Elucidating cellular response to treatment with viral immunotherapies in pediatric high-grade glioma and medulloblastoma

Abstract

HSV G207, a double-stranded, DNA virus, and the polio:rhinovirus chimera, PVSRIPO, a single positive-strand RNA virus, are viral immunotherapies being used to treat pediatric malignant brain tumors in clinical trials. The purpose of this work is to elucidate general response patterns and putative biomarkers of response. Multiple pediatric high-grade glioma and medulloblastoma cell lines were treated with various multiplicities of infection of G207 or PVSRIPO. There was a significant inverse correlation between expression of one HSV cellular receptor, CD111, and the lethal dose of 50% of cells (LD50) of cells treated with G207 (r = -0.985, P<0.001) but no correlation between PVSRIPO cellular receptor expression (CD155) and LD50. RNA sequencing of control cells and cells treated for 8 and 24 h revealed that there were few shared differentially expressed (DE) genes between cells treated with PVSRIPO and G207: GCLM, LANCL2, and RBM3 were enriched whilst ADAMTS1 and VEGFA were depleted. Likewise, there were few shared DE genes enriched between medulloblastoma and high-grade glioma cell lines treated with G207: GPSM2, CHECK2, SEPTIN2, EIF4G2, GCLM, GDAP1, LANCL2, and PWP1. Treatment with G207 and PVSRIPO appear to cause disparate gene enrichment and depletion suggesting disparate molecular mechanisms in malignant pediatric brain tumors.

Keywords: Cell response; G207; Medulloblastoma; Oncolytic virus; PVSRIPO; Pediatric high-grade glioma; RNA sequencing; Viral immunotherapy.

Fig. 1

Dose response curves of pHGG and medulloblastoma cell lines to viral immunotherapy. (A) Cell viability after treatment with PVSRIPO for 24 h. Note the separation of cell line viability at MOI 1; the cell lines D324, 8A, D425, D283 grouped together as more “resistant”. (B) Cell viability after treatment of cells with G207 for 72 h. There was no clear grouping (“resistant” vs “sensitive”) viability at MOI 1. (C) PCA plot of mRNA sequencing data of untreated medulloblastoma and pHGG cells demonstrates clustering of pHGG cells (D456, X21584). D324 or DAOY cells were distinct from other pHGG and medulloblastoma cells. (D) Unsupervised hierarchical clustering of D283, D324, D341, D556, D456, and X21584 cells using the geneset GOBP_DEFENSE_RESPONSE_TO_VIRUS, demonstrates grouping of D556/D341 cells and D283/D324 cells, similar to the grouping of the dose response curve at MOI 1 (A). MOI: multiplicity of infection.

Fig. 2

Cell surface receptor expression is correlated with cytotoxicity for CD111 but not CD155. (A) Dot plot of Nectin-1 (CD111) receptor expression and Poliovirus receptor (CD155) expression by cell type as percentage of cells expressing receptor. (B) Scatter plot of flow cytometry data demonstrating Lack CD111 expression in 8A cells with robust CD155 expression vs. D556 cells demonstrating both robust CD111 and CD155 expression. (C) There is a significant inverse correlation between CD111 expression and LD50 from G207 treatment not seen in CD155 expression and LD50 from PVSRIPO treatment (D).

Fig. 3

The most common enriched and depleted hallmark pathways. Bar graphs of hallmark gene sets with at least eight significantly (P ≤ 0.05) depleted (black) or enriched (gray) treatment comparisons. Treatment comparisons are on the y axis, NES on the x axis. NES: normalized enrichment score.

Fig. 4

Medulloblastoma cells response differently to treatment with G207 and PVSRIPO. (A) Venn diagram demonstrating minimal overlap in G207 and PVSRIPO of significantly differentially expressed genes. PCA plots of medulloblastoma cells treated with G207 (B), PVSRIPO (C), and the two graphs combined (D). (E) Volcano plot of differentially expressed genes of cells treated with PVSRIPO vs. G207 demonstrating the interaction effect between time point and treatment, controlling for both.

Fig. 5

Genetic activation is preferentially increased in “sensitive” medulloblastoma cell lines. (A) Volcano plot of untreated “resistant” medulloblastoma cells (D283 and D324) vs. “sensitive” (D341 and D556) cells demonstrates many more significantly more enriched differentially expressed cells in the sensitive group. (B) Line graphs demonstrating all significantly depleted genes in sensitive vs. resistance medulloblastoma cells. There were significantly more gene depletions in sensitive cells. (C) Line graphs demonstrating all significantly enriched genes in sensitive vs. resistant medulloblastoma cells with a trend toward more genes in the sensitive group.

Fig. 6

Volcano plot of top 50 differentially expressed genes in < and ≥ 1 year survival glioblastoma patient samples. Genes in blue are depleted in patients living < 1 year (N = 28). Genes in red are enriched in patients living ≥ 1 year (N = 29).