Sensitivity to inhibition of DNA repair by Olaparib in novel oropharyngeal cancer cell lines infected with Human Papillomavirus

Abstract

The incidence of Human Papillomavirus (HPV)-associated oropharyngeal squamous cell carcinoma (OPSCC) is increasing rapidly in the UK. Patients with HPV-positive OPSCC generally show superior clinical responses relative to HPV-negative patients. We hypothesised that these superior responses could be associated with defective repair of DNA double strand breaks (DSB). The study aimed to determine whether defective DNA repair could be associated with sensitivity to inhibition of DNA repair using the PARP inhibitor Olaparib. Sensitivity to Olaparib, and induction and repair of DNA damage, were assessed in a panel of 8 OPSCC cell-lines, including 2 novel HPV-positive lines. Effects on cell cycle distribution and levels of PARP1 and p53 were quantified. RNA-sequencing was used to assess differences in activity of DNA repair pathways. Two HPV-positive OPSCC lines were sensitive to Olaparib at potentially therapeutic doses (0.1-0.5 μM). Two HPV-negative lines were sensitive at an intermediate dose. Four other lines, derived from HPV-positive and HPV-negative tumours, were resistant to PARP inhibition. Only one cell-line, UPCISCC90, showed results consistent with the original hypothesis i.e. that in HPV-positive cells, treatment with Olaparib would cause accumulation of DSB, resulting in cell cycle arrest. There was no evidence that HPV-positive tumours exhibit defective repair of DSB. However, the data suggest that a subset of OPSCC may be susceptible to PARP-inhibitor based therapy.

Fig 1. Quantification of HPV transcription.

mRNA levels of HPV encoded genes were quantified in HPV-positive OPSCC cell lines. Reads Per Kilobase of transcript per million Mapped reads (RPKM) are plotted for the individual HPV open reading frames (ORF) of the four HPV-positive cell lines.

Fig 2. Determination of HPV integration state.

Circos plots were used to illustrate fusion transcripts between HPV16 and human genome. HPV16 sequences and human chromosomes are not to scale. Fusion transcripts are represented as grey lines indicating where read ends mapped. Thicker/darker lines indicate multiple reads detected.

Fig 3. Gene expression in HPV positive and negative cell lines.

The heatmap represents differential gene expression between HPV-positive and negative cell lines. Gene expression was normalised across the dataset: black represents the median, green represents expression below the median and red above median. After correction for multiple testing, 211 transcripts were significantly differentially expressed between the two groups (Benjamini–Hochberg FDR p<0.05).

Fig 4. mRNA levels of p53, p21, p16 and PARP1 in HPV positive and negative cell lines.

The number of reads, expressed as RPKM, for p53, p21, p16 and PARP1 was determined. Data are depicted as histograms with tabulated values showing fold change between the HPV-positive and negative groups, and significance by T-test with and without fdr adjustment for multiple tests.

Fig 5. PARP1 protein in HPV-positive and negative cell lines.

Protein was extracted from untreated cells and probed for PARP1 and β-actin (loading control). Image analysis was used to estimate relative levels of PARP1, and showed expression at >3-fold above the group median in UPCISCC90, and at >2-fold above the median in UMSCC19.

Fig 6. Surviving fractions of OPSCC cell lines in response to Olaparib treatment.

Panel A: HPV-positive cells. At all doses between 0.1 & 10 μM Olaparib, PCOC2 and UPCISCC90 showed significantly lower surviving fractions compared to PCOC3 and UMSCC47 (p<0.05; two way ANOVA with Bonferroni post-test). Panel B: HPV-negative cells. At doses of 0.5 & 1 μM Olaparib, HEKn and UMSCC74a showed significantly lower surviving fractions compared to UMSCC19 and UMSCC4 (p<0.05; two way ANOVA with Bonferroni post-test). The results are representative of at least 3 experiments, all performed in triplicate. Mean and standard deviation are shown.

Fig 7. Induction of γ-H2AX following Olaparib treatment.

Induction of γ-H2AX foci was assessed 24 & 48 hrs after treatment with 10 μM Olaparib. γ-H2AX levels were quantified using flow cytometry and are depicted as Mean Fluorescent Intensity (MFI) normalised to vehicle for each cell line. Data are representative of at least 3 independent experiments (ANOVA and Tukey post test, * P < 0.05, ** P<0.01, *** P<0.001). Error bars indicate standard deviation. Baseline levels of γ-H2AX in untreated cells are shown in S6 Fig.

Fig 8. Cell cycle distribution following Olaparib treatment.

The proportion of cells in G1, S and G2 for HPV-positive and negative cell lines 24 & 48 hrs post treatment with vehicle and 10 μM Olaparib are shown. Data are representative of at least 3 independent experiments (ANOVA and Tukey post test, * P < 0.05, ** P<0.01, *** P<0.001). Error bars indicate standard deviation.

Fig 9. Protein levels of PARP1, cleaved PARP1, p53 and phospho-p53 48hrs following Olaparib treatment.

Results for untreated and Olaparib dosed cells are shown. Cell lines treated with 10Gy ionising radiation, collected 24hrs post irradiation, were also included. Two membranes were used: one for PARP1 and cleaved PARP1 and β-actin 1; and the other for p53 and β-Actin 2.