Downloaded from www.microbiologyresearch.org by

IP: 54.39.17.49

On: Thu, 12 Apr 2018 12:20:35

J. gen. Virol. (1987), 68, 3081-3089. Printed in Great Britain

Key words: HPV/capsid antigens~expression in bacteria

3081

Expression of Human Papillomavirus Type 6 and Type 16 Capsid Proteins in Bacteria and Their Antigenic Characterization

By L. B A N K S , 1. G. M A T L A S H E W S K I , -~ D. P IM, 3 M. C H U R C H E R , 4 C. R O B E R T S 1 AND L. C R A W F O R D 3

~ Department of Biochemical Virology, The Wellcome Research Laboratories, Langley Court, Beckenham, Kent BR3 3BS, U.K., "-Institute of Parasitology, MacDonald College, McGill

University, St. Anne De Bellevue, Quebec, Canada, 3Molecular Virology Laboratory, Imperial Cancer Research Fund, Lincoln's Inn Fields, London and 4Division of Virology, Department of

Pathology, University of Cambridge, Addenbrooke's Hospital, Cambridge, U.K.

(Accepted 7 August 1987)

SUMMARY

The L1 and L2 capsid proteins encoded by human papillomavirus types 6 and 16 (HPV-6 and HPV-16) have been synthesized in bacteria. Antisera were raised against the HPV-6 L1- and L2-fl-galactosidase fusion proteins and against an HPV-16 L1 C-terminal peptide which was 14 amino acids long. The HPV-16 L1 peptide antibodies have been shown to be highly reactive with the HPV-16 Ll-fl-galactosidase fusion protein but not against the equivalent HPV-6 Ll-fl-galactosidase fusion protein. The effectiveness of these antibodies was compared with commercially available anti- bovine papillomavirus type 1 (BPV-1) antibodies and the results demonstrated that the anti-BPV-1 antibodies reacted well against HPV-6 Ll-fl-galactosidase but not against HPV-16 Ll-fl-galactosidase. In addition, the L2 portion of the HPV-6 L2- fl-galactosidase fusion protein appeared particularly immunogenic, since antibodies raised against this fusion protein were predominantly reactive with the L2 moiety. The HPV-16 L1 peptide antibodies described here will be preferred reagents for the specific detection of HPV-16 capsid antigens, which may be particularly important in early diagnosis of HPV-16 infection.

INTRODUCTION

Papillomaviruses are found in a number of higher organisms including man, in which over 40 different types have now been identified. Of these many types, most cause only mild if inconvenient lesions of skin and mucosa (zur Hausen, 1977). Recently however, a small number of these viruses have been shown to be probable aetiological agents in a number of human cancers of which cervical cancer is by far the most prevalent. Over 90~ of cervical cancer biopsies contain human papillomavirus (HPV) DNA of types 16, 18, 31, 33 or 35, of which types 16 and 18 are most common (zur Hausen, 1985). Cell lines derived from these tumours have been shown to contain HPV DNA (Diirst et al., 1983; Boshart et al., 1984), RNA (Schwarz et al., 1985; Schneider-Gadicke & Schwarz, 1986) and a number of HPV early proteins (Smotkin & Wettstein, 1986; Banks et al., 1987; Androphy et al., 1987). In contrast, HPV-6 and HPV-11 infections tend to be associated with benign lesions of the genital tract. Hence early identification of the HPV type in an infection is particularly important, and immunologically based assays seem to offer the most convenient means of diagnosis.

To date most studies on HPV capsid antigens in infected tissue have relied upon commercial anti-bovine papillomavirus type 1 (BPV-1) antisera (Dako, Copenhagen, Denmark) raised

0000-7857 © 1987 SGM

Downloaded from www.microbiologyresearch.org by

IP: 54.39.17.49

On: Thu, 12 Apr 2018 12:20:35

3082 L. BANKS AND OTHERS

against detergent-disrupted BPV-1 virions (Jenson et al., 1980; McDougall et al., 1986; Fey et al., 1986). Recently however it has been shown that the L2 open reading frame (ORF) of HPV-la encodes a type-specific viral capsid protein easily detectable in HPV-l-induced warts (Komly et al., 1986). Thus viral capsid proteins appear to be the most likely targets for immunological diagnosis. Here we report the expression of the viral capsid proteins in Escherichia coli from the L1 and L20RFs of both HPV-6 and HPV-16. Antisera have been raised to these proteins and these have been used to investigate their antigenic properties.

METHODS

Production ofHPV-6 capsidproteins in E. coll. From sequence information of HPV-6 (Schwarz et al., t983) we located the L1 and L 2 0 R F s , and identified suitable restriction fragments for cloning into the pUR vectors (Rfither & M/iller-Hill, 1983). For HPV-6 L1, a XbaI to HindlII fragment extending from nucleotides 5903 through 7902 to 2307, coding for amino acids 39 to 501, was purified by gel electrophoresis and then ligated into XbaI/HindllI-cut pUR288. For HPV-6 L2, a BamHI to XbaI fragment extending from nucleotides 4722 to 5903, coding for amino acids 94 to 460, was purified by gel electrophoresis and then ligated into BamHI/XbaI-cut pUR278. The resulting plasmids were used to transform E. coil H4.4 and ampiciUin-resistant clones were grown in the presence of 100 ~tg/ml isopropyl-fl-D-thiogalactoside (IPTG). Cells were grown to saturation at 37 °C, collected by centrifugation then washed in PBS (150 mM-NaCI, 10 mM-sodium phosphate, pH 7-2). Bacterial cells were then lysed in SDS-PAGE sample buffer (50mM-Tris-HCl pH 6.8, 2% SDS, 5 ~ glycerol, 0-1 M-dithiothreitol), subjected to SDS-PAGE (Laemmli, 1970) and stained with Coomassie Brilliant Blue.

The recombinant plasmid pAS-6L2 expressing HPV-6 L2 alone was made as follows. The BamHI to XbaI fragment extending from nucleotides 4722 to 5903, was made blunt-ended at the XbaI site with T4 DNA polymerase and purified by gel electrophoresis. It was then ligated into BamHI- and SalI-cut plasmid vector pAS 1 (Rosenberg et al., 1983) which was also made blunt-ended at the SalI site. The ligation mixture was used to transform E. coil ARI20 which contains the ~. CI gene. Ampicillin-resistant clones were grown at 37 °C to an A s s0 of 1 and then in the presence of 40 ttg/ml nalidixic acid for an additional 4 h at 37 °C. Cells were collected and analysed as described above.

ProductionofHPV-16 capsidproteins. From sequence information of HPV-16 (Seedorf et al., 1985) the L1 and L2 ORFs were located. For HPV-16 L1, a BamHl to DdeI fragment extending from nucleotides 6150 through 7904 to 24 coding for amino acids 196 to 531 was made blunt-ended with T4 DNA polymerase, a BamHI linker was inserted, the fragment was purified by gel electrophoresis and then ligated into the BamHI site of pUR289. For HPV-16 L2 a StuI to BamHI fragment extending from nucleotides 4466 to 6150, coding for amino acids 78 to 473, was made blunt-ended with T4 DNA polymerase, a BamHI linker was inserted, the fragment was purified by gel electrophoresis and then ligated into the BarnHI site of pUR278. The resulting plasmids were then used to transform E. coli H4.4 and resulting clones were analysed by SDS-PAGE as described above.

In addition, HPV-16 L1 sequences were also cloned into pEX expression vectors (Stanley & Luzio, 1984). Two constructs coding for HPV-16 L1-/~-galactosidase fusion proteins from the pEX vectors were produced: pHX-I containing a HphI to BamHI fragment coding for HPV-16 L1 amino acids 1 to 195 and pHX-2 containing a BamHI to Sphl fragment coding for amino acids 196 to 531. These were induced by growing bacterial cultures overnight at 30 °C followed by temperature shift to 42 °C for 2 b. Proteins were then analysed as described above. The differences between pHX-2 and pUR-16 L1 are additional 3' non-coding sequences in pUR-16 L1 and pHX-2 is induced by temperature shift whereas pUR-16 L1 is induced with IPTG.

Production of anti-HPV-16 L1 antibodies and analysis by Western blotting. BALB/c mice were inoculated twice subcutaneously with 30 gg peptide coupled to thyroglobulin in incomplete Freund's adjuvant. The 14 amino acid peptide from the carboxy-terminal region of L1 was synthesized by the solid-phase technique and then coupled through its N terminus to thyroglobulin as described (Nelson et al., 1984). Mice were given an intravenous boost with a further 30 ~tg peptide-thyroglobulin conjugate 2 weeks after the initial inoculation. One week later, sera were collected and tested for anti-L1 activity by Western blotting as follows. Bacterial cells containing HPV-6 or HPV-16 L1-/3-galactosidase fusion protein were lysed in SDS-PAGE sample buffer and subjected to SDS-PAGE. Proteins were electrophoretically transferred onto nitrocellulose filters as described previously (Banks et al., 1985). Blots were then incubated in blocking buffer (10~ calf serum in PBS) for 2 h at 42 °C. Antiserum was diluted in blocking buffer and the blots were incubated at room temperature for 3 h. Blots were washed three times in PBS then reacted with goat anti-mouse-biotin conjugate (Sigma) followed by avidin-peroxidase conjugate (Sigma). Blots were developed in the presence of 0.1~ hydrogen peroxide and 1.5 mg/ml Hanker-Yates reagent (Polysciences).

Production o f anti HP [I-6 capsid antibodies. Mice were inoculated as described above with 3 ~tg per inoculation of gel-purified HPV-6 Ll-fl-galactosidase or HPV-6 L2-fl-galactosidase fusion proteins. Sera were collected and tested for either anti-L1 or anti-L2 activity by Western blotting as described above.

Downloaded from www.microbiologyresearch.org by

IP: 54.39.17.49

On: Thu, 12 Apr 2018 12:20:35

HP V capsid proteins XbaI

( a ~ ~ HindlII

BamHI BamHI

( c ~ L2

~ / S a l I

BamHI Stul/BamHI (d) ~ LI ( ~ 2

BamHI Fig. 1. (a) Recombinant pUR288 (pUR-6L1) coding for HPV-6 Ll-fl-galactosidase fusion protein. The solid boxed area represents the Ll-coding region and the open boxed area represents 3' non-coding sequences. Arrows indicate the direction of transcription. The resulting construct coded for a fusion protein containing HPV-6 L1 amino acids 39 to 501. (b) Recombinant pUR278 (pUR-6L2) coding for HPV-6 L2-fl-galactosidase fusion protein. The solid boxed area represents the L2-coding region and the open boxed area represents 3' non-coding sequences. The resulting construct coded for a fusion protein containing HPV-6 L2 amino acids 94 to 460. (c) Recombinant pAS-1 (pAS-6L2) coding for HPV-6 L2 protein. The solid boxed area represents the L2-coding region and the open boxed area represents 3' non-coding sequences. The resulting construct coded for a protein containing HPV-6 L2 amino acids 94 to 460. (d) Recombinant pUR289 (pUR-16LI) coding for HPV-16 Ll-fl-galactosidase fusion protein. The solid boxed area represents the L 1-coding region and the open boxed area represents 3' non-coding sequences. The resulting construct coded for a fusion protein containing HPV-16 L1 amino acids 196 to 531. (e) Recombinant pUR278 (pUR-16L2) coding for HPV-16 L2-fl-galactosidase fusion protein. The solid boxed area represents the L2-coding region and the open boxed area represents 3' non-coding sequences. The resulting construct coded for a fusion protein containing HPV-16 L2 amino acids 78 to 473. The diagram is not to scale.

3083

RESULTS

Expression of HPV-6 capsid proteins in E. coli

For HPV-6 L1 the XbaI to HindIII f ragment containing coding sequence for amino acids 39 to 501 was cloned into the prokaryot ic expression vector pUR288 as described in Methods. This resulted in a construct coding for the HPV-6 L 1 0 R F linked to the C terminus of the fl- galactosidase gene (Fig. 1 a). Similarly, for HPV-6 L2 a BamHI to XbaI fragment containing coding sequence for amino acids 94 to 460 was cloned into pUR278. This resulted in a construct coding for the HPV-6 L 2 0 R F linked to the C terminus of the fl-galactosidase gene (Fig. 1 b). Upon I P T G induction the HPV-6 Ll-f l -galactosidase fusion protein synthesized had a mol.wt. of about 153K (Fig. 2a, lane 2), which was somewhat smaller than would have been est imated from the L1 amino acid sequence l inked to fl-galactosidase, although it was apparent that there were a number of degradat ion products. The HPV-6 L2-fl-galactosidase fusion protein had a mol.wt, of about 160K (Fig. 2a, lane 3), which was similar to that which would be est imated from the L2 amino acid sequence l inked to fl-galactosidase.

Expression of HPV-16 capsid proteins in E. coli

For HPV-16 L1 a BamHI to DdeI fragment containing coding sequence for amino acids 196 to 531 was cloned into pUR289. This construct coded for the HPV-16 L 1 0 R F l inked to the C terminus of the fl-galactosidase gene (Fig. ld) . Finally, for HPV-16 L2 a StuI to BamHI

Downloaded from www.microbiologyresearch.org by

IP: 54.39.17.49

On: Thu, 12 Apr 2018 12:20:35

3084 L. BANKS A N D OTHERS

(a) 1 2 3 (b) 1 2

200 - -

,,,liiN 97.4-- ,~. m

68--

43--

200--

4

97"4-- ~ ~ * ~

68-- ~ ~

4 3 - -

Fig. 2. SDS-PAGE of (a) HPV-6 L1- and L2-fl-galactosidase and (b) HPV-16 L1- and L2-fl- galactosidase fusion proteins synthesized in E. coli H4.4. Total protein was extracted in SDS-PAGE sample buffer, run on a gel and stained with Coomassie Brilliant Blue. Lanes 1, cultures containing plasmid pUR278; lanes 2, cultures containing plasmids pUR-6L1 (a) and pUR-16L1 (b); lanes 3, cultures containing ptasmids pUR-6L2 (a) and pUR-16L2 (b). All cultures were grown in the presence of IPTG. The mol.wt, standards were myosin (200K), phosphorylase B (97.4K), bovine serum albumin (68K) and ovalbumin (43K). Arrows indicate the appropriate L1 and L2 fusion proteins.

fragment containing coding sequence for amino acids 78 to 473 was cloned into pUR278 as described in Methods. This construct coded for the HPV-16 L 2 0 R F linked to the C terminus of fl-galactosidase (Fig. 1 e). Upon I P T G induction the HPV-16 L l-fl-galactosidase fusion protein synthesized had a mol.wt, of about 125K (Fig. 2b, lane 2) which was considerably smaller than would be estimated from the L1 amino acid sequence, suggesting that a large amount of degradation was taking place. The HPV-16 L2-fl-galactosidase fusion protein synthesized had a mol.wt, of about 130K (Fig. 2 b, lane 3) which was again somewhat smaller than the estimate of 160K mol.wt. Both of the HPV-16 fusion proteins appeared to be unstable, thus explaining the low levels of expression. However, as shown below, an HPV-16 Ll-fl-galactosidase fusion protein of the expected size was detectable by Western blotting, although the protein was not visible on a Coomassie Brilliant Blue-stained gel.

Production of antibodies against HPV-16 L1

From the HPV-16 sequence information it was possible to predict the pr imary structure of an L1 C-terminal peptide. This was chosen since from our previous experience, antibodies to peptides derived from the C terminus of a protein are most likely to react with the native protein. A peptide was made which was 14 amino acids long and ended eight amino acids from the C terminus of L1 (Fig. 3a). This was linked to thyroglobulin and used to inoculate mice. The anti- L1 peptide antiserum was then reacted on Western blots with the Ll-fl-galactosidase fusion

Downloaded from www.microbiologyresearch.org by

IP: 54.39.17.49

On: Thu, 12 Apr 2018 12:20:35

(a) N

HP V capsid proteins 3085

Lys-Arg-Lys-Ala-Thr-Pro-Thr -Thr -,.%r-Ser -Thr -,.%r-Thr-Thr C

(b) 1 2 3 4 5 (c) 1 2 3 4 5

--200

--97.4

-68

--43

Fig. 3. Immunoreactivity of HPV-16 L1 antisera against Ll-fl-galactosidase fusion proteins synthesized in E. coli H4.4. (a) Amino acid sequence used as an immunogen to raise anti-L 1 antibodies. (b) Coomassie Brilliant Blue stain of total protein. (c) Western blot analysis of total protein with antiserum raised against LI peptide. Lanes 1, induced pUR278; lanes 2, pUR-6L1 ; lanes 3, pHX-I ; lanes 4, pHX-2; lanes 5, pUR-16L1.

proteins. Since the Ll-fl-galactosidase fusion protein generated above tended to become degraded, two additional Ll-fl-galactosidase fusion proteins were used. These proteins consisted of the amino-terminal half of HPV-16 L1 (pHX-1 coding for amino acids 1 to 195) and the carboxy-terminal half of HPV-16 L1 (pHX-2 coding for amino acids 196 to 531) and were found to be more stable than the pUR-derived L1 fusion protein. The results obtained are shown in Fig. 3. Clearly the antiserum to the L1 peptide reacted strongly with two of the HPV-16 Ll-fl- galactosidase fusion proteins (Fig. 3 c, lanes 4 and 5) and, as expected, no reaction was observed against the pHX-1 amino-terminal fusion protein (Fig. 3c, lane 3). No reaction was observed against either fl-galactosidase alone or the HPV-6 Ll-fl-galactosidase fusion protein (Fig. 3c, lanes 1 and 2 respectively). This result confirms the correct translation of the HPV-16 Ll-fl- galactosidase fusion protein. The peptide antiserum detected a band of approximately 145K mol.wt. (Fig. 3 c, lane 5) although only a weak Coomassie Brilliant Blue stained band was visible in this position (Fig. 3b, lane 5). However, we believe that this stained protein represents a background polypeptide, confirming the extensive degree of degradation of this fusion protein.

Production of antibodies against HPV-6 capsid proteins

The HPV-6 L1- and L2-/~-galactosidase fusion proteins were excised from preparative polyacrylamide gels and used to inoculate mice over a number of weeks as described previously (Banks et al., 1987). The activity of the antibodies was then determined by Western blot analysis of E. coli H4.4 cells expressing the appropriate fusion protein. The results obtained are shown in Fig. 4. Anti-HPV-6 L1-/~-galactosidase fusion protein antibodies clearly reacted strongly with both fusion proteins (Fig. 4a), demonstrating antibody against at least the /~-galactosidase portion of the L 1 fusion protein. The anti-HPV-6 L2-/~-galactosidase fusion protein antibodies

Downloaded from www.microbiologyresearch.org by

IP: 54.39.17.49

On: Thu, 12 Apr 2018 12:20:35

3086

(a) 1 2 (b) 1

L. BANKS AND OTHERS

2 (a) 1 2 (b) 1 2 (c) 1 2

!m wi, ,gm L 2 ~ ' ~ ~ ~ • :

M

|

Fig. 4 Fig. 5

Fig. 4. ImmunoreactivityofHPV-6capsid antisera against Ll-fl-galactosidase fusion protein (lanes 1) and L2-~-galactosidase fusion protein (lanes 2). (a) Antibodies against the L l-/~-galactosidase protein, (b) antibodies against the L~-galactosidase protein.

Fig. 5. Immunoreactivity of anti-HPV-6 L2-fl-galactosidase antibody against HPV-6 L2 protein synthesized in E. coli AR120 from plasmid pAS-6L2. Cultures were grown in the presence (lanes 1) or in the absence of nalidixic acid (lanes 2). (a) Coomassie Brilliant Blue stain of total protein. (b) Western blot analysis of total protein with antiserum raised against L2-fl-galactosidase fusion protein. (c) Western blot analysis of total protein with preimmune mouse serum.

reacted with the L2-fl-galactosidase fusion protein and degradation products thereof, but only very weakly with the Ll-fl-galactosidase fusion protein (Fig. 4b, lane 1). Only a very weak reaction was observed against pUR-fl-galactosidase alone (data not shown), and the strong reaction obtained against the protein of lower mol.wt. (Fig. 4b, lane 2) represented reaction against residual L2 sequences on the degraded L2-fl-galactosidase fusion protein. This result was particularly surprising, since it demonstrated that virtually the entire humoral response in these mice was directed against the L2 portion of the L2-fl-galactosidase fusion protein. The specificity of these antibodies was investigated further by Western blot analysis on E. coli AR120 cells expressing HPV-6 L2 not fused to fl-galactosidase. This construct was generated as described in Methods (Fig. 1 c), and upon nalidixic acid induction of these cells a 50K mol.wt. protein was induced (Fig. 5a). When these cells were reacted with anti HPV-6 L2-fl- galactosidase fusion protein antibody, a strong reaction was obtained (Fig. 5b).

Antigenic analysis of the HPV capsid proteins

From the above data it is apparent that the anti-HPV-16 L1 peptide antiserum would be an ideal reagent for differentiating between HPV-6 and HPV-16. However, we were also interested in determining how they compared with commercially available antisera to papillomavirus eapsid proteins. Hence a series of Western blots were done with the papillomavirus capsid proteins using antisera to BPV capsid antigens (Jenson et al., 1980) now marketed by Dako,

Downloaded from www.microbiologyresearch.org by

IP: 54.39.17.49

On: Thu, 12 Apr 2018 12:20:35

(a) 1 2 3 4

H P V capsid proteins

5 6 (b) 1 2 3 4 5 6

-200

-97'4

3087

3 4 5 6 (d) 1 2 (c) 1

. --200

W -97.4

Fig. 6, Immunoreactivity of anti-papillomavirus capsid antibodies against capsid proteins synthesized in E. coli. Cultures were grown in the presence of IPTG and represent the following: lanes 1, pUR- 16L2; lanes 2, pUR-16L1 ; lanes 3, pHX-1 ; lanes 4, pHX-2; lanes 5, pUR-6L2; lanes 6, pUR-6L1. (a) Coomassie Brilliant Blue stain of total protein. (b) Western blot analysis of total protein with antiserum raised against fl-galactosidase. (c) Western blot analysis of total protein with antiserum raised against HPV-16 LI carboxy-terminal peptide. (d) Western blot analysis of total protein with antiserum raised against detergent-disrupted BPV-1 virions (Dako).

which is routinely used with biopsy samples to diagnose H P V infection. The results obtained are shown in Fig. 6, and demonstrate a number of interesting points. First, the BPV antiserum (Fig. 6d) had no reactivity against any of the L2-fl-galactosidase fusion proteins (lanes 1 and 5), confirming that the L2 protein is not conserved between different papillomaviruses, and is thus an ideal target for immunological typing of an HPV infection. Second, although the anti-BPV serum reacted very strongly against the HPV-6 Ll-fl-galactosidase fusion protein (lane 6), very little activity was observed against the HPV-16 L 1-fl-galactosidase fusion proteins (lanes 2 to 4). This result indicates that this antiserum was not really effective against HPV-16, and would not be an efficient or reliable reagent for diagnosing HPV-16 infection even where capsid proteins were being produced. Finally, it is apparent that the anti-BPV sera reacted against only the pHX-2 fusion protein (lane 4) and not pHX-1 (lane 3). This indicates that most of the cross- reactive epitopes which exist on the L1 protein are contained primarily within the carboxy- terminal half of the ORF.

D I S C U S S I O N

Here we have described the expression of the capsid proteins of HPV-6 and HPV-16 in E. co~i, and the production of antisera specific for the HPV-6 L1 and L2 proteins and the HPV-16 L1 protein. These reagents should prove invaluable for early diagnosis of HPV-16 infection in

Downloaded from www.microbiologyresearch.org by

IP: 54.39.17.49

On: Thu, 12 Apr 2018 12:20:35

3088 L. BANKS AND OTHERS

samples where this protein is being expressed. These antibodies should also permit detailed analysis of HPV capsid antigen expression in a variety of different cell systems.

Western blot analysis Of these proteins has revealed a number of interesting points. Although polyclonal antibodies raised against detergent-disrupted BPV-1 detect papillomavirus genus- specific antigens shared by all papillomaviruses (Orth et al., 1978; Jenson et al., 1980) it is clear that this cross-reactivity varies to a considerable extent. Here it is apparent that antiserum raised to detergent-disrupted BPV-1 reacts very well against HPV-6 L1 protein, but very poorly against HPV-16. Hence the efficacy of using such antiserum for detection of HPV-16 antigens must be questionable. It is notable that the small reactivity observed with the anti-BPV-1 serum is against the carboxy-terminal half of the L1 protein. This is significant from the point of possible vaccine development, since sequences from the carboxy-terminal half of the L1 protein are quite highly conserved between HPV types. Because of the weakness of this reaction with HPV-16 L1, studies on genital dysplasias with anti-BPV serum could give rise to misleading results. Samples that contain HPV-16 L1 may be recorded as negative, whereas those recorded as positive are most likely to contain the relatively harmless HPV-6. The discrimination between HPV-6 and HPV-16 is extremely important in view of the very different prognoses of the two types of lesion. Numerous studies have shown that HPV-6 is associated with rather benign lesions, whereas HPV-16 is associated with malignant carcinomas and advanced premalignant lesions. The failure to detect HPV-16 could therefore be serious.

It is also apparent from the above experiments that the BPV-1 antiserum is highly specific for L1 proteins of the viral capsid, and not L2, confirming that the L2 protein is type-specific as shown previously (Komly et al., 1986). Interestingly, this protein appears to be very immunogenic since antibodies raised to the HPV-6 L2-~-galactosidase fusion protein are predominantly directed against the L2 portion of the fusion protein. Again this appears particularly significant for possible vaccine development.

The antibodies described here are currently being used to analyse HPV proteins expressed in a variety of lesions. The HPV-6 Ll-/~-galactosidase fusion protein antibodies react with a protein of approximately 60K mol.wt, in Western blots of plantar warts (data not shown) and some reactivity has been observed with anti-HPV-16 LI peptide antibodies on thin sections of cervical biopsies. However, high background reactions have been observed in some cases, and for these reasons the reagents are being refined by the generation of monoclonal antibodies. Ultimately, these reagents should offer the best means of typing a given HPV infection. Also, the fusion proteins described here can also be used to screen sera from cervical cancer patients to determine whether they produce anti-HPV antibodies, and these studies are currently in progress.

We would like to thank Tony Minson for providing us with the pHX-1 and pHX-2 plasmids and Jonathan Rothbard for the peptide synthesis. We would also like to thank Pat Ward for help in preparing this manuscript . G.M. acknowledges support by N S E R C of Canada.

R E F E R E N C E S

ANDROPHY, E. J., HUBBERT, N., SCHILLER, J. ~f. LOWY, D. (1987). Identification of an HPV-16 E6 protein in mammal i an cells carrying HPV-16 D N A . EMBO Journal 6, 989-992.

BANKS, L. M., HALLIBURTON, I. W., PURIFOY, D. J. M., KILLINGTON, R. A. & POWELL, K. L. (1985). Studies on the herpes simplex virus alkaline nuclease: detection of type-common and type-specific epitopes on the enzyme. Journal of General Virology 66, 1-14.

BANKS, L., SPENCE, P., ANDROPHY, E., HUBBERT, N., MATLASHEWSKI, G., MURRAY, A. & CRAWFORD, L. (1987). Identification of h u m a n papillomavirus type 18 E6 polypeptide in cells derived from human cervical carcinomas. Journal of General Virology 68, 1351-1359.

BOSHART, M., GISSMANN, L., IKENBERG, H., KLEIN'HEINZ, A., SCHEURLEN, W. & ZUR HAUSEN, H. (1984). A new type of papillomavirus D N A , its presence in genital cancer biopsies and in cell lines derived from cervical cancer. EMBO Journal 3, 1151-1157.

DURS'r, M., GISSMANN, L., IKENBERG, H. & ZUR HAUSEN, H. (1983). A papillomavirus D N A from a cervical carcinoma and its prevalence in cancer biopsy samples from different geographic regions. Proceedings of the National Academy of Sciences, U.S.A. 80, 3812-3815.

Downloaded from www.microbiologyresearch.org by

IP: 54.39.17.49

On: Thu, 12 Apr 2018 12:20:35

H P V capsid proteins 3089

FEY, S. J., HANSEN, K., FRANDSEN, K., VETNER, M., GISSMANN, L. & LARSEN, V. M. (1986). The type of human papillomavirus present in cervical infections can be determined by the occurrence of specific marker proteins. Cell Biology International Reports 10, 905-913.

JENSON, A. B., ROSENTHAL, J. D., OLSEN, C., PASS, F., LANCASTER, W. D. & SHAH, K. (1980). immunologic relatedness of papillomaviruses from different species. Journal of the National Cancer Institute 64, 495-500.

KOMLY, C. A., BREITBURD, F., CROISSANT, O. & STREECK, R. E. (1986). The L2 open reading frame of human papillomavirus type 1 a encodes a minor structural protein carrying type-specific antigens. Journal of Virology 60, 813-816.

LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, London 227, 680-685.

McDOUGALL, I. K., BECKMANN, A. i . & KIVIAT, N. B. (1986). Methods for diagnosing papiUomavirus infection. In Papillomaviruses, Ciba Foundation Symposium 120, pp. 86-103. Edited by D. Evered & S. Clark. Chichester: John Wiley & Sons.

NELSON, L. i . , ROTHBARD, N. B., SHIVRBA, A. R., LINKE, K. H., SCHOOLNIK, G. K. & CLAYTON, A. D. (1984). All eight unassigned reading frames of mouse mitochondrial D N A are expressed. EMBO Journal 3, 3165 3175.

ORTH, G., BREITBURO, F. & FAVRE, i . (1978). Evidence of antigenic determinants shared by the structural polypeptides of (Shope) rabbit papillomavirus and h uman papillomavirus type 1. Virology 91, 243-255.

ROSENBERG, M., HO, Y. & SHATZMAN, A. (1983). The use of pKC30 and its derivatives for controlled expression of genes. Methods in Enzymology 101, 123-138.

Rf.3TI-IER, V. & MULLER-HILL, B. (1983). Easy identification of c D N A clones. EMBO Journal 2, 1791-1794. SCHNEIDER-GADICKE, A. & SCHWARZ, E. (1986). Different human cervical carcinoma cell lines show similar

transcription patterns of h u m a n papillomavirus type 18 early genes. EMBO Journal 5, 2285-2292. SCHWARZ, E., Df3RST, M., DEMARKOWSKI, C., LATTERMANN, O., ZECH, R., WOLFSPERGER, E., SUHAI, S. & ZUR HAUSEN,

H. (1983). D N A sequence and genome organisation of genital h u m a n papillomavirus type 6b. EMBO Journal 2, 2341-2348.

SCHWARZ, E., FREESE, U. K., GISSMANN, L., MAYER, W., ROGGENBUCK, B., STREMLAU, A. & ZUR HAUSEN, H. (1985). Structure and transcription of h u m a n papillomavirus sequences in cervical carcinoma cells. Nature, London 314, 111-114.

SEEDORF, K., KRAMMER, G., Df..)RST, i . , SUHAI, S. & ROWEKAMP, M. (1985). H u m a n papillomavirus type 16 D N A sequence. Virology 145, 181 185.

SMOTKIN, D. & WETrSTEIN, F. O. (1986). Transcription of h uman papillomavirus type 16 early genes in a cervical cancer and a cancer derived cell line and identification of the E7 protein. Proceedings of the National Academy of Sciences, U.S.A. 83, 4680-4684.

STANLEY, K. K. & LUZIO, J. P. (1984). Construction of a new family of high efficiency bacterial expression vectors: identification of c D N A clones coding for h u m a n liver proteins. EMBO Journal 3, 1429 1434.

ZUR HAUSEN, H. (1977). H u m a n papillomaviruses and their possible role in squamous cell carcinomas. Current Topics in Microbiology and Immunology 78, 1-30.

ZUR HAUSEN, H. (1985). Genital papillomavirus infections. In Viruses and Cancer, Society for General Microbiology Symposium vol. 37, pp. 83-90. Edited by P. W. J Rigby & N. M. Wilkie. Cambridge: Cambridge Universi ty Press.

(Received 15 May 1987)