Carcinogenesis Advance Access originally published online on September 14, 2006
Carcinogenesis 2006 27(11):2341-2353; doi:10.1093/carcin/bgl172
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Interferon-ß treatment of cervical keratinocytes naturally infected with human papillomavirus 16 episomes promotes rapid reduction in episome numbers and emergence of latent integrants
1 Medical Research Council Cancer Cell Unit, Cambridge CB2 2XZ, UK
2 Department of Oncology, University of Cambridge CB2 2XZ, UK
3 Department of Pathology, University of Cambridge CB2 1QP, UK
*To whom correspondence should be addressed at: Medical Research Council Cancer Cell Unit, Hills Road, Cambridge, CB2 2XZ, UK. Tel: +44 1223 763285; Fax: +44 1223 763284; Email: nc109{at}cam.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Following integration of human papillomavirus (HPV) into the host genome, overexpression of the viral oncogenes E6 and E7 requires loss of the transcriptional repressor functions of E2. A key step in HPV-related carcinogenesis is therefore clearance of residual viral episomes, which encode E2. As spontaneous loss of HPV-16 episomes in vitro is associated with increased expression of antiviral genes inducible by type I interferon (IFN), we used the W12 model to examine the effects of exogenous IFN-ß on cervical keratinocytes containing HPV-16 episomes as a result of ‘natural’ infection in vivo. In contrast to studies of cells transfected with HPV-31 or bovine papillomavirus, IFN-ß caused rapid reduction in numbers of HPV-16 episomes. This was associated with the emergence of cells bearing previously latent integrants, in which there was increased expression of E6 and E7. Our data indicate that integrated HPV-16 can exist in a minority of cells in a mixed population without exerting a selective advantage until episome numbers are reduced. The kinetics of cell death and changes in viral transcription and translation that we observed support a model where integrants are initially present in cells also containing episomes, with generalized episome clearance by IFN-ß resulting in integrant de-repression. We conclude that IFN-ß can hasten the transition from episomal to integrated HPV-16 in naturally infected cervical keratinocytes. Greater emphasis should be placed on episome loss in models of HPV-related carcinogenesis. We provide the strongest evidence to date that treating HPV-16 lesions by inducing an IFN response may cause clinical progression.
Abbreviations: HPV, human papillomavirus; HMBS, hydroxymethylbilane synthase; HR-HPV, high-risk HPV; IFN, interferon; ORF, open reading frame; QPCR, quantitative polymerase chain reaction.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Integration of high-risk human papillomavirus (HR-HPV) is a key step in cervical neoplastic progression (1,2). Whereas most subclinical and low-grade infections harbour viral episomes, carcinomas (typically squamous cell carcinomas) usually contain integrated fragments of the viral genome, and have frequently lost all episomal copies (3–5). The overall frequency of integration in potentially premalignant squamous intraepithelial lesions is not certain but appears to increase with lesion grade (4–6). Integrants retain a variable segment of the viral genome, including the oncogenes E6 and E7, but generally show loss or disruption of the open reading frame (ORF) of the viral transcription/replication factor E2 (7,8). Altered expression of viral genes appears to be the principal source of the selective advantage conferred by integration. In cells containing integrated HR-HPV, E6 and E7 proteins are derived from transcripts stabilized by host polyadenylation signals (9,10). Moreover, among the functions of the E2 protein is a well-characterized capacity to repress the transcription of the E6 and E7 genes from integrated viral DNA; in the context of integration, therefore, loss of E2 has a de-repressive effect upon the viral oncogenes (11,12).
Integration must be understood in its cellular context. Cervical keratinocyte cell lines established from productive viral lesions indicate that HR-HPV episomes are maintained at 
100 copies per cell in basal epithelial cells (13–15). Should an integration event occur in a cell containing 100 episomes, a further 99 will remain, and will continue to express E2. Therefore, freeing E6 and E7 from repression requires not only disruption of the E2 ORF of the integrant itself, but also clearance of residual intact E2 ORFs by loss of the remaining episomes from the integrant-bearing cell (16). Therefore, integration and episomal loss are best regarded as discrete and equally important steps in the progression of viral physical state.
Our laboratory is investigating the transition in physical state of HPV-16 in cervical keratinocytes using the unique W12 model system. W12 is a polyclonal culture generated by one of us (M.A.S.) following explant culture of a low-grade cervical squamous intraepithelial lesion (LSIL). Therefore, the clinical lesion from which W12 was derived resulted from ‘natural’ infection of cervical keratinocytes with HPV-16—the HR-HPV type most frequently associated with cervical squamous cell carcinoma (13,17). At low-passage numbers, polyclonal W12 stably maintains episomes at a load of 100–200 per cell and no integrants are detectable by Southern analysis. The cells are genomically stable and in organotypic tissue culture they recapitulate the phenotypic properties of the LSIL from which they were derived (13,14,18). Over the course of prolonged cultivation, spontaneous integration and episomal loss are observed in W12, with ensuing high-level chromosomal instability and the acquisition of a more aggressive phenotype (18,19). Polyclonal W12 should be distinguished from the clone W12E, which was derived from W12 during a period when cells containing integrants only were spontaneously emerging and is reported to contain 1000 episomes per cell (9,19).
We have recently observed that spontaneous loss of episomes from W12 cells is associated with increased expression of antiviral genes that are inducible by type I interferon (IFN) (16). The notion that the IFN response may participate in the clearance of papillomavirus episomes is supported by evidence that cells transfected with either bovine papillomavirus (BPV) or HPV-31 episomes and cultivated in the presence of exogenous type I IFN undergo a gradual diminution of episomal load over the course of numerous passages (20,21). It has been inferred from such studies that type I IFN may have the capacity to cure some HPV infections. However, while IFN preparations have been shown to have some efficacy against benign HPV-related lesions, their overall role in the management of squamous intraepithelial lesions remains controversial (22,23).
The present study uses W12 to examine the effects of exogenous IFN-ß on cervical keratinocytes naturally infected with HPV-16. We reasoned that responses to exogenous type I IFN might differ in cervical keratinocytes naturally infected with HR-HPV, as these cells would be expected to display heterogeneity in viral copy number and physical state that mirror the circumstances that exist in vivo. We report the novel and significant observation that IFN-ß can dramatically accelerate the progression from an ostensibly episomal population to one in which only integrants remain. Our observations indicate that greater emphasis should be placed on episomal loss in models of HR-HPV-related carcinogenesis, and have substantial implications for clinical strategies that involve inducing an IFN response in treating HR-HPV-related disease in vivo.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Cell culture and growth analysis
All cells were grown in monolayer culture in order to model events that occur in the basal cells of cervical epithelium—the key site of dysregulated expression of viral oncogenes in HR-HPV-related carcinogenesis (24). W12 cells were cultivated in the presence of lethally irradiated G3T3 murine fibroblast feeder cells as described previously (13). Cells were treated with 1000 U/ml of recombinant human IFN-ß (Merck Biosciences, Nottingham, UK) The dose of 1000 U/ml IFN-ß was shown in preliminary experiments (M. T. Herdman and M. R. Pett, unpublished data) to induce IFN-stimulated genes to levels seen in W12 during spontaneous episome loss (16). The dose is also consistent with previous studies of the cellular effects of IFN-ß (21,25).
Passaging was undertaken at approximately 90% confluence, seeding the new passage with 5 x 105 W12 cells and 2 x 106 G3T3 feeders to a 9 cm dish. W12 clone MP4 (W12.MP4) was derived from the polyclonal W12 cell line by dilution cloning at Passage 11 as described previously (26). Studies of the short-term effects of IFN-ß were undertaken at Passage 17 for both polyclonal W12 and W12.MP4. In addition, we studied the long-term effects of IFN-ß by continuous treatment of W12.MP4 during serial passaging from Passage 17 to 22. All extractions of DNA, RNA and protein were performed at subconfluence, following removal of feeder cells. The HPV-negative cervical carcinoma cell line C33A (LGC Promochem, Teddington, UK) was cultivated as described previously (27), and served as a negative control for western immunoblotting.
Growth curves were initiated 2 days after seeding in order to allow identifiable colonies to emerge. At each time-point, images of 0.22 mm2 fields were captured at 25 points on a grid of 1 cm intervals across the growth surfaces of two replicates, using an Olympus IX50 inverted phase contrast microscope workstation (Olympus, London, UK). The cells in each field were then counted manually using AnalySIS software v.3.2 (Soft Imaging Systems, Münster, Germany). In this way, 1% of the growth surface was assessed at each time-point.
DNA preparation and Southern analysis
Total DNA extracts were prepared from W12 cells as described previously (14). Each DNA sample was digested with PstI or BamHI (New England Biolabs, Hitchin, UK), and electrophoresed through 0.8% agarose gel. DNA was blotted onto a Hybond N+ nylon membrane (Amersham Biosciences, Chalfont St. Giles, UK), and probed with [
-32P]dCTP-labelled full-length HPV-16 genomic probe as described previously (13).
Absolute quantitative PCR analysis of HPV-16 DNA copy number
All real-time quantitative PCR (QPCR) experiments were performed using adaptations of standard SYBR green protocols (28,29). HPV-16 DNA copy number was determined from a standard curve generated by subjecting a reference DNA construct to exactly the same amplification procedure as the test DNA samples. The reference DNA construct contained single copies of amplicons from the viral E2 and E6 ORFs and the housekeeping gene hydroxymethylbilane synthase (HMBS), and was prepared using the Original TA Cloning kit (Invitrogen, Paisley, UK). This was quantified by spectrophotometry and subjected to serial 10-fold dilutions to give reference standards ranging from 2 ng to 2 fg, which were used to establish primer efficiencies and calibration curves. The W12 DNA samples were quantified using the E2 and E6 primers described by Peitsaro et al. (30) and the HMBS primers described by Moberg et al. (29), in 25 µl reactions comprising SYBR Green JumpStart mix (Sigma-Aldrich, Gillingham, UK), 500 nM primer pair, and 10 ng template DNA. QPCR reactions were carried out using an MJR Opticon thermal cycler (MJ Research, Bio-Rad Laboratories, Waltham, MA). The following cycling conditions were employed: initial denaturation 94°C for 2 min; then 40 cycles of 94°C, 15 s; 58°C, 20 s; 72°C, 15 s; 76.5°C, 5 s; plate-read; then a final extension of 78°C, 8 min; then a melting curve from 65°C to 90°C. Absolute copy numbers of E2, E6, and HMBS genes in W12 DNA extracts were determined by comparison to the reference standards. All QPCR reactions were performed in triplicate, and each DNA sample was analysed in three technical replicates. The average number of episomes was determined from the mean of the E2 copy number divided by half the mean of the HMBS copy number [as the polyclonal cells used were diploid (18)].
Annexin-V fluorescence-activated cell sorting (FACS) analysis
W12 cells were analysed for phosphatidylserine externalization using annexin-V-FITC antibody (BD Biosciences, Erembodegem, Belgium) and propidium iodide (PI; Sigma-Aldrich), following the manufacturers' instructions. After labelling, the cells were passed through a 70 µm filter and sorted using a BD LSR II flow cytometer (BD Biosciences) and FACSDiVaTM software v.4.0 (BD Biosciences). For each time-point, populations were gated on the basis of side scatter and forward scatter to exclude cellular debris, and 10 000 gated events were plotted according to annexin-V-FITC and PI signal intensity. Quadrant gating and compensation were established independently for each time-point using unstained and single-stained reference samples. The ratio of FITC-positive PI-negative cells (quadrant 4) to FITC-negative PI-negative cells (quadrant 3) was taken as an index of early apoptosis (31). PI-positive cells were not included in the analysis because no distinction can be drawn between cells in this fraction that have undergone apoptosis and cells that have undergone necrosis during sample preparation.
Relative QPCR analysis of HPV-16 gene expression
Cells were lysed in situ by application of Bio-RNA Xcell2 solution (BioGene, Kimbolton, UK), and total RNA extracts were prepared according to the manufacturer's instructions. Residual DNA was removed by Turbo DNase treatment (Ambion, Huntingdon, UK), and messenger RNA was purified using the Oligotex mRNA Minikit (QIAgen, Crawley, UK). mRNA was reverse-transcribed using the RT–PCR system (Promega, Southampton, UK) with oligo(dT)15 primers. RT-negative samples were analysed to confirm the absence of DNA contamination.
QPCR of cDNA for levels of HPV transcripts was performed using a protocol adapted from Ponchel and colleagues (28), and calculations derived from Pfaffl (32). Using the primer sequences specified by Vandesompele and colleagues (33), four housekeeping genes—glyceraldehyde-3-phosphate dehydrogenase (GAPDH), HMBS, hypoxanthine guanine phosphoribosyltransferase 1 (HPRT1) and tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein, zeta isoform (YWHAZ)—were found to be unresponsive to IFN-ß in W12 cells (M. T. Herdman, unpublished data) and used as reference genes. Primers for the HPV-16 test transcripts were as follows: E2-Fwd (5'-ACCAAGATCAGAGCCAGACAC-3'), E2-Rev (5'-AGTGAGGATTGGAGCACTGTC-3'); E6-Fwd (5'-TGTTTCAGGACCCACAGGAGC-3'), E6-Rev (5'-CGCAGTAACTGTTGCTTGCAG-3'); E7-Fwd (5'-AGGAGGATGAAATAGATGGTCCAG-3') and E7-Rev (5'-CTTTGTACGCACAACCGAAGC-3'). The efficiency of each primer pair was calculated from the slope of a 6-point dilution series, as described previously (32).
Each IFN-ß-treated or untreated W12 cDNA sample was subjected to QPCR for the four housekeeping reference genes and three viral test genes, performing each reaction in triplicate. Reactions comprised 1x SYBR Green JumpStart mix (Sigma-Aldrich), 500 nM primer pairs and 250 pg cDNA in a final volume of 50 µl. Amplification was detected in real-time using the MJR Opticon thermal cycler under the same cycling conditions used for DNA (see above), but extending the quantitation process to 45 cycles. To eliminate any non-specific fluorescence signal, background fluorescence (taken as the minimum signal between cycles 10 and 45) was subtracted from each reaction curve. For each gene, the fluorescence threshold was optimized to coincide with the period of exponential amplification (two cycles before the curve's second derivative maximum). A crossing point (CP) was derived for each curve, and the median of the triplicate CP values was used for the calculation. The following adaptation of Pfaffl's equation was used to assess the change in expression induced by IFN-ß for each viral gene, compared with each of the four reference genes in turn, where E = primer efficiency, 
CP = (CPIFN negative – CPIFN-ß), and subscripts v and r refer to viral and reference genes respectively:
 |
For a given viral gene, expression ratios were calculated in relation to each of the four reference genes, and the mean and standard deviation of these values were established. Changes in expression ratios were plotted on a log scale, in order to give equal weight to gains and losses in expression. 95% confidence intervals (CIs) for significant changes in expression were calculated from internal comparisons between the four reference genes.
Western immunoblotting
Levels of HPV-16 E7 protein were analysed by western immunoblotting of total protein extracts as described previously (14), using sheep anti-E7 antibody (a kind gift of S. Inglis, Cantab Pharmaceuticals, Cambridge, UK). C33A protein extract was used as a negative control and protein from polyclonal W12 at p51—shown previously to contain high levels of E7 (14)—was used as a positive control. Levels of HPV-16 E2 protein were assessed in enriched nuclear extracts, prepared as described previously (12), using mouse anti-E2 monoclonal antibody (a kind gift of M. Hibma, University of Otago, New Zealand). C33A nuclear extract was used as a negative control and recombinant HPV-16 E2 expressed in insect cells (a kind gift of N. Maitland, University of York, UK) was used as a positive control. To confirm equal loading and transfer, membranes were stripped and re-blotted with mouse anti-ß-actin monoclonal antibody (Abcam, Cambridge, UK).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
IFN-ß promotes very rapid transition in the physical state of HPV-16 from episomal to integrated
We first examined the effect of exogenous IFN-ß on the physical state of HPV-16 in cells of polyclonal W12 at Passage 17 that contained HPV-16 episomes but showed no evidence of integrated viral DNA by Southern analysis at the beginning of the experiment (Figure 1A). The untreated cells retained episomes at a constant level throughout the 8-day time-course. In contrast, IFN-ß caused a brisk reduction in the intensity of the episomal bands over the same period, indicating rapid reduction in the amount of HPV-16 DNA. Strikingly, by the time the treated cells had reached near-confluence (Day 18), two new bands had arisen (Figure 1A; blue arrows), representing virus–host junctions of an integrant that had come to be present in a large proportion of the cells. Thus, over the course of a single passage of W12, IFN-ß caused reduction in episome numbers and the emergence of a previously undetectable HPV-16 integrant.
|
Next, low-passage polyclonal W12 cells were subjected to dilutional cloning at Passage 11. We can deduce that the resulting clonal populations were completely free of integrants at the time of cloning, because had any integrants existed in the single cell from which the clone was derived, those integrants would be ubiquitous in the daughter population. One episome-containing clone isolated in this manner—W12.MP4—was cultivated for six additional passages prior to further investigation. W12.MP4 at Passage 17 behaved in a similar manner to polyclonal W12, maintaining episomes in IFN-negative medium, but showing a rapid reduction in episome numbers over 8 days in the presence of IFN-ß (Figure 1B). By the end of the following passage of W12.MP4 (Passage 18) an integrant previously undetectable by Southern analysis came to dominate the treated population (Figure 2A; Series 1, lane 5). We concluded that the integration event must have occurred between Passage 11 (when the W12.MP4 clone was generated from a single, episome-only cell) and 18. To date, we have observed a rapid transition from episomal to integrated HPV-16 DNA following IFN-ß treatment in nine separate experiments, using both polyclonal and clonal W12 (M. T. Herdman, unpublished data).
|
To achieve a clearer understanding of the kinetics of the reduction in episome numbers, the number of episomes per cell in treated and untreated W12 cells was determined by QPCR, analysing the copy number of the HPV-16 E2 gene [which is generally deleted from W12 integrants that are selected (16,18)]. Figure 1C and D correspond to the Southern blots of Figure 1A and B respectively, and demonstrate that both polyclonal W12 and the W12.MP4 clonal series rapidly underwent considerable reduction in episome numbers, even within the first 24 h of IFN-ß treatment. For both series, the treated cells showed a steady reduction in episome numbers (blue bars), which by 8 days had fallen in both populations by
80%. Copy numbers of the HPV-16 E6 gene were also quantified, and were found to follow the same trend as E2 (data not shown); this observation was not surprising, as the E6 DNA copy number in the lost episomes was much greater than the E6 copy number in the few integrants that were present during the first 8 days of treatment (which was prior to the emergence of integrants by Southern blot analysis). In our second set of experiments, W12.MP4 cells were serially passaged in two identical long-term cultures, diverging from a common population at Passage 16, and treated continuously with IFN-ß from Passage 17 to 19 (Figure 2A, Series 1 and 2). By the end of the second passage of treatment (Passage 18), both series showed a reduction in episome numbers of >99%. Significantly, the banding pattern of the Southern blot indicates that the same integrant had come to dominate both populations (indicated by blue arrows in Figure 2A, with white arrows showing bands shared by episomes and integrants). Cells bearing the integrant must therefore have been present in the starting population at Passage 16, below the threshold of detection of Southern blotting (an observation that argues against the view that IFN treatment generates new integration events). Despite the presence of the integrant in a minority of cells of the starting population of W12.MP4 at Passage 16, the untreated cells in both series retained a full episomal load for at least four passages. These findings refute the notion that HR-HPV integration must immediately be followed by episomal loss. Instead, they show that integrants can be present in some cells in a population for lengthy periods before they exert a selective advantage.
In a third serial passaging experiment, using W12.MP4 cells cultivated from Passage 17 (Figure 2B), an integrant emerged spontaneously in IFN-negative cells after several passages (lanes 1–5). This was associated with spontaneous episome loss, as described previously (16). In cells continuously treated with IFN-ß (lanes 6–10), episome numbers were rapidly reduced and multiple integrants were detected at the end of Passage 17 (including the integrant that arose in the IFN-negative series; lane 6). By Passage 22 of the treated series, all but two of the virus–host junctional bands had faded (lane 10), indicating that a single integrant (which was different to that seen in the IFN-negative series) had come to dominate the population, presumably by virtue of having the greatest growth advantage in IFN-ß-containing medium. As a further component of this experiment, some cells were treated with IFN-ß for the first passage (Passage 17) only, then returned to IFN-negative medium for three additional passages (lanes 11–13). In these cells, episome numbers again reduced quickly, but multiple junctional bands persisted. Importantly, these bands included those that dominated both the untreated series (lane 5) and the continuously treated series (lane 10). This finding supports results from the second serial passaging experiment, and indicates that IFN-ß caused selection of integrant-containing cells already present in the starting population.
These data emphasize that multiple integrants can exist latently in a population that appears to contain only episomes. As long as episomes persist in the population, cells in which integrants are present have no selective advantage over cells with episomes only. In contrast, when episome numbers reduce, integrants quickly become apparent and co-exist until one outgrows the others. The continued presence of IFN-ß provides a strong selective pressure, resulting in the rapid dominance of the integrant-containing cells with the greatest growth advantage.
Changes in cell growth correspond to change in the physical state of HPV-16
IFN-ß dramatically reduced the rate of growth of polyclonal W12 for the first 8 days of treatment (Figure 3A), coinciding with the period of reduction in episome numbers (cf. Figure 1C). After this point, however, there was an increase in the growth rate (although this did not attain levels seen in the absence of IFN-ß). Analysis of our Southern data strongly suggest that the cells that have a relative growth advantage in IFN-ß-containing medium once episome numbers have reduced are those that harbour integrants (Figure 1A). The clonal W12.MP4 series exhibited a longer period of impeded growth (lasting until approximately Day 19) before a similar increase in growth rate was observed (Figure 3B). This pattern mirrors the relatively late appearance of integrants in W12.MP4 (Figures 1B and 2A). The delay in demonstrable physical state transition in W12.MP4 compared with polyclonal W12 is consistent with the notion that fewer cells capable of relatively IFN-resistant growth (i.e. those containing integrants) are available to gain a selective advantage following reduction in episome numbers.
|
In all experiments, varied morphological appearances were seen within individual dishes of IFN-treated cells, where flourishing colonies were often visible alongside colonies growing poorly or eventually undergoing morphological changes consistent with senescence or apoptosis (Figure 4). This supports the premise that the response to IFN-ß is variable, with some colonies growing well in the presence of the cytokine, while others grow poorly and eventually apoptose. A more detailed assessment of the pattern of cell death observed is discussed below.
|
When IFN-ß-treated W12.MP4 cells were returned to IFN-negative medium at Passage 18 (the passage after episomes were lost; Figure 2A), the growth rate (Figure 3C; purple) was essentially the same as that of untreated cells at Passage 17 (pink). In contrast, with continuing cultivation in IFN-ß at Passage 18, the growth rate (green) was dramatically increased from that of treated cells at Passage 17 (pale blue), further illustrating the relative IFN resistance of the emergent population. Taken together, our growth data suggest that the greatest IFN-ß-induced growth inhibition is seen when episomes are detectable, consistent with previous reports that the presence of E2 can confer sensitivity to the growth-inhibiting effects of IFN-ß (21). Once episomes are lost, cells with integrated HPV-16 grow at an increased rate and come to predominate, at a pace that depends on the prevalence of cells containing integrants in the starting population.
Mechanism of reduction in episome numbers following interferon-ß treatment
We considered two models to explain the very rapid reduction in episome numbers and emergence of integrants following IFN-ß treatment: one cytolytic and the other non-cytolytic. In the cytolytic model, IFN-ß induces early death of all episome-containing cells. This model presumes that there must also be rare integrant-only cells, which have already cleared their episomes, yet have no selective advantage over episome-only cells that make up the majority of the population. When IFN-ß is added, only these integrant-bearing cells survive, and they predominate by default.
The second model involves non-cytolytic episome clearance. In this scenario, cells containing only episomes would lose the proliferative drive from episomally expressed E6 and E7. These cells would eventually undergo senescence or apoptosis, but only after a lag phase required for episome loss to occur. In contrast, for cells containing both episomes and integrants (i.e. an intracellular mixture), loss of episomes—and therefore E2—would de-repress the integrants and lead to increased levels of integrant-derived E6 and E7. These changes would in turn provide the integrant-only cells with a relative resistance to apoptosis and confer a growth advantage, again after a lag phase required for episome loss.
These two models are not necessarily mutually exclusive. While some cells (including those bearing integrants) may clear their episomes non-cytolytically, others may succumb to the pro-apoptotic drive of the IFN response. Indeed, such a polymorphous response would be consistent with the multi-faceted nature of antiviral pathways. We investigated which mechanism makes the predominant contribution to the reduction in episome numbers that we observed. While some data of use in resolving this question were available from the previous work of our group (16,18) and others (19), we sought further evidence by analysing how changes in the physical state of HPV-16 following IFN-ß treatment related to cell death and to changes in viral gene transcription and translation.
Reduction in episome numbers is not preceded by a peak in apoptosis
We first looked at apoptosis in relation to the reduction in episome numbers. We predicted that if the cytolytic model predominated, there would be a major initial surge of apoptosis, coinciding with the period of reduction in episome numbers (which approached 50% over the first 2 days; Figures 1C, 1D). In contrast, if episomes were largely cleared by a non-cytolytic mechanism, there would not be a major initial surge of apoptosis. Instead, there might be some relatively late apoptosis in this setting, due to the death of episome-only cells that had lost the proliferative and anti-apoptotic effects of viral E6 and E7 following episome clearance. In this case, apoptosis would be at its highest when episome clearance was almost complete.
The kinetics of apoptosis over the course of IFN-ß treatment of W12.MP4 cells were assessed using the annexin-V assay of phosphatidylserine externalization—a marker of the inversion of cellular membranes in the early stages of apoptosis (31). While apoptosis was increased in the presence of IFN-ß throughout the time-course (Figure 5), there was no initial peak, with little apoptosis during the 50% reduction in episome numbers over the first 2 days of IFN-ß treatment (Figures 1D and 6B). Instead, apoptosis was markedly greater between Days 4 and 8 (i.e. peaking relatively late in the course of episomal loss) when 75% of the episomes in the population as a whole had already been lost (Figure 5). This late surge in apoptosis opposes a predominantly cytolytic model and suggests a role for a non-cytolytic mechanism of episome clearance. Given that wild-type p53 is present in W12 cells (12), the late surge in apoptosis is consistent with a loss of E6 protein from those cells lacking integrants, leading to accumulation of p53 and an increase in pro-apoptotic effects.
|
|
Reduction in episome numbers is not preceded by reduced viral gene transcription or translation
We next studied the relationship between the change in the physical state of HPV-16 and levels of viral mRNA and protein, assessed by QPCR (Figure 6) and western immunoblotting (Figure 7), respectively. We had two objectives in doing so. Our first aim was to investigate whether IFN-ß treatment might achieve non-cytolytic clearance of HPV-16 episomes by interfering with the transcription or translation of viral genes, as is seen in a range of other viral infections (34,35). Given that the HPV proteins E1 and E2 are critical for episomal replication, we hypothesized that IFN-ß could achieve episomal clearance by selectively inhibiting transcription or translation of these genes. In the absence of E1 and E2, HPV replication would cease, while host cells would continue to divide, diluting the remaining episomes among their progeny. In this scenario, levels of viral mRNA and/or protein would be predicted to show a substantial initial fall, which would precede the reduction in episome numbers.
|
Our second aim was to investigate whether the integrant that emerged following IFN-ß treatment was characterized by conspicuously elevated levels of E6 and E7, which we and others have shown to provide a strong selective advantage compared with episome-only cells (18,19). If so, cells containing such an integrant would not be expected to remain a minority population for extended periods alongside W12 cells in which only episomes are detectable, as cells highly expressive of viral oncogenes would soon come to predominate. We reasoned that if cells with these features did emerge in our experiments it would be very unlikely that they had such features prior to IFN-ß treatment. Instead, a more credible scenario would be that such integrants were previously repressed by episomes coexisting within the same cell, and only came to be highly expressive of E6 and E7 when episome clearance was achieved.
Messenger RNA levels
We used relative QPCR to determine the changes in levels of HPV-16 mRNA in polyclonal W12 and W12.MP4. In order to provide direct comparisons with relative changes in viral load in the same populations, we used the data shown in Figure 1C and D to determine the fold change in episome load for IFN-ß-treated versus untreated cells (shown in Figure 6A and B, respectively. Fold change was plotted on a log scale to give equal weight to gains and losses in episome load; the steady loss of episomes between Days 0 and 8 is best seen in Figure 1C and D). IFN-ß did not induce a sudden fall in viral transcripts in polyclonal W12 (Figure 6C), nor in W12.MP4 (Figure 6D). In polyclonal W12, over the first 4 days of treatment, no statistically significant decline in the expression of any viral ORF was observed. E2, E6 and E7 mRNA levels all showed a downward trend from Day 2 to 6, accompanying—but not preceding—the reduction in episome numbers (Figure 6A). Beyond Day 6, levels of E2 mRNA continued to decline as episome numbers did the same. In contrast, E6 and E7 followed an upward trend as integrants emerged.
The initial decline in total levels of E6 and E7 mRNA in polyclonal W12 is most likely explained by the reduced levels of DNA template, as the episomes were lost (Figures 1C and 6A). The rate of decline in E6 and E7 mRNA levels after Day 4 was not as great as for E2 mRNA, and after Day 6 both E6 and E7 mRNA levels showed an upward trend as cells containing integrants emerged. We revised the expression ratios (Figure 6E) to depict changes in E6 and E7 mRNA levels in polyclonal W12 (Figure 6C) as a function of the DNA copy numbers of the viral ORF templates (Figures 1C and 6A) over the same time-course. DNA copy numbers were derived from PCR assessment of levels of the viral E6 gene, which is present with E7 in both integrants and episomes (data not shown). As the episomal load fell to 10% of that of untreated polyclonal W12 cells—between Days 0 and 6—expression of E6 and E7 per copy of DNA template remained similar to that of untreated cells (Figure 6E). In contrast, between Days 6 and 18 (when the number of remaining episomes approached zero and integrants accounted for an increasing number of E6 and E7 DNA copies) the E6 and E7 gene expression ratios per DNA template showed a gradual rise, with 12-fold increases by Day 18, when <5% of episomes remained (Figure 6E). This up-regulated expression of integrant-derived E6 and E7 is consistent with transcriptional de-repression of integrants, associated with E2 loss following a reduction in episome numbers.
In clone W12.MP4 the changes in mRNA levels were measured over a short time-course of 8 days only, during which time episome numbers reduced substantially (Figures 1D, 6B) but the integrant-containing cells had not yet emerged to be detectable by Southern analysis (Figure 1B). We observed reduced levels of E2, E6 and E7 mRNA over 8 days, consistent with the loss of episomal DNA template (Figure 6D). Expression of E6 and E7 adjusted for template number did not change dramatically during this short time-course (Figure 6F), consistent with the fact that integrants emerged later in W12.MP4 than in polyclonal W12. Messenger RNA was not available from later in the W12.MP4 time-course, when integrants eventually emerged. However, changes in E7 protein were assessed over this longer time-course (see below).
Protein levels
In polyclonal W12 the trends observed in viral protein levels matched those seen for the viral transcripts. E2 protein underwent a gradual decline (Figure 7A), coinciding with the steady reduction in numbers of viral DNA template (Figure 1C) and the fall in E2 mRNA level (Figure 6C). The overall concentration of E7 protein in the polyclonal W12 population declined temporarily but increased again by Day 18 of the time-course (Figure 7B), when there were negligible levels of E2 DNA (Figure 6A) and mRNA (Figure 6C); there were insufficient cells to confirm that E2 protein was present at negligible levels at Day 18. In the clonal W12.MP4 series, levels of E7 protein were also assessed over a much longer time-course. In keeping with the mRNA data, E7 protein levels in the population declined over 8 days, as episome numbers reduced (Figure 7C). However, by Passage 18—the point at which integrant-bearing cells were first apparent in W12.MP4 by Southern analysis—E7 protein was strongly expressed (Figure 7D). Importantly, over the four succeeding passages—when episomes were undetectable and only integrant-containing cells were seen—levels of E7 protein per cell were considerably greater than in the starting population.
These data indicate that the reduction in episome numbers in W12 following IFN-ß treatment is not achieved by silencing viral gene transcription or translation, for such a mechanism would require a substantial reduction in E2 levels prior to the reduction in episomal load. Rather, they suggest that the slow fall in E2 mRNA and protein levels is secondary to the loss of template episome DNA. Meanwhile, in the population as a whole, E6 and E7 expression undergoes an initial decline, corresponding with (but not preceding) the reduction in episome numbers, until highly expressive, integrant-bearing cells come to dominate the population when E2 has declined to negligible levels.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
We show for the first time that IFN-ß can dramatically hasten the transition from episomal to integrated HPV-16 in naturally infected cervical keratinocytes. Interestingly, this finding parallels our recent identification of endogenous activation of antiviral genes that are inducible by type I IFN during spontaneous progression from episome-only cells to integrant-only cells in W12 (16). Our present data also show that integrated HPV-16 DNA can exist for long periods in a minority of cells in a heterogeneous population without exerting a selective growth advantage. However, once the reduction in episome numbers is initiated, the integrant-bearing cells quickly dominate the population.
Our findings permit re-evaluation of earlier studies examining the effects of IFN-ß on cells containing other papillomaviruses. Turek and colleagues (20) transfected BPV episomes into immortalized murine fibroblasts, and observed gradual episome loss in response to treatment with murine IFN. BPV episome numbers were reduced much more slowly than we have observed for HPV-16, requiring >7 passages to achieve complete loss. The emergence of integrant-bearing cells was not observed [although integration appears to be a relatively rare occurrence in this system (36–38)].
In a study by Chang and colleagues (21) of normal human foreskin keratinocytes transfected with HPV-31 and containing detectable integrated and episomal viral DNA, IFN-ß caused growth inhibition by 3 days, with the emergence of IFN-ß-resistant cells at between Days 10 and 14. However, the nature of the IFN-resistant population remained uncertain. Changes in viral physical state occurred over a very different time-scale, with episomes being lost gradually over 8 passages and integrants actually decreasing (albeit to a more modest degree than the episomes). Nevertheless, data from cervical carcinoma cell lines and cells over-expressing HPV-31 E2 led Chang and colleagues (21) to infer that IFN-ß treatment could provide a selective advantage to cells containing integrated HR-HPV. Our current data provide novel, direct evidence demonstrating that this is the case. The W12 model is particularly well suited to such investigations, as it represents a polyclonal population of cells derived from cervical squamous epithelium naturally infected with HPV-16. In such a population the response to IFN-ß is dramatic, with rapid emergence of IFN-resistant cells in close temporal association with reduction in episome numbers and selection of cells containing integrated HPV-16.
There has been sustained interest in using IFN and other immunomodulators to treat benign and premalignant HPV-associated lesions of the cervix and other sites (22,23). Many trials have been too small, poorly controlled, and varied in their design to permit a detailed meta-analysis. Nevertheless, studies of patients with premalignant squamous intraepithelial lesions typically show partial success, with most patients achieving regression, but with some persistent or progressive lesions (23,39). Assuming the population dynamics observed here in vitro also apply in vivo, our data provide the most direct and strongest evidence to date that IFN treatment of HR-HPV-associated lesions may be detrimental. While cure may be achieved in lesions containing episomes only, treating a lesion containing a mixture of episomal and integrated HR-HPV DNA might be ineffective or counterproductive. Although decreasing the number of episomes would reduce the probability of new integration events occurring, the emergence of integrant-containing cells with elevated levels of viral oncogene expression and a selective growth advantage would be expected to cause clinical progression.
We considered two non-exclusive models to explain the observed reduction in episome numbers and emergence of integrant-only cells following IFN-ß treatment: one cytolytic and the other non-cytolytic. Firstly, it was possible that IFN-ß induced death of episome- containing cells but did not kill cells containing integrants only, which emerged from latency to dominate the population. This model requires that W12 populations where only episomes are detectable by Southern analysis also contain a minority of cells containing integrants only. The second model holds that latent integrants exist in cells that also contain episomes as part of an intracellular mixed state. In these cells, integrants are repressed by episome-derived E2, until IFN-ß reduces episome numbers by a non-cytolytic mechanism, with consequent integrant de-repression.
In our opinion, three lines of evidence support an important role for the latter, non-cytolytic model. First, the integrant-only cells that emerged following IFN-ß treatment showed up-regulated transcription of E6 and E7 per gene DNA copy, at up to 12 times the levels of untreated, episome-bearing cells (see Figure 6E). In cells followed for an adequate time-course (treated W12.MP4 cells at Passages 21 and 22; see Figure 7D), we observed substantially greater levels of E7 protein per cell than in the untreated episome-only cells. Such integrant-only, E7-overexpressing cells have been shown previously to have a much higher clonogenicity than episome-only cells (18) and a large selective growth advantage in mixed cultures [(19) and M. R. Pett, unpublished observations]. However, in multiple long-term cultures of polyclonal W12 populations not treated with IFN-ß, episomes can be maintained for >40 passages without the emergence of integrants (16), strongly suggesting that the E7-overexpressing integrant-only cells are not present. Emergence of integrant-only cells in polyclonal W12 only occurs when there has been a rapid fall in episome numbers, occurring either spontaneously (16) or following IFN-ß treatment.
Second, the surge in apoptosis that we observed did not occur until after the episome load was in steep decline. Episome numbers were reduced after 24 hours, and levels had fallen by 50% at 48 h (see Figure 1C and D). In clear contrast, the peak of apoptosis in the same population did not occur until Day 6. Were cytolysis to be the principal mechanism of episome loss, a high level of apoptosis would be anticipated over the period of steepest decline in episome numbers. The observed lag before the surge in apoptosis argues for cell death occurring following non-cytolytic clearance of episomes, when the proliferative and anti-apoptotic effects of the viral proteins have been lost. The apoptosis we observed may be p53-dependent, as the TP53 gene is wild-type in W12 cells, and p53 protein is detectable by western blotting [(12) and X.-Y. Zang and M. T. Herdman, unpublished data]. However, the associations between p53 expression and the other parameters measured in the present study are likely to be complex. As well as being affected by changing levels of HPV-16 E6, p53 protein is also independently inducible by type I IFN (40).
Third, we obtained supporting data from our recent investigation of events seen when integrant-only cells spontaneously emerge in untreated polyclonal W12 cells, in association with spontaneous reductions in episome numbers (16). Single cell cloning by limiting dilution during the period of episome loss demonstrated the presence of cells containing an intracellular mixture of the integrant being selected and episomes (16). Moreover, we found evidence for the up-regulation of genes inducible by type I IFN in these cells during spontaneous reduction in episome numbers. This suggests that integrants exist in the context of intracellular mixtures and are repressed until episome numbers are reduced, either spontaneously by endogenous induction of an antiviral response, or by the application of exogenous IFN-ß. In support of our findings in the present study, we observed that spontaneously emerging integrants have elevated levels of E6 and E7 mRNA per cell (16). Interestingly, during the emergence of these cells, residual E2 at 2% of mRNA levels in polyclonal W12 cells containing only episomes by Southern analysis still showed partial repression of integrants, with full de-repression only being seen when there was complete loss of E2.
When taken together, our data therefore support a role for episome clearance by non-cytolytic means. At first, most cells retain enough E6 and E7 to overcome the pro-apoptotic effect of IFN-ß (23), with episome-derived E2 conferring sensitivity to the growth-inhibiting effects of IFN-ß (21). As episome copy numbers fall, most cells revert to an HPV-negative state in which their susceptibility to apoptosis is increased. At the same time, rare cells that contain integrants as well as episomes lose E2, up-regulate E6 and E7, and divide at an increased rate, with relative resistance to IFN-ß. The length of the interval before integrant-containing cells come to predominate may depend on the prevalence of such cells, as evidenced by our observation of delayed acceleration of the growth of clonal W12.MP4 cells compared with polyclonal W12. The growth advantage that integrant-containing cells acquire is likely to result from loss of E2 and consequent de-repression of E6 and E7. E2 also has an independent capacity to inhibit proliferation and survival (through both p53-dependent and -independent mechanisms), even in the absence of viral oncogenes (41,42). In the context of viral integration, E2 will not only repress transcription of E6 and E7 from the integrant (12), but also interact directly with the E6 and E7 proteins, causing their cellular relocalization and functional inhibition (43,44).
Many of the non-cytolytic antiviral effector mechanisms associated with type I IFN rely on undermining viral replication (34,35). The kinetics of the loss of HPV-16 mRNA and protein suggest that the reduction in HPV-16 episome numbers caused by IFN-ß is not preceded by silencing of viral transcription or translation. While transcript and protein levels for all HPV-16 genes examined do initially fall with IFN-ß treatment, they do so only after the clearance of viral template DNA is well underway. Moreover, adjusting for the viral load, expression of E6 and E7 mRNA per viral gene actually increases when integrants emerge following IFN-ß treatment. It remains a possibility that episomal DNA could be cleared non-cytolytically by sequestration away from the cellular replicative machinery. Type I IFNs have been shown to reorganize nuclear domains including promyelocytic leukaemia (PML) bodies (45), the location of several viral proteins in papillomavirus-infected cells, and possibly the site of viral replication (46-48). PML bodies may be involved in halting the replication of some DNA viruses through sequestration (49), though the details of this relationship in papillomavirus infection have yet to be elucidated. It will be of interest in future studies to examine the subcellular location of HPV-16 DNA following IFN treatment.
In conclusion, the transition in the physical state of HR-HPV has been identified as an important step in cervical neoplastic progression in vivo (3–5) and in vitro (18,19,50), yet these studies have frequently highlighted the role of viral integration without addressing the concomitant process of episomal loss. While integration of HR-HPV is undoubtedly a critical step, our results indicate that it occurs frequently and does not confer an immediate selective advantage. Rather, it is the subsequent process of episomal loss that permits cells with integrants to predominate. Our data support a model where IFN-ß depletes E2 from the initially rare cells that also contain integrants, thus de-repressing integrant-derived E6 and E7 gene expression. Generalized episomal loss would also remove the proliferative drive from cells lacking integrants, thereby enhancing the selective advantage of integrant-containing cells.
|
Acknowledgments |
|---|
This study was funded by the Medical Research Council and Cancer Research UK. M.T.H. was supported by the Comyns Berkeley Research Studentship from Gonville and Caius College, Cambridge.
Conflict of Interest Statement. None declared.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
- Münger K., Baldwin A., Edwards K.M., Hayakawa H., Nguyen C.L., Owens M., Grace M., Huh K. (2004) Mechanisms of human papillomavirus-induced oncogenesis. J. Virol. 78:11451–11460.
[Free Full Text] - Longworth M.S. and Laimins L.A. (2004) Pathogenesis of human papillomaviruses in differentiating epithelia. Microbiol. Mol. Biol. Rev. 68:362–372.
[Abstract/Free Full Text] - Dürst M., Kleinheinz A., Hotz M., Gissman L. (1985) The physical state of human papillomavirus type 16 DNA in benign and malignant genital tumours. J. Gen. Virol 66:Pt 71515–1522.
[Abstract/Free Full Text] - Klaes R., Woerner S.M., Ridder R., Wentzensen N., Dürst M., Schneider A., Lotz B., Melsheimer P., von Knebel Doeberitz M. (1999) Detection of high-risk cervical intraepithelial neoplasia and cervical cancer by amplification of transcripts derived from integrated papillomavirus oncogenes. Cancer Res. 59:6132–6136.
[Abstract/Free Full Text] - Tonon S.A., Picconi M.A., Bos P.D., Zinovich J.B., Galuppo J., Alonio L.V., Teyssie A.R. (2001) Physical status of the E2 human papilloma virus 16 viral gene in cervical preneoplastic and neoplastic lesions. J. Clin. Virol. 21:129–134.[CrossRef][ISI][Medline]
- Lehn H., Villa L.L., Marziona F., Hilgarth M., Hillemans H.G., Sauer G. (1988) Physical state and biological activity of human papillomavirus genomes in precancerous lesions of the female genital tract. J. Gen. Virol. 69:Pt 1187–196.
[Abstract/Free Full Text] - Schwarz E., Freese U.K., Gissmann L., Mayer W., Roggenbuck B., Stremlau A., zur Hausen H. (1985) Structure and transcription of human papillomavirus sequences in cervical carcinoma cells. Nature 314:111–114.[CrossRef][Medline]
- Choo K.B., Pan C.C., Han S.H. (1987) Integration of human papillomavirus type 16 into cellular DNA of cervical carcinoma: preferential deletion of the E2 gene and invariable retention of the long control region and the E6/E7 open reading frames. Virology 161:259–261.[CrossRef][ISI][Medline]
- Jeon S. and Lambert P.F. (1995) Integration of human papillomavirus type 16 DNA into the human genome leads to increased stability of E6 and E7 mRNAs: implications for cervical carcinogenesis. Proc. Natl Acad. Sci. USA 92:1654–1658.
[Abstract/Free Full Text] - Wentzensen N., Ridder R., Klaes R., Vinokurova S., Schaefer U., Doeberitz M.K. (2002) Characterization of viral-cellular fusion transcripts in a large series of HPV16 and 18 positive anogenital lesions. Oncogene 21:419–426.[CrossRef][ISI][Medline]
- Dowhanick J.J., McBride A.A., Howley P.M. (1995) Suppression of cellular proliferation by the papillomavirus E2 protein. J. Virol. 69:7791–7799.[Abstract]
- Bechtold V., Beard P., Raj K. (2003) Human papillomavirus type 16 E2 protein has no effect on transcription from episomal viral DNA. J. Virol. 77:2021–2028.
[Abstract/Free Full Text] - Stanley M.A., Browne H.M., Appleby M., Minson A.C. (1989) Properties of a non-tumorigenic human cervical keratinocyte cell line. Int. J. Cancer 43:672–676.[ISI][Medline]
- Alazawi W., Pett M., Arch B., Scott L., Freeman T., Stanley M.A., Coleman N. (2002) Changes in cervical keratinocyte gene expression associated with integration of human papillomavirus 16. Cancer Res. 62:6959–6965.
[Abstract/Free Full Text] - Lowy D.R. and Howley P.M. (2001) Papillomaviruses. In Fields B.N., Knipe D.M., Howley P.M., Griffin D.E. (Eds.). Fields' Virology(Lippincott Williams & Wilkins, Philadelphia) pp. 2231–2256.
- Pett M.R., Herdman M.T., Palmer R.D., Yeo G.S.H., Shivji M.K., Stanley M.A., Coleman N. (2006) Selection of cervical keratinocytes containing integrated HPV16 associates with episome loss and an endogenous antiviral response. Proc. Natl Acad. Sci. USA 103:3822–3827.
[Abstract/Free Full Text] - Muñoz N., Bosch F.X., de Sanjosé S., Herrero R., Castellsagué X., Shah K.V., Snijders P.J., Meijer C.J. (2003) Epidemiologic classification of human papillomavirus types associated with cervical cancer. N. Engl. J. Med. 348:518–527.
[Abstract/Free Full Text] - Pett M.R., Alazawi W.O., Roberts I., Dowen S., Smith D.I., Stanley M.A., Coleman N. (2004) Acquisition of high-level chromosomal instability is associated with integration of human papillomavirus type 16 in cervical keratinocytes. Cancer Res. 64:1359–1368.
[Abstract/Free Full Text] - Jeon S., Allen-Hoffmann B.L., Lambert P.F. (1995) Integration of human papillomavirus type 16 into the human genome correlates with a selective growth advantage of cells. J. Virol. 69:2989–2997.[Abstract]
- Turek L.P., Byrne J.C., Lowy D.R., Dvoretzky I., Friedman R.M., Howley P.M. (1982) Interferon induces morphologic reversion with elimination of extrachromosomal viral genomes in bovine papillomavirus-transformed mouse cells. Proc. Natl Acad. Sci. USA 79:7914–7918.
[Abstract/Free Full Text] - Chang Y.E., Pena L., Sen G.C., Park J.K., Laimins L.A. (2002) Long-term effect of interferon on keratinocytes that maintain human papillomavirus type 31. J. Virol. 76:8864–8874.
[Abstract/Free Full Text] - Cirelli R. and Tyring S.K. (1994) Interferons in human papillomavirus infections. Antiviral Res. 24:191–204.[CrossRef][ISI][Medline]
- Koromilas A.E., Li S., Matlashewski G. (2001) Control of interferon signaling in human papillomavirus infection. Cytokine Growth Factor Rev. 12:157–170.[CrossRef][ISI][Medline]
- Higgins G.D., Uzelin D.M., Phillips G.E., McEvoy P., Marin R., Burrell C.J. (1992) Transcription patterns of human papillomavirus type 16 in genital intraepithelial neoplasia: evidence for promoter usage within the E7 open reading frame during epithelial differentiation. J. Gen. Virol. 73:2047–2057 (Pt 8).
[Abstract/Free Full Text] - Der S.D., Zhou A., Williams B.R., Silverman R.H. (1998) Identification of genes differentially regulated by interferon alpha, beta, or gamma using oligonucleotide arrays. Proc. Natl Acad. Sci. USA 95:15623–15628.
[Abstract/Free Full Text] - Freshney R.I. (2000) Culture of Animal Cells: A Manual of Basic Technique., Wiley-Liss, New York.
- Yee C., Krishnan-Hewlett I., Baker C.C., Schlegel R., Howley P.M. (1985) Presence and expression of human papillomavirus sequences in human cervical carcinoma cell lines. Am. J. Pathol. 119:361–366.[Abstract]
- Ponchel F., Toomes C., Bransfield K., et al. (2003) Real-time PCR based on SYBR-Green I fluorescence: an alternative to the TaqMan assay for a relative quantification of gene rearrangements, gene amplifications and micro gene deletions. BMC Biotechnol. 3:18.[CrossRef][Medline]
- Moberg M., Gustavsson I., Gyllensten U. (2003) Real-time PCR-based system for simultaneous quantification of human papillomavirus types associated with high risk of cervical cancer. J. Clin. Microbiol. 41:3221–3228.
[Abstract/Free Full Text] - Peitsaro P., Johansson B., Syrjänen S. (2002) Integrated human papillomavirus type 16 is frequently found in cervical cancer precursors as demonstrated by a novel quantitative real-time PCR technique. J. Clin. Microbiol. 40:886–891.
[Abstract/Free Full Text] - Koopman G., Reutelingsperger C.P., Kuijten G.A., Keehnen R.M., Pals S.T., van Oers M.H. (1994) Annexin V for flow cytometric detection of phosphatidylserine expression on B cells undergoing apoptosis. Blood 84:1415–1420.
[Abstract/Free Full Text] - Pfaffl M.W. (2001) A new mathematical model for relative quantification in real-time RT–PCR. Nucleic Acids Res. 29:e45.
[Abstract/Free Full Text] - Vandesompele J., De Preter K., Pattyn F., Poppe B., Van Roy N., De Paepe A., Speleman F. (2002) Accurate normalization of real-time quantitative RT–PCR data by geometric averaging of multiple internal control genes. Genome Biol. 3:RESEARCH0034.1–0034.11.[CrossRef]
- Biron C.A. and Sen G.C. (2001) Interferons and Other Cytokines. In Knipe D.M. and Howley P.M (Eds.). Fields' Virology(Lippincott Williams & Wilkins, Philadelphia) pp. 321–344.
- Samuel C.E. (2001) Antiviral actions of interferons. Clin. Microbiol. Rev. 14:778–809.