Transcriptomic, Proteomic, and Genomic Mutational Fraction Differences Based on HPV Status Observed in Patient-Derived Xenograft Models of Penile Squamous Cell Carcinoma

Abstract

Metastatic penile squamous cell carcinoma (PSCC) has only a 50% response rate to first-line combination chemotherapies and there are currently no targeted-therapy approaches. Therefore, we have an urgent need in advanced-PSCC treatment to find novel therapies. Approximately half of all PSCC cases are positive for high-risk human papillomavirus (HR-HPV). Our objective was to generate HPV-positive (HPV+) and HPV-negative (HPV-) patient-derived xenograft (PDX) models and to determine the biological differences between HPV+ and HPV- disease. We generated four HPV+ and three HPV- PSCC PDX animal models by directly implanting resected patient tumor tissue into immunocompromised mice. PDX tumor tissue was found to be similar to patient tumor tissue (donor tissue) by histology and short tandem repeat fingerprinting. DNA mutations were mostly preserved in PDX tissues and similar APOBEC (apolipoprotein B mRNA editing catalytic polypeptide) mutational fractions in donor tissue and PDX tissues were noted. A higher APOBEC mutational fraction was found in HPV+ versus HPV- PDX tissues (p = 0.044), and significant transcriptomic and proteomic expression differences based on HPV status included p16 (CDKN2A), RRM2, and CDC25C. These models will allow for the direct testing of targeted therapies in PSCC and determine their response in correlation to HPV status.

Keywords: APOBEC mutations; human papillomavirus-positive penile squamous cell carcinoma; patient-derived xenograft; penile squamous cell carcinoma.

Figure 1

H&E and p16 IHC images for original patient tissue (donor) and passage 2 of the PDX model (P2).

Figure 2

Differentially expressed proteins based on HPV status. (a) Normalized linear expression of RRPA data of CDKN2A, RRM2, and CDC25C depicted in a bar blot with false discovery rate (FDR) adjusted p-value given for each protein per data set. (b) Immunoblot of lysates for HPV+ (red bars) and HPV− (black bars) PDX models for RRM2, HPV16-E7, and GAPDH depicted. (c) Immunoblot of CDC25C and GAPDH. (d) Immunoblot of p16 and GAPDH. The uncropped blots are shown in File S1.

Figure 3

RNA-Seq Data Analysis. (a) MCP-counter deconvolution revealed differences in the population of microenvironment cells such as B cells, cancer-associated fibroblast, monocytes, macrophages, and activated CD8+ T cells in patient tumor tissues labeled Pe9, Pe10, Pe13, Pe16, Pe18, Pe20, and Pe25 compared to PDX tissues labeled XPe3, XPe9, XPe10, XPe13, XPe16, and XPe821. (b) Box plots of different cell populations found to be highly expressed in patient tumor samples compared to PDX tissues. p-values were obtained from t-test with FDR adjustment.

Figure 4

(a) Mutations in tier 1 genes from the cancer gene census database. (b) Somatic variants found only in the PDX tissue (yellow), only in the patient tumor tissue (blue), or shared between the patient tumor tissue and the PDX tissue (red).

Figure 5

(a) APOBEC mutation (C-to-T and C-to-G mutation in TCW context) fraction percentage in original patient tumor (x-axis) and PDX tissue (y-axis) plotted. Pink circles are HPV+ tissues while blue circles are HPV− tissues. (b) The APOBEC mutation fraction in the PDX tissues alone are shown in the box and whiskers plot.