|Home | About | Journals | Submit | Contact Us | Français|
Mucosal high-risk (HR) human papillomaviruses (HPVs) that cause cervical and other anogenital cancers also are found in ~25% of head and neck carcinomas (HNCs), especially those arising in the oropharynx and the tonsils. While many HR HPV types are common in anogenital cancer, over 90% of HPV-positive HNCs harbor HPV type 16 (HPV-16). Using a quantitative colony-forming assay, we compared the ability of full-length mucosal HPV genomes, i.e., the low-risk HPV-11 and HR HPV-16, -18, and -31, to persist in and alter the growth of primary human keratinocytes from the foreskin, cervix, and tonsils. The HR HPV types led to the formation of growing keratinocyte colonies in culture independent of the site of epithelial origin. However, HPV-18 induced colony growth in all keratinocytes >4-fold more effectively than HPV-16 or HPV-31 and >20-fold more efficiently than HPV-11 or controls. HPV-11-transfected or control colonies failed to expand beyond 32 to 36 population doublings postexplantation. In contrast, individual HR HPV-transfected clones exhibited no apparent slowdown of growth or “crisis,” and many maintained HPV plasmid persistence beyond 60 population doublings. Keratinocyte clones harboring extrachromosomal HR HPV genomes had shorter population doubling times and formed dysplastic stratified epithelia in organotypic raft cultures, mirroring the pathological features of higher-grade intraepithelial lesions, yet did not exhibit chromosomal instability. We conclude that, in culture, the HR HPV type, rather than the site of epithelial origin of the cells, determines the efficacy of inducing continued growth of individual keratinocytes, with HPV-18 being the most aggressive mucosal HR HPV type tested.
Human papillomaviruses (HPVs) are small DNA tumor viruses that infect, persist in, and cause proliferative lesions in the epithelial cells of the skin, ectoderm-derived mucosae, and their adnexa. Mucosal HPV types are associated with most if not all carcinomas of the uterine cervix, many anogenital cancers, and ~25% of head and neck cancers (HNCs) (reviewed in reference 55). HNCs arising in the oropharynx or tonsils are HPV positive in over 50% of the cases. Although the same oncogenic or “high-risk” (HR) HPV types are found in the cervix, in the anogenital area, and in HNC, their prevalence is strikingly different: in cervical and anogenital cancers, HPV type 16 (HPV-16) is found in ~50% of the tumors, followed by HPV-18 (~20%), while the remaining cases harbor over 15 other HR HPV types, including HPV-31 and other well-characterized as well as additional, novel HPV types. In contrast, over 90 to 95% of HPV-positive HNCs contain HPV-16, whereas HPV-18, HPV-31, or other HR HPVs are infrequent (10, 18, 33, 61). Furthermore, HR HPVs appear to vary in oncogenic potential: HPV-18 is associated with more advanced cervical disease that progresses more rapidly and typically has a worse prognosis than that caused by other HR HPV types (7, 34, 68). In contrast, other mucosal, “low-risk” (LR) HPV types (HPV-6 or -11) are not commonly associated with malignant progression even though they also give rise to clinically significant disease, such as genital warts and recurrent respiratory papillomatosis in the larynx and trachea that require repeated therapeutic interventions.
In vivo, HPV persists in the nuclei of epithelial cells in the form of unintegrated, supercoiled circular plasmid DNA. Spontaneous immortalization of human keratinocytes in culture occurs rarely, yet keratinocyte lines from primary skin (HaCaT) (5) and the cervix (NIKS) (1) have been described. Early experiments demonstrated that HR HPV genomes can expand the life span of, or “immortalize,” primary human foreskin keratinocytes in culture (2, 14, 32, 53, 58, 71). However, the culture conditions precluded the establishment and maintenance of persistently replicating, extrachromosomal HPV plasmids in these previous studies. In contrast to persistently infected cells, most HPV-associated cervical, anogenital, and head and neck carcinomas contain disrupted, integrated HR HPV DNA (19, 30, 59). The integrated HR HPV genomes invariably express two viral early proteins, E6 and E7, that bind to and disrupt the functions of the tumor suppressor gene products p53 and pRB, respectively, and interact with a number of additional cellular targets (70). The E6 and E7 gene products of both LR and HR mucosal HPV types are required for persistence, yet only those of HR HPV strains effectively immortalize epithelial cells in culture when ectopically expressed under the control of heterologous or retroviral promoters (55).
As initially demonstrated by Laimins and coworkers (27, 43, 46), recircularized HR HPV genomes can establish persistent replication in human primary foreskin keratinocytes. Under optimal conditions, such cells harbor extrachromosomal plasmid HPV genomes, exhibit an extended life span in vitro, undergo dysplastic differentiation in organotypic “raft” cultures, and have been used to dissect early and late events of the viral life cycle, including viral gene expression, vegetative viral DNA replication, and virion synthesis (16, 17, 26, 35, 46, 47). Similar analyses have been extended to cervical keratinocytes (3, 63) and airway epithelial cells (39). However, previous studies that exploited HPV persistence in primary human epithelial cells primarily investigated the viral life cycle and have not examined individual virus genome-cell interactions.
In this study, we directly compared the capacity of HR HPV-16, -18, and -31 as well as the LR HPV-11 to establish persistence and lead to growth alterations in primary human epithelial cells derived from the three main mucosal sites where HPV-associated cancers arise: the cervix, the tonsillar fossa, and foreskin. In addition to assessing early gene promoter activity and initial genome amplification (23), we utilized a quantitative colony formation assay followed by subculturing multiple individual cell clones stemming from individual virus genome-cell interactions (36). In contrast to previous studies, this approach allowed us to quantify the viral genome capacity to extend the cellular life span (“immortalize”) in culture and to alter their growth properties and genomic stability in relation to the viral HPV genome status and viral gene transcription, all parameters which could otherwise be masked in heterogeneic mass cultures.
We found that, in culture, tonsillar epithelial cells did not reflect the apparent preference for HPV-16 observed in patients. The efficiency of extended-growth clone formation was solely a function of HR HPV type: HPV-18 was strikingly more effective than HPV-16 or -31 in establishing growing cell clones harboring extrachromosomal plasmid HPV genomes in all keratinocytes tested regardless of their site of origin. Furthermore, in contrast to HPV-16, which exhibits considerable functional differences among its geographic variant genomes harboring altered control sequences (26, 38), we found that the similarly diverse HPV-18 upstream regulatory region sequences (URRs) did not confer the same variant-dependent activity.
Despite shortened population doubling (PD) times and dysplastic differentiation of the cells in organotypic culture, HPV persistence did not lead to chromosomal instability. Because of their apparent long-term stability in culture, these HR HPV-containing keratinocyte clones can serve as a representative model to study cellular and viral changes in the progression from early persistent HR HPV infection to cancer.
Primary human foreskin (HFK), cervical (HCK), and tonsillar epithelial (HTE) keratinocyte cultures were prepared from neonatal foreskins and hysterectomy- and tonsillectomy-derived tissues as described previously (31, 39, 57, 63). The HPV-16 W12E subclone harboring extrachromosomal HPV-16 plasmids was a gift from Paul Lambert (30), while the CIN612 cell line was a gift from Laimonis Laimins (46). All keratinocyte cultures and SCC13 cells, an HPV-negative squamous cell carcinoma line (56), were grown on irradiated J2 fibroblast feeder cells in E medium, containing 0.5 μg/ml hydrocortisone, 0.1 nM cholera toxin, 5 μg/ml transferrin, 5 μg/ml insulin, 2 nM 3,3′-5-triodo-l-thyronine, and 5 ng/ml epidermal growth factor (27). HeLa cells were cultured as described previously (28).
The replication-competent HPV-16 W12E genome was a gift from Paul Lambert (30). The HPV-31 plasmid (40) was a gift from Attila Lörincz, Wolfson Institute, London, United Kingdom. The HPV-18 (4) and HPV-11 (20) plasmids were gifts from Harald zur Hausen, DKFZ, Heidelberg, Germany. Total DNA was phenol-chloroform extracted from HPV-18-positive cervical carcinoma biopsy specimens (squamous cell carcinomas and one adenocarcinoma, i.e., isolate M29) and the integrated physical status of the HPV DNA determined by Southern blotting in the laboratory of Attila Lörincz (data not shown). HPV-18 URRs (spanning nucleotides [nt] 7404 to +180) were amplified from each DNA sample using the following primers: nt 7404 to 7424, 5′-GCGCAGATCTACTGCACACCTTACAGCATCC-3′, and nt 180 to 160, 5′-CCGGAAGCTTGCAGTGAAGTGTTCAGTTCCG-3′. The HPV-18 URRs were then cloned into a pUC chloramphenicol acetyltransferase (CAT) reporter plasmid by standard methods as described previously (38). URR nucleotide alterations in recombinant plasmid constructs were verified by automated sequencing (University of Iowa DNA Core). The phylogenetic relationship of the altered HPV-18 URRs to defined HPV-18 variants (E, European; Af, African; and As/Ai, Asian/American Indian) was determined with a rooted dendrogram using ClustalX.1 software (data not shown). The most variable segments of the altered HPV-18 URR segments were then subcloned into the HPV-18 reference genome via AscI and BamHI digestion.
CAT assays were performed as described previously (66). Prior to transfection, the wild-type (wt) HPV genomes were first released from the bacterial pUC vectors by restriction digestion (HPV-16-W12 and HPV-11 with BamHI; HPV-18 and HPV-31 with EcoRI) and then religated and purified over plasmid purification columns (MaxiKit; Qiagen, Valencia, CA) as described previously (37). For RNase protection assays, 3 μg of recircularized DNA was transiently transfected into HeLa or SCC13 cell cultures. Total RNA was harvested 24 h later using RNAqueous kits (Ambion, Austin, TX) and analyzed as described previously (66) in RNase protection assays as described previously (28), using PCR-generated probes illustrated in Fig. Fig.1A.1A. For initial HPV genome amplification (also referred to as transient-replication) assays, 3 μg of recircularized DNA was transfected into SCC13 cells with Effectene (Qiagen, Valencia, CA) as described previously (37). Total DNA was isolated from transfected cells (QIAamp DNA blood kit; Qiagen, Valencia, CA) at 5 days posttransfection. The samples were digested with DpnI and linearized with endonucleases that recognize a single restriction site in each HPV genome, i.e., BamHI for HPV-16 and -11 or EcoRI for HPV-31 and -18, before Southern blotting. For colony formation assays, 2 × 106 low-passage (8 to 10 PDs postexplant) primary human keratinocytes per 100-mm dish were transfected with 2 μg of recircularized HPV DNA and 1 μg of pRSV-neo. Control cultures transfected with pCMV-βgal showed reproducible transfection efficiencies of 15 to 25%. Cells were transferred at dilutions onto irradiated J2 fibroblast feeder cells 1 day later, selected in 100 to 200 μg G418/ml E medium for 5 days, and allowed to grow for another 15 to 20 days without selection. After total colony numbers per transfection were counted, individual colonies were subcultured using cloning cylinders from dishes with <40 colonies. Total cellular DNA as well as total RNA was harvested from clonal cultures to assess HPV persistence and transcription.
A total of 5 μg of DpnI-resistant, transiently transfected DNAs or 2 μg of whole-cell DNAs of HPV-induced clonal cultures was resolved on 1.0% agarose gels, depurinated in 0.25 M HCl, and blotted directly onto positively charged nylon membranes (Hybond-XL; Amersham Biosciences Corp., Piscataway, NJ) by alkaline transfer with 0.4 N NaOH. The blots were then hybridized at 65°C with probes (1.5 × 106 cpm/ml hybridization buffer) containing a representative single fragment or an equimolar cocktail of PCR-amplified segments of HPV-16 (nt 6226 to 3873/4471 to 6000), HPV-31 (nt 6161 to 872), HPV-18 (nt 2813 to 5709/3941 to 7565/7404 to 297), or HPV-11 (nt 7080 to 225) and [α-32P]dATP/dCTP labeled by random priming (HotPrime kit; GenHunter Corp., Nashville, TN). A titration of the respective linearized HPV DNA was included to determine viral copy numbers per cell. Growth phenotypes were determined by plating 7 × 105 to 1 × 106 cells onto 60-mm dishes with irradiated fibroblasts. The cells were counted after 3 and 7 days to determine PD times.
Organotypic (“raft”) cultures were grown in E medium using collagen inserts (Biocoat; Becton Dickinson, Bedford, MA), sectioned, and hematoxylin-eosin stained to assess squamous differentiation as described previously (38). Karyotypes were determined by arresting cells grown without feeders for 24 to 36 h with colcemide prior to G-banding of chromosomes and digital analysis of structural and numerical abnormalities of representative mitoses by the Cytogenetics Laboratory, Department of Pediatrics, University of Iowa Hospitals and Clinics.
To compare the activities of different mucosal HPV genomes immediately after introduction into the cells, we first assessed transcription initiating at the viral major early gene promoters in RNase protection assays at 24 h posttransfection. Recircularized HPV-16, -31, and -11 genomes were all actively transcribed upon transient transfection in HeLa cells, which do not support HPV replication, (Fig. (Fig.1A,1A, lanes 1 to 3), and all four mucosal HPV types, including HPV-18, in the HPV-negative squamous cell carcinoma line SCC13 displayed comparable transcription levels (Fig. (Fig.1A,1A, lanes 4 to 9).
Next, we compared the initial amplification of the HPV genomes in standard transient-replication experiments with SCC13 cells (26, 27). The assay measures the presence of HPV DNA replicated in mammalian cells and thus rendered DpnI-resistant by Southern blotting (Fig. (Fig.1B).1B). At 5 days posttransfection, the HPV-16, -31, -18, and -11 plasmids all underwent initial amplification. However, there was a striking disparity in their transient-replication levels. The LR HPV-11 amplified 10-fold more efficiently than the HR HPV-18 or -31. Furthermore, the most prevalent HR HPV, HPV-16, was over 3-fold less efficient than the other HR HPVs and ~30-fold less efficient than HPV-11. Therefore, while the major early gene promoters in all HPVs tested were active at comparable levels, initial plasmid amplification varied significantly among the four mucosal HPV types.
To see whether the levels of initial plasmid amplification of the four mucosal HPVs correlated with their abilities to alter the growth of primary human keratinocytes, we employed a colony formation assay that quantitatively evaluates the induction of discrete colonies growing in monolayer cultures on plastic (in the presence of irradiated fibroblast feeders) that can be counted under an inverted microscope. To determine whether the colony-forming cells would continue to divide in culture, we subcultured multiple individual colonies to determine the extent of their “immortalization” and characterized them further.
Primary foreskin keratinocytes transfected with neo only as controls formed colonies at a low rate (1 to 10 colonies/culture) under these conditions but did not divide beyond 26 to 32 PDs postexplantation upon subculturing. The colony-forming efficiencies of mucosal HPV genomes were calculated by averaging the increase in colony counts in cultures transfected with recircularized HPV DNA over neo-transfected controls in four independent experiments using three independent donor foreskins, as we had demonstrated in previous studies (36, 38). As expected from previous work by Laimins and colleagues (65), the LR HPV-11 did not enhance colony counts beyond those of the neo controls despite high initial amplification of its plasmid genome. In contrast, the HR HPV-16 plasmid, which initially amplified at the lowest levels, led to colony formation with an equal if not better efficiency than HR HPV-31 (Fig. (Fig.2A).2A). HR HPV-18 induced four- to fivefold more colonies than either HPV-16 or HPV-31 (Fig. (Fig.2A).2A). Most of the HR HPV-transfected colonies expanded into individual clonal cultures that exhibited no apparent slowdown of growth or “crisis” and continued growing beyond 36 PDs postexplantation (the numbers of surviving colonies are given as ratios in Fig. Fig.22).
Regardless of the foreskin isolate used, the relative ability of each viral plasmid to induce colony growth was consistent in all transfections, arguing that the phenotypes observed were not cell donor specific but rather were a measure of the capacity of each HPV type to induce virus-cell interactions that result in extended growth. A single clone was isolated from HFKs transfected with the LR HPV-11 in these four experiments (Fig. (Fig.2A).2A). As this HPV-11-positive clone continued to grow beyond 36 PDs postexplantation, it was included in further analyses.
To assess the susceptibility of primary keratinocytes from different epithelial sites associated with HPV infection to the different mucosal HPV genomes, we also tested the HPV-16, -31, -18, and -11 genomes in primary human epithelial cells from the uterine cervix (HCK) (Fig. (Fig.2C)2C) and tonsillar epithelium (HTE) (Fig. (Fig.2D).2D). As in the foreskin keratinocytes, HPV-18 formed colonies more efficiently than HPV-16 or HPV-31, while HPV-11 did not increase the colony counts beyond that with neo controls. The continued survival of the HPV-18/HCK colonies was similar to that for HPV-16/HCK, but only a single HPV-31/HCK clone continued to grow. No HPV-11-transfected colonies survived. In HTE cells from three distinct donors, HPV-18 again induced the most colonies that were capable of extended growth (Fig. (Fig.2D).2D). Interestingly, no colonies formed with HPV-31 or HPV-11 in repeated assays. These results show that HPV-18 is capable of efficiently immortalizing all primary keratinocytes tested regardless of their epithelial site of origin, and they identify HPV-18 as the most aggressive HR HPV type in its ability to form keratinocyte lines with extended life spans in culture.
Next, we examined available independent, proliferating clones of keratinocytes derived from different mucosal sites for HPV DNA status by Southern blotting. Representative examples are illustrated in Fig. Fig.2B.2B. Interestingly, the viral genomes in 95% of the HPV-18/HFK clones were extrachromosomal, compared to 56% of HPV-16/HFKs. Similarly, in HCKs, 80% of the HPV-18 clones contained extrachromosomal genomes (compared to 87% in HTEs), while none of the HPV-16 cells in these experiments did. We have shown previously that that HPV-16 could lead to persistent extrachromosomal viral copy numbers in HCKs from an alternate cervical cell donor (63). Four to 32 clones from different virus-cell site interactions that were found to harbor extrachromosomal HR HPV genomes were characterized further (Table (Table1).1). HPV-16 and HPV-31 clonal HFK cultures contained ~10 virus genome copies per cell (ranging from 3 to 20 copies per cell), similar to the HPV-16 W12E cells. At 34 PDs postexplantation, the unique growing LR HPV-11 clone originally harbored mostly extrachromosomal plasmid genomes at ~10 copies/cell with traces of novel restriction fragments suggestive of integrated HPV-11 DNA (Fig. (Fig.2B).2B). Upon further growth and reanalysis, the cells were found to contain ~10 copies of integrated HPV-11 DNA in an apparent head-to-tail arrangement (type 2 integration ) upon DNA linearization with BamHI. HPV-18-harboring clones displayed a higher average number of ~20 to 25 copies/cell (ranging from 5 to 61) regardless of their epithelial site of origin (Table (Table11).
Examination of steady-state transcript levels initiated at the major viral early gene promoters comprising mRNAs encoding the HPV E6 and E7 gene products revealed that they varied widely between individual clones (Fig. (Fig.3).3). The mRNA expression did not appear to correlate strongly with the HPV type or its ability to induce immortalized colony formation. Consistent with recent reports (21, 38), viral steady-state mRNA levels in clones harboring integrated rather than extrachromosomal HR HPV genomes (as seen in the syngeneic set of HPV-16 HFK clones in Fig. Fig.3B)3B) did not necessarily give rise to increased mRNA expression (Fig. (Fig.3A,3A, lanes 1 and 2, and B, clones 1 and 2).
To characterize the potential growth advantage of cells with persistent, extrachromosomal HR HPV genomes over virus-free primary keratinocytes in culture, we measured the PD times of representative clones stemming from different transfections compared to the source primary cells (Table (Table1).1). While the primary keratinocyte populations doubled every 3 (foreskin), 3.5 (tonsil), and 3.7 (cervix) days, the PD times of the HR HPV-16 and -18-harboring cells were 1.4- to nearly 2-fold shorter. HPV-31/HFK and CIN612 cells, which also harbor extrachromosomal HPV-31 plasmids, displayed a more modest increase in growth rate of ~1.3-fold over the source cells (Table (Table11).
Furthermore, because the E7 proteins of HR HPVs can lead to abnormal centrosome duplication soon after they are expressed (11, 13), we examined the karyotypes of representative clones of keratinocytes from different sites immortalized by the plasmid persistence of different HPV types. We found that all remained diploid (2N) through further passaging (>26 PDs posttransfection or PDs of ≥36 postexplantation [Table [Table1]).1]). Some of the individual HR HPV clones exhibited remarkable stability in culture, as they retained exclusively extrachromosomal plasmid HPV genomes and apparently intact, diploid chromosomal complements for over 60 PDs (M. J. Lace, J. R. Anson, T. H. Haugen, and L. P. Turek, unpublished data). In contrast, the karyotype of the W12E clone harboring an extrachromosomal HPV-16, originally explanted from a cervical lesion (64) and subcloned (30), showed a tetraploid (4N) karyotype accompanied by multiple chromosomal alterations, including intrachromosomal rearrangements, chromosomal amplifications, and deletions. Similarly, the CIN612 cell line, explanted from a high-grade cervical lesion and harboring stably replicating HPV-31 plasmids, also displayed a near-tetraploid karyotype with multiple chromosomal abnormalities (Table (Table11).
We then examined the ability of representative cell clones from diverse epithelial sites harboring extrachromosomal HR HPV genomes to undergo orderly stratification in organotypic (“raft”) cultures. The HPV-16, -31, and -18 foreskin-derived (HFK) cultures generated dysplastic, partially differentiated stratified epithelia resembling higher-grade squamous epithelial lesions that are associated with HR HPV infection in the cervix (Fig. 4 A, C, and D, respectively). In contrast, the HPV-16 W12E cells exhibited greater atypia and poor stratification under the same conditions (Fig. (Fig.4B),4B), consistent with their origin from an intraepithelial cervical lesion. Similar to results obtained with the HPV-18-immortalized HFK cultures (Fig. (Fig.4D),4D), cervical (HCK) and tonsillar (HTE) clones harboring extrachromosomal HPV-18 also displayed a partially dysplastic phenotype characterized by minimal basal thickening and nearly normal maturation (Fig. 4E and F). Taken together, these results strongly suggest that differences in the immortalization efficiencies between the HR HPV types are due to critical early events in the virus genome-cell interaction rather than to differences in viral DNA status, viral gene expression, or the growth phenotypes of the established HPV-positive cells.
Finally, to test whether the reference HPV-18 genome (classified as an As/Ai variant) used in this and other studies could inherently be more active in early viral functions than other defined HPV-18 variants, we amplified HPV-18 URRs containing defined HPV-18 variant sequences (48). The URR is the most variable sequence component of the HPV genome, yet it serves as the critical control region modulating replication and early gene expression (e.g., expression of the E1 replicase and the transforming genes E6 and E7). We initially tested the major early promoter activities of these diverse HPV-18 URRs in reporter expression assays and further compared the abilities of these altered HPV-18 URRs to influence initial plasmid amplification and immortalization capacity in keratinocytes (Fig. (Fig.5).5). In contrast to identical assays comparing altered HPV-16 URRs derived from cervical carcinomas (38), the HPV-18 As/Ai reference and related genomes did not consistently display enhanced early promoter activities (Fig. (Fig.5A),5A), amplify more efficiently (Fig. (Fig.5B),5B), or form colonies more readily than the E or Af URRs (Fig. (Fig.5C).5C). Therefore, in contrast to the functional differences between defined HPV-16 sequence variants, these results demonstrate that HPV-18 in general (not solely the As/Ai reference genome) exhibited greater immortalization capacity than HR HPV-16 or -31.
In this study, we compared early events in the establishment of HPV persistence and altered growth in primary human epithelial cells. In contrast to previous reports, we tested the capacities of HPV-11, -16, -18, and -31 to form immortalized clonal cultures harboring extrachromosomal plasmid HPV genomes in primary human keratinocytes explanted from various host sites that are targets for HPV-associated lesions. This novel quantitative approach identified HPV-18 as the most aggressive HPV type tested in establishing persistence and altering the cellular life span.
One of the aims of the study was to compare the susceptibilities of primary human keratinocytes from epithelial sites associated with type-specific mucosal HPV persistence and resulting disease. While the clinical prevalence of different mucosal HR HPVs in anogenital sites is the same as in the cervix (55), HPV-16 represents more than 90 to 95% of all mucosal HPV types identified in HPV-positive HNC tumor tissues or in the oral cavity and oropharynx of HNC patients as well as controls (10, 18, 61). Interestingly, we found that the HR HPV-18 was strikingly more effective than HPV-16 in establishing extended-growth cell clones from any of the epithelial sites, including tonsillar (HTE), cervix (HCK), or foreskin (HFK) cells derived from multiple donors. It is not clear which factors in the local microenvironment or immune response in vivo may favor HPV-16 or impair other mucosal HPV infections, yet any site differences in susceptibility are not apparent in explanted cells grown under uniform culture conditions. At the same time, the capacity of primary human keratinocytes from multiple epithelial sites to continue dividing in culture and to become immortalized by HR HPVs varied. Neonatal primary foreskin keratinocytes (HFKs) showed the least variability and supported persistence of all the HR HPVs compared to adolescent or adult tissue-derived cells from the cervix (HCK) or the tonsillar epithelium (HTE).
The immunogenic profiles of host keratinocytes in the oropharynx compared to those in the genital tract will likely play an important role in mediating type-specific HPV infection, immunosubversion/evasion, and tumorigenic progression of the HPV-immortalized keratinocyte host in vivo. HR HPV pseudovirions have also been shown to employ alternate endocytotic pathways to achieve entry into the host cell (6, 51); therefore, potential variation in the innate infectivity of HR HPV particles could also contribute to a type-specific differential in HR HPV persistence in vivo. Our results, however, demonstrate that keratinocytes derived from various epithelial sites associated with HPV infection are equally susceptible, under the same culture conditions, to HPV-18 immortalization following initial viral DNA amplification.
The capacity of HPV-18 to immortalize primary human keratinocytes more efficiently than the other mucosal HPV types tested could be caused by potential qualitative or quantitative differences in the capacity of viral oncoproteins to interact with key regulators of the epithelial host which modulate cellular proliferation (reviewed in reference 23). Both the E6 and E7 oncoproteins are multifarious: E6 influences epithelial host immortalization and tumor progression by interacting with an array of critical mediators of apoptosis (29) and cellular growth (49), including components of the p53 (41, 44) and p14/ARF (60) pathways. E6 also binds via its conserved PDZ domain to Notch1 (8, 67) as well as the hDLG and hScrib gene products (42). We replaced the HPV-16 E6 3′ PDZ domain with the same domain from HPV-18 E6 to determine if a type-specific variation in PDZ activity could contribute to the increased immortalization capacity of HPV-18. However, no significant difference in initial amplification levels or the colony-forming capacity of the HPV-16 wt versus the hybrid HPV-16 E6 3′ PDZ mutant genome was noted. Furthermore, clonal cultures derived from transfection of this altered HPV-16 E6 genome exhibited viral copy numbers and growth rates identical to those for the wt (data not shown). The E7 oncogenes of HR HPV types, in addition to binding and inactivating the RB protein family, have been shown to upregulate the Akt pathway (45) and target the p21 as well as the RB proteins to subvert cell cycle control in order to escape senescence (24).
Alternatively, the HPV-18 genome may establish itself in more cells by the virtue of higher viral gene expression, higher levels of initial genome amplification, or both. Transcriptional involvement was originally suggested by early, semiquantitative studies where the ability of HPV-18 to immortalize foreskin keratinocytes more frequently than HPV-16 had been mapped to the HPV-18 enhancer/control region (58). We recently defined a similar role for cellular factors interacting with the early gene promoter in the viral persistence of bovine papillomavirus type 1 (22). In this study, the HPV-18 genome was also more efficient in extending the life span of primary human cervical (HCK) and tonsillar (HTE) keratinocytes. Furthermore, a higher percentage of individual HPV-18-induced clones contained only extrachromosomal, unintegrated viral genomes compared to clones induced by HPV-16 or -31.
The underlying cause of high HPV-18 amplification is not apparent: none of the mucosal HPV genomes tested was defective in any of the viral early gene functions required for transient or persistent replication, and each genome was capable of amplification and persistence per se. The varying capacity of the HPV genomes to amplify could be due to differences in how the viral E1 and E2 proteins physically interact and/or modify the structure of the viral origin of replication (25, 69). In addition, initial amplification as well as colony formation is likely to be controlled at the level of early viral gene expression. However, when we compared steady-state mRNAs originating at the homologous major early (E6 and E7) gene promoters of each circular virus genome, their levels were comparable for the four mucosal HPV types tested in two cell types. A possible technical limitation of these experiments is the fact that the mRNA expression was measured in established carcinoma cell lines and not in primary human keratinocytes. Additionally, mRNAs encoding the viral replication proteins E1 and/or E2 may originate from alternate promoters at the initial plasmid amplification stage and/or are expressed at limiting levels not readily detectable in these assays. Thus, it remains possible that type-specific initial plasmid amplification could also be a function of differential E1 and/or E2 expression.
Numerous variants of the prototypical reference sequences have been classified for many of the mucosal HPVs as a function of geographical distribution and phylogeny. The reference HPV-18 genome (9), used in this and previous studies, is related to the As/Ai HPV-18 variant (48). In HPV-16, non-European variants have been found to be associated with more aggressive clinical disease (48), increased early gene transcription, augmented plasmid genome amplification (26), and enhanced immortalization capacity in keratinocyte cultures (38). Recent studies have suggested that some non-HPV-18 variants (54), including the As/Ai reference genome, may be more strongly associated with cervical carcinomas than European variants (72). However, in contrast to the case for HPV-16, the HPV-18 reference and reconstituted HPV-18 genomes harboring altered URRs that resemble those of As/Ai, Af, or E variants exhibited transcription, replication, and immortalization capacity phenotypes that were independent of their phylogenetic origin. However, we do exclude the possibility that additional nucleotide variations present in other segments of the variant genomes may also contribute to variations in viral functions throughout the HPV life cycle.
In contrast to the HR types, the mucosal LR HPV-11 did not immortalize primary keratinocytes despite comparable early gene expression and high initial amplification levels. This result mirrored a previous study where a pooled culture of HPV-11-transfected foreskin human keratinocytes maintained replicating HPV-11 plasmids that extended the life span of the host twofold (65). In the entire series of experiments, we isolated a single growing HFK clone growing beyond 32 PDs postexplantation which initially contained extrachromosomal HPV-11 plasmid genomes that subsequently integrated in an apparent head-to-tail tandem arrangement (integration type 2 ) upon further passaging. The contribution of HPV-11 gene expression to the extended growth phenotype of this cell line is not clear. The ability of the HR HPV E6 and E7 to more effectively interact with the same cell cycle machinery targeted by the LR HPV-11 E6 and E7 may increase the probability of immortalization directly and/or indirectly by extending the life span of the cells in culture, thus increasing the likelihood of additional cellular mutations influencing cell cycle control (41).
Previous studies have suggested that early genomic instability in HR HPV-harboring keratinocytes may play a role in the progression to immortalization. The E7 protein can cause centrosome duplications and chromosomal abnormalities in keratinocytes immortalized by HPV-16 E6 and E7 proteins expressed from viral fragments or retroviral vectors (12, 15, 50, 52, 62) or from persistent plasmid HPV-16 genomes introduced into a previously immortalized keratinocyte cell line (11). In contrast, the karyotypes of the clonal keratinocyte cultures immortalized by HR HPV plasmid genomes in this study remained diploid, harbored no significant structural chromosomal abnormalities, and retained moderate levels of squamous differentiation in organotypic “raft” cultures resembling mid- to higher-grade intraepithelial lesions. Thus, genomic instability is apparently not a prerequisite to HPV immortalization under these conditions. We have not, however, ruled out the potential contribution of subtle genomic alterations not reflected in the karyotype analysis. In comparison, the extrachromosomal HPV-16-positive W12E and HPV-31-positive CIN612 cells, originally explanted from cervical lesions, exhibited cellular abnormalities and distorted growth patterns in organotypic cultures that resembled carcinoma in situ and were found to carry significant chromosomal alterations presumably acquired during the course of disease development before explantation and/or in long-term culture.
In contrast to HR HPV cell lines derived from cervical lesion explants, the syngeneic sets of individual clones from different epithelial sites derived in this study resemble cells of early clinical dysplastic lesions induced by HR HPV persistence. They will permit the analysis of a variety of virus-cell interactions that take place in persistent HPV infection and can serve as a sensitive model to detect influences that promote HR HPV integration and other viral and cellular changes leading to carcinogenic progression.
We thank Shivanand R. Patil for the cytogenetic analysis of HPV-immortalized cultures generated in this study, Attila Lörincz for HPV-18 DNAs, Christina Isacson and Greg Thomas for their assistance with plasmid constructions, and William Bonnez and Denise Galloway for discussions on HR HPV variants.
This study was supported by Department of Veterans Affairs Merit Awards to L.P.T. and T.H.H.
Published ahead of print on 9 September 2009.