic amino acids, regions A, B and C, at E 1 amino acids 84 to 86, 105 to 108 and 112 to 114, respectively. Point mutations made in all of these clusters abolished the nuclear transport of E1 protein in COS-7 cells (Lentz et al., 1993). Individual mutations in regions A or B seemed to have the major effects, and mutations in region C enhanced the nuclear transport defects of E1 caused by mutations in region B when mutations were made in both C and B regions. To evaluate the effect of point mutations in these three regions in the context of our E l - l a c Z fusion construct, a 2 5 9 b p fragment (nucleotides 1088 to 1347), containing point mutations in regions A, B and C, was PCR-amplified from clone NL7 (kindly provided by M. Lentz). B a m H I restriction sites were added on both ends of the fragment via the P C R primer, and the PCR-amplified mutant sequence was cloned into pE1-259 to replace the wild-type E1-259 fragment and create pE1-259M. The E1-259M fragment was subcloned from pE1-259 to p C H l l 0 to generate pCHE1-259M. The original point mutations from N L 7 were confirmed in pCHE1-259M by D N A sequencing as described above. The expression of the El-flgalactosidase fusion protein from pCHE1-259M in COS1 cells was demonstrated on Western blots (see Fig. 1,

lane 7), and was comparable to the parental pCHE1-259. The subcellular localization of the mutant fusion protein was examined by indirect immunofluorescence as indicated above. As shown in Fig. 4(b), pCHE1-259M was totally defective for nuclear localization (b) in contrast to the pCHE1-259 clone (a). In this study, by generating the El-fl-galactosidase fusion constructs, the contribution o f various regions o f BPV-1 E1 to nuclear transport was examined. The results presented here correlated closely with those from the previous mutational study and confirmed the requirement for the clusters of basic amino acids (Lentz et al., 1993). Furthermore, these amino-terminal N L S sequences were not only found to be necessary, but the region of E1 containing the basic amino acid clusters (fusion clone pCHE1-259) was also sufficient for the nuclear localization of the E1 fl-galactosidase fusion proteins. Each of the other five El-fl-galactosidase fusion proteins remained cytoplasmic, indicating that they lacked functional nuclear localization signals. In sum, the six fusion constructs comprised 87 % of the E1 O R F (see Fig. 2). The portions of the E 1 0 R F not represented in these clones were distributed in six small clusters encoding from eight to 22 amino acids. N o n e of these six uncloned regions resembled an N L S ; two of the regions lacked any basic amino acids, two regions had only a single basic amino acid, and the two largest regions (19 and 22 amino acids, respectively) had only two basic amino acids each. It is unlikely that these small uncloned regions contributed to nuclear localization. Therefore, E 1 appears to have a single N L S confined to a triplex of basic residues in the amino-terminal region of the protein. We would like to thank Dr Michael R. Lentz for providing the NL7 clone that contains point mutations in the E1 gene, and Dr Robert C. Burghardt for helping carry out immunocytochemistryexperiments in his laboratory. This work was supported by a grant from the Texas Advanced Research Program and by NIH grant CA 56699.

References BERG, L., LUSKY, M., STENLtrND,A. & BOTCHAN,M.R. (1986). Repression of bovine papilloma virus replication is mediated by a virally encoded transacting factor. Cell 46, 753-762. BLITZ,1. L. & LAIMINS,L. A. (1991). The 68-kilodalton E1 protein of bovine papillomavirus is a DNA binding phosphoprotein which associates with the E2 transcriptional activator in vitro. Journal of Virology 65, 649-656. DE ROaERTtS, E.M., LONGTHORNE,R. F. & GURDON,J.B. (1981). Intracellular migration of nuclear proteins in Xenopus oocytes. Nature, London 272, 254-256. DINGWALL,C. & LASK~Y,R. A. (1986). Protein import into the cell nucleus. Annual Review of Biology 2, 367-390. DINGWALL,C., SHKRNICK,S. V. & LASKEY,R. A. (1982). A polypeptide domain that specifiesmigration of nucleoplasmin into the nucleus. Cell 30, 449-458. GARCIA-BUSTOS,J., HEITMAN,J. & HALL,M. (1991). Nuclear protein localization. Biochimica et biophysica acta 1071, 83 101.

Short communication

GEARING, D. P., KING, J. A., GOUGH, N. M. & NICOLA,N. A. (1989). Expression cloning of a receptor for human granulocyte-macrophage colony-stimulating factor. EMBO Journal 8, 3667 3676. GLUZMAN, Y. (1981). SV40-transformed simian cells support the replication of early SV40 mutants. Cell 23, 175 182. HALL, C.V., JACOB, P.E., RINGOLD, G.M. & LEE, F. (1983). Expression and regulation of Escherichia coli lacZ gene fusions in mammalian ceils. Journal of Molecular and Applied Genetics 2, 101-109. KALDERON,D., RICHARDSON,W. D., MARKHAM,A. F. & SMITH,A. E. (1984 a). Sequence requirements for nuclear location of simian virus 40 large-T antigen. Nature, London 311, 33-38. KALDERON, D., ROBERTS,B. L., RICHARDSON,W. D. & SMITH, A. E. (1984b). A short amino acid sequence able to specify nuclear location. Cell 39, 499-509. LAEMMLI, U.K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, London 277, 680-685. LENTZ, M. R., PAK, D., MOHR, I. & BOTCHAN,M. R. (1993). The E1 replication protein of bovine papillomavirus type-1 contains an extended nuclear localization signal that includes a p34 ~c~ezphosphorylation site. Journal of Virology 67, 1414-1423. LIM, K. & CHAE,C. B. (1989). A simple assay for DNA transfection by incubation of the cells in culture dishes with substrates for flgalactosidase. Biotechniques 7, 576-579. LUSKY, M. & BOTCHAN, M. R (1985). Genetic analysis of bovine papillomavirus type 1 trans-acting replication factors. Journal of Virology 53, 955 965. LUSKY, M. & BOTCHAN,M. R. (1986). Transient replication of bovine papilloma virus type 1 plasmids: cis and trans requirements. Proceedings of the National Academy of Sciences, U.S.A. 83, 3609-3613. MILLER, M., PARK, M.K. & HANOVER, J.A. (1991). Nuclear pore complex: structure, function, and regulation. Physiological Reviews 71, 90%949. NEWMEYER, D. D., LucocQ, J. M., BURGLIN, T. R. & DE ROBERTIS,

2467

E. M. (1986). Assembly in vitro of nuclei active in nuclear protein transport: ATP is required for nucleoplasmin accumulation. EMBO Journal 6, 3333-3340. NEWPORT, J.W. & FORBES, D.J. (1987). The nucleus: structure, function, and dynamics. Annual Review of Biochemistry 56, 535-565. RABSON,M. S., YEE, C., YANG, Y. C. & HOWLEY,P. M. (1986). Bovine papillomavirus type 1 3' early region transformation and plasmid maintenance functions. Journal of Virology 60, 624-634. RICHARDSON,W. D., MILLS,A. D., DILWORTH,S. D., LASKEY,R. A. & DINGWALL, C. (1988). Nuclear migration involves two steps: rapid binding at the nuclear envelope followed by slower translocation through nuclear pores. Cell 52, 655-664. SANTUCCI,S., ANDROPHY,E. J., BONNE-ANDREA,C. & CLERTANT,P. (1990). Proteins encoded by the bovine papillomavirus E1 open reading frame: expression in heterologous systems and in virally transformed cells. Journal of Virology 64, 6027-6039. SCHEKMAN, R. (1985). Protein localization and membrane traffic. Annual Review of Cell Biology 1, 115-143. STEIN, L. S., STOICA,G., TILLEY,R. & BURGHARDT,R. C. (1991). Rat ovarian granulosa cell culture: a model system for the study of cell-cell communication during multistep transformation. Cancer Research 51,696-706. STENLUND,A. • USTAV,M. (1991). Cis and trans requirements for BPV DNA replication. Journal of Cellular Biochemistry 15 (supplement E), 76. SuN, S., THORNER,L., LENTZ, M., MAcPI-IERSON,P. & BOTCHAN,M. R. (1990). Identification of a 68-kilodalton nuclear ATP-binding phosphoprotein encoded by bovine papillomavirus type 1. Journal of Virology 64, 5093 5105. USTAV, M. & STENLUND,A. (1991). Transient replication of BPV-1 requires two viral polypeptide encoded by the El and E2 open reading frames. EMBO Journal 10, 449457. WILSON, V. G. & LUDES-MEVERS,J. (1992). Partite expression of the bovine papillomavirus E1 open reading frame in Escherichia coll. Biochemica biophysica acta 1129, 215-218. (Received 26 January 1994; Accepted 8 April 1994)