189, 132-l 40 (1992)
VIROLOGY
Human Papillomavirus Type 16 (HPV 16) Gene Expression and DNA Replication in Cervical Neoplasia: Analysis by in Situ Hybridization MAl-THIAS DURST,’ DAGMAR GLITZ, ACHIM SCHNEIDER,* AND HARALD ZUR HAUSEN Deutsches
Krebsforschungszentrum,
Forschungsschwerpunkt Angewandte Tumorvirologie, Im Neuenheimer and *Frauenklinik, Universitat U/m, Prittwitzstrasse 43, 7900 Ulm Received December
26, 199 1; accepted
Feld 280, D-6900 Heidelberg
1;
March 20, 1992
We have analyzed human papillomavirus (HPV) type 16 RNA expression in premalignant cervical lesions of different severity and in squamous cervical cancers by RNA-RNA in situ hybridization in order to find differences in the topographic distribution of viral RNA, which might correlate with the severity of disease. In the basal layer of low-grade squamous intraepithelial lesions (SIL) only weak transcription of viral early genes was observed. Signal intensity increased strongly in the more differentiated cells accompanied by high levels of HPV DNA replication. This pattern of viral gene expression, together with the onset of viral late transcription in the upper differentiated layer of the epithelium, most likely reflects the productive phase of viral infection. In contrast, in high-grade SIL viral transcription was comparatively strong in basal cells and evenly distributed throughout the undifferentiated epithelium. This difference of viral transcription in the basal layer of the respective lesions points to an altered regulation of viral gene expression which may be causally linked to the progression of precursor lesions. Evidence for disrupted expression of 3’ early genes (EZ, E4, and E5), analogous to the situation in HPV-DNA containing cervical carcinoma-derived cell lines, was not found in any of the HPV-16-positive premalignant lesions nor in the majority of cancers. The similarity of the viral transcription pattern of high-grade SIL and cancers suggests that additional host gene alterations are necessary for malignant progression. 0 1992 Academic Press, Inc.
INTRODUCTION
netic dissection of the viral genome has mapped the immortalizing function to the viral genes E6 and E7 (Barbosa et a/., 1989; Hawley-Nelson et a/., 1989; Milnger et al., 1989a; Barbosa et a/., 1991). Transformation is believed to result in part from the specific interaction of the viral E6 and E7 gene products with the cell regulatory protein ~53 and the product of the retinoblastoma tumor suppressor gene (Rb), respectively (Dyson et al., 1989; Mijnger et al., 1989b; Werness et a/., 1990). Compared to HPV 16 and 18 the E6 and E7 proteins of HPV 6 and 11 show reduced complex formation with these cellular proteins (Munger et al., 1989b; Gage et a/., 1990; Scheffner et al., 1990). The importance of viral E6 and E7 proteins for transformation is further substantiated by the fact that these viral ORFs are conserved and expressed in HPV-16positive premalignant and malignant tumors (Shirasawa et al., 1988; Crum et al., 1989; Cornelissen et al., 1990; Stoler et a/., 1990a; Higgins et al., 1991) despite frequent integration of the viral genome into the host DNA in some cancers (DOrst et al., 1985, 1986). AdditTonally, integration often leads to the disruption of the viral E2 gene (Schwarz et al., 1985), whose gene product was shown to function in part as a repressor of E6-E7 transcription (Romanczuk eta,!, 1990). Inactivation of the repressor function of E2 may therefore be responsible for the high steady state levels of E6-E7 RNA in tumors. Moreover, the importance of viral E6-
Human papillomaviruses (HPV) are epitheliotrophic viruses known to infect both human skin and mucosa. A subset of these viruses infects the anogenital tract. HPV 6 and 11 usually induce benign proliferations of the genital skin and mucosa. Others such as HPV 16, 18, 31, 33, and 35 have been linked to premalignant lesions and cancer of the anogenital tract in particular of the uterine cervix and are therefore considered to be high-risk viruses (zur Hausen and Schneider, 1987). Because of the lack of a suitable cell culture system or animal models for the cultivation of HPV our current understanding of the biology of these viruses is mainly based on experiments performed with molecularly cloned viral DNA. To some extent the clinical association of different HPV types correlates remarkably well with the biological activity of the respective viral DNA in cell culture systems. Most pertinent in this regard is the immortalization of human primary keratinocytes in response to transfection with carcinoma-associated HPV types (Diirst et a/., 1987; Pirisi et al., 1987; Kaur and McDougall et al., 1989). This property is not shared by HPV 6 or 1 1 (Schlegel et al., 1988; Pecoraro et al., 1989; Woodworth et a/., 1989). Subsequent ge’ To whom reprint requests should be addressed. * Frauenklinik, Universitat Ulm. Prittwitzstrasse 43, 7900 Ulm. 0042.6822192
$5.00
Copyright Q 1992 by Academrc Press, inc. All rights of reproduction !n any form reserved.
132
HPV TRANSCRIPTION
AND REPLICATION
E7 in the maintenance of the transformed phenotype was demonstrated by downmodulating viral E6-E7 expression in a HPV-positive cell line, which resulted in reduced cell growth and loss of tumorigenicity (von Knebel-Doeberitz et al., 1988). Our working model of HPV-linked carcinogenesis is primarily based on the above-described experimental systems as well as on the biochemical properties of the viral gene products. The obvious necessity to correlate this data with corresponding molecular events that take place in naturally occurring HPV-associated tumors, prompted us to analyze HPV 16 transcription and viral DNA replication in the context of the different histopathological alterations characteristic for a range of squamous intraepithelial lesions (SIL) of the cervix and squamous cervical cancer by in situ hybridization. MATERIALS
AND METHODS
Biopsies Punch biopsies were taken from 60 patients under colposcopic guidance. Biopsies were immediately frozen in cold (-70”) 2-methylbutane for subsequent storage in liquid nitrogen. Histological examination revealed 14 low-grade SIL and 24 high-grade SIL (SIL = Squamous intraepithelial lesion, from the proposed Bethesda terminology (Hum. Pathol. 21,704-708, 1990) which includes cervical intraepithelial neoplasia (CIN) or other epithelial abnormalities of the cervix). The remaining biopsies were either histologically normal or consisted mainly of connective tissue. Tissues from 19 cervical squamous cell carcinomas were taken from radical hysterectomy specimens at the time of surgery and were frozen and stored in an identical fashion. RNA-RNA
in situ hybridization
Serial cryostat sections mounted on 3-aminopropyltriethoxysilane-coated slides were fixed in 4% paraformaldehyde in 2X SSPE (0.3 n/l NaCI, 20 mM NaH,PO,, 2 mM EDTA) for 1O-l 5 min at room temperature, digested with proteinase K (0.5 pglml) for 10 min at 37”, and hybridized with strand-specific RNA probes of both orientations spanning part of the upstream regulatory region (URR) as well as parts of the early and late region (Fig. 1) (Seedorf et a/., 1985). Briefly, radioactively labeled RNA probes were generated in the transcription vector Bluescribe using either T3 or T7 RNA polymerase, 100 &i of 32P-UTP (800 Ci/mmol from Amersham), and 0.25 mM of each remaining precursor as cold substrate. This yielded an RNA with a specific activity of 10’ cpm/pg. After DNase treatment, the probes were subjected to limited alkaline hydrolysis. Hybridization was performed overnight at 42” in a
IN CERVICAL
NEOPLASIA
133
solution containing 50% formamide, 2X SSPE, 10% w/v dextran sulfate, 10 mM Tris, pH 7.5, 1X Denhardt’s solution (0.02% each bovine serum albumin, Ficoll, and polyvinylpyrollidone), 500 pg/ml tRNA, 100 yglml herring sperm DNA (sonicated), 0.1 o/oSDS, and 1O5 cpm probe/PI. These conditions approximately equal Tm -25” to -30” for RNA-RNAin situ hybridization (Cox et a/., 1984; Bodkin and Knudson, 1985). Sections were washed in 50% formamide, 2X SSPE, 0.1% SDS for 30 min at 50”, treated with RNase A (50 pg/ml) for 30 min at 37”, and subsequently washed in 50% formamide, 0.1 X SSC, 0.1 O/oSDS for 30 min at the same temperature (Tm -10” to -1 So). Slides were dehydrated in graded alcohols containing 300 mMammonium acetate, dried, and dipped in Kodak NTB 2 emulsion diluted 1 :l in 600 mM ammonium acetate. After storage for 1-14 days, the slides were developed in Kodak D-l 9 developer, fixed, counterstained with haematoxylin and eosin, and coverslipped. The sections were examined and photographed under a Zeiss Axiophot microscope equipped with a rotatable lightfield and darkfield condenser to reveal tissue histology and to enable good signal visualization, respectively. RNA-DNA
in situ hybridization
HPV DNA was detected in tissue sections by using a modification of the RNA-RNA in situ hybridization protocol. Best results were achieved by denaturing the target DNA prior to fixation. In brief, serial cryostat sections mounted on 3-aminopropyl-triethoxysilanecoated slides were air dried only. The slides were then incubated in 50% formamide, 0.5~ SSC for 15 min at 70”, followed by fixation in 4% paraformaldehyde, 2x SSPE for 15 min at room temperature. Thereafter the slides were dipped several times in 2x SSC and subjected to proteinaseK digestion (0.5 pg/ml in 2x SSC for 10 min at 37”) and subsequent RNaseA digestion (50 pg/ml in 2X SSC for 30 min at 37”). Enzymatic activity was neutralized by incubation in glycin. Preceding hybridization, both probe (strand-specific RNA probes spanning to entire HPV 16 genome) and target DNA were denatured for 15 min at 95”. After incubation overnight at 42” the slides were washed in 50% formamide, 2X SSPE, 0.1% SDS for 30 min at 42”, treated with RNase A (as for RNA-RNA hybridization) and subsequently washed in 50% formamide, 0.5x SSC, 0.1% SDS for 30 min at 37”. RESULTS All biopsies were screened for the presence of HPV 16 DNA by RNA-DNA in situ hybridization. Three of 14 low-grade SIL and 13 of 24 high-grade SIL but none of the normal tissues were HPV-16-positive. Thirteen of
134
DijRST
w
al6 E&S7
s?aa
ET AL.
so3
CT84
52s
%a6 7lm IJRR
l2
E244 26&l
cm
a153
E2-E5-I2
nlw Ll
FIG. 1. The organization of the HPV 16 genone (Seedorf eta/., 1985) Including adjustments m the E 1 ORF {Matsukura et al., 1986), the E5 ORF (Bubb et a/., 1988; Halbert and Galloway, 1988) and the URR E2 protein binding site (Ronanczuk eta/., 1990) IS shown on top. ORFs are referred to by open boxes. The indicated subgenomic HPV 16 fragments were cloned into a transcription vector in order to generate strand-specific riboprobes for in situ hybridization. -
nineteen squamous cervical cancers were also positive. Theviral transcription pattern of each HPV-16-positive lesion or cancer was then examined in detail. Hybridization probes were derived from the URR and from the early and late region of the HPV 16 genome (Fig. 1). Low-grade
SI L
In all lesions viral early gene expression increased significantly in differentiated cells accompanied by high levels of HPV DNA replication. Moreover, transcripts encoding structural proteins were evident. Compared to early transcripts, the proportion of cells expressing late transcripts was smaller and the signals were confined to the more terminally differentiated cells (Fig. 2, upper panel). This pattern of gene expression was also observed in adjacent sections of the same lesions. High-grade
SIL
Similar to low-grade lesions high-grade SIL also expressed the entire early region of the viral genome, but signal distribution was more uniform. Late viral genes were not expressed and viral DNA replication was generally low (Fig. 2, lower panel).
Squamous
cervical
cancer
In undifferentiated cancers the viral expression pattern was similar to that of high-grade SIL (Fig. 3, lower panel). In some cancers viral gene expression was heterogeneous, e.g., expression of late transcripts which correlated with marked viral DNA replication (Fig. 4, upper panel). In view of the frequently observed inactivation of the viral E2 ORF in carcinoma-derived cell lines, expression of this particular ORF in cancer biopsies was of particular interest. Only in two cancer biopsies were no hybridization signals obtained with a hybridization probe corresponding to the immediate 3’ part of the E2 ORF (probe E2-E5-L2 in Fig. 1) (data not shown). In addition to viral mRNA, some cancers also expressed viral RNA of antisense orientation (Higgins et al., 1991; Vormwald-Dogan et al., 1992). Differences in viral transcriptional basal layer of SIL
activity
in the
Differences in hybridization signal intensity for viral early transcripts were observed consistently in the basal layer of different SIL. In high-grade lesions signals for E6-E7, E2-E4, and E2-E5-L2 probes (Fig. 1) were
FIG. 2. In siru hybridization analysis of a low-grade and a high-grade SIL (CIN I and CIN Ill). Photomicrographs of serial sections of each lesion are shown. Sections a and f are hematoxylin- and eosin-stained only. Viral mRNA was detected in sections b-d and g-i using strand-specific probes as indicated (see also Ftg. 1). Small arrowheads in band g mark the epithelial basement membrane. In sections e and j vrral DNA only was detected. Specrfic hybridization signals are seen as white spots by darkfield microscopy. Magnification Xi 60.
HPV TRANSCRIPTION
AND REPLICATION
IN CERVICAL
NEOPLASIA
135
136
DijRST
distributed uniformly throughout the undifferentiated epithelium. In comparison, viral transcription in the basal layer of two low-grade lesions was markedly reduced (compare in Fig. 2b and g and in Fig. 4b and d). This difference in the level of viral early gene expression in the basal layer of various premalignant lesions was confirmed by repeated examination of these lesions under identical experimental conditions. DISCUSSION ln situ hybridization permits detection of viral DNA and RNA in the context of a highly complex tissue architecture. This is particularly important when correlating viral gene expression with pathological alterations. In situ hybridization complements a wide range of biochemical analyses that enable the identification of individual viral mRNA species but which depend on extraction of total RNA from tissue. From these analyses an increasingly more complex pattern of viral transcripts is emerging (Dilts eta/., 1990; Doorbar eta/., 1990; Sherman et al., 1992). Because several mRNA transcripts show considerable overlap especially in the 3’ early region, differentiation between individual mRNA species by in situ hybridization is not always feasible. Our analysis therefore does not attempt to differentiate between individual mRNA species but is restricted to viral transcripts of the 5’- and 3’-early region and to the late region of HPV 16. Low-grade SlLs There was an overall increase in transcription of the entire early region concomitant with increasing cellular differentiation (Crum et a/., 1988; Stoler et a/., 1992). This type of expression pattern was also described for HPV-11-induced condylomata (Stoler et a/., 1989, 1990b). An increase in viral transcription could be explained on the basis of a strong increase in viral replication, which provides a high level of DNA template for transcription, or as a result of derepression of viral transcription or both (Dijrst et a/,, 1991). Whether an increase in viral early transcription actually correlates with higher levels of viral protein has so far only been shown in the case of E4. Recent experiments by Crum et a/. (1990) have shown that the viral E4 protein (most likely derived from a polycistronic mRNA comprising the E6*I, E7, El,E4, and E5 ORF) is expressed at high levels in the upper differentiated layer of low-grade SIL.
ET AL.
The HPV 16 E4 protein interacts and destabilizes the cytokeratin network which may be essential for ensuring efficient virus particle release (Doorbar et al., 199 1). In our study, transcription of the Ll region was detected in low grade SIL only and seems to be a hallmark for productive viral lesions. Several mechanisms of regulation of late viral transcription are possible: Transcription downstream of the early polyadenylation site could depend on a particular state of differentiation Alternatively, late transcripts could be subject to degradation unless otherwise protected (Kennedy et al., 1990, 1991). Also the use of differentiation-dependent viral intragenomic promoters has been implicated but this requires further verification. By employing the extremely sensitive technique of reverse transcription combined with PCR, viral transcripts of the late region could be detected in most cervical neoplasias irrespective of histological grading (V. Vormwald-Dogan, personal communication). This suggests that very low levels of viral late transcripts are indeed present in most lesions. However, it is not clear whether these transcripts are partially degraded or biologically inactive. High-grade SIL and cervical cancers High-grade lesions and undifferentiated cervical cancers showed a uniform distribution of hybridization signals throughout the thickness of the epithelium for all probes derived from the early region of the HPV 16 genome. In cancers transcription of the 3’ part of the early region is of particular interest since it contrasts the situation in HPV-16-positive cervical carcinomaderived cell lines (Schneider-Gadicke and Schwarz, 1986; Smotkin and Wettstein, 1986; Baker et al., 1987). In these cell lines viral integration into the host DNA has uncoupled the E2 ORF from the major viral promoter (P97). The function of the HPV 16 E2 protein in the natural context of HPV infection is believed to be in part that of a transcriptional repressor for P97. Therefore, lack of E2 expression could lead to increased steady state levels of viral early transcripts and possibly to increased amounts of transforming protein. However, breakpoints for viral integration other than those of the El/E2 region are known (Durst et al., 1986) and several HPV-16-positive tumors harbor not only integrated but also either both integrated and episomal viral molecules or episomal molecules only (Diirst et
FIG. 3. In situ hybridization of a moderately differentiated (Ca. 34) and an undifferentiated (Ca. 1 12) squamous cervical cancer. Photomicrographs of serial sections of each lesion are shown. Sections a and f are hematoxylin- and eosin-stained only. Viral mRNA was detected in sections b-d and g-i using strand-specific probes as indicated (see also Fig. 1). In sectrons e and j viral DNA only was detected. Specific hybridization signals are seen as white spots by darkfield microscopy. Magnification X 160.
HPV TRANSCRIPTION
d W @J W
5 CD W
w I’
: m
0
AND REPLICATION
IN CERVICAL
NEOPLASIA
137
138
DijRST
ET AL.
E6-E7
CIN
I
GIN m
FIG. 4. Detection of viral E6-E7 mRNA in a low-grade and a high-grade SIL (CIN I and CIN Ill) by RNA-RNA in situ hybridization. Brightfield (a and c) and darkfield (b and d) views of one section is shown for each lesion. Small arrowheads in b and d mark the epithelial basement membrane. Magnification X220
al., 1985, Matsukura et al., 1989; Fukushima et al., 1990). Thus, the attributed role of viral E2 in transformation may be somewhat oversimplified (Nasseri eta/., 1991). The detection of viral transcripts spanning the 3’ part of the E2 ORF in 1 1 of 13 cancers examined in this study supports this notion. A model for HPV-linked carcinogenesis Our analyses of HPV transcription in preneoplastic lesions are in support of a postulated malfunctional control of viral transcription as a necessary component for neoplastic progression (zur Hausen, 1986, 1991). The low levels of viral early transcripts observed in the basal layer of two low-grade SIL are likely to be due to specific downregulation of viral transcription by an as yet unknown cellular control mechanism. Progression of high-grade SIL may result from an unscheduled increase in E6-E7 gene expression in basal cells. Evidence in support of suppression of viral transcription in
HPV-infected cells as a consequence of failing intracellular regulation is also provided by experimental systems. Inoculation of HPV-16-immortalized keratinocytes or nontumorigenic somatic cell hybrids into nude mice resulted in a remarkable reduction of E6-E7 gene expression (Bosch eta/., 1990; Dijrst eta/., 1991). This contrasts the behavior of malignant cells which continue to transcribe E6-E7 genes at a high level. Evidence for the existence of the postulated repressor molecule is given by the suppressive effect of host cell genes on the viral URR which controls E6-E7 gene expression. Under tissue culture conditions constitutive expression of an indicator gene (CAT) under the control of the HPV 18-URR is obtained in malignant cervical carcinoma cells after fusion of these cells with a nonmalignant cells (R&l et a/., 1991). Suppression occurs at the level of initiation of transcription and is eliminated after cycloheximide treatment. Dysregulation of viral transcription therefore seems to be an early step in neoplastic progression. Clearly,
HPV TRANSCRIPTION
AND REPLICATION
this could also be brought about by integration of the viral genome into the host DNA, which may not only disrupt the regulation of viral gene expression by viral factors but may also subject the viral genome to the influence of host-specific c&-acting elements (von Knebel Doeberitz et al., 1991). However, this does not seem to be the case for HPV-16-associated SIL since with few exceptions most of the precancers harbor episomal viral DNA only (DOrst et al., 1985; Cullen et a/., 1991). In contrast, the majority of HPV 18 positive SIL do not express the 3’ part of the early viral region possibly as a consequence of viral integration (Stoler et a/., 1992). There is little difference between the HPV 16 transcriptional pattern of high grade SIL and undifferentiated squamous cell carcinoma. This suggests that for malignant conversion of high-grade SIL, additional host gene alterations are necessary, possibly unrelated to papillomavirus gene expression. ACKNOWLEDGMENTS This work was supported by the Deutsche Forschungsgemeinschaft (DO 162/l -2). We are grateful to Vera Vormwald-Dogan and Renate Webler for their assistance in the preparation of the artwork.
REFERENCES BAKER, C. C., PHELPS. W. C., LINDGREN, V., BRAUN, M. J., GONDA, M. A., and HOWLEY. P. M. (1987). Structural and transcriptional analysis of human papillomavirus type 16 sequences in cervical carcinoma cell lines. /. viral. 61, 962-971. BARBOSA, M. S., and SCHLEGEL, R. (1989). The E6 and E7 genes of HPV-18 are sufficient for inducing two-stage in vitro transformation of human keratinocytes. Oncogene 4, 1529-l 532. BARBOSA,M. S., VASS, W. C.. Lowv, D. R.. and SCHILLER,1. T. (1991). In vitro biological activittes of the E6 and E7 genes vary among human papillomavlruses of different oncogenic potential. /. v;ro/. 65,292-298. BODKIN, D. K., and KNUDSON, D. L. (1985). Sequence relatedness of polyam virus genes to cognates of the polyam serogroup viruses by RNA-RNA blot hybridization. Virology 143, 55-62. BOSCH, F. X.. SCHWARZ,E., BOUKAMP. P., FUSENIG,N. E., BARTSCH,D., and ZUR HAUSEN, H. (1990). Suppression in vivo of human papillomavirus type 18 E6-E7 gene expression In nontumorigenic HeLa X fibroblast hybrid cells. /. Viral. 64, 4743-4754. BUBE, V., MCCANCE, D. J., and SCHLEGEL,R. (1988). DNA sequences of the HPV 16 E5 ORF and the structural conservation of its encoded protein. Virology 164, 234-346. CORNELISSEN,M. T. E., SMITS, H. L., BRIET, M. A., VAN DEN TWEEL, J. G., STRUYK,A. P. H. B.. VAN DER NOORDAA, 1.. and TER SCHEGGET, I. (1990). Uniformity of the splicing pattern of the E6/E7 transcripts in human papillomavirus type 16.transformed human fibroblasts, human cetvical premalignant lesions and carcinomas. /. Gen. viral. 71, 1243-1246. Cox, K. H., DELEON, D. V., ANGERER,L. M., and ANGERER,R. C. (1984). Detection of mRNAs in sea urchin embryos by in situ hybridization using asymmetnc RNA probes. Dev. Biol. 101, 485-502. CRUM, C. P., SYMBULA, M., and WARD, B. E. (1989). Topography of
IN CERVICAL
NEOPLASIA
139
early HPV 16 transcription in high-grade genital precancers. Am. J. Pathol. 134, 1183-l 188. CRUM, C. P., Nuovo, G., FRIEDMAN, D., and SILVERSTEIN,S. J. (1988). Accumulation of RNA homologous to human papillomavirus type 16 open reading frames in genital precancers. /. Viral. 62, 84-90. CRUM. C. P., BARBER,S., SYMBULA, M., SNYDER, K., SALEH. A. M.. and ROCHE, I. K. (1990). Coexpresslon of the human papillomavirus type 16 E4 and Ll open reading frames in early cewical neoplasia. Virology 178, 238-246. CULLEN, A. P., REID, R.. CAMPION, M., and L~RINCZ, A. T. (1991). Analysis of the physical state of different human papillomavirus DNAs in lntraepithelial and invasive cervical neoplasm. /. Viral. 65, 606-612. DILTS, D. P., BROKER,T. R., and CHOW, L. T. (1990). The structures of the human papillomavlrus type 1 and type 16 messenger RNAs determined by polymerase chain reaction. in “Papillomaviruses” (P. M. Howley and T. R. Broker, Eds.), pp. 533-540. A. R. Liss, New York. DOORBAR.J., PARTON, A., HARTLEY, K., BANK, L.. CROOK, T., STANLEY, M., and CRAWFORD,L. (1990). Detection of novel splicing patterns in a HPV 16.containing keratinocyte cell line. Virology 178, 254262. DOORBAR, J., ELY, S., STERLING, J., MCLEAN, C., and CRAWFORD, L. (1991). Specific interaction between HPV 16 El -E4 and cytokeratins results in collapse of the eplthelial cell Intermediate filament network. Nature 352, 824-827. DORS~, M., KLEINHEINZ,A., HOTZ, M., and GISSMANN, L. (1985). The physical state of human papillomavlrus type 16 DNA In benign and mallgnant genital tumours. /. Gen. Viroi. 66, 1515-l 522. DORST, M., SCHWARZ, E., and GISSMANN, L. (1986). Integration and persistence of human papillomavlrus DNA In genital tumors. In “Viral etiology of cervical cancer” (R. Peto and H. zur Hausen, Eds.), pp. 273-280. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. DORST, M., DZARLIEVA-PETRUSEVSKA,R. T., BOUKAMP, P., FUSENIG, N. E., and GISSMANN, L. (1987). Molecular and cytogenetic analySIS of Immortalized human primary keratlnocytes obtained after transfectlon with human paplllomavirus type 16 DNA. Oncogene 1, 251-256. DORST, M., BOSCH, F. X., GLITZ, D., SCHNEIDER,A., and ZUR HAUSEN, H. (1991). Inverse relationship between HPV 16 early gene expression and cell differentiation In nude mice eplthelial cysts and tumors Induced by HPV positive human cell lines. I. L&o/. 65, 796804. DYSON, N., HOWLEY. P. M., MONGER, K., and HARLOW, E. (1989). The human papillomavlrus 16 E7 oncoproteln is able to bind to the retinoblastoma gene product. Science 243, 934-937. FUKUSHIMA, M., YAMAKAWA, Y., SHIMANO, S., HASHIMOTO, M., SAWADA. Y., and FUJINAGA, K. (1990). The physlcal state of human paplllomavlrus 16 DNA in ceNlcal carcinoma and cervical Intraepithellal neoplasla. Cancer 66, 2 155-2 16 1. GAGE, 1. R., MEYERS,C., and WETTSTEIN,F. 0. (1990). The E7 proteins of the nononcogenlc human papillomavlrus type 6b (HPV-6b) and of the oncogenic HPV-16 differ in retinoblastoma protein binding properties. /. viral. 64, 723-730. HALBERT. C. L., and GALLOWAY, D. A. (1988). Identification of the E5 open reading frame of human papillomavlrus 16. /. Viral. 62, 1071-1075. HAWLEY-NELSON,P., VOUSDEN, K. H., HUBBERT.N. L., Lowv, D. R., and SCHILLER. J. T. (1989). HPV 16 E6 and E7 proteins cooperate to immortalize human foreskin keratlnocytes. E/I&30 /. 8, 39053910. HIGGINS, G. D., UZELIN, D. M., PHILLIPS, G. E., and BURRELL, C. J. (1991). Presence and distribution of human paplllomavirus sense
140
DijRST
and antisense RNA transcripts in genital cancers. J. Gen. Viral. 72, 885-895. KAUR, P., and MCDOUGALL, J. K. (1989). HPV-18 immortalization of human keratinocytes. Virology 173, 302-310. KENNEDY,I. M., HADDOW, J. K., and CLEMENTS,J. B. (1990). Analysis of human papillomavirus type 16 late mRNA 3’ processing signals in vitro and in vivo. J. Viral. 64, 1825-l 829. KENNEDY,I. A., HADDOW, J. K., and CLEMENTS, J. B. (1991). Anegative regulatory element in the human papillomavirus type 16 genome acts at the level of late mRNA stability. J. Viral. 65, 2093-2097. MATSUKUFIA,T., KANDA, T., FURUNO, A., YOSHIKAWA, H., KAWANA, T., and YOSHIIKE, K. (1986). Cloning of monomeric human papillomavirus type 16 DNA integrated within cell DNAfrom a cervical carcinoma. J. Viral. 58, 979-982. MATSUKURA,T.. KOI, S., and SUGASE, M. (1989). Both episomal and integrated forms of human papillomavirus type 16 are involved in invasive cancer. Virology 172, 63-72. MONGER, K., PHELPS,W. C., BUBB, V., HOWLEY, P. M., and SCHLEGEL, R. (1989a). The E6 and E7 genes of the human papillomavirus type 16 together are necessary and sufficient for transformation of primary human keratinocytes. J. Viral. 63, 4417-442 1. MONGER, K., WERNESS,B. A., DYSON, N., PHELPS,W. C., HARLOW, E., and HOWLEY, P. M. (1989b). Complex formation of human papillomavirus E7 proteins with the retinoblastoma tumor suppressor gene product. EMBO /. 8,4099-4105. NASSERI, M., GAGE, J. R.. LORINCZ, A., and WEITSTEIN, F. 0. (1991). Human papillomavirus type 16 immortalized cervical keratinocytes contain transcripts encoding E6, E7, and E2 initiated at the P97 promoter and express high levels of E7. virology 184, 131-140. PECORARO,G.. MORGAN, D., and DEFENDI, V. (1989). Differential effects of human papillomavirus type 6, 16, and 18 DNAs on immortalization and transformation of human cervical epithelial cells. Proc. Nat/. Acad. Ski. USA 86, 563-567. PIRISI, L., YASUMOTO. S.. FELLER, M., DONIGER, J., and DIPAOLO, J. A. (1987). Transformation of human fibroblasts and keratinocytes with human papillomavirus type 16 DNA. J. Viral. 61, 1061-l 066. ROMANCZUK, H., THIERRY, F., and HOWLEY, P. M. (1990). Mutational analysis of cis elements involved in E2 modulation of human papillomavirus type 16 P,, and type 18 P,,, promoters. /. Viral. 64, 2849-2859. R&L, F., ACHTSTXITER, T., BAUKNECHT,T.. HUITER, K.-J., FUT~ERMAN, G.. and ZUR HAUSEN, H. (1991). Extinction of HPV 18 upstream regulatory region in cervical carcinoma cells after fusion with nonmalignant keratinocytes under non-selective conditions. fI\/IBO 1. 10,1337-1345. SCHEFFNER,M., WERNESS,B. A., HUIBREGTSE,J. M., LEVINE,A. J., and HOWLEY, P. M. (1990). The E6 oncoprotein encoded by human papillomavirus types 16 and 18 promotes the degradation of ~53. Cell 63, 1129-l 136. SCHLEGEL, R., PHELPS,W. C., ZHANG, Y.-L., and BARBOSA, M. (1988). Quantitative keratinocyte assay detects two biological activities of human papillomavirus DNA and identifies viral types associated with cervical carcinoma. EMBO 1. 7, 3181-3187. SCHNEIDER-GADICKE,A., and SCHWARZ, E. (1986). Different human cervical carcinoma cell lines show similar transcription patterns of human papillomavirus type 18 early genes. EM80 J. 5, 22852292. SCHWARZ, E., FREESE,U. K., GISSMANN, L., MAYER, W., ROGGENBUCK, B., STREMLAU, A., and ZUR HAUSEN, H. (1985). Structure and transcription of human papillomavirus sequences in cervical carcinoma cells. Nature 314, 1 1 l-l 14.
ET AL. SEEDORF, K., K~~MMER, G., DORST, M., SUHAI, S.. and R~WEKAMP, W. G. (1985). Human papillomavirus type 16 DNA sequence. viral0gy 145,iai-185. SHERMAN, L., GOLAN, I., ALLOUL, N., DURST. M., and BAFLAM,A. (1992). Expression and splicing pattern of human papillomavirus type 16 mRNAs in preinvasive and invasive carcinomas of the cervix, in human keratinocytes immortalized by HPV 16, and in cell lines established from cervical cancers. Int. J. Cancer 50, 356-364. SHIRASAWA,H., TOMITA, Y.. KUBOTA, K., KASAI. T., SEKIYA,S., TAKAMIZAWA, H., and SIMIZU, B. (1988). Transcriptional differences of the human papillomavirus type 16 genome between precancerous lesions and invasive carcinomas. /. Viral. 62, 1022-l 027. SMOTKIN, D., and WETTSTEIN, F. 0. (1986). Transcription of human papillomavirus type 16 early genes in a cervical cancer and a cancer-derived cell line and identification of the E7 protein. Proc. Nat/. Acad. Sci. USA 83,4680-4684. STOLER, M. H., WOLINSKY, S. M.. WHITBECK, A., BROKER,T. R., and CHOW, L. T. (1989). Differentiation-linked human papillomavirus types 6 and 11 transcription in genital condylomata revealed by in situ hybridization with message-specific RNA probes. Virology 172, 331-340. STOLER, M. H., RHODES, C. R., WHITBECK, A., CHOW, L. T., and BROKER,T. R. (1990a). Gene expression of HPV 16 and 18 in cervcal neoplasia. In “Papillomaviruses” (P. M. Howley and T. R. Broker, Eds.), pp. l-l 1. Wiley-Liss, New York. STOLER. M. H., WHITBECK,A., WOLINSKY, S. M., BROKER,T. R., CHOW, L. T., HOWET, M. K., and KREIDER,1. W. (1990b). Infectious cycle of human papillomavirus type 11 in human foreskin xenografts in nude mice. /. Viral. 64, 3310-3318. STOLER. M. H., RHODES,C. R., WHITBECK.A., WOLINSKY,S. M., CHOW, L. T., and BROKER,T. R. (1992). Human papillomavirus type 16 and 18 gene expression in cervical neoplasias. Hum. Pat/Jo/. 23, 1 17128. VON KNEBEL DOEBERITZ,M., OLTERSDORF,T., SCHWARZ, E., and GISSMANN, L. (1988). Correlation of modified human papillomavirus early gene expression with altered cell growth in C4-1 cewical cancer cells. Cancer Res. 48, 3780-3786. VON KNEBELDOEBERITZ,M., BAUKNECHT,T., BARTSCH,D., andzu~ HAUSEN, H. (1991). The influence of chromosomal integration on glucocorticoid regulated transcription of the growth stimulating papillomavirus E6-E7 early genes in human cervical cancer cells. Proc. Nat/. Acad. Sci. USA 88, 141 l-1 415. VORMWALD-DOGAN, V., FISCHER, B., BLUDAU, H., FREESE,U. K.. GISSMANN, L., GLITZ. D., SCHWARZ,E.. and DORST, M. (1992). Sense and antisense transcripts of human papillomavirus type 16 in cervical cancers. /. Gen. Viral. 73, in press. WERNESS,B. A., LEVINE,A. J., and HOWLEY, P. M. (1990). Association of human papillomavirus type 16 and 18 E6 proteins with ~53. Science 248, 76-79. WOODWORTH,C. D., DONIGER.J., and DIPAOLO, J. A. (1989). Immortalization of human foreskin keratinocytes by various human papillomavirus DNAs corresponds to their association with cervical carcinomas. /. Viral. 63, 159-l 64. zur HAUBEN, H. (1986). Intracellular surveillance of persisting viral infections: Human genital cancer results from deficient cellular control of papillomavirus gene expression. Lancer ii, 489-491. ZUR HAUSEN, H., and SCHNEIDER,A. (1987). The role of papillomaviruses in human anogenital cancer. In “The Papovaviridae: The Papillomaviruses” (P. M. Howley and N. Salzman. Eds.), pp. 245263. Plenum, New York. ZUR HAUSEN, H. (1991). Human papillomaviruses in the pathogenesis of anogenital cancer. Virology 184, 9-l 3.
VIROLOGY
Human Papillomavirus Type 16 (HPV 16) Gene Expression and DNA Replication in Cervical Neoplasia: Analysis by in Situ Hybridization MAl-THIAS DURST,’ DAGMAR GLITZ, ACHIM SCHNEIDER,* AND HARALD ZUR HAUSEN Deutsches
Krebsforschungszentrum,
Forschungsschwerpunkt Angewandte Tumorvirologie, Im Neuenheimer and *Frauenklinik, Universitat U/m, Prittwitzstrasse 43, 7900 Ulm Received December
26, 199 1; accepted
Feld 280, D-6900 Heidelberg
1;
March 20, 1992
We have analyzed human papillomavirus (HPV) type 16 RNA expression in premalignant cervical lesions of different severity and in squamous cervical cancers by RNA-RNA in situ hybridization in order to find differences in the topographic distribution of viral RNA, which might correlate with the severity of disease. In the basal layer of low-grade squamous intraepithelial lesions (SIL) only weak transcription of viral early genes was observed. Signal intensity increased strongly in the more differentiated cells accompanied by high levels of HPV DNA replication. This pattern of viral gene expression, together with the onset of viral late transcription in the upper differentiated layer of the epithelium, most likely reflects the productive phase of viral infection. In contrast, in high-grade SIL viral transcription was comparatively strong in basal cells and evenly distributed throughout the undifferentiated epithelium. This difference of viral transcription in the basal layer of the respective lesions points to an altered regulation of viral gene expression which may be causally linked to the progression of precursor lesions. Evidence for disrupted expression of 3’ early genes (EZ, E4, and E5), analogous to the situation in HPV-DNA containing cervical carcinoma-derived cell lines, was not found in any of the HPV-16-positive premalignant lesions nor in the majority of cancers. The similarity of the viral transcription pattern of high-grade SIL and cancers suggests that additional host gene alterations are necessary for malignant progression. 0 1992 Academic Press, Inc.
INTRODUCTION
netic dissection of the viral genome has mapped the immortalizing function to the viral genes E6 and E7 (Barbosa et a/., 1989; Hawley-Nelson et a/., 1989; Milnger et al., 1989a; Barbosa et a/., 1991). Transformation is believed to result in part from the specific interaction of the viral E6 and E7 gene products with the cell regulatory protein ~53 and the product of the retinoblastoma tumor suppressor gene (Rb), respectively (Dyson et al., 1989; Mijnger et al., 1989b; Werness et a/., 1990). Compared to HPV 16 and 18 the E6 and E7 proteins of HPV 6 and 11 show reduced complex formation with these cellular proteins (Munger et al., 1989b; Gage et a/., 1990; Scheffner et al., 1990). The importance of viral E6 and E7 proteins for transformation is further substantiated by the fact that these viral ORFs are conserved and expressed in HPV-16positive premalignant and malignant tumors (Shirasawa et al., 1988; Crum et al., 1989; Cornelissen et al., 1990; Stoler et a/., 1990a; Higgins et al., 1991) despite frequent integration of the viral genome into the host DNA in some cancers (DOrst et al., 1985, 1986). AdditTonally, integration often leads to the disruption of the viral E2 gene (Schwarz et al., 1985), whose gene product was shown to function in part as a repressor of E6-E7 transcription (Romanczuk eta,!, 1990). Inactivation of the repressor function of E2 may therefore be responsible for the high steady state levels of E6-E7 RNA in tumors. Moreover, the importance of viral E6-
Human papillomaviruses (HPV) are epitheliotrophic viruses known to infect both human skin and mucosa. A subset of these viruses infects the anogenital tract. HPV 6 and 11 usually induce benign proliferations of the genital skin and mucosa. Others such as HPV 16, 18, 31, 33, and 35 have been linked to premalignant lesions and cancer of the anogenital tract in particular of the uterine cervix and are therefore considered to be high-risk viruses (zur Hausen and Schneider, 1987). Because of the lack of a suitable cell culture system or animal models for the cultivation of HPV our current understanding of the biology of these viruses is mainly based on experiments performed with molecularly cloned viral DNA. To some extent the clinical association of different HPV types correlates remarkably well with the biological activity of the respective viral DNA in cell culture systems. Most pertinent in this regard is the immortalization of human primary keratinocytes in response to transfection with carcinoma-associated HPV types (Diirst et a/., 1987; Pirisi et al., 1987; Kaur and McDougall et al., 1989). This property is not shared by HPV 6 or 1 1 (Schlegel et al., 1988; Pecoraro et al., 1989; Woodworth et a/., 1989). Subsequent ge’ To whom reprint requests should be addressed. * Frauenklinik, Universitat Ulm. Prittwitzstrasse 43, 7900 Ulm. 0042.6822192
$5.00
Copyright Q 1992 by Academrc Press, inc. All rights of reproduction !n any form reserved.
132
HPV TRANSCRIPTION
AND REPLICATION
E7 in the maintenance of the transformed phenotype was demonstrated by downmodulating viral E6-E7 expression in a HPV-positive cell line, which resulted in reduced cell growth and loss of tumorigenicity (von Knebel-Doeberitz et al., 1988). Our working model of HPV-linked carcinogenesis is primarily based on the above-described experimental systems as well as on the biochemical properties of the viral gene products. The obvious necessity to correlate this data with corresponding molecular events that take place in naturally occurring HPV-associated tumors, prompted us to analyze HPV 16 transcription and viral DNA replication in the context of the different histopathological alterations characteristic for a range of squamous intraepithelial lesions (SIL) of the cervix and squamous cervical cancer by in situ hybridization. MATERIALS
AND METHODS
Biopsies Punch biopsies were taken from 60 patients under colposcopic guidance. Biopsies were immediately frozen in cold (-70”) 2-methylbutane for subsequent storage in liquid nitrogen. Histological examination revealed 14 low-grade SIL and 24 high-grade SIL (SIL = Squamous intraepithelial lesion, from the proposed Bethesda terminology (Hum. Pathol. 21,704-708, 1990) which includes cervical intraepithelial neoplasia (CIN) or other epithelial abnormalities of the cervix). The remaining biopsies were either histologically normal or consisted mainly of connective tissue. Tissues from 19 cervical squamous cell carcinomas were taken from radical hysterectomy specimens at the time of surgery and were frozen and stored in an identical fashion. RNA-RNA
in situ hybridization
Serial cryostat sections mounted on 3-aminopropyltriethoxysilane-coated slides were fixed in 4% paraformaldehyde in 2X SSPE (0.3 n/l NaCI, 20 mM NaH,PO,, 2 mM EDTA) for 1O-l 5 min at room temperature, digested with proteinase K (0.5 pglml) for 10 min at 37”, and hybridized with strand-specific RNA probes of both orientations spanning part of the upstream regulatory region (URR) as well as parts of the early and late region (Fig. 1) (Seedorf et a/., 1985). Briefly, radioactively labeled RNA probes were generated in the transcription vector Bluescribe using either T3 or T7 RNA polymerase, 100 &i of 32P-UTP (800 Ci/mmol from Amersham), and 0.25 mM of each remaining precursor as cold substrate. This yielded an RNA with a specific activity of 10’ cpm/pg. After DNase treatment, the probes were subjected to limited alkaline hydrolysis. Hybridization was performed overnight at 42” in a
IN CERVICAL
NEOPLASIA
133
solution containing 50% formamide, 2X SSPE, 10% w/v dextran sulfate, 10 mM Tris, pH 7.5, 1X Denhardt’s solution (0.02% each bovine serum albumin, Ficoll, and polyvinylpyrollidone), 500 pg/ml tRNA, 100 yglml herring sperm DNA (sonicated), 0.1 o/oSDS, and 1O5 cpm probe/PI. These conditions approximately equal Tm -25” to -30” for RNA-RNAin situ hybridization (Cox et a/., 1984; Bodkin and Knudson, 1985). Sections were washed in 50% formamide, 2X SSPE, 0.1% SDS for 30 min at 50”, treated with RNase A (50 pg/ml) for 30 min at 37”, and subsequently washed in 50% formamide, 0.1 X SSC, 0.1 O/oSDS for 30 min at the same temperature (Tm -10” to -1 So). Slides were dehydrated in graded alcohols containing 300 mMammonium acetate, dried, and dipped in Kodak NTB 2 emulsion diluted 1 :l in 600 mM ammonium acetate. After storage for 1-14 days, the slides were developed in Kodak D-l 9 developer, fixed, counterstained with haematoxylin and eosin, and coverslipped. The sections were examined and photographed under a Zeiss Axiophot microscope equipped with a rotatable lightfield and darkfield condenser to reveal tissue histology and to enable good signal visualization, respectively. RNA-DNA
in situ hybridization
HPV DNA was detected in tissue sections by using a modification of the RNA-RNA in situ hybridization protocol. Best results were achieved by denaturing the target DNA prior to fixation. In brief, serial cryostat sections mounted on 3-aminopropyl-triethoxysilanecoated slides were air dried only. The slides were then incubated in 50% formamide, 0.5~ SSC for 15 min at 70”, followed by fixation in 4% paraformaldehyde, 2x SSPE for 15 min at room temperature. Thereafter the slides were dipped several times in 2x SSC and subjected to proteinaseK digestion (0.5 pg/ml in 2x SSC for 10 min at 37”) and subsequent RNaseA digestion (50 pg/ml in 2X SSC for 30 min at 37”). Enzymatic activity was neutralized by incubation in glycin. Preceding hybridization, both probe (strand-specific RNA probes spanning to entire HPV 16 genome) and target DNA were denatured for 15 min at 95”. After incubation overnight at 42” the slides were washed in 50% formamide, 2X SSPE, 0.1% SDS for 30 min at 42”, treated with RNase A (as for RNA-RNA hybridization) and subsequently washed in 50% formamide, 0.5x SSC, 0.1% SDS for 30 min at 37”. RESULTS All biopsies were screened for the presence of HPV 16 DNA by RNA-DNA in situ hybridization. Three of 14 low-grade SIL and 13 of 24 high-grade SIL but none of the normal tissues were HPV-16-positive. Thirteen of
134
DijRST
w
al6 E&S7
s?aa
ET AL.
so3
CT84
52s
%a6 7lm IJRR
l2
E244 26&l
cm
a153
E2-E5-I2
nlw Ll
FIG. 1. The organization of the HPV 16 genone (Seedorf eta/., 1985) Including adjustments m the E 1 ORF {Matsukura et al., 1986), the E5 ORF (Bubb et a/., 1988; Halbert and Galloway, 1988) and the URR E2 protein binding site (Ronanczuk eta/., 1990) IS shown on top. ORFs are referred to by open boxes. The indicated subgenomic HPV 16 fragments were cloned into a transcription vector in order to generate strand-specific riboprobes for in situ hybridization. -
nineteen squamous cervical cancers were also positive. Theviral transcription pattern of each HPV-16-positive lesion or cancer was then examined in detail. Hybridization probes were derived from the URR and from the early and late region of the HPV 16 genome (Fig. 1). Low-grade
SI L
In all lesions viral early gene expression increased significantly in differentiated cells accompanied by high levels of HPV DNA replication. Moreover, transcripts encoding structural proteins were evident. Compared to early transcripts, the proportion of cells expressing late transcripts was smaller and the signals were confined to the more terminally differentiated cells (Fig. 2, upper panel). This pattern of gene expression was also observed in adjacent sections of the same lesions. High-grade
SIL
Similar to low-grade lesions high-grade SIL also expressed the entire early region of the viral genome, but signal distribution was more uniform. Late viral genes were not expressed and viral DNA replication was generally low (Fig. 2, lower panel).
Squamous
cervical
cancer
In undifferentiated cancers the viral expression pattern was similar to that of high-grade SIL (Fig. 3, lower panel). In some cancers viral gene expression was heterogeneous, e.g., expression of late transcripts which correlated with marked viral DNA replication (Fig. 4, upper panel). In view of the frequently observed inactivation of the viral E2 ORF in carcinoma-derived cell lines, expression of this particular ORF in cancer biopsies was of particular interest. Only in two cancer biopsies were no hybridization signals obtained with a hybridization probe corresponding to the immediate 3’ part of the E2 ORF (probe E2-E5-L2 in Fig. 1) (data not shown). In addition to viral mRNA, some cancers also expressed viral RNA of antisense orientation (Higgins et al., 1991; Vormwald-Dogan et al., 1992). Differences in viral transcriptional basal layer of SIL
activity
in the
Differences in hybridization signal intensity for viral early transcripts were observed consistently in the basal layer of different SIL. In high-grade lesions signals for E6-E7, E2-E4, and E2-E5-L2 probes (Fig. 1) were
FIG. 2. In siru hybridization analysis of a low-grade and a high-grade SIL (CIN I and CIN Ill). Photomicrographs of serial sections of each lesion are shown. Sections a and f are hematoxylin- and eosin-stained only. Viral mRNA was detected in sections b-d and g-i using strand-specific probes as indicated (see also Ftg. 1). Small arrowheads in band g mark the epithelial basement membrane. In sections e and j vrral DNA only was detected. Specrfic hybridization signals are seen as white spots by darkfield microscopy. Magnification Xi 60.
HPV TRANSCRIPTION
AND REPLICATION
IN CERVICAL
NEOPLASIA
135
136
DijRST
distributed uniformly throughout the undifferentiated epithelium. In comparison, viral transcription in the basal layer of two low-grade lesions was markedly reduced (compare in Fig. 2b and g and in Fig. 4b and d). This difference in the level of viral early gene expression in the basal layer of various premalignant lesions was confirmed by repeated examination of these lesions under identical experimental conditions. DISCUSSION ln situ hybridization permits detection of viral DNA and RNA in the context of a highly complex tissue architecture. This is particularly important when correlating viral gene expression with pathological alterations. In situ hybridization complements a wide range of biochemical analyses that enable the identification of individual viral mRNA species but which depend on extraction of total RNA from tissue. From these analyses an increasingly more complex pattern of viral transcripts is emerging (Dilts eta/., 1990; Doorbar eta/., 1990; Sherman et al., 1992). Because several mRNA transcripts show considerable overlap especially in the 3’ early region, differentiation between individual mRNA species by in situ hybridization is not always feasible. Our analysis therefore does not attempt to differentiate between individual mRNA species but is restricted to viral transcripts of the 5’- and 3’-early region and to the late region of HPV 16. Low-grade SlLs There was an overall increase in transcription of the entire early region concomitant with increasing cellular differentiation (Crum et a/., 1988; Stoler et a/., 1992). This type of expression pattern was also described for HPV-11-induced condylomata (Stoler et a/., 1989, 1990b). An increase in viral transcription could be explained on the basis of a strong increase in viral replication, which provides a high level of DNA template for transcription, or as a result of derepression of viral transcription or both (Dijrst et a/,, 1991). Whether an increase in viral early transcription actually correlates with higher levels of viral protein has so far only been shown in the case of E4. Recent experiments by Crum et a/. (1990) have shown that the viral E4 protein (most likely derived from a polycistronic mRNA comprising the E6*I, E7, El,E4, and E5 ORF) is expressed at high levels in the upper differentiated layer of low-grade SIL.
ET AL.
The HPV 16 E4 protein interacts and destabilizes the cytokeratin network which may be essential for ensuring efficient virus particle release (Doorbar et al., 199 1). In our study, transcription of the Ll region was detected in low grade SIL only and seems to be a hallmark for productive viral lesions. Several mechanisms of regulation of late viral transcription are possible: Transcription downstream of the early polyadenylation site could depend on a particular state of differentiation Alternatively, late transcripts could be subject to degradation unless otherwise protected (Kennedy et al., 1990, 1991). Also the use of differentiation-dependent viral intragenomic promoters has been implicated but this requires further verification. By employing the extremely sensitive technique of reverse transcription combined with PCR, viral transcripts of the late region could be detected in most cervical neoplasias irrespective of histological grading (V. Vormwald-Dogan, personal communication). This suggests that very low levels of viral late transcripts are indeed present in most lesions. However, it is not clear whether these transcripts are partially degraded or biologically inactive. High-grade SIL and cervical cancers High-grade lesions and undifferentiated cervical cancers showed a uniform distribution of hybridization signals throughout the thickness of the epithelium for all probes derived from the early region of the HPV 16 genome. In cancers transcription of the 3’ part of the early region is of particular interest since it contrasts the situation in HPV-16-positive cervical carcinomaderived cell lines (Schneider-Gadicke and Schwarz, 1986; Smotkin and Wettstein, 1986; Baker et al., 1987). In these cell lines viral integration into the host DNA has uncoupled the E2 ORF from the major viral promoter (P97). The function of the HPV 16 E2 protein in the natural context of HPV infection is believed to be in part that of a transcriptional repressor for P97. Therefore, lack of E2 expression could lead to increased steady state levels of viral early transcripts and possibly to increased amounts of transforming protein. However, breakpoints for viral integration other than those of the El/E2 region are known (Durst et al., 1986) and several HPV-16-positive tumors harbor not only integrated but also either both integrated and episomal viral molecules or episomal molecules only (Diirst et
FIG. 3. In situ hybridization of a moderately differentiated (Ca. 34) and an undifferentiated (Ca. 1 12) squamous cervical cancer. Photomicrographs of serial sections of each lesion are shown. Sections a and f are hematoxylin- and eosin-stained only. Viral mRNA was detected in sections b-d and g-i using strand-specific probes as indicated (see also Fig. 1). In sectrons e and j viral DNA only was detected. Specific hybridization signals are seen as white spots by darkfield microscopy. Magnification X 160.
HPV TRANSCRIPTION
d W @J W
5 CD W
w I’
: m
0
AND REPLICATION
IN CERVICAL
NEOPLASIA
137
138
DijRST
ET AL.
E6-E7
CIN
I
GIN m
FIG. 4. Detection of viral E6-E7 mRNA in a low-grade and a high-grade SIL (CIN I and CIN Ill) by RNA-RNA in situ hybridization. Brightfield (a and c) and darkfield (b and d) views of one section is shown for each lesion. Small arrowheads in b and d mark the epithelial basement membrane. Magnification X220
al., 1985, Matsukura et al., 1989; Fukushima et al., 1990). Thus, the attributed role of viral E2 in transformation may be somewhat oversimplified (Nasseri eta/., 1991). The detection of viral transcripts spanning the 3’ part of the E2 ORF in 1 1 of 13 cancers examined in this study supports this notion. A model for HPV-linked carcinogenesis Our analyses of HPV transcription in preneoplastic lesions are in support of a postulated malfunctional control of viral transcription as a necessary component for neoplastic progression (zur Hausen, 1986, 1991). The low levels of viral early transcripts observed in the basal layer of two low-grade SIL are likely to be due to specific downregulation of viral transcription by an as yet unknown cellular control mechanism. Progression of high-grade SIL may result from an unscheduled increase in E6-E7 gene expression in basal cells. Evidence in support of suppression of viral transcription in
HPV-infected cells as a consequence of failing intracellular regulation is also provided by experimental systems. Inoculation of HPV-16-immortalized keratinocytes or nontumorigenic somatic cell hybrids into nude mice resulted in a remarkable reduction of E6-E7 gene expression (Bosch eta/., 1990; Dijrst eta/., 1991). This contrasts the behavior of malignant cells which continue to transcribe E6-E7 genes at a high level. Evidence for the existence of the postulated repressor molecule is given by the suppressive effect of host cell genes on the viral URR which controls E6-E7 gene expression. Under tissue culture conditions constitutive expression of an indicator gene (CAT) under the control of the HPV 18-URR is obtained in malignant cervical carcinoma cells after fusion of these cells with a nonmalignant cells (R&l et a/., 1991). Suppression occurs at the level of initiation of transcription and is eliminated after cycloheximide treatment. Dysregulation of viral transcription therefore seems to be an early step in neoplastic progression. Clearly,
HPV TRANSCRIPTION
AND REPLICATION
this could also be brought about by integration of the viral genome into the host DNA, which may not only disrupt the regulation of viral gene expression by viral factors but may also subject the viral genome to the influence of host-specific c&-acting elements (von Knebel Doeberitz et al., 1991). However, this does not seem to be the case for HPV-16-associated SIL since with few exceptions most of the precancers harbor episomal viral DNA only (DOrst et al., 1985; Cullen et a/., 1991). In contrast, the majority of HPV 18 positive SIL do not express the 3’ part of the early viral region possibly as a consequence of viral integration (Stoler et a/., 1992). There is little difference between the HPV 16 transcriptional pattern of high grade SIL and undifferentiated squamous cell carcinoma. This suggests that for malignant conversion of high-grade SIL, additional host gene alterations are necessary, possibly unrelated to papillomavirus gene expression. ACKNOWLEDGMENTS This work was supported by the Deutsche Forschungsgemeinschaft (DO 162/l -2). We are grateful to Vera Vormwald-Dogan and Renate Webler for their assistance in the preparation of the artwork.
REFERENCES BAKER, C. C., PHELPS. W. C., LINDGREN, V., BRAUN, M. J., GONDA, M. A., and HOWLEY. P. M. (1987). Structural and transcriptional analysis of human papillomavirus type 16 sequences in cervical carcinoma cell lines. /. viral. 61, 962-971. BARBOSA, M. S., and SCHLEGEL, R. (1989). The E6 and E7 genes of HPV-18 are sufficient for inducing two-stage in vitro transformation of human keratinocytes. Oncogene 4, 1529-l 532. BARBOSA,M. S., VASS, W. C.. Lowv, D. R.. and SCHILLER,1. T. (1991). In vitro biological activittes of the E6 and E7 genes vary among human papillomavlruses of different oncogenic potential. /. v;ro/. 65,292-298. BODKIN, D. K., and KNUDSON, D. L. (1985). Sequence relatedness of polyam virus genes to cognates of the polyam serogroup viruses by RNA-RNA blot hybridization. Virology 143, 55-62. BOSCH, F. X.. SCHWARZ,E., BOUKAMP. P., FUSENIG,N. E., BARTSCH,D., and ZUR HAUSEN, H. (1990). Suppression in vivo of human papillomavirus type 18 E6-E7 gene expression In nontumorigenic HeLa X fibroblast hybrid cells. /. Viral. 64, 4743-4754. BUBE, V., MCCANCE, D. J., and SCHLEGEL,R. (1988). DNA sequences of the HPV 16 E5 ORF and the structural conservation of its encoded protein. Virology 164, 234-346. CORNELISSEN,M. T. E., SMITS, H. L., BRIET, M. A., VAN DEN TWEEL, J. G., STRUYK,A. P. H. B.. VAN DER NOORDAA, 1.. and TER SCHEGGET, I. (1990). Uniformity of the splicing pattern of the E6/E7 transcripts in human papillomavirus type 16.transformed human fibroblasts, human cetvical premalignant lesions and carcinomas. /. Gen. viral. 71, 1243-1246. Cox, K. H., DELEON, D. V., ANGERER,L. M., and ANGERER,R. C. (1984). Detection of mRNAs in sea urchin embryos by in situ hybridization using asymmetnc RNA probes. Dev. Biol. 101, 485-502. CRUM, C. P., SYMBULA, M., and WARD, B. E. (1989). Topography of
IN CERVICAL
NEOPLASIA
139
early HPV 16 transcription in high-grade genital precancers. Am. J. Pathol. 134, 1183-l 188. CRUM, C. P., Nuovo, G., FRIEDMAN, D., and SILVERSTEIN,S. J. (1988). Accumulation of RNA homologous to human papillomavirus type 16 open reading frames in genital precancers. /. Viral. 62, 84-90. CRUM. C. P., BARBER,S., SYMBULA, M., SNYDER, K., SALEH. A. M.. and ROCHE, I. K. (1990). Coexpresslon of the human papillomavirus type 16 E4 and Ll open reading frames in early cewical neoplasia. Virology 178, 238-246. CULLEN, A. P., REID, R.. CAMPION, M., and L~RINCZ, A. T. (1991). Analysis of the physical state of different human papillomavirus DNAs in lntraepithelial and invasive cervical neoplasm. /. Viral. 65, 606-612. DILTS, D. P., BROKER,T. R., and CHOW, L. T. (1990). The structures of the human papillomavlrus type 1 and type 16 messenger RNAs determined by polymerase chain reaction. in “Papillomaviruses” (P. M. Howley and T. R. Broker, Eds.), pp. 533-540. A. R. Liss, New York. DOORBAR.J., PARTON, A., HARTLEY, K., BANK, L.. CROOK, T., STANLEY, M., and CRAWFORD,L. (1990). Detection of novel splicing patterns in a HPV 16.containing keratinocyte cell line. Virology 178, 254262. DOORBAR, J., ELY, S., STERLING, J., MCLEAN, C., and CRAWFORD, L. (1991). Specific interaction between HPV 16 El -E4 and cytokeratins results in collapse of the eplthelial cell Intermediate filament network. Nature 352, 824-827. DORS~, M., KLEINHEINZ,A., HOTZ, M., and GISSMANN, L. (1985). The physical state of human papillomavlrus type 16 DNA In benign and mallgnant genital tumours. /. Gen. Viroi. 66, 1515-l 522. DORST, M., SCHWARZ, E., and GISSMANN, L. (1986). Integration and persistence of human papillomavlrus DNA In genital tumors. In “Viral etiology of cervical cancer” (R. Peto and H. zur Hausen, Eds.), pp. 273-280. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. DORST, M., DZARLIEVA-PETRUSEVSKA,R. T., BOUKAMP, P., FUSENIG, N. E., and GISSMANN, L. (1987). Molecular and cytogenetic analySIS of Immortalized human primary keratlnocytes obtained after transfectlon with human paplllomavirus type 16 DNA. Oncogene 1, 251-256. DORST, M., BOSCH, F. X., GLITZ, D., SCHNEIDER,A., and ZUR HAUSEN, H. (1991). Inverse relationship between HPV 16 early gene expression and cell differentiation In nude mice eplthelial cysts and tumors Induced by HPV positive human cell lines. I. L&o/. 65, 796804. DYSON, N., HOWLEY. P. M., MONGER, K., and HARLOW, E. (1989). The human papillomavlrus 16 E7 oncoproteln is able to bind to the retinoblastoma gene product. Science 243, 934-937. FUKUSHIMA, M., YAMAKAWA, Y., SHIMANO, S., HASHIMOTO, M., SAWADA. Y., and FUJINAGA, K. (1990). The physlcal state of human paplllomavlrus 16 DNA in ceNlcal carcinoma and cervical Intraepithellal neoplasla. Cancer 66, 2 155-2 16 1. GAGE, 1. R., MEYERS,C., and WETTSTEIN,F. 0. (1990). The E7 proteins of the nononcogenlc human papillomavlrus type 6b (HPV-6b) and of the oncogenic HPV-16 differ in retinoblastoma protein binding properties. /. viral. 64, 723-730. HALBERT. C. L., and GALLOWAY, D. A. (1988). Identification of the E5 open reading frame of human papillomavlrus 16. /. Viral. 62, 1071-1075. HAWLEY-NELSON,P., VOUSDEN, K. H., HUBBERT.N. L., Lowv, D. R., and SCHILLER. J. T. (1989). HPV 16 E6 and E7 proteins cooperate to immortalize human foreskin keratlnocytes. E/I&30 /. 8, 39053910. HIGGINS, G. D., UZELIN, D. M., PHILLIPS, G. E., and BURRELL, C. J. (1991). Presence and distribution of human paplllomavirus sense
140
DijRST
and antisense RNA transcripts in genital cancers. J. Gen. Viral. 72, 885-895. KAUR, P., and MCDOUGALL, J. K. (1989). HPV-18 immortalization of human keratinocytes. Virology 173, 302-310. KENNEDY,I. M., HADDOW, J. K., and CLEMENTS,J. B. (1990). Analysis of human papillomavirus type 16 late mRNA 3’ processing signals in vitro and in vivo. J. Viral. 64, 1825-l 829. KENNEDY,I. A., HADDOW, J. K., and CLEMENTS, J. B. (1991). Anegative regulatory element in the human papillomavirus type 16 genome acts at the level of late mRNA stability. J. Viral. 65, 2093-2097. MATSUKUFIA,T., KANDA, T., FURUNO, A., YOSHIKAWA, H., KAWANA, T., and YOSHIIKE, K. (1986). Cloning of monomeric human papillomavirus type 16 DNA integrated within cell DNAfrom a cervical carcinoma. J. Viral. 58, 979-982. MATSUKURA,T.. KOI, S., and SUGASE, M. (1989). Both episomal and integrated forms of human papillomavirus type 16 are involved in invasive cancer. Virology 172, 63-72. MONGER, K., PHELPS,W. C., BUBB, V., HOWLEY, P. M., and SCHLEGEL, R. (1989a). The E6 and E7 genes of the human papillomavirus type 16 together are necessary and sufficient for transformation of primary human keratinocytes. J. Viral. 63, 4417-442 1. MONGER, K., WERNESS,B. A., DYSON, N., PHELPS,W. C., HARLOW, E., and HOWLEY, P. M. (1989b). Complex formation of human papillomavirus E7 proteins with the retinoblastoma tumor suppressor gene product. EMBO /. 8,4099-4105. NASSERI, M., GAGE, J. R.. LORINCZ, A., and WEITSTEIN, F. 0. (1991). Human papillomavirus type 16 immortalized cervical keratinocytes contain transcripts encoding E6, E7, and E2 initiated at the P97 promoter and express high levels of E7. virology 184, 131-140. PECORARO,G.. MORGAN, D., and DEFENDI, V. (1989). Differential effects of human papillomavirus type 6, 16, and 18 DNAs on immortalization and transformation of human cervical epithelial cells. Proc. Nat/. Acad. Ski. USA 86, 563-567. PIRISI, L., YASUMOTO. S.. FELLER, M., DONIGER, J., and DIPAOLO, J. A. (1987). Transformation of human fibroblasts and keratinocytes with human papillomavirus type 16 DNA. J. Viral. 61, 1061-l 066. ROMANCZUK, H., THIERRY, F., and HOWLEY, P. M. (1990). Mutational analysis of cis elements involved in E2 modulation of human papillomavirus type 16 P,, and type 18 P,,, promoters. /. Viral. 64, 2849-2859. R&L, F., ACHTSTXITER, T., BAUKNECHT,T.. HUITER, K.-J., FUT~ERMAN, G.. and ZUR HAUSEN, H. (1991). Extinction of HPV 18 upstream regulatory region in cervical carcinoma cells after fusion with nonmalignant keratinocytes under non-selective conditions. fI\/IBO 1. 10,1337-1345. SCHEFFNER,M., WERNESS,B. A., HUIBREGTSE,J. M., LEVINE,A. J., and HOWLEY, P. M. (1990). The E6 oncoprotein encoded by human papillomavirus types 16 and 18 promotes the degradation of ~53. Cell 63, 1129-l 136. SCHLEGEL, R., PHELPS,W. C., ZHANG, Y.-L., and BARBOSA, M. (1988). Quantitative keratinocyte assay detects two biological activities of human papillomavirus DNA and identifies viral types associated with cervical carcinoma. EMBO 1. 7, 3181-3187. SCHNEIDER-GADICKE,A., and SCHWARZ, E. (1986). Different human cervical carcinoma cell lines show similar transcription patterns of human papillomavirus type 18 early genes. EM80 J. 5, 22852292. SCHWARZ, E., FREESE,U. K., GISSMANN, L., MAYER, W., ROGGENBUCK, B., STREMLAU, A., and ZUR HAUSEN, H. (1985). Structure and transcription of human papillomavirus sequences in cervical carcinoma cells. Nature 314, 1 1 l-l 14.
ET AL. SEEDORF, K., K~~MMER, G., DORST, M., SUHAI, S.. and R~WEKAMP, W. G. (1985). Human papillomavirus type 16 DNA sequence. viral0gy 145,iai-185. SHERMAN, L., GOLAN, I., ALLOUL, N., DURST. M., and BAFLAM,A. (1992). Expression and splicing pattern of human papillomavirus type 16 mRNAs in preinvasive and invasive carcinomas of the cervix, in human keratinocytes immortalized by HPV 16, and in cell lines established from cervical cancers. Int. J. Cancer 50, 356-364. SHIRASAWA,H., TOMITA, Y.. KUBOTA, K., KASAI. T., SEKIYA,S., TAKAMIZAWA, H., and SIMIZU, B. (1988). Transcriptional differences of the human papillomavirus type 16 genome between precancerous lesions and invasive carcinomas. /. Viral. 62, 1022-l 027. SMOTKIN, D., and WETTSTEIN, F. 0. (1986). Transcription of human papillomavirus type 16 early genes in a cervical cancer and a cancer-derived cell line and identification of the E7 protein. Proc. Nat/. Acad. Sci. USA 83,4680-4684. STOLER, M. H., WOLINSKY, S. M.. WHITBECK, A., BROKER,T. R., and CHOW, L. T. (1989). Differentiation-linked human papillomavirus types 6 and 11 transcription in genital condylomata revealed by in situ hybridization with message-specific RNA probes. Virology 172, 331-340. STOLER, M. H., RHODES, C. R., WHITBECK, A., CHOW, L. T., and BROKER,T. R. (1990a). Gene expression of HPV 16 and 18 in cervcal neoplasia. In “Papillomaviruses” (P. M. Howley and T. R. Broker, Eds.), pp. l-l 1. Wiley-Liss, New York. STOLER. M. H., WHITBECK,A., WOLINSKY, S. M., BROKER,T. R., CHOW, L. T., HOWET, M. K., and KREIDER,1. W. (1990b). Infectious cycle of human papillomavirus type 11 in human foreskin xenografts in nude mice. /. Viral. 64, 3310-3318. STOLER. M. H., RHODES,C. R., WHITBECK.A., WOLINSKY,S. M., CHOW, L. T., and BROKER,T. R. (1992). Human papillomavirus type 16 and 18 gene expression in cervical neoplasias. Hum. Pat/Jo/. 23, 1 17128. VON KNEBEL DOEBERITZ,M., OLTERSDORF,T., SCHWARZ, E., and GISSMANN, L. (1988). Correlation of modified human papillomavirus early gene expression with altered cell growth in C4-1 cewical cancer cells. Cancer Res. 48, 3780-3786. VON KNEBELDOEBERITZ,M., BAUKNECHT,T., BARTSCH,D., andzu~ HAUSEN, H. (1991). The influence of chromosomal integration on glucocorticoid regulated transcription of the growth stimulating papillomavirus E6-E7 early genes in human cervical cancer cells. Proc. Nat/. Acad. Sci. USA 88, 141 l-1 415. VORMWALD-DOGAN, V., FISCHER, B., BLUDAU, H., FREESE,U. K.. GISSMANN, L., GLITZ. D., SCHWARZ,E.. and DORST, M. (1992). Sense and antisense transcripts of human papillomavirus type 16 in cervical cancers. /. Gen. Viral. 73, in press. WERNESS,B. A., LEVINE,A. J., and HOWLEY, P. M. (1990). Association of human papillomavirus type 16 and 18 E6 proteins with ~53. Science 248, 76-79. WOODWORTH,C. D., DONIGER.J., and DIPAOLO, J. A. (1989). Immortalization of human foreskin keratinocytes by various human papillomavirus DNAs corresponds to their association with cervical carcinomas. /. Viral. 63, 159-l 64. zur HAUBEN, H. (1986). Intracellular surveillance of persisting viral infections: Human genital cancer results from deficient cellular control of papillomavirus gene expression. Lancer ii, 489-491. ZUR HAUSEN, H., and SCHNEIDER,A. (1987). The role of papillomaviruses in human anogenital cancer. In “The Papovaviridae: The Papillomaviruses” (P. M. Howley and N. Salzman. Eds.), pp. 245263. Plenum, New York. ZUR HAUSEN, H. (1991). Human papillomaviruses in the pathogenesis of anogenital cancer. Virology 184, 9-l 3.
