Transcriptional Patterns of High-Risk Human Papillomavirus Types 16, 18, 45, 68b Genes

MATERIALS AND METHODS

Study area: The study was carried out at Department of Microbiology, Faculty of Medicine, Chulalongkorn University, Thailand and Department of Computational Biology and Medical Sciences, Graduate School of Frontier Sciences, The University of Tokyo, Japan. This research project was conducted from April, 2017-September, 2018.

Cervical cancer cell lines: Five cervical cancer cell lines containing different HR-HPV types were used in the present study including CaSki (containing approximately 600 copies of integrated HPV16 per cell, (CRL-1550 Lot No. 3794357)), SiHa (containing approximately 1-2 copies of integrated HPV16 per cell (HTB-35 Lot No. 4031219)), HeLa (containing integrated HPV18), MS751 (containing integrated HPV45 (HTB-34 Lot No. 58078726)) and ME180 (containing integrated HPV68b (HTB-33 Lot 57758662)). All cell lines were maintained in Dulbecco's Modified Eagle's Medium (DMEM) (HyClone, South Logan, Utah, USA) with 10% fetal bovine serum in a CO₂ incubator at 37°C. This study has been approved (COA No. 677/2015) by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University.

RNA preparation for RNA sequencing: Total RNA was extracted from cervical cancer cell lines using an RNeasy^® Mini Kit (QIAGEN, Valencia, CA, USA) according to the manufacturer's instructions. Approximately 10,000,000 cells were used. The RNA was eluted twice in a total volume of 100 μL of RNase-free H₂O. RNA was quantified by Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA) and samples with RIN value >8.0 were used for RNAseq analysis. Illumina TrueSeq Stranded mRNA Sample Preparation Kit was used. Samples were sequenced on an Illumina HiSeq2500 (Illumina, Inc., San Diego, CA, USA) using a paired-end run.

Data analysis: The Hisat2 v2.1.0 aligner was used to map the raw data reads onto the viral reference sequences as follows, HPV16 (NC_001526.2), HPV18 (NC_001357.1), HPV68b (FR751039.1) and HPV45 (X74479.1).

Following the alignment, the viral reads were assembled into transcripts using Cufflinks v. 2.2.1. The transcripts were merged into full spliced transcripts (Cuffmerge) and quantified viral isoforms (Cuffdiff) as Fragments per Kilobase per Million (FPKM). The analyzed data were visualized, using the Integrated Genomic Viewer (IGV) to identify the expressed transcripts and splicing junction of each HPV type.

Validation of HPV transcripts by real-time polymerase chain reaction (real time-PCR): Primers were designed to detect the viral transcripts in five cell lines by real-time RT-PCR. cDNA was synthesized using Superscript III Reverse Transcriptase (Invitrogen, Carlsbad, CA, USA). Real-time PCR was performed using SsoAdvanced Universal SYBR Green Supermix (Bio-Rad, Hercules, CA, USA). Primer sequences were shown in Table 1. The PCR conditions were: initial denaturing at 98°C for 30 sec, followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min. GAPDH was used as a housekeeping gene and delta-delta Ct was calculated to investigate differences in gene expression.

RESULTS

Analysis of expressed transcripts: The reads were mapped onto the viral reference sequences to identify the viral transcripts that were expressed in each cervical cancer cell lines (CaSki (HPV16), SiHa (HPV16), HeLa (HPV18), MS751 (HPV45) and ME180 (HPV68b)). Transcripts containing truncated E6 and full-length E7 genes were highly detected in all four HR-HPV types 16, 18, 45 and 68b. Full-length E6, full-length E1 and truncated E1 transcripts were also detected but in less amount (Fig. 1a-e). CUFF.1.1 (CaSki, Fig. 1a), CUFF.1.2 (SiHa, Fig. 1b), CUFF.1.1 (HeLa, Fig. 1c), CUFF.3.1 (MS751, Fig. 1d) and CUFF.3.2 (ME180, Fig. 1e) were highly expressed as shown by the coverage and FPKM of transcripts. The FPKM values of viral transcripts were shown in supplementary data S1. Transcript containing full-length E2 and truncated E2 ORF were found in HPV16 positive CaSki cell line (Fig. 1a). The amplification of papillomavirus oncogene transcripts (APOTs) and sanger sequencing assays were performed and revealed that the HPV16 E2/E4 gene was retained in the CaSki cell line (supplementary data S2).

The splicing junctions of all four HPV types were found within E6 and E1 regions. For the E6 region, one Splicing Donor (SD) at the 5 end and different Splicing Acceptor (SA) positions at the 3 end were found as follow; three splicing junctions were found in HPV16 positive cervical cancer cell lines, CaSki and SiHa, SD226^SA409(E6*I), SD226^SA526(E6*II) and SD226^SA742 (E6*X). Two splicing junctions were found in HPV18 (HeLa), SD233^SA416 (E6*I), SD233^SA635, HPV45 (MS751); SD230^SA412 (E6*I), SD230^SA640 and HPV68b (ME180); SD129^SA311 (E6*I), SD129^SA406.

Splicing junctions within E1 region found in CaSki and SiHa were SD880^SA3358, SD880^SA3361, SD880^SA3391, SD880^SA1726, SD880^SA2405, SD880^SA2582(E1C), SD880^SA2709 (E2*), SD880^SA3020, SD880^SA3078, SD880^SA3329, SD577^SA6810, SD898^SA1725, SD1302^SA2709, SD1302^SA3358 (E2C), SD1760^SA3391, SD1263^SA3391, SD2309^SA3461 and the other forms were SD96^SA1063, SD226^ SA2709 (E6*IV), SD226^SA3329, SD226^SA3358 (E6*III), SD226^SA3361, SD226^SA3391 and SD579^SA6809.

Fig. 1(a-e):

Read depth maps and splicing events visualized by IGV of four HR-HPV types 16, 18, 45 and 68b transcripts obtained from five different cervical cancer cell lines (a) CaSki (HPV16), (b) SiHa (HPV16), (c) HeLa (HPV18), (d) MS751 (HPV45) and (e) ME180 (HPV68b). The fragments per kilobase per million (FPKM) of CUFF and other viral transcripts were shown in supplementary S1 excel

Fig. S2:

HPV16 sequence transcript (S2A) obtained from CaSki cell line detected by APOT assay and direct sequencing was presented by BLAST analysis (S2B and S2C) and proposed transcript (S2D)

Splicing junctions within E1 region in HPV18 (HeLa) were SD929^SA2779, SD977^SA1836, SD1342^SA1436, SD1987^SA2047 and one splicing event within E7 region, SD599^SA619. HPV68b (ME180) were SD839^SA2586, SD683^SA2586, SD839^SA2586 and no E1 splicing junctions were found in HPV45 (MS751). Transcripts found in four HR-HPV types 16, 18, 45 and 68b in five cervical cancer cell lines were expressed as biscistronic and polycistronic as shown in Fig. 2a-e.

Major transcripts of HR-HPV types 16, 18, 45 and 68b: According to coverage and FPKM values visualized by IGV software, major transcripts found in CaSki (HPV16), SiHa (HPV16), HeLa (HPV18), MS751 (HPV45) and ME180 (HPV68b) were SD226^SA409-SD880^SA3358 (Fig. 2a), SD226^SA409-SD880^SA2709 (Fig. 2b), SD233^SA416 (Fig. 2c), SD230^SA412 (Fig. 2d) and SD129^SA331-SD839^SA2586 (Fig. 2e), respectively. All major transcripts of four HR-HPVs contain disruption within the E6 gene.

Validation of HPV oncogene expressions by real-time PCR: HPV Transcripts Were Validated To Confirm the RNAseq results, primers specific for a full length and spliced HPV transcript types were designed as shown in Table 1. It has been shown that full-length E7 transcript of all 4 HR-HPV types was highly expressed. A spliced transcript was the second most expressed, however, full-length E6 transcript was less expressed (Fig. 3).

Fig. 2(a-e):

Map of HPV16, 18, 45 and 68b transcripts detected in five cervical cancer cell lines (a) CaSki (HPV16), (b) SiHa (HPV16), (c) HeLa (HPV18), (d) MS751 (HPV45) and (e) ME180 (HPV68b)

Fig. 3:

Validation of HPV16, 18, 45 and 68b oncogene expressions (full-length E7 (FLE7), spliced E6 (E6*1) and FLE6) detected in five cell lines by real-time PCR

DISCUSSION

Four HR-HPV types 16, 18, 45 and 68 mRNA obtained from five different cervical cancer cell lines were used for RNAseq analysis to identify major transcripts necessary for the expression of viral oncogenes. Expression of viral oncogenes E6 and E7 are detected as bicistronic and polycistronic, however, it has been reported that a bicistronic transcript encoding the full-length E6 and E7 ORFs translates only E6 protein^14,15. The mechanism is that the spacing between the E6 ORF and E7 ORF within this full-length transcript is only two nucleotides, thus, the scanning ribosome does not have enough time to discard its releasing factors and to efficient reinitiation of E7 protein translation. To efficiently translate the E7 gene, a monocistronic transcript containing the full-length E7 ORF transcribed from a promoter located within the E6 region should be expressed^17,18. Deep RNA sequencing was performed to identify and quantitate the major transcripts that were expressed in five cervical cancer cell lines containing four different HR-HPV types included HPV16, 18, 45 and 68b. The RNAseq results of four HR-HPV types in the present study revealed less mRNA transcripts encoding the full-length E6 and monocistronic full-length E7 transcripts (Fig. 1-2). Thus, these two transcript types were not required for increased E6 and E7 protein expression. The major transcripts that were detected in all four HR-HPV types were transcripts with disruption of the E6 intronic region where a splicing event occurred that consistent with previously published papers that E7 oncoprotein was efficiently expressed from mRNA with splicing event occurs within E6 gene^14,19,20,23. The explanation is that splicing of E6 intron creates a frameshift, resulting in the spliced mRNA which has enough nucleotide space for the translation termination of truncated E6 protein and re-initiation of full-length E7 translation¹⁴. However, it has been reported that the distance between the splicing acceptor at the E6 region and the start codon of the E7 gene must be optimal to efficiently translate E7 protein^15,23. Quantitative analysis as shown by coverage and FPKM values of major transcripts revealed that transcript types 226^409 (HPV16), 233^416 (HPV18), 230^412 (HPV45) and 129^311(HPV68b) with full-length E7 ORF were mostly expressed (Fig. 2a-e) and these transcripts were named E6*I¹⁴. The HPV16, for example, the distance between the SA409 to E7 start codon at nucleotide position 562 is 153 bp that is the optimal space for strongly translate E7 gene when compared to SA526 to E7 start codon 562 (distance 36 bp). The transcript with splicing acceptor within the E7 gene and unspliced within the E1 gene were proposed to encode for E1 protein^24-26. It has been reported that splicing event within the E6 gene was found only in HR-HPVs but not Low risk HPVs²². The level of E6*I transcript found in CaSki and SiHa were not largely different as shown by RNAseq (FPKM 664,626 and 510,559 in CaSki and SiHa, respectively) and real-time PCR (Fig. 3), although there is the difference in copy number of integrated HPV16 within each cell line, that may be due to high methylation (>90%) and closed chromatin were found in the Long Control Region (LCR) of HPV16 in CaSki cell lines (600 copies per cell), however, no methylation was found in HPV16 LCR in SiHa cell line (1-2 copies per cell)^27-29.

It has been shown by APOTs assay that the HPV16 E2/E4 gene was retained in the CaSki cell line and splicing acceptor was also found in the E2/E4 gene at position 3358 (supplementary data S2). In accordance with RNAseq results, in which the major transcript of HPV16 found in the CaSki cell line was SD226/SA409-SD880/SA3358. The importance of SA3358 for HPV16 E6 and E7 translation was reported³⁰. When comparing HPV16 transcripts found in cervical cancer cell lines, CaSki and SiHa in the present study with previous report transcripts of the W12 cell line (W12 cell lines derived from LSIL cells), it was found that transcript encoding the full-length E6 and E7 ORFs with splicing donor at 880 was found only in cervical cancer cell lines^7,12. This transcript type was previously reported to be expressed only in CIN3 and Squamous Cell Carcinoma (SCC)¹⁹. Transcript type E6*I was mostly detected in cervical cancer^9,19,22. In the present study, HPV16 oncogene transcripts E6*I containing truncated E6 and full-length E7 were present in high levels compared to E6*II and E6*X in both CaSki and SiHa cervical cancer cell lines. E6*I and E7 transcripts were also highly expressed in HR-HPV types 18, 45 and 68b. One study showed that E7 oncoprotein of HPV16 and HPV18 was highly expressed in CaSki and HeLa, respectively, that correlated with the major transcript for E7 gene expression (E6*I) detected in the present study³¹.

The cervical cancer cell line MS751 revealed transcripts of truncated E6, E7 and part of E1 ORFs that consistent with the previous report on the characterization of HPV45 DNA in MS751 cell line in which E6, E7 and part of E1 ORFs were retained after viral integration into host chromosome³². Apart from E6*I, E7 and E1 transcripts, both HPV18 and 45 (Both HPVs are in the same group, Alpha 7 papillomavirus) expressed non-coding region, however, their significance is not well understood.

In summary, full-length HPV E7 ORF was found in most expressed transcripts, followed by truncated E6 (E6*1). Full-length E6 ORF was less detected. E2 transcript was also found in CaSki cell lines. HR-HPV16, 18, 45, 68b expressed common major transcripts with splicing donor and acceptor sites that occurred within the E6 gene (E6*I) that facilitate E7 oncogene translation. Thus, major transcripts encoded for truncated E6*I isoform and E7 proteins expressed from HR-HPV types may be involved in carcinogenesis³³ could be used as biomarkers for diagnostics and prognostics of HPV related diseases. The role of truncated proteins that are transcribed from various HR-HPV transcripts in carcinogenesis is limited, further functional experiments of mentioned proteins is of interest. The transcripts might be used as the target for cervical cancer therapy which may help to improve the treatment. Although, the present study did not performed RNAseq in clinical samples due to very low amount of RNA obtained from cervical scrapes, the data presented in the present study would be useful for functional analysis of the role of truncated HR-HPV proteins in cervical cancer cell lines.

CONCLUSION

RNA sequencing analysis of four HR-HPV types including 16, 18, 45 and 68b obtained from five cervical cancer cell lines revealed various transcript types, however, common major transcripts with splicing donor and acceptor sites that occurred within the E6 gene (E6*I) were found in all four HR-HPV types. The major transcript could be used as biomarkers for diagnostics and prognostics of HPV related diseases. The important of truncated E6 protein (E6*I) in carcinogenesis need to be further investigated.

SIGNIFICANCE STATEMENT

This study discovers the common major transcripts as shown in Fig. 1a-e named as CUFF that were highly expressed from HR-HPV types 16, 18, 45 and 68b as presented by FPKM values (supplementary data S1). The transcript data can be beneficial for researchers to design primers for cloning of various viral genes especially truncated forms such as E6*I, E6*II for further study the role and the involvement of these viral proteins expressed from different HR-HPV types in carcinogenesis and viral-host protein interactions. Viral gene knock down experiments using cervical cancer cell lines as a model could be applied. This study will help the researcher to uncover the critical role of truncated HR-HPV E6 proteins that limited data were reported so far. Thus, a new molecular pathway and the specific role of truncated E6 proteins driven carcinogenesis may be arrived at.

ACKNOWLEDGMENTS

This work was supported by Chulalongkorn University Office of International Affairs Scholarship for Short-term Research, Ratchadapiseksompoch Fund, Faculty of Medicine, Chulalongkorn University, grant number RA59/013 and was partly supported by the TRF Grant for a new researcher from Thailand Research Fund Grant (Grant number TRG6080008). We thank Vanida Mungmee for her kindly help in cell culture preparation, Megumi Kombu for her help with the RNAseq library construction and Yuuta Kuze for his help with the data analyses. We thank the Suzuki lab at the University of Tokyo for providing facilities necessary for the RNAseq experiment.