Transcriptional analysis of metastatic uveal melanoma survival nominates NRP1 as a therapeutic target

Abstract

Uveal melanoma is a rare form of melanoma with particularly poor outcomes in the metastatic setting. In contrast with cutaneous melanoma, uveal melanoma lacks BRAF mutations and demonstrates very low response rates to immune-checkpoint blockade. Our objectives were to study the transcriptomics of metastatic uveal melanoma with the intent of assessing gene pathways and potential molecular characteristics that might be nominated for further exploration as therapeutic targets. We initially analyzed transcriptional data from The Cancer Genome Atlas suggesting PI3K/mTOR and glycolysis as well as IL6 associating with poor survival. From tumor samples collected in a prospective phase II trial (A091201), we performed a transcriptional analysis of human metastatic uveal melanoma observing a novel role for epithelial-mesenchymal transition associating with survival. Specifically, we nominate and describe initial functional validation of neuropillin-1 from uveal melanoma cells as associated with poor survival and as a mediator of proliferation and migration for uveal melanoma in vitro. These results immediately nominate potential next steps in clinical research for patients with metastatic uveal melanoma.

Trial registration: ClinicalTrials.gov NCT01835145 NCT03565445.

Conflict of interest statement

J.J.L. declares Scientific Advisory Board: (no stock) 7 Hills, Spring bank (stock) Actym, Alphamab Oncology, Arch Oncology, Kanaph, Mavu, Onc.AI, Pyxis, Tempest. Consultancy with compensation: Abbvie, Aligos, Array, Bayer, Bristol-Myers Squibb, Checkmate, Cstone, Eisai, EMD Serono, KSQ, Janssen, Merck, Mersana, Novartis, Partner, Pfizer, RefleXion, Regeneron, Ribon, Rubius, Silicon, Tesaro, Werewolf, Xilio, Xencor. Research Support: (all to institution for clinical trials unless noted) AbbVie, Agios (IIT), Array (IIT), Astellas, Bristol-Myers Squibb (IIT & industry), Corvus, EMD Serono, Immatics, Incyte, Kadmon, Macrogenics, Merck, Spring bank, Tizona, Xencor. Travel: Bristol-Myers Squibb, Janssen, Mersana, Pyxis. Patents: (both provisional) Serial #15/612,657 (Cancer Immunotherapy), PCT/US18/36052 (Microbiome Biomarkers for Anti-PD-1/PD-L1 Responsiveness: Diagnostic, Prognostic and Therapeutic Uses Thereof). RB: None declared, Patents: (all provisional) PCT/US15/612657 (Cancer Immunotherapy), PCT/US18/36052 (Microbiome Biomarkers for Anti-PD-1/PD-L1 Responsiveness: Diagnostic, Prognostic and Therapeutic Uses Thereof), PCT/US63/055227 (Methods and Compositions for Treating Autoimmune and Allergic Disorders). For the remaining authors, there are no conflicts of interest.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc.

Figures

Fig. 1

Transcriptional programs associated with overall survival in primary uveal melanoma (pUM) from The Cancer Genome Atlas (TCGA). (a) Kaplan–Meier survival curves of MTORC1 signaling and IL6/JAK/STAT3 signaling gene expression in pUM, split by median expression of each signature (high vs. low). Survival risk table is shown below the Kaplan–Meier plot in each panel. (b) Forest plots showing the hazard ratio and P-values in Cox proportional hazards (PHs) multivariable model of the signaling pathways with demographic and clinical covariates. n = 80 patients in the TCGA primary UM cohort were shown for (a) and (b). Log-rank test was used in (a), and Cox PH multivariable model was used in (b).

Fig. 2

Epithelial-mesenchymal transition (EMT) signature is significantly associated with overall survival (OS) in patients with metastatic uveal melanoma. (a) Expression of the EMT gene signature in patient survival groups split by 1-year OS; n = 14 in patients with OS ≤ 1 year, n = 5 in patients with OS > 1 year. (b) Kaplan–Meier survival analysis of OS in patients with tumors EMThigh and EMTlow, split by median expression of the EMT gene signature. Nineteen patients were shown in (a) and (b), split by two different metrics [1-year OS in (a), and median EMT expression in (b)]. Survival risk table is shown below the plot. Two-sided Student’s t-test was used in (a). Cox proportional hazards univariable model was used in (b).

Fig. 3

Differentially expressed genes of metastatic uveal melanoma in patient survival groups split by 1-year overall survival (OS). (a) Expression heatmap of 76 differentially expressed gens (DEGs) comparing tumors from patients who lived less than 1 year to those who lived longer. Genes were filtered by P < 0.005 (unadjusted) and fold change ≥2.0 or ≤–2.0. Samples were clustered on the column with dendrogram shown above the heatmap. Annotation bar labels patient groups with OS ≤ or >1 year. Genes are shown in the boxes to the right of the heatmap, following the same order as the gene dendrogram to the left side of the heatmap. (b) Expression of 22 DEGs upregulated in tumors from patients with OS ≤ 1 year relative to those with OS > 1 year. FC = expression fold of change calculated by comparing patients with OS >1 year to patients with OS ≤1 year. 1 yr = one year. The limma voom regression model with precision weights was used in (a) and (b).

Fig. 4

Knockdown of neuropillin-1 (NRP1) gene expression in vitro induces G1 arrest and inhibits cell proliferation and invasion. (a) NRP1 mRNA expression in uveal melanoma cell lines Mel290, Mel285, OMM1.3, and OCM3. n = 2 in each of the four cell lines, quantitative PCR was performed in triplicates, experiment was repeated two times. (b) Western blots of Mel290, Mel285, OMM1.3, and OCM3 cells transfected with NRP1 small interfering RNA (siRNA) shows inhibition of NRP1 expression and induction of p27Kip1 expression. (c) Cell cycle analysis of uveal melanoma cells transfected with NRP1 siRNA for 72 h shows a significant increase in G1 population in Mel290 (P < 0.0001) and Mel285 (P = 0.035) cells but not OMM1.3 and OCM3 cells; n = 3 in each of the four cell lines, with two groups each, experiment was repeated three times. (d and e) The siRNA knockdown of NRP1 expression significantly inhibits cell viability (d) after 72 h (Mel290 P = 0.009, Mel285 P = 0.003, OMM1.3 P = 0.18, OCM3 P < 0.001) and cell invasion (e) after 24 h of uveal melanoma cells that highly express NRP1 (Mel290 and Mel285). In (d), n = 3 in each of the four cell lines, with two groups each, experiment was repeated three times. (f) Quantitation of migrated cells shows the selective inhibition of invasion by NRP1 siRNA knockdown in Mel290 (P = 0.046) and Mel285 (P = 0.007) cells but not OMM1.3 (P = 0.40) and OCM3 (P = 0.42) cells. n = 3 in each of the four cell lines, with two groups each, experiment was repeated three times. In (c), (d), and (f), each bar is shown as mean ± S.E.M, with standard error calculated as the SD divided by the square root of the number of samples. Two-sided Student’s t-test was used in (c), (d), and (f); ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05.