VIROLOGY

184,460-464

(199 1)

Type-Specific

and Cross-Reactive

Epitopes in Human Papillomavirus

Type 16 Capsid Proteins

BARBARA K. BEISS,* EDGAR tiEiMER,t ARTHUR FELIX,t ROBERT D. BuRK,+§ DIANE B. RI-ITER,§ ROBERT G. MALLON, 11AND ANNA S. KADISH*~’ Departments of *Pathology, *Pediatrics, and Microbiology and Immunology, and OGynecology and Obstetrics, Albert Einstein College of Medicine, Bronx, New York 1046 1; tPeptide Research Department, Hoffmann-La Roche, Nutley, New Jersey 07110; and )IDepartment of Pathology, New York University School of Medicine, New York, New York 100 16 Received February 6, 199 I; accepted June 3, 199 I Genital human papillomavirus (HPV) 18 infection is frequently associated with cancer of the uterine cervix, as well as with precancerous lesions. In order to generate serologic reagents which might be useful in the diagnosis of HPV 16 infection, rabbit polyclonal and mouse monoclonal antisera were raised to carboxy terminal peptides from the HPV 16 Ll and L2 open reading frames (ORFs). Anti-L1 and -L2 peptide sera recognized HPV 16 Ll and L2 fusion proteins in Western blots and by immunoprecipitation. In Western blot analysis of Ll proteins from different HPV types, antisera to the Ll peptide reacted only with HPV 16, thus identifying an HPV 18 type-specific linear epitope. Anti-L2 peptide sera reacted with L2 fusion proteins from HPVs 6 and 16, but not from BPV, thus identifying a partially cross-reactive epitope in the HPV 16 L2. Computer analysis of carboxy terminal amino acid sequences of the Ll and L2 ORFs of multiple HPV types supported the Western blot findings. Despite the HPV 16 type specificity found in Western blots, anti-L1 peptide sera identified nuclear antigen by immunocytochemistry in cervical biopsies infected with HPV 16, as well as other genital HPV types. Anti-L2 peptide sera failed to recognize antigen in infected tissue. Q 1991 Academic Press, Inc.

Human papillomaviruses (HPV) are small DNA viruses which have been associated with benign and malignant proliferative lesions of squamous epithelium (reviewed in 1, 2). Among the HPV types known to infect the human genital tract, HPVs 6 and 11 have been associated with benign epithelial lesions, whereas HPVs 16, 18, 31, and 33 and other types have been found in invasive cervical carcinomas and in their precursor lesions (3). Since only small numbers of mature virions are produced in HPV-induced genital lesions, viral capsid antigens cannot be readily prepared from infected tissue, nor are they expressed in HPV-transformed cell lines or in cancers (4). Using recombinant DNA technology, HPV proteins have been expressed in bacteria and have been used to generate antibodies which recognize relevant viral antigens in tissue and in Western blots (5-9). Synthetic peptides have provided an additional source of HPV antigens for study (70). Our goals in this study were to generate antisera which might be useful in detecting HPV type-specific epitopes expressed in human lesions. Because of the oncogenic potential of HPV 16 and its relatively high prevalence in the populations screened (I 1, 12), we elected to study HPV 16. We concentrated our initial studies on the Ll and L2 proteins, which have been ’ To whom 0042-6822/91

requests

for reprints

$3.00

Copyright 0 1991 by Academic Press, Inc. All rights of reproduction m any form reserved.

should

shown to contain both type-specific and cross-reactive domains. Potentially antigenic sequences were selected on the basis of hydrophilicity and secondary structure. Peptides were synthesized by Merrifield solid phase methodology (13) and purified to homogeneity by HPLC. A cysteine residue was introduced at the amino terminus of each peptide for coupling to carrier, and keyhole-limpet hemocyanin (KLH) was coupled using N-succinimidyl-3-maleimidopropionate as cross-linker. Since antibodies to terminal peptides have been shown to be superior to those against internal regions in binding native proteins (14) carboxy terminal peptides from the HPV 16 Ll (amino acids 508-531, LGKRKATPTTSSTSTTAKRKKRKL) and L2 (453-473, YMLRKRRKRLPYFFSDVSL) were chosen for production of antisera. High titered rabbit and mouse antisera to the HPV 16 Ll and L2 peptides were generated. Peptide antibodies were detected by enzymelinked immunoadsorbant assay (ELISA) using unconjugated Ll or L2 peptide as antigen. Antisera were affinity purified on columns prepared by coupling the amino terminus of the peptides to CH-Sepharose 4B with lethyl-3-(3-dimethylaminopropyl) carbodiimide as coupling agent. In Western blots, absorbed as well as affinity-purified antisera to the HPV 16 Ll peptide specifically recognized two different recombinant HPV 16 Ll fusion proteins which include the carboxy terminus of the Ll (Fig.

be addressed.

460

SHORT a

b

9768 -

97-

COMMUNICATIONS

684343252518PLl

all

QLla

QBPV

aL2a

C

d

9768--'

9 7 -

aL1

3. :

lr

yi.

r

43-

f; j: k ii t

d r I

25-

i

68-,

16 43-2

.: “” .,

25--3

PL2

PLl

aL2

aL2a

aLla

aL2a

alla

a12a

QLlaQ8PV

a8PV

FIG. 1. Western blot of HPV 16 Ll and L2 fusion proteins. Bacterial cultures expressing HPV 16 Ll and L2 recombinant proteins plGLlBN1, p6L2NXl. and pl6L2XB5,clonedinpATHlO, weregenerously provided by Dr. Denise Galloway (5). An HPV 16 Ll construct cloned in pRAl0 (pH 16Ll) was obtained from Dr. Robert Mallon (6, 7). Western blot analysis was done usrng total lysates of induced cultures. Sera were preabsorbed with lysates of bacterial cultures not expressing HPV proteins. (a) HPV 16 Ll , pl6Ll BNl (97 kDa). (b) HPV 16 Ll, pH16Ll (39 kDa). (c) HPV 16 L2, p16L2XB5 (62 kDa) (four left lanes); HPV 6 L2, p6L2NXl (100 kDa) (last lane on right). (d) PATH 10 (no PV insert). Sera tested are indicated below each lane as follows: PLl and PL2, prebleed of rabbits immunized wrth Ll or L2; (Y, anti-L1 or -L2 peptide serum; a, affinity-purified anti-peptide serum; BPV, polyspecific anti-PV serum (Dako). Arrows indicate specific Ll or L2 bands.

1). When HPV 16 Ll fusion protein pl6Ll BNl, cloned in pATHlO (5) was tested (Fig. 1a), the reactive band was seen at 97 kDa, as expected. Western blots of HPV 16 Ll fusion protein pH16L1, cloned in pRAl0 (6), generally showed three bands (Fig. 1 b): an upper band at 39 kDa, the expected molecular weight of this fusion protein, and two lower molecularweight bands, a weak one at 33 kDa and a stronger one at 28 kDa, presumed to be degradation products of the Ll protein. Reactivity of anti-L1 peptide sera with HPV 16 Ll fusion proteins was specific in that excess Ll peptide (3-4 pg/ml), but not L2 peptide, reduced the staining of the HPV 16 Ll fusion protein and its degradation products (data not shown). Polyspecific anti-BPV serum consistently gave a weak reaction with HPV 16 Ll fusion proteins (Figs. la and lb). In Western blot analysis, anti-L1 peptide sera were type-specific in that they recognized HPV 16 Ll fusion proteins which include the carboxy terminus, but failed to recognize Ll proteins from HPV 6, 6b, 1 1, 18, and 1

461

or BPV 1 and 2 (data not shown). Anti-L1 sera did not recognize L2 proteins or lysates of cultures without PV inserts (PATH 10) (Fig. 1 a and 1d). Preimmune sera did not react with Ll or L2 fusion proteins. Thus our antiLl peptide sera recognized an HPV 16 type-specific linear antigen in Western blots. In addition, the antiHPV 16 Ll peptide serum failed to recognize an HPV 6 carboxy terminal Ll synthetic peptide of comparable size in ELISA (kindly provided by C. C. Li, The Johns Hopkins University School of Hygiene and Public Health, Baltimore, MD) (10). Type-specific antigens in HPV Ll proteins have been found by Doorbar and Gallimore using a serum against an HPV 1 recombinant protein (15) and by Li using an anti-HPV 6 Ll carboxy terminal peptide serum (10). Type-common epitopes in the Ll protein, as defined by reactivity with anti-BPV genus specific sera, have been mapped to a colinear position in HPV 6 and HPV 16, both in the amino half of the proteins (16). In Western blots, anti-L2 peptide sera recognized recombinant L2 proteins from HPV 16 (p16L2XB5, 62 kDa) and HPV 6 (p6L2NX1, 100 kDa, plus lower molecular weight degradation product) which include the carboxy terminus of the L2 protein (Fig. 1c), but did not react with Ll fusion proteins (Fig. la), with BPV 1 L2 (data not shown), or with lysates of bacteria without PV insert (pATHlO, Fig. 1d). In addition, anti-L2 peptide sera reacted with L2 fusion proteins from HPV 6b (data not shown). To further characterize the amino acid sequences of the carboxy terminal ends of the Ll and L2 proteins of HPV 16 and other HPV types and BPV, the computer program DFASTP with ktup = 2 (17) was used. The 25 carboxy terminal amino acids of the Ll proteins of the HPV and BPV types included in Genbank (HPVs 1 a, 6, 8, 11, 16, 18, and 33 and BPV 1) were compared with the carboxy terminal sequences of the HPV 16 Ll protein (Fig. 2). The HPV 16 Ll carboxy terminal sequence was not similar to any other Ll carboxy terminal sequences tested (Fig. 2A). Of all the sequences tested, BPV 1 Ll and L2 and HPV 33 Ll had the highest similarity scores with HPV 16 Ll , which were not significant There was, however, significant homology between the carboxy terminal amino acid sequences of the L2 proteins of HPV 16 and of genital HPV types 6, 1 1, 18, and 33 (Fig. 2B), but not those of HPV 1 a, HPV 8, or BPV 1 (data not shown). HPV 33 L2 was the most homologous to HPV 16 L2, with 72% identity in a 25amino acid overlap. Computer analysis of the carboxy terminal sequences of different HPV and BPV types thus supported the Western blot data, demonstrating at least one linear type-specific epitope in the carboxy terminus of HPV 16 Ll , and one partially cross-reactive epitope in the carboxy terminus of the HPV 16 L2.

462

SHORT

COMMUNICATIONS

B

INITIAL

HPV-33 1' 1"

OPTIMIZ

L2p

I

I

HPSYYMLRKRRKRLPYFFSDVSLAA .. .. .. .. . . . . . . . . . . . . . . . . . . . . . LHPSYFILRRRRKRFPYFFTDVRVAA

122

122

90

93

90

91

87

88

40

40

I HPV-18 1' 1"

A

INITIAL

BPV-1 1' 1"

HPV-33 1' 1"

1' 1"

Llp 31

43

1' 1"

I

I

I

I

HPSYYMLRKRRKRLPYFFSDVSLAA ::..:. ::::::.: :::::. FYLHPAWYFARKRRKRIPLFFSDVAA

30

37

1' 1"

L2p

I

I

I

I

HPSYYMLRKRRKRLPYFFSDVSLAA ::..:. :.::::.: ::.::. FYLHPTWYFARRRRKR IPLFFTDVAA

HPV-la

I

I

L2p

HPV-11

Llp

I TLGKRKATPTTSSTSTTAKRKKRKL :. ::. . ...::.... :.: AGCSTVRKRRISQKTSSKPAKKKKK I I

1' 1"

I I TLGKRKATPTTSSTSTTAKRKKRKL ..:.::...::.. ..: .: KAKPKLKRAAPTSTRTSSAKRKKVKK I I

BPV-1

32

.::

I HPV-6

32

I

I

HPSYYMLRKRRKRLPYFFSDVSLAA : ::...:.:::.::::.: YLWPLYYFIPKKRKRVPYFFADGFVAA

OPTIMIZ

L2p

I I TLGKRKATPTTSSTSTTAKRKKRKL .. .. .. .. .. .. HTVDLYSSNYTLHPSLLRKRKKRKHA I I

L2p

L2p

I

I

I

I

HPSYYMLRKRRKRLPYFFSDVSLAA .. .. .. .. .. .. .. IALNNSTGDFELHPSLRKRRKRAYV

FIG. 2. Alignment of PV library Ll and L2 sequences with HPV 16 Ll and HPV 16 L2. PV library sequences tested consisted of the 25 carboxy terminal amino acids of Ll and L2 proteins of HPV 1 a, 6, 8, 11, 16, 18, and 33 and BPV 1. Only PV types with the highest similarity scores are shown. Scores under 41 represent nonidentity. Scores are directly comparable for sequences of identical length. Initial scores (INITIAL) represent actual alignment, while optimized scores (OPTIMIZ) allow for insertions and deletions. (:) Denotes absolute identity, whereas (-) denotes a conservative change. (I) Denotes the boundaries of the initial alignment, (A) The test sequence HPV 16 Ll p (TLGKRKATPTTSSTSTTAKRKKRKL) is the sequence on top of each pair (1’) and is compared to the sequence below (1”) which is being evaluated. The identity score of the HPV 16 Ll sequence with itself was 102. (B) The test sequence HPV 16 L2p (HPSYYMLRKRRKRLPYFFSDVSLAA) is the sequence on top of each pair (1’) and is compared to the sequence below (1”). The identity score of the HPV 16 L2 sequence with itself was 142.

Thus, despite the evidence of considerable amino acid homology between Ll proteins from different PV types (18, 19) the carboxy terminal end of the HPV 16 protein was unique among the PVs tested. In addition, although the L2 proteins of different PVs are generally less homologous than Ll proteins, our studies have documented the presence of at least one shared epitope at the carboxy terminal ends of L2 proteins of genital HPV types. A potential use of anti-peptide sera is the isolation and identification of HPV proteins from tissue or cell extracts, or from recombinant bacterial lysates. Using our rabbit anti-HPV 16 Ll and L2 peptide sera, HPV 16 Ll and HPV 16 L2 fusion proteins were specifically immunoprecipitated from [35S]methionine-labeled bac-

terial culture lysates (Fig. 3). We were able to immunoprecipitate HPV 16 Ll protein from labeled induced culture lysates (pH16Ll) using three different antisera (Fig. 3A). An antiserum raised against a peptide (provided by Dr. R. Mallon) consisting of the first 13 amino acids of the X phage cll gene product (MVRANKRNQALRI), and comprising the leader sequence contributed by vector pRA10 in the pH16Ll fusion protein, recognized the fusion protein at its amino terminus. Our anti-HPV 16 Ll peptide serum and anti-BPV (Dako) recognized epitopes at the carboxy terminus of the HPV 16 Ll insert. Similarly, anti-HPV 16 L2 peptide serum successfully immunoprecipitated L2 fusion protein from induced bacterial lysates (p16L2XB5, Fig. 3B, Lane 4, 100 kDa).

SHORT A

COMMUNICATIONS

B 1

2

3

4

5

kD - 200

1 kD

-9,

97-

-68

CM-

2

3

4

5

*

FIG. 3. lmmunoprecipitation of HPV 16 Ll and L2 fusion proteins. Bacterial cultures expressing HPV 16 Ll , L2, and E7 proteins were generously provided by Drs. Robert Mallon (pH 16L1.6) and Denise Galloway (p16L2XB5 and pl6E7NP1,5). Cultures were grown in M56 minimal medium, induced for production of the fusion proteins, labeled with [35S]-methionine, and treated with lysozyme. Labeled lysates were incubated with relevant antisera (1:5 or 1: 10 dilutions) overnight at 4’. Washed Protein A-agarose beads were added, and the mixture was incubated 2 hr at RT. Complexes were washed extensively in RIPA buffer, precipitated, resuspended In sample buffer with dithiotriol (DTT), boiled, and loaded on gels. Gels were developed by autoradiography. (A) Total lysate of 35S-labeled culture expressing HPV 16 Ll (pH 16L1, 39 kDa) (Lane l), HPV 16 Ll lysate IP with NRS (Lane 2) IP with anti-HPV 16 Ll peptide serum (Lane 3), IP with anti-cl1 serum (Lane 4) and IP with anti-BPV (Lane 5). Arrow marks the IP HPV 16 Ll band at 39 kDa, seen in lanes 3.4. and 5. (B) 35S-labeled uninduced bacterial lysate (Lanes 1 and 2) and induced bacterial lysates expressing HPV 16 L2 (pH 16L2XB5) (Lanes 3 and 4); IP wrth NRS (Lanes 1 and 3); IP with anti-HPV 16 L2 peptide serum (Lanes 2 and 4). Large arrowhead marks the IP HPV 16 L2 band at 100 kDa in Lane 4. Culture lysate expressing HPV 16 E7 protein (p16E7NPl) IP with monoclonal anti-HPV 16 E7 antibody (Triton) (Lane 5. small arrow at 59 kDa).

In order to determine if anti-HPV 16 Ll peptide sera would detect viral capsid proteins in HPV-infected cervical lesions, anti-L1 sera were tested on formalin fixed, paraffin-embedded cervical biopsies using standard immunoperoxidase techniques. HPV DNA types infecting these women were previously identified by Southern blot analysis of cervicovaginal lavage or cervical biopsy samples (1 1, 20). Anti-L1 peptide sera gave positive nuclear staining for viral capsid antigen in areas of superficial koilocytosis in biopsies from patients infected with HPVs 6, 11, 16, 18, and 31 as did the polyspecific anti-BPV serum (Dako) used as positive control (data not shown). Biopsies that failed to stain with the polyspecific serum did not react with anti-L1 peptide sera. A plantar wart, positive with polyspecific anti-PV serum, did not stain with anti-L1 peptide serum. Despite the exquisite HPV 16 type specificity of our anti-L1 peptide serum in Western blot analysis, detec-

463

tion of Ll protein in tissue using immunocytochemical techniques gave a cross-reactive pattern in that anti-L1 peptide sera stained HPV-infected koilocyte nuclei in lesions infected with multiple genital HPV types (including 6, 1 1, 16, and 18) but not with HPV 1. The finding of cross-reactivity on tissue sections using antisera that display type-specific reactivity in Western blot is not fully understood. Doorbar et a/. found similar cross-reactivity between HPV 1 and 2 and BPV 1 Ll proteins on immunofluorescence testing of antiHPV 1a Ll sera on wart tissue, despite type-specific reactivity of the same sera on Western blots of recombinant proteins (15). Whereas epitopes in Western blots are linear and denatured, secondary and tertiary structure may be maintained in tissues. The cross-reactive staining pattern seen with anti-HPV 16 Ll peptide sera on tissues infected with different genital HPV types may possibly be due to conformational epitopes present in HPV virions in formalin-fixed tissues. In addition, the role of fixation in altering antigen expression is unclear. Five monoclonal antibodies generated to our HPV 16 Ll peptide displayed Western blot reactivity comparable to that seen with polyclonal anti-L1 sera, in that they were HPV 16 type specific (data not shown). One monoclonal antibody (of IgM subtype) reacted both on Western blot and in tissue. Tissue reactivity with this antibody was cross-reactive, as was seen with the polyclonal anti-L1 peptide sera. Thus we were unable to produce monoclonal antibodies which discriminated between Western blot and immunocytochemistry reactivity. lmmunostaining of HPV-infected cervical biopsies with anti-L2 peptide sera did not yield specific nuclear or cytoplasmic reactivity in any of the biopsies tested, suggesting either that the L2 protein is present in insufficient amounts for detection or that the specific epitope(s) detected by our sera is not exposed in formalin-fixed tissue. In summary, in using anti-peptide sera we have identified an HPV 16 type-specific epitope in the HPV 16 Ll protein and an epitope(s) shared by several genital HPV types in the L2 protein. Antisera to these antigens did not prove useful as reagents for HPV typing in tissues by immunocytochemistry. These anti-peptide sera may, however, ultimately prove useful for isolation and characterization of HPV 16 capsid proteins from human tissues. ACKNOWLEDGMENTS We thank Dr. Dentse Galloway for generously contributing fusion proteins and Chou-Chi Li for providing synthetic peptide and antiserum. We thank Dr. Peter Davies for his advice and encouragement, Julie Masch and Peter Kanetsky for their excellent technical

464

SHORT

COMMUNICATIONS

assistance, and Craig Werner for help with the computer studies. This study was supported in part by Grants CA-47630 and CA13330, from the National Cancer Institute, National Institute of Health, and a grant from the Brookdale Foundation. B.K.B. and D.B.R. were supported by National Institutes of Health training Grant CA-09173. The data in this paper are from a thesis submitted by B.K.B. in partial fulfillment of the requirements forthe degree of Doctor of Philosophy in the Sue Golding Graduate Division of Medical Sciences, Albert Einstein College of Medicine, Yeshiva University.

10.

REFERENCES

13.

1. VOUSDEN, K. H., Cancer Cells 1,43-50 (1989). 2. ARENDS, M. I., WYLLIE, A. H., and BIRD, C. C., Human Parho/. 21, 686-698 (1990). 3. WILLED, G. D., KURMAN, R. J., REID, R., GREENBERG, M., and JENSON, A. B., Int. J. Gynaecol. farhol. 8, 18-25 (1989). SMOTKIN, D.. and WETSTEIN, F. O., froc. Nat/. Acad. SC;. USA 83,4680-4684 (1986). FIRZLAFF, 1. M., HSIA, C.-N. L., HALBERT, C. P. H. L., JENISON, S. A., and GALLOWAY, D. A. Cancer Cells 5, 105-l 13 (1987). MALLON, R. G., SISK, W., and DEFENDI, V., Gene 42, 241-251 (1986). MALLON, R. G., WOJCIECHOWICZ, D., and DEFENDI, V., 1. Viol. 61, 1655-1660 (1987).

8. 9.

11. 12.

14. 15. 16. 17. 18. 19. 20.

BANKS, L., MATLASHEWSKI. G., PIM, D., CHURCHER, M., ROBERTS, C., and CRAWFORD, L., 1. Gen. Viral. 68, 3081-3089 (1987). PATEL, D., SHEPHERD, P. S., NAYLOR, J. A., and MCCANCE, D. J.,f. Gen. Viral. 70, 69-77 (1989). LI, C. C., SHAH, K. V., SETH, A., and GILDEN, R., J. l&o/. 61, 2684-2690 (1987). KADISH, A. S., BURK, R. D., KRESS, Y., CALDERIN, S., and ROMNEY, S. L., Human Parho/. 17, 384-392 (1986). RITTER, D. B., KADISH, A. S., VERMUND, S. H., ROMNEY, S. L., VILLARI, D., and BURK, R. D., Am. 1. Obsfef. Gynecol. 159, 1517-1525 (1988). MARGOLIN, A., and MERRIFIELD, R. B., Annu. Rev. Biochem. 39, 841-866 (1970). FRIEDRICH, U., SCHEIDTMANN, K. H., and WALTER, G., lmmunol. Len 12,207-215 (1986). DOORBAR, J., and GALLIMORE, P. H., f. l&o/. 61, 2793-2799 (1987). JENISON, S. A., FIRZLAFF, J. M., LANGENBERG, A., and GALLOWAY, D. A., /. viral. 62, 2115-2123 (1988). LIPMAN, D. J., and PEARSON, W. R., Science 227, 1435-1441 (1985). BROKER, T. R., and CHOW, L. T., Cancer Cells 4,589-594 (1986). NAKAI, Y., LANCASTER, W. D., LIM, L. Y., and JENSON, A. B., Infervirology 25, 30-37 (1986). BURK, R. D., KADISH, A. S., CALDERIN, S., and ROMNEY, S., Am. J. Obstet. Gynecol. 154, 982-989 (1986).

Type-specific and cross-reactive epitopes in human papillomavirus type 16 capsid proteins.

Genital human papillomavirus (HPV) 16 infection is frequently associated with cancer of the uterine cervix, as well as with precancerous lesions. In o...
585KB Sizes 0 Downloads 0 Views