STAT3 activation by E6 is essential for the differentiation-dependent HPV18 life cycle

Abstract

Human papillomaviruses (HPV) activate a number of host factors to control their differentiation-dependent life cycles. The transcription factor signal transducer and activator of transcription (STAT)-3 is important for cell cycle progression and cell survival in response to cytokines and growth factors. STAT3 requires phosphorylation on Ser727, in addition to phosphorylation on Tyr705 to be transcriptionally active. In this study, we show that STAT3 is essential for the HPV life cycle in undifferentiated and differentiated keratinocytes. Primary human keratinocytes containing high-risk HPV18 genomes display enhanced STAT3 phosphorylation compared to normal keratinocytes. Expression of the E6 oncoprotein is sufficient to induce the dual phosphorylation of STAT3 at Ser727 and Tyr705 by a mechanism requiring Janus kinases and members of the MAPK family. E6-mediated activation of STAT3 induces the transcription of STAT3 responsive genes including cyclin D1 and Bcl-xL. Silencing of STAT3 protein expression by siRNA or inhibition of STAT3 activation by small molecule inhibitors, or by expression of dominant negative STAT3 phosphorylation site mutants, results in blockade of cell cycle progression. Loss of active STAT3 impairs HPV gene expression and prevents episome maintenance in undifferentiated keratinocytes and upon differentiation, lack of active STAT3 abolishes virus genome amplification and late gene expression. Organotypic raft cultures of HPV18 containing keratinocytes expressing a phosphorylation site STAT3 mutant display a profound reduction in suprabasal hyperplasia, which correlates with a loss of cyclin B1 expression and increased differentiation. Finally, increased STAT3 expression and phosphorylation is observed in HPV positive cervical disease biopsies compared to control samples, highlighting a role for STAT3 activation in cervical carcinogenesis. In summary, our data provides evidence of a critical role for STAT3 in the HPV18 life cycle.

Fig 1. HPV18 induces the activation of STAT3 in primary human keratinocytes.

A) Representative western blots of normal human keratinocytes (NHK) and HPV18-containing keratinocytes subjected to high calcium differentiation and analysed for STAT3 phosphorylation using specific antibodies against pS727 STAT3, pY705 STAT3 and total STAT3. Involucrin expression is a marker of differentiation and E7 demonstrates the presence of HPV18 in the samples. GAPDH serves as a loading control. B) Quantification of the protein band intensities in A) standardised to GAPDH levels. Bars represent the means ± standard deviation from 4 independent biological repeats using 2 donor cell lines. *P<0.05 (Student’s t-test). Representative sections of organotypic raft cultures from C) NHK and D) HPV18-containing keratinocytes stained with antibodies specific for pS727 STAT3 and total STAT3 (green) and counterstained with DAPI to highlight the nuclei (blue in the merged panels). Images were acquired using identical exposure times. Scale bar, 40 μm. White dotted lines indicate the basal cell layer.

Fig 2. HPV E6 promotes the dual phosphorylation and activation of STAT3.

A) Representative western blot of C33a cells transfected with GFP tagged HPV18 oncoproteins E5, E6 or E7 and analysed for STAT3 activation using specific antibodies detecting phosphorylated and total STAT3. Expression of HPV oncoproteins was confirmed using a GFP antibody. GAPDH served as a loading control. B) Representative luciferase reporter assays from C33a cells co-transfected with GFP tagged HPV18 oncoproteins E5, E6 or E7 and either a β-casein promoter reporter plasmid or a Pom^C reporter plasmid, which contain STAT3 binding sites, and promoter activity measured using a dual-luciferase system. Data are presented as relative to the GFP transfected control cells. Bars represent the means ± standard deviation from at least three independent biological repeats. **P<0.01, N.S. = not significant (Student’s t-test). C) Representative luciferase reporter assays from C33a cells co-transfected with GFP tagged E6 and a β-casein promoter reporter plasmid treated with the STAT3 inhibitors cryptotanshinone and S3I-201, and the STAT5 inhibitor pimozide. Promoter activity was measured using a dual-luciferase system. Data are presented as relative to the GFP transfected control. Bars represent the means ± standard deviation from at least three independent biological repeats. **P<0.01 (Student’s t-test). D) Representative western blot of C33a cells transfected with GFP or GFP tagged HPV18 E6, untreated or treated with the STAT3 inhibitors as above and analysed for total and phosphorylated STAT3 (Y705 and S727), the expression of cyclin D1 and Bcl-xL and GFP to demonstrate expression of the GFP-18E6 fusion protein. GAPDH expression was used as a loading control. Data shown are representative of at least three biological repeats. E-H) C33a cells were transiently transfected with GFP or GFP tagged HPV18 E6 and left untreated or treated with the STAT3 inhibitors as above and RNA was extracted for qRT-PCR analysis of the indicated STAT3 dependent genes. Samples were normalized against U6 mRNA levels. Representative data are presented relative to the GFP transfected control. Bars are the means ± standard deviation from at least three biological repeats. *P<0.05, **P<0.01, ***P<0.001 (Student’s t-test).

Fig 3. STAT3 is phosphorylated in cells expressing E6 defective for E6AP binding, p53 degradation and PDZ domain binding.

A) Representative western blot of C33a cells transfected with GFP or GFP tagged HPV18 E6 wild type, E6 ΔPDZ, E6 F4V or E6 L52A and analysed using antibodies detecting total and phosphorylated STAT3 and p53. GAPDH expression was used as a loading control and GFP confirmed expression of the E6 proteins. Data presented are representative of at least three independent experiments. B) Organotypic raft sections were stained with an antibody specifically detecting STAT3 pS727 (green) and counterstained with DAPI to highlight the nuclei (blue—in merged panels). Images were acquired with identical exposure times. The dotted line indicates the basal cell layer. Scale bar, 20 μm.

Fig 4. Janus kinases (JAK) are responsible for mediating Y705 phosphorylation in E6 and HPV18 containing keratinocytes.

A) Representative western blot of keratinocytes subjected to high calcium differentiation analysed with antibodies specific for the total and phosphorylated forms of JAK2 (Y1007/1008). GAPDH served as a loading control. Data shown are representative of at least three independent biological repeats. B) Representative western blot of C33a cells transfected with GFP and GFP tagged HPV18 E6 and analysed using specific antibodies detecting phosphorylated (Y1007/1008) and total JAK2. Expression of E6 was confirmed using a GFP antibody. GAPDH served as a loading control. C) Representative western blots of HPV18 containing keratinocytes incubated with the JAK1/2 inhibitor Ruxolitinib or JAK2 inhibitor Fedratinib analyzed with antibodies detecting total and phosphorylated forms of JAK2 and STAT3. GAPDH served as a loading control. Data shown are representative of at least three independent biological repeats.

Fig 5. Identifying the protein kinases responsible for the S727 phosphorylation of STAT3 in HPV18 containing keratinocytes.

A) Representative western blot of keratinocytes subjected to high calcium differentiation analysed with antibodies specific for the total and phosphorylated forms of ERK1/2 (Thr202/Tyr204), p38 (Thr180/Tyr182) and JNK (Thr183/Tyr185). GAPDH served as a loading control. Data shown are representative of at least three independent biological repeats. B) Representative western blot of C33a cells transfected with GFP and GFP tagged HPV18 E6 and analysed using specific antibodies detecting phosphorylated and total forms of ERK1/2 (Thr202/Tyr204), p38 (Thr180/Tyr182) and JNK (Thr183/Tyr185). GAPDH served as a loading control. C) Representative western blots of HPV18 containing keratinocytes incubated with specific inhibitors of STAT3 (Crypto), MSK1 (SB-747651A; SB), p38 (VX-745; VX)), JNK (JNK-IN-8), MKK1/2 (UO126) or in combination as described in the methods and materials and examined with antibodies specific for phosphorylated and total STAT3. The phosphorylation status of substrate proteins pMSK1 (T581), pcJun (S73), pMAPKAPK2 (T334), and ERK1/2 (T202/Y204) demonstrated inhibitor efficacy and specificity. D) Quantification of the protein band intensities of (C) standardised to GAPDH and shown relative to the DMSO control. Bars represent the means ± standard deviation from at least three independent biological repeats. *P<0.05, **P<0.01, ***P<0.001 (Student’s t-test).

Fig 6. STAT3 is required for viral gene expression in undifferentiated keratinocytes.

A) Left: Representative western blots of HPV18 containing keratinocytes transfected with a panel of 4 STAT3 specific siRNA and a scrambled control and analysed by western blotting using a total STAT3 antibody and antibodies detecting the E6 and E7 HPV18 early proteins. GAPDH served as loading control. Right: Quantification of the STAT3 protein band intensities standardised to GAPDH levels. Data are expressed as fold reduced compared to scramble siRNA treated cells. Bars represent the means ± standard deviation of at least three biological repeats. **p <0.01 (Student’s t-test). B) Samples treated as in (A) and RNA extracted for qRT-PCR determination of the stat3 and HPV early gene expression levels. Samples were normalized against U6 mRNA levels and data are presented relative to the scrambled control. Bars represent means ± standard deviation of at least three biological repeats. *P<0.05 (Student’s t-test). C) Representative western blots from HPV18 containing keratinocytes treated with increasing doses of the STAT3 inhibitor cryptotanshinone and analysed with antibodies specific for phosphorylated and total STAT3, and E6 and E7 HPV18 early proteins. GAPDH served as a loading control. Data are representative of at least three biological repeats. D) RNA was extracted from HPV18 containing keratinocytes treated with 10 μM cryptotanshinone and qRT-PCR analysis performed for levels of HPV E6 and E7 mRNA. Samples were normalized against the U6 mRNA levels and expressed relative to the DMSO control. Bars represent means ± standard deviation of at least three biological repeats. *P<0.05 (Student’s t-test). E) Representative western blots from HPV18 containing keratinocytes transduced with a lentivirus encoding a dominant negative Y705F phospho-site mutant or transiently transfected with a dominant negative STAT3 S727A mutant and analysed by blotting with antibodies specific for phosphorylated and total STAT3 and the E6 and E7 HPV18 early proteins. An antibody detecting the FLAG epitope confirmed expression of the Y705F STAT3 mutant and GAPDH served as a loading control. Data shown are representative of at least three biological repeats. F) Association of STAT3 with HPV18 genomes was assessed in HPV18 containing keratinocytes by ChIP with control antibody (rabbit IgG) or STAT3-specific antibody. Co-precipitating DNA was analysed by qPCR. The x axis represents the position in the HPV genome amplified and each data point represents the central point in each amplicon. Enrichment of STAT3 binding is expressed as fold over IgG negative control calculated by the ΔΔC_T method. The data represent the mean and standard deviation of three independent experiments. A graphical representation of the HPV18 genome is shown below the data, linearized for ease of presentation (E; early promoter, L; late promoter, LCR; long control region).

Fig 7. Suppression of STAT3 impairs cell cycle progression and HPV genome maintenance in undifferentiated keratinocytes.

A) Representative western blots from HPV18 containing keratinocytes treated with increasing doses of the STAT3 inhibitor cryptotanshinone and analysed with antibodies specific for cyclin D1 and p21. GAPDH served as loading control. Data are representative of at least three biological independent repeats. B) RNA was extracted from HPV18 containing keratinocytes incubated with 10 μM cryptotanshinone and analysed by qRT-PCR for levels of ccnd1 and p21 mRNA. Samples were normalised against U6 mRNA levels and presented relative to the DMSO control. Bars represent means ± standard deviation of at least three biological repeats. **P<0.01 (Student’s t-test). C) Representative western blots of HPV18 containing keratinocytes transfected with a panel of 4 STAT3 specific siRNA and a scrambled control and analysed by western blotting using antibodies specific for cyclin D1 and p21. GAPDH served as loading control. Data are representative of at least three biological independent repeats. D) Representative western blots from HPV18 containing keratinocytes transduced with a lentivirus encoding a dominant negative Y705F phospho-site mutant or transiently transfected with a dominant negative STAT3 S727A mutant and analysed by blotting with antibodies specific for cyclin D1 and p21. GAPDH served as loading control. Data are representative of at least three biological independent repeats. Representative flow cytometric analysis of the cell cycle in E) HPV18 containing keratinocytes or F) normal human keratinocytes treated with 10 μM cryptotanshinone. G) As in E except cells treated with a pool of four STAT3 specific siRNAs. Data are expressed as percentage of cells at each stage of the cell cycle. Bars represent means ± standard deviation of at least three biological repeats. *P<0.05, ***P<0.001 (Student’s t-test) compared to DMSO or scrambled control cells. H) Representative Southern blot analysis of undifferentiated HPV18 containing keratinocytes treated with 10 μM cryptotanshinone or a pool of four STAT3 specific siRNAs.

Fig 8. STAT3 is important for cell cycle progression and HPV18 genome amplification in differentiated keratinocytes.

Representative western blots of phosphorylated and total STAT3, HPV18 E6 and E7, cyclin D1, p21, involucrin and filaggrin in A) calcium or B) methylcellulose-differentiated HPV18 containing keratinocytes in the presence or absence of 10 μM cryptotanshione. Data shown represent at least three biological repeats. C) Southern blot analysis of HPV18 episomes in keratinocytes treated with 10 μM cryptotanshinone and differentiated in methylcellulose for 120 hours. DNA was linearized with EcoRI, producing a single band running at approximately 8 kbp, demonstrating a differentiation-dependent increase in viral episome copy number in untreated control cells and a reduction in episome copy number in cells treated with cryptotanshinone. Digestion with BglII shows a lack of detectable multimeric/integrated HPV genomes in all experimental conditions. Data shown are representative of two donor cell lines. D) Signal intensity was quantified using ImageJ software. Bars represent means ± standard deviation of two biological repeats from two donor cell lines.

Fig 9. STAT3 is necessary for delayed differentiation and increased keratinocyte proliferation in a stratified epithelium.

A) Representative organotypic raft cultures of NHK and HPV18-containing keratinocytes transduced with empty lentivirus and lentivirus expressing a dominant negative phosphorylation null Y705F mutant STAT3 fixed at day 14 and stained with haematoxylin and eosin (H&E) to assess morphology. Sections were also stained with antibodies specific for B) cyclin B1 C) involucrin and D) E1^E4. Nuclei are visualised with DAPI (blue) and white dotted lines indicate the basal cell layer. E) The number of cells positive for cyclin B1 and E1^E4 in the empty vector and Y705F STAT3 transduced HPV18-containing keratinocyte sections was counted in 5 fields of vision from sections of three independent raft cultures from two donor lines. Bars represent means ± standard deviation. ***P<0.001 (Student’s t-test) compared to empty vector control.

Fig 10. STAT3 expression and phosphorylation is increased in HPV-associated cervical disease.

A) Representative immunofluorescence analysis of sections from organotypic raft cultures of NHK and a W12 cell line presenting with HSIL morphology detecting pS727 STAT3 levels (green). Nuclei were visualized with DAPI (blue) and the white dotted line indicates the basal layer. Images were acquired with identical exposure times. B) Representative western blots from cytology samples of CIN lesions of increasing grade analysed with antibodies specific for phosphorylated (Y705 and S727) and total STAT3 levels. GAPDH served as a loading control. C-E) Scatter dot plot of densitometry analysis of a panel of cytology samples. 20 samples from each clinical grade (neg, CIN I-III) were analysed by western blot and densitometry analysis was performed using ImageJ. Phosphorylated STAT3 levels were first normalised against total STAT3 levels before normalising against protein levels using GAPDH as a loading control. F) Representative immunofluorescence analysis of tissue sections from cervical lesions of increasing CIN grade. Sections were stained for pS727 STAT3 levels (green) and nuclei were visualized with DAPI (blue). Images were acquired with identical exposure times. Scale bar, 20 μm.