VIROLOGY

185,

625-632

Identification

(1991)

of the Nuclear JIAN ZHOU,

Localization

Signal of Human Papillomavirus

Type 16 l-1 Protein

JOHN DOORBAR,* XIAO YI SUN, LIONEL V. CRAWFORD,* CORNELIA S. McLEANt AND IAN H. FRAZER’

Lions Human Immunology Laboratories, Department of Medicine, University of Queensland, Princess Alexandra Ho.$pitd~ Brisbane QLD 4 102 Australia; and ‘ICRF Tumour Virus Group and tDepartment of Immunology, Department of Pathology, University of Cambridge, Tennis Court Road, Cambridge CB2 1QP, United Kingdom Received Human cells. For direct the HPV16 Ll HPV16 Ll (KRKKRK aa 510 to localization NLS could databases sequences

June

17, 199 I; accepted

a/., 1986). More recent work (Robbins et al., 1991) has established another NLS consisting of two shorter but interdependent basic aa sequences separated by a series of 1 O-l 2 spacer aa. HPV16 Ll has basic aa sequences resembling both of these NLS within its carboxy terminal 22 amino acids, and we therefore constructed a range of substitution and deletion mutants of HPV16 Ll expressed by rVV to determine the role of these putative NLS in the nuclear localization of this and other HPV capsid proteins. VV recombinant Ll (recL1) was used for this study because rW can produce Ll protein which is easily detected by immunofluorescence, and Ll protein produced by rW in eukaryotic cells should undergo the same post-translational modification as natural Ll and be subject to the same protein transport mechanisms.

Infection with human papillomavirus (HPV), especially of genotypes 16 and 18, is an important factor in the development of anogenital carcinoma, and DNA sequences from these viruses are found in a majority of invasive cervical carcinoma biopsies and derived cell lines (Durst et al., 1983; Fuchs et a/., 1988). HPV16 proteins cannot be produced by standard cell culture techniques, but recombinant proteins have been produced in a number of expression systems, including recombinant vaccinia virus (rVV) (Browne et a/., 1988; Zhou et al., 1990). The major structural protein Ll of HPV16 is a polypeptide of 58 kDa which is exclusively localized in the nucleus of infected cells when it is expressed from a rVV (Browne et al., 1989; Zhou et al., 1990). Nuclear proteins are thought to accumulate in the cell nucleus because they contain nuclear location signals (NLS) (Dingwall and Laskey 1986; Dingwall et al., 1988; Goldfarb et a/., 1986), which in some way facilitate their selective entry (Dingwall et al., 1982) through the nuclear pore complex (Feldherr et a/., 1984). The first NLS sequence to be analyzed at the amino acid (aa) level was that of SV40 large T-antigen, for which a sequence of 5 basic aa KKKRK is sufficient to direct the virus protein or a reporter protein to the nucleus (Lanford and Butel 1984; Kalderon et a/., 1984a; Lanford et reprint

9, 199 1

papillomavirus type 16(HPV16) Ll and L2 capsid proteins can be detected only in the nucleus of infected other nuclear proteins, specific sequences of basic amino acids(aa) termed nuclear localization signals (NLS) protein from the cytoplasm to the nucleus. We used a series of deletion and substitution mutations of the protein, produced by recombinant vaccinia virus (rVV), to identify NLS within HPV16 Ll and showed that contains two NLS sequences, each containing basic aa clusters. One NLS consisted of 6 basic amino acids from aa 525 to 530) at the carboxy terminal end of Ll . The other NLS contained 2 basic aa clusters(KRK from 512 and KR at aa 525, 526) separated by 12 amino acids. Mutations in either NLS did not alter nuclear of Ll when the other remained intact, but mutations to both prevented nuclear localization of Ll. The Ll be overridden by introduction of a membrane binding sequence at the amino terminal end of the protein. A search showed that all sequenced papillomaviruses are predicted to have Ll and L2 capsid proteins with of basic amino acids homologous with one or both NLS of HPV16 Ll. o 1991 Academic press, ~nc.

INTRODUCTION

’ To whom

August

requests

should

MATERIAL Plasmid

AND

METHODS

construction

Construction of point mutations of HPV16 L 1. The method for the production of point mutations was based on Sayers eta/. (1988). Briefly, synthetic oligonucleotides with various mutation sequences (Table 1) were annealed to single-stranded HPV16 Ll DNA in M13mpl9 cloned from pHPV16 and extended by Klenow DNA polymerase (Amersham). After removal of the nonmutant strand by treatment with Neil and exonuclease III, the mutant strand was then used as a

be addressed. 625

0042-6822/91

$3.00

Copyright @ 1991 by Academic Press, Inc. All rights of reproduction in any form resewed.

ZHOU

626

ET AL.

TABLE THE NUCLEOTIDE

SEQUENCE

OF

THE PCR PRIMERS USED

CONSTRUCT

HPV16

Ll

MUTANTS

5’GCCCGGGl-l-A-llTGGCCl-TCAATCCTGClTGTAGTAAAAAl-l-i-GC 5’GCCCGGGlTATGTAGCl-riIXGTlTTCCTAATGTAAATTll-GGTITGGCC 5’GGCCCGGGlTAAGCAGTrGTAGAGGTAGATGAGG 5’CCTCTACAACTGCTAAATAGCAAAAAACGTAAGCTGTAAG 5’CCTCTACAACTGCTAAACGCTAAAAACGTAAGCTGTAAG 5’CTACAACTGCTAAACGCAAATAACGTAAGCTGTAAG BCTGCTAAACGC AAAAAATAGTAAGCTGTAAGTATTG 5’GGCCCGGGTTAACGTT’TrlTGCGTTTAGCAGTTG 5’CGCCCGGGClTACAGCTT’ACGTTTllTGCGAlTAGCAGTTGTAGAGG 5’CGCCCGGGTTlACAGC~ACGlTllTTGTTlll-AGCAGTTGTAGAGG S’CGCCCGGGClTACAGCTTACGTl-TAlTGCGTTTAGCAG~G 6CGCCCGGGCAATACTTACAGClTACGA7TmGCGTTTAGCAG GCGmAGCAG 5’CGCCCGGGCAATACl-rACAGCTTAI 5’CGCCCGGGCATACAATAClTACAGATTACGTrlTTTGCGTlTAGC 5’CAnAGGAAAACGAGAAGCTACACCCACCACC 50.XAAATlTACA~AGGAAAACCPAAAGCTACACCCACC 5’CCAAAA~ACA-l-TAGGAGAACGAAAAGCTACACCC 5’AGCAGTTGTAGAGGTAGATGAGGTGGTGGGTGTAGCAmCGTmCCTAATGT~~ 5’AGCAGTTGTAGAGGTAGATGAGGTGGTGGGTGTAGCTmGGTmCCTAATGTAAA~ 5’AGCAG~GTAGAGGTAGATGAGGTGGTGGGTGTAGC~CGATTTCCTAATGTAAAm 5’AGCAG~GTAGAGGTAGATGAGGTGGTGGGTGTAGCAmCCTAATGTAAAm

M-5 M-4 M3-0 M3-1 M3-2 M3-3 M3-4 M3-5 Kl-N R2-N K3-N K4-N R5-N K6-N M3-l-l M3-1-2 M3-1-3 Kl N-minus R2P-minus KBN-minus KRK-NPN

template to reconstruct double-stranded molecules. The sequence was checked by the dideoxy-termination method (Sanger et al., 1977) using T7 DNA polymerase. The Ll sequences with various point mutations were cut from Ml 3mpl9 with BgllllHindIII and inserted into RK19/16Ll (Zhou et a/., 1990) to replace the wild-type Ll sequence of HPV16. The VV 4b promoter and the HPVl6 Ll gene with point mutations were transfered from RKl9/16Ll into pSX3 (Zhou et al., 1990) to produce vaccinia intermediate vectors (Fig. 1). Construction of deletion mutations of HPV16 L 1. Deletion mutations of Ll were produced using the polymerase chain reaction (PCR). First, a plasmid pUCl8/ 4bLl was constructed, containing the VV 4b promoter

~7.5

TO

Sequence

Primer

824

1

E. coli gpt

p4b

Ll mutations

824 .,

FIG. 1. Plasmid vectors used to construct recombinant vaccinia viruses. The selectable marker gene gpt with its 7.5K promoter and the HPV16 Ll mutations with a 4b late W promoter are flanked by the vaccinia 824 gene. Promoters are marked by a solid box, Ll mutations represents 23 different mutations (see Table 2).

(Kent, 1988) and the HPV16 Ll ORF, by cutting the 4b promoter and the HPV16 Ll ORF from pLC18 (Zhou et a/., 1990) with XballSphl and inserting the resulting fragment into the XballSphl site of pUC18. The 17-mer reverse sequence primer (M 13RSP) was used as a common 5’ PCR primer, and primers for various deletions derived from the 3’sequence of the Ll ORF, each containing a Smal site for cloning (Table l), were used for the PCR reaction. PCR amplification was performed in 100 ~1 reaction mixture with 0.1 pg pUC18/4bLl plasmid DNA, 1 PM primers, a mixture of dNTPs at 200 PM, 5 mM KCI, and 2 U Taq DNA polymerase (Promega). The reaction mixture was overlaid with mineral oil and subjected to 30 cycles of amplification. Each amplification cycle involved denaturation at 92” for 2 min, annealing at 54” for 2 min, and elongation at 72” for 2 min. The amplified 2-kb fragment was extracted with phenol and purified by 1% agarose gel electrophoresis. Ll deletion mutations were isolated together with the 4b promoter by digestion with Smal and cloned into pSX3 for construction of recombinant viruses (Fig. 1). Construction of insertion mutation of L 1. A 13 hydrophobic amino acid cluster MAALALLAALLW was added to the N-terminal end of Ll by two rounds of PCR procedures. First, a 99-mer 5’ primer 5’ClTCTCGCACTACTGGCGGCGCTGCTAGTTGTCTCTCTTTGGCTGCCTAGTGAGGCCACTGTCTACTTGCCTCCTGTCCCAGTATCTAAGGlTGTAAGC-3’anda34-mer3’

HPV16 Ll NUCLEAR LOCALIZATION

primer 5’GGCCCGGGlTAClTACGllllTTGCGllTAGCAG-3’were used to amplify an Ll ORF which would code for a 10 amino acid insertion at the N-terminus of Ll (LLALIAALLW). This extended Ll ORF fragment was used as template for a second round of PCR amplification with the same 3’ primer and a new 5’ primer 5’-GCGGATCCATGGCTGCCCTTCTCGCACTACTGGCGGCGCTGCTAGTTGTC-3’which added another two alanine codons and a methionine codon, plus a BarnHI cloning site. The Ll ORF with the added putative membrane binding domain was cloned into the RK19 plasmid (Kent, 1988) to produce the RKl SLl/meb plasmid. Then, the 4b promoter and Ll with the inserted mutation were cloned into pSX3 for construction of rVV (Fig. 1). Construction of recombinant vaccinia virus. The methods described previously (Zhou et a/., 1991a) were followed. Briefly, the modified pSX3 plasmids, comprising HPV16 Ll genes with various mutations driven from a VV 4b late promoter, the Escherichia co/i gpt gene (Falkner and Moss 1988; Boyle and Coupar, 1988) as a selectable marker, and flanking fragments of the serine protease inhibitor gene II (B24R) from vaccinia virus (Kotwal and Moss, 1989; Smith eta/., 1989) were transfected into W WR strain-infected (0.05 PFU/ cell) CV-1 cells by calcium phosphate precipitation. Virus plaques were purified twice in CV-1 cells in the presence of mycophenolic acid at a concentration of 25 pg/ml. Virus stocks were frozen and used for immunofluorescence studies. Immunofluorescence. CV-1 cells were grown on eight-chamber slides (Nunc Inc., Naperville, IL) and infected at 30 PFU/cell with rVV. At 18 hr postinfection cells were fixed with 80% ethanol and treated with 5% BSA in PBS for 1 hr at 37”. The monolayers were reacted with Camvir 1 hybridoma supernatant (McLean et al., 1990) at 1:20 dilution with 5% BSA in PBS for 2 hr at room temperature followed by three washes with PBS. Slides were then incubated with HPLC-purified fluorescein isothiocyanate-conjugated anti-mouse IgG (Amersham) at 1:50 dilution for 1 hr at room temperature and examined by uv epifluorescence microscopy with appropriate filters. RESULTS The location of rVV-expressed HPV16 Ll protein (recL1) was examined using indirect immunofluorescence with an Ll -specific MAb. The unmodified recL1 was found exclusively in the nucleus of infected CV-1 cells (Fig. 2b). CV-1 cells infected with wild-type W showed no staining with Camvir 1 (Fig. 2a). To determine a defined segment of HPV16 Ll which was responsible for targeting the polypeptide into the nucleus

SIGNAL

627

of infected cells, stepwise deletion mutants of the carboxy terminus were first introduced. Deletion of the last 27, 17, or 7 amino acids (pSXM-5, M-4, M3-0) all prevented specific nuclear localization of recL1, which was instead distributed evenly throughout the cell (Table 2, Figs. 2c, 2d). Deletion of the last 5, 4, 3, or 2 amino acids (pSXM3-2, M3-3, M3-4, M3-5) allowed nuclear localization similar to that seen with unmodified recL1, while deletion of the last 6 amino acids (pSXM3-1) gave an intermediate result with predominantly nuclear but a little cytoplasmic localization of recL1 (Fig. 2e). Next, a series of point mutations were introduced in the Ll ORF, each of which resulted in the substitution of 1 of the 6 basic amino acids, lysine or arginine, in the carboxy terminal end of Ll by asparagine (pSXKl-N, K2-N, K3-N, K4-N, K5-N, K6-N). None of these single aa substitutions altered the nuclear location of recL1 (Table 2). To determine whether the sequence of basic amino acids at positions 5 10 to 5 12 was important for nuclear localization, we introduced, using oligonucleotide-directed mutagenesis, single and multiple point mutations in the Ll ORF which resulted in mutation of Lys510, Arg5”, Lys512 to asparagine, proline, asparagine either individually or simultaneously (Kl N-minus, R2P-minus, K3N-minus, KRK-NPN), and none of these mutations changed the nuclear location of recL1 (Table 2, Fig. 29). Then, point mutations and a stop codon were introduced to the Ll ORF to produce a recL1 truncated of the terminal 6 amino acids, and with substitution of Lys510, Arg5”, Lys512 by glutamic acid or proline (SXM3-l-l, M3-1-2, M3-1-3). Each of these mutations abolished specific nuclear localization of recL1 (Table 2; Fig. 2f), in contrast to the recL1 protein (pSXM3-1) which was similarly carboxy truncated but without further mutation and which was predominantly localized in the nucleus (Table 2. Fig. 2e). Finally, a putative membrane binding sequence was added to the N-terminal end of Ll to examine which signal sequence was dominant. A 13 amino acid hydrophobic sequence (MWLLAALLW) believed to be a transmembrane domain (Bargmann et al., 1986) was added by two rounds of PCR. This recL1 (M3-6meb) was located exclusively in cytoplasm but unlike the other cytoplasmic recL1 mutants, which were evenly distributed through the cytoplasm, M3-6meb was distributed as small clumps, suggesting that it was bound to membrane structures in infected cells. (Table 2. Fig. 2h). When the predicted sequences for the Ll and L2 proteins of all papillomaviruses listed in Genebank were examined for nuclear localizing signals, we found comparable sequences of basic amino acids at the

ZHOU

ET AL.

HPV16

carboxy terminal (Table 3).

Ll

NUCLEAR

end of each of the Ll and L2 proteins

DISCUSSION The nuclear and cytoplasmic compartments of the eukaryotic cells contain largely distinct sets of proteins (Bonner, 1975). Studies of the fate of endogenously synthesized and exogenously introduced proteins suggest that all nuclear proteins are synthesized in the cytoplasm, from which the mature polypeptides migrate rapidly to the nucleus (De Robertis et a/., 1978). Thus the steady-state segregation of proteins between nucleus and cytoplasm appears to be governed by an intrinsic property of mature polypeptides. In viva many small proteins diffuse freely between nucleus and cytoplasm and accumulate in either compartment due to selective retention or differential degradation (Davey et al., 1985). The size limit for free diffusion of a spherical protein between nucleus and cytoplasm is estimated at about 60 kDa, so some specific transport mechanism to the nucleus other than diffusion must exist for large proteins. HPV16 Ll protein is a viral structural protein of approximately 58 kDa as estimated by SDSPAGE (Browne et a/., 1988; Zhou et al., 1990). The protein is synthesized in the cytoplasm, but must migrate into the nucleus rapidly for assembly with L2 into mature virions (Zhou et a/., 199 1b), because Ll protein is located exclusively in the nucleus in immunolocalization studies. Our results concerning the NLS of the Ll protein of HPV16 can be simply explained only if there are two sequences within the carboxy amino acids of the HPV16 Ll protein which are important for nuclear localization of this protein. The first sequence, which we have termed the carboxy terminal NLS, lies within the seven carboxy terminal aa of the Ll protein, and the second NLS, which we term the bipartite NLS after Robbins et a/., (1991), comprises the aa from 510 to 512, together with the aa at 525 and 526. The removal of both NLS of HPV16 Ll abolished the specific nuclear translocation of the protein and the protein diffused evenly between nucleus and cytoplasm. The cytoplasmic localization of the recL1 produced from pSXM3-0 argues that while there are other sequences of basic aa within HPV16 Ll , they do not contribute to the nuclear localization of this protein. Considering first the carboxy terminal NLS, this can be defined in those mutants in which the bipartite se-

LOCALIZATION

SIGNAL

629

quence is destroyed by mutations from aa 5 10 to 5 12. Comparing Kl N-minus, R2P-minus, and K3N-minus with SXM3-l-1, SXM3-l-2, and SXM3-1-3, it is clear that the carboxy 7 aa contribute a nuclear localization signal. These 6 aa include a sequence of basic amino acids (RKKRKR) similar to the characterized NLS in SV40 large T-antigen in which a short sequence (KKKRK) is sufficient to target reporter proteins to the nucleus. A single amino acid in this sequence is crucial for nuclear location of SV40 large T-antigen (Kalderon et al., 1984b; Lanford and Butel, 1984; Goldfarb et a/., 1986) and a lack of effect of single point mutations in the comparable sequence in the HPV16 Ll protein suggested the existence of a second NLS within this protein. Data from the predicted aa sequences of the Ll and L2 capsid proteins of other papillomaviruses (Table 3) show that all papillomavirus Ll and L2 proteins possess similar carboxy terminal basic sequences:- the sequence in HPV18 Ll is degenerate, but all other sequenced papillomavirus Ll proteins have at least 5 basic aa within a sequence of 6, close to the carboxy terminus of the protein. Considering the bipartite NLS, this can be defined in those mutants which lack the 6 carboxy terminal amino acids which include the carboxy terminal NLS. Mutant pSXM3-1 is mostly localized to the nucleus, although as pSXM3-2 is completely localized to the nucleus the arginine and lysine at position 525, 526 probably both contribute to the bipartite NLS, as it is unlikely that the two basic aa at 525 and 526 alone could constitute the carboxy terminal NLS. Further mutations of pSXM3-1 showed that each of the aa from 5 10 to 5 12 is also critical to this NLS. These results are directly comparable to those obtained for the nucleoplasmin protein by Robbins et a/. (1991) who found that two short clusters of basic amino acids separated by 1O-l 2 amino acids tolerant of substitution could act as a NLS. We did not explore by mutation the role of the amino acids adjacent to the basic aa from 5 10 to 5 12, and between aa 512 and aa 525, but searched for similar basic amino acid sequences within the sequences of the Ll capsid protein of other papillomaviruses. The data (Table 3) showed that most Ll proteins contained such bipartite NLS motifs, with no conserved aa sequences close to or between the two short basic amino acid sequences. If the necessary and sufficient NLS of this type is 2 basic amino acids separated by any 8-l 2 other amino acids from another 2 basic amino acids,

FIG. 2. Location of the Ll gene product in recombinant W-infected cells. CV-1 cells were infected at 30 PFU/cell with Ll recombinant W. Cells were fixed and incubated with anti-L1 monoclonal antibodies diluted 1:20 followed by FITC-conjugated antimouse IgG and visualized by uv microscopy. (a) Wild-type W; (b) unmodified Ll pSX5VV(Zhou et a/., 1990); (c) deletion mutation pSXM-5; (d) deletion mutation pSXM3-0; (e) deletion mutation pSXM3-1; (f) deletion and point mutation combination SXM3-1-2; (g) point mutation KRK-NPN; (h) insertion mutation M3-6meb.

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 l 0

0 0 0 0 0 0 0 0 0 0 0 a 0 0 0 0 0 0 0 0 0

named on the left, the predicted protein N, nuclear; C = N, uniform throughout

0 0 0

l

0 0 0 0 0 N 0 0

0

0 0 0

NUCLEAR TARGETING SEQUENCE

l 0

Ll

0 0

0 0 0

0 0

0 0 0

0 0 0

0 0

l

0

OF THE HPV16

2

Note. The carboxy terminal amino acid sequence of HPV16 Ll IS shown at the top. Beneath, for each mutation right. the subcellular location of the correspondmg recL1 protein is indicated according to the followlng code: predominantly nuclear with some cytoplasmic staining; C, granular cytoplasmic. * With a MAALLALIAALLVV sequence at the N-terminal end of Ll polypeptide.

psx5 pSXM-5 pSXM-4 pSXM3-0 pSXM3- 1 pSXM3-2 pSXM3-3 pSXM3-4 pSXM3-5 pSXK1 -N pSXR2-N pSXK3-N pSXKCN pSXR5-N pSXK6-N SXM3-1-1 SXMB-l-2 SXMB-1-3 K 1 N-minus R2P-minus KBN-minus KRK-NPN M3-6Meb*

MUTATIONS

TABLE

0 0 0 0 0

0 0 0 0 0 N 0

0

ON 0

l 0

0 0 0

0

l

0

ON ON

0 ON

ON ON

ON

l N

ON l

ON

0

0

c

N

C=N C=N C=N

N

C=N C=N C=N N&C N N N N

sequence is shown, and, on the nucleus and cytoplasm; N $ C,

0 0

0 0

l

0 0 0 0 0 N 0 0

0

F’

1

P

2

HPVl6 TABLE THE PREDICTED

AA SEWENCES

OF PAPILLOMAVIRUS L2 proteins HPVl L2 HPVl6 L2 HPVll L2 HPV8 L2 HPV6b L2 HPV5 L2 HPV18 L2 HPV33 L2 HPV41 L2 HPV57 L2 RPVl L2 BPVl L2 Ll proteins C-terminus and bipartite NLY HPVlG Ll HPVl Ll BPVl Ll HPV57 Ll RPV Ll HPV41 Ll HPVll Ll HPV6b Ll HPV33 Ll C-terminus NLS only HPV5 Ll HPV8 Ll Bipartite NLS only HPV18 Ll

Ll NUCLEAR LOCALIZATION

3 OF THE C TERMINUS LATE PROTEINS.

viralnnstgdfelhpslRKRRKRayv. tiiadagdfylhpsyymlRKRRKRlpyffsdvslaa. pvfrtgsdfylhptwyfaRRRRKRiplfftdvaa. virhthdnsgdfflhpslRRRKRKRKy1. pvfitgsgfylhpawyfaRKRRKRiplffsdvaa. iihphdstgdfylhpslhRRKRKRKy1. igihgthyylwp@yfipKKRKRvpyHadgfvaa.

tiwdgadMhp.syfilRRRRKRfpyfftdvRvaa. wdlysgsmdydihpslIRRKRKKRKRvyfsdgRvasRpk. vyivggdyyllpsyvlwpKRRKRvhyffadgyvaa. vfifegnadgtyyleeplRKKRRKstfIladgsvavyae.

ghtvdlyssnytlhpsllRKRKKRKha.

tlgKRKatpttsststtaKRKKRK1. SttKRKtvRvstsaKRRRKa.

StvRKRRisqKtssKpaKKKKK. PvsRKRaaataaaptaKRKKvRR.

SsnKRvstqstalttyRRptKRRRKa. ssnKRvstqstalttyRRptKRRRKa. tglKRpavsKpstapKRKRtKtKK. tgvKRpavsKasaapKRKRaKttKR

pKlKRaaptstRtssaKRKKvKK.

ngtKavsyKgsnRgtKRKRKn. ngtKsisRgsvRgtKRKRKn.

igpRKRsapsattssKpaKRvRvRaRK.

a Sequence of the C-terminus of papillomavirus late proteins from Genbank. Basic amino acids are indicated in bold and stop codons are shown as dots. For definitions of the nuclear localization sequences, see Discussion.

as suggested by our data and that of the mutation experiments by Robbins et a/. (1991) we would predict that HPV18 Ll would be dependent for nuclear localization on the bipartite NLS, whereas HPV5 and HPV8 Ll would depend on the C-terminus NLS; the other Ll proteins, like HPV16 Ll , could use either. No L2 proteins contained a bipartite NLS sequence, suggesting that the L2 protein is dependent on the carboxy NLS for its nuclear targetting and any single aa should prove crucial for its nuclear location. A bipartite NLS motif is also present in other virus nuclear proteins, including the polymerase basic protein of influenza virus (Nath and Nayak 1990) in which two clusters of 4 basic amino acids with a spacer segment of 16 amino acids composes the NLS. Analysis has shown that a sequence of 2 basic amino acids followed by any 10 amino acids and then 3 more basic amino acids out of the next 5 is found in half (275/492) of all sequenced nuclear proteins, but only in 4.2% of

SIGNAL

631

sequenced cytoplasmic proteins (Robbins et al., 199 1). Most (1901251) nonnuclear proteins containing a NLS could also contain a putative membrane directing hydrophobic amino acid cluster, and one current hypothesis is that such proteins are secreted or targeted to the plasma membrane, mitochondria, or chloroplast, as the hydrophobic membrane insertion sequence is “dominant” over the NLS. Our insertion mutation of Ll in which a hydrophobic aa cluster at the N-terminal end could override the NLS of the protein clearly supports this hypothesis. The significance of the double NLS sequences in HPV16 Ll and most other papillomavirus Ll is not clear. It could have evolved as a security sequence because the loss of a few basic amino acids in the carboxy terminal NLS of these proteins would not affect their function of targeting Ll to the nucleus for virus assembly. It could also represent a gene duplication or exchange during the evolution of these viruses: the lack of sequence homology at the nucleotide level between the basic aa sequences seen in the two NLS sequences of Ll argues against internal duplication. The two basic regions of NLS probably bind to the same receptor in the nuclear transport mechanism during Ll entry into nucleus, because synthetic peptides with the NLS sequences of nucleoplasmin (bipartite), El A, and SV40 large T-antigen (single basic sequence) all interact with the same group of putative receptor proteins (Yamasaki et al., 1989). Recent evidence indicates that while the NLS determines the specificity of protein transport, the rate of transport is controlled by amino acids flanking the NLS. A casein kinase II site [serine”’ serine”‘] is an important factor in the enhancement of the transport of SV40 T-antigen to the nucleus (Rihs et al., 1991) and similar casein kinase II sites are observed in the NLS region of other nuclear proteins. Phosphorylation at a serine or threonine close to the NLS may enhance nuclear transportation by altering NLS conformation and thus the affinity of the NLS sequence for its receptor. In HPV16 Ll and some other papillomavirus Ll and L2 proteins, serine and threonine residues resembling casein kinase II sites are present close to the NLS of these proteins (Table 3). Further studies are needed to determine the effect of phosphorylation on nuclear targeting of Ll and L2 proteins. ACKNOWLEDGMENTS We thank Dr. L. Gissmann for providing HPVl6 DNA plasmid, Professor Laskey for communication of results before publication, and Ms. Kul Singh for secretarial work. This work was supported by grants from the Queensland Cancer Fund, National Health and Medical Research Foundation of Australia, the Mayne bequest, and the Princess Alexandra Hospital Development Foundation.

ZHOU

632

REFERENCES BARGMANN. C. I., HUNG, M. C., and WEINBERG, R. A. (1986). The neu oncogene encodes an epidermal growth factor-receptor-related protein. Nature 319, 226-230. BONNER, W. M. (1975). Protein migration Into nuclei. II. Frog oocyte nuclei accumulate a class of mlcroinjected oocyte nuclear proteins and exclude a class of microinjected oocyte cytoplasmic proteins. 1. Cell Viol. 64, 431-437. BOYLE, D. B., and COUPAR, B. E. H. (1988). A dominant selectable marker for the construction of recombinant poxviruses. Gene 65, 123-128. BROWNE, H. M., CHURCHER, M. J., STANLEY, M. A., SMITH, G. L., and MINSON, A. C. (1988). Analysis of the Ll gene product of human papillomavirus type 16 by expression in a vaccinia virus recombinant. /. Gen. v;ro/. 69, 1263-l 273. DAVEY, J., DIMMOCK, N. J., and COLMAN, A. (1985). Identification of the sequence responsible forthe nuclear accumulation of the influenza virus nucleoprotein rn Xenopus oocytes. Ceil 40, 667-675. DE ROBERTIS, E. M., LONGTHORNE, R. F., and GURDON, J. B. (1978). Intracellular migration of nuclear proteins in Xenopus oocytes. Nature 272, 254-256. DINGWALL, C., and LASKEY, R. A. (1986). Protein import into the cell nucleus. Annu. Rev. Ceil. Biol. 2, 367-390. DINGWALL, C., ROBBINS, J., DILWORTH, S. M., ROBERTS, B., and RICHARDSON, W. D. (1988). The nucleoplasmin nuclear localization sequence is larger and more complex than that of SV40 large T antigen. J. Cell Biol. 107, 84 l-849. DINGWALL, C., SHARNICK. S. V., and LASKEY, R. A. (1982). A polypeptide domain that specifies migration of nucleoplasmin into the nucleus. Cell 30, 449-458. DURST. M., GISSMANN. L., IKENBERG. H., and ZUR HAUSEN, H. (1983). A papil!omavirus DNA from a cetvlcal carcinoma and its prevalence in cancer samples from different geographic regions. Proc. Nat/. Acad. Sci. USA 80, 3812-3815. FALKNER, F. G., and Moss, B. (1988). Escherichia co/i gpt gene provides domrnant selection for vaccinla virus open reading frame expression vectors. J. Viral. 62, 1849-l 854. FELDHERR, C. M., KALLENBACH, E. J.. and SCHULTZ, N. (1984). Movement of a karyophllic protein through the nuclear pores of oocytes. 1. Cell Biol. 99, 2216-2222. FUCHS, P. G., GIRARDI, F., and PFISTER, H. (1988). Human papillomavirus DNA in normal, metaplastic, preneoplastic and neoplastic eplthelia of the cervix uteri. Int. /. Cancer 41, 41-45. GOLDFARB, D. S.. GARIEPY, J., SCHOOLNIK, G., and KORNBERG, R. D. (1986). Synthetic peptides as nuclear location signals, Nature 322, 641-644. KALDERON, D., RICHARDSON, W. D., MARKHAM, A. F., and SMITH, A. E. (1984a). Sequence requirements for nuclear locatlon of simian virus 40 large T-antigen. Nature 311, 33-38. KALDERON, D., ROBERTS, B. L., RICHARDSON, W. D., and SMITH, A. E. (1984b). A short amino acid sequence able to specify nuclear location. Ceil 39, 499%509.

ET AL KENT, R. K. (1988). “The lsolatton and Analysis of the Vaccinia Virus 4b Promoter.” Ph.D. Thesis, Univ. of Cambndge. KOTWAL, G. J., and Moss, B. (1989). Vaccinia virus encodes two proteins that are structurally related to members of the plasma serine protease inhibitor superfamily. /. Viral. 63, 600-606. LANFORD, R. E., and BUTEL, 1. S. (1984). Construction and characterization of an SV40 mutant defective In nuclear transport of T antigen. Cell37, 801-813. LANFORD. R. E.. KANDA, P.. and KENNEDY, R. C. (1986). Induction of nuclear transport with a synthetic peptide homologous to the SV40 T antigen transport signal. Cell 46, 575-582. MCLEAN, C. S., CHURCHER, M. J., MEINKE. J., SMITH, G. L., HIGGINS, G., STANLEY. M.. and MINSON, A. C. (1990). Production and characterization of a monoclonal antibody to human paptllomavirus type 16 using a recombinant vaccinia virus. 1. C/in. Pathol. 43, 488-492. NATH, S. T., and NAYAK, D. P. (1990). Function of two discrete regions is required for nuclear localization of polymerase basic protein 1 of A/WSN/33 influenza virus (Hl Nl). Mol. Cell. Bioi. 10, 4139-4145. RIHS, H-P., JANS, 0. A., FAN, H., and PETERS. R. (1991). The role of nuclear cytoplasmic protein transport is determined by the casein kinase II site flanking the nuclear localization sequence of SV40 T-antigen. EMBO J. 10, 633-639. ROBBINS, J., DILWORTH, S. M., LASKEY, R. A., and DINGWALL, C. (1991). Two interdependent basic domains In nucleoplasmin nuclear targetlng sequence: Identification of a class of bipartite nuclear targeting sequences. Cell 64, 615-623. SANGER, F., NICKLEN, S., and COULSON, A. R. (1977). DNA sequencing with chain-terminating Inhibitors. Proc. Nat/. Acad. SC;. USA 74, 5463-5467. SAYERS, 1. R., SCHMIDT, W., and ECKSTEIN, F. (1988). 5’.3’ exonucleases in phosphorothioate-based oligonucleotide-directed mutagenesis. Nucleic Acids Res. 16, 791-802. SMITH, G. L., HOWARD, S. T., and CHAN. Y. S. (1989). Vaccinia virus encodes a family of genes with homology to senne protease inhibitors. /. Gen. \/irol. 70, 2333-2343, YAMASAKI, L., KANDA. P., and LANFORD, R. E. (1989). Identification of four nuclear transport signal-binding proteins that interact with diverse transport signals. Mol. Cell. Biol. 9, 3028-3036. ZHOU, J.. CRAWFORD, L., MCLEAN, L., SUN, X. Y., STANLEY, M., ALMOND, N., and SMITH. G. L. (1990). Increased antibody response to human papillomavirus type 16Ll protein expressed by recombinant vaccinia virus lacking serine protease inhibitor genes. /. Gen. viral. 71, 218552190. ZHOU, J., MCINDOE, A., DAVIES, H., SUN, X. Y., and CRAWFORD, L (199 1 a). The induction of cytotoxlc T-lymphocyte precursor cells by recombinant vaccinia virus expressing human papillomavirus type 16Ll. Virology 181, 203-210. ZHou, J., SUN, X. Y., STENZEL, D. J., and FRAZER. I. H. (1991 b). Co-expression of vaccinia virus recombinant Ll and 12 capsid proteins of HPV16 in epithelial cells is sufficient for assembly of HPVvirionlike particles. Virology. In press.