Palbociclib and dsRNA sensor co-operate to enhance anti-cancer effects through ER stress and modulation of immune evasion

Abstract

Cytoplasmic pattern recognition receptors (PRR) for double-stranded RNA, such as RIG-I/MDA5, are key mediators of anti-viral responses. Here we screen for synergistic drug-virotherapy combinations and find that the reovirus type III Dearing strain (Rt3D)-palbociclib combination augments oncolytic virus-induced stress responses and increases interferon production and signaling. Data from RIG-I agonist and ER stress-inducing agents further confirms the crosstalk between RNA-sensing and ER stress in inducing cancer cell death and interferon production. Combined Rt3D-palbociclib also increases innate immune activation and IFN-induced HLA expression within tumor cells, with accompanying alterations in the epigenetic landscape and endogenous retroviral (ERV) elements. Analysis of the immunopeptidome in treated cells further reveals changes to HLA-captured peptides, including altered expression of peptides from cancer or testis antigens and ERVs. Our findings thus highlight the crosstalk between stress signaling and PRR activation for mediating enhanced anti-cancer efficacy.

Fig. 1. Palbociclib enhances Rt3D cytotoxicity and virus defence response.

A A375 were treated with Rt3D (MOI 1) and 80 compounds at various doses. Low Z scores (below −2, dashed line) indicate sensitisers (palbociclib is shown in red, with relative Z scores that represent PD-0332991 doses 1 μM (−2.041) and PF-332991 doses at 1 μM (−5.076) and 0.5 μM (−5.570)). B Cell viability in A375 cells treated with Rt3D ± palbociclib (1 μM) by MTT (left) and crystal violet (right) 72 h after treatment (data are presented as mean values ± SEM, n = 3 biologically independent experiments). C Cell viability in melanoma cells MeWo (BRAF/RAS wild-type), and DO4 (Ras mutant) treated with Rt3D ± palbociclib (1 μM) by MTT (72 h). D Cell viability measured at 72 h by MTT assay in 4434 cells treated ∓ palbociclib (1 μM) (data are presented in (C, D) as mean values ± SEM, n = 3 biologically independent experiments). E Volcano plot of RNA sequencing showing upregulated and downregulated genes for A375 cells treated with palbociclib (1 μM), Rt3D (MOI 0.1) or the combination compared to basal at 48 h. Highlighted genes in red belong to the GSEA set ‘reactome interferon signalling’. F RT-qPCR of IFNa, IFNb, IL-28, IL-29, ISG15 and OASL in A375 cells treated with Rt3D ± palbociclib (1 μM) at 48 h. Data presented are mean values ± SEM, n = 3 biologically independent experiments. P values were determined by one-way ANOVA corrected for multiple comparisons. G RT-qPCR of mouse IFNα and IFNβ in murine BRAF-mutant melanoma 4434 cells treated with Rt3D ± palbociclib (1 μM) at 48 h. Data are presented as mean values ± SEM, n = 3 biologically independent experiments. P values were determined by one-way ANOVA corrected for multiple comparisons. H Proteomic analysis of A375 cells treated with Rt3D (0.1) in combination with palbociclib (1 μM) reveal upregulated (red) and downregulated (blue) proteins categorised under the GO term ‘defence response to virus’ (24 h, data are from 1 independent experiment). I IFNα and IFNβ, measured in cell-free supernatant from Rt3D-treated samples (MOI 0.05-0.5) ± palbociclib (1 μM) by ELISA. Results from 2 independent experiments. The lines joining the open and closed circles indicate the data points derived from the same experiment. J Tumour volumes for animals bearing A375 tumours treated with either a sham injection with vehicle (n = 13), an intra-tumoural injection of Rt3D (1 × 10⁶ pfu, n = 11), palbociclib (n = 10), or the combination (n = 9). Tumour volumes for C57BL/6 mice bearing murine BRAF-mutant 4434 tumours were treated with a sham injection with vehicle (n = 6), an intra-tumoural injection of Rt3D (1 × 10⁶ pfu, n = 9), palbociclib (n = 6), or the combination (n = 8). Data are presented as mean values ± SEM, and p values were derived from one-way ANOVA adjusted for multiple comparisons from area under curve values from individual mice, where p > 0.05 ns, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. Source data are provided as a Source Data file.

Fig. 2. Rt3D-palbociclib increases antigen-processing machinery and immunogenicity.

A Proteomic analysis of A375 cells treated with Rt3D (0.1–1) ± palbociclib (1 μM) reveal upregulated (red) and downregulated (blue) proteins categorised under the GO term ‘antigen processing and presentation’ (48 h, data are from 1 independent experiment). B Expression of HLA-A,B,C (class I) and HLA-DR,DP,DQ (class II) in A375 cells treated with Rt3D ± palbociclib by FACS analysis. Data are representative of 3 independent biological experiments (shown in Fig. S2D). C Expression of HLA-A,B,C (class I) and HLA-DR,DP,DQ (class II) in A375 cells treated with Rt3D plus palbociclib ± ruxilitinib (RUX). Data are representative of 2 independent biological experiments (shown in Fig. S2E). D Proteomic analysis of A375 cells treated with Rt3D (0.1–1) ± palbociclib (1 μM) reveal upregulated (red) and downregulated (blue) proteosome subunits. PSMB8, 9 and 10 (highlighted) are incorporated into the proteosome to switch the standard proteasome into an immunoproteasome (24 h, above, 48 h, below, data are from 1 independent experiment for each time point). E Fold change in CTA petides from HLA-I with Rt3D-palbociclib versus basal conditions. Peptides belonging to the MAGEA4 protein are highlighted (1–8). Data represent 1 independent experiment. F Proteomic analysis of A375 cells treated with Rt3D (0.1–1) ± palbociclib (1 μM) reveal upregulated (red) and downregulated (blue) MAGEA4 protein (48 h, top). RT-qPCR of MAGEA4 in A375 cells treated with Rt3D ± palbociclib, ∓ruxilitinib (RUX) (1 μM) at (48 h). Proteomics data represent 1 independent experiment and PCR data are presented as mean values ± SEM, n = 3 independent biological experiments. G ATP release by cell titre glo assay in supernatants from A375 cells treated with indicated Rt3D doses ∓ palbociclib (1 μM). Data are presented as mean values ± SEM, n = 3 independent biological experiments. P values were determined by a two-way ANOVA with adjustments made for multiple comparisons. H HMGB1 release by western blot of supernatants of treated A375 cells. Data are representative of n = 2 independent biological experiments (available in source file). I Calreticulin (CRT, white arrows) cell surface expression by confocal microscopy of treated A375 cells, quantified for 3 fields of view per experiment, for 3 independent biological repeats. J Pictomicrographs of macrophages stained with CD11b-FITC co-cultured with treated, phrodo-stained tumour cells (red), showing engulfed (yellow arrows) and non-engulfed (white arrows) cells, enlarged examples are shown on the left. Percentages of macrophage population that dual-stain for phrodo are shown. Data are presented as mean values ± SEM, n = 3 independent biological experiments. P values were determined by ANOVA with adjustments made for multiple comparisons. P values were derived where p > 0.05 ns, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. Source data are provided as a Source Data file.

Fig. 3. Immune cell activation in vivo with Rt3D-palbociclib treatment.

A Mice bearing 4434 tumours were treated with sham injection with vehicle (n = 6), palbociclib (100 mg/kg daily by oral gavage, n = 5), a single injection of 5  × 10⁶ pfu Rt3D (n = 6), or the combination (n = 6). Tumours were harvested at day 6 after injection and TILs stained for activation markers. Data show proportions of CD69+ or GzmB+ cells in CD8, CD4 T or NK cells. A one-way ANOVA was used with the p value corrected for multiple comparisons. B Kaede mice bearing 4434-mcherry tumours were treated with palbociclib (100 mg/kg daily by oral gavage, blue arrows) followed by a single injection of 5 ×10⁶ pfu Rt3D. The following day, tumours were photoconverted and, 48 h later, collected and stained for activation markers. Data are shown for 4 gated populations as shown: unlabelled, antigen-free DCs (negative staining for both kaede red and mCherry); DCs from tumour origin (kaede red+); tumour antigen-loaded DCs (mCherry+ DCs); and DCs from tumour origin, also loaded with tumour antigen (kaede red+ and mCherry+). Created in BioRender. Nenclares, P. (2025) https://BioRender.com/rwram82. C Data for experiment as describe in B, display type I conventional DCs, cDC1, and type II conventional DCs, cDC2 in lymph nodes. D. Data for type I conventional DCs, cDC1, and type II conventional DCs, cDC2 in tumours. For data in C and D, a one-way ANOVA was used with the p value corrected for multiple comparisons (PBS/vehicle, n = 7, PBS/pal n = 6, Rt3D/vehicle n = 7, Rt3D/pal n = 7). P values were derived where p > 0.05 ns, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. Source data are provided as a Source Data file.

Fig. 4. Rt3D-palbociclib increases RNA sensor expression and RNA ERV species expression that is dependent on JAK/STAT signalling. ERV peptides are found in the immunoproteome.

A Volcano plot of RNA sequencing data showing upregulated and downregulated genes for A375 cells treated with Rt3D (MOI 0.1) plus palbociclib (1 μM), compared to Rt3D alone, at 48 h. Data are from n = 3 independent biological replicates. B Proteomic analysis of A375 cells treated with Rt3D (0.1–1) ± palbociclib (1 μM) reveal upregulated (red) and downregulated (blue) histone proteins (48 h, data are from 1 independent experiment). C Proteomic analysis of A375 cells treated with Rt3D (0.1–1) ± palbociclib (1 μM) reveal upregulated (red) and downregulated (blue) categorised under the GO term ‘DNA modification’ (48 h, data are from 1 independent experiment). D RT-qPCR of RIG-I (DDX58), MDA-5 (IFIH1) and TLR3 in A375 cells treated with Rt3D plus palbociclib (1 μM) at 48 h. Data presented are mean values ± SEM, n = 3 biologically independent experiments. P values were determined by one-way ANOVA corrected for multiple comparisons. E siRNA silencing of RNA sensors or scrambled control (SCR) in A375 cells treated with Rt3D (0.1) ∓ palbociclib (1 μM) by sulforhodamine B assay. Data are representative of 3 independent repeats. F RT-qPCR of Rt3D genome in A375-infected cells ± palbociclib (1 μM) at 48 h. Data presented are mean values ± SEM, n = 5 biologically independent experiments. G Replication of Rt3D by one-step virus growth assay for A375 ± palbociclib (1 μM). Data presented are mean values ± SEM, n = 3 biologically independent experiments. H RT-qPCR for the expression of a panel of 26 previously described ERVs in A375, treated with combination therapy, or Rt3D, respectively, from 1 independent biological experiment. I RT-qPCR of the ERVs MER21C, MLTA10, MLT1C627 or MER57B1, in A375 cells treated with Rt3D (MOI 0.1) ± palbociclib at 48 h. Data presented are mean values ± SEM, n = 3 biologically independent experiments. P values were determined by one-way ANOVA corrected for multiple comparisons. J MLT1C49, MER21C, MLTA10, MLT1C627 or MER57B1 expression in A375 cells treated with Rt3D (MOI 0.1) and/or palbociclib (1 μM) ± the JAK/STAT inhibitor, ruxolitinib (RUX, 1 μM) at 48 h. Data presented are mean values ± SEM, n = 3 biologically independent experiments. K Volcano plot of RNA sequencing data showing upregulated and downregulated ERVs in A375 cells treated with Rt3D ∓ palbociclib compared to basal at 48 h (using TETool kit). L Fold change in ERV peptides from captured HLA-I with Rt3D-palbociclib versus basal conditions. ERV Peptides Filtered: 5% Peptide FDR, Confirmed TMT Quant, Single Genomic Loci, Predicted as MHC Binding. P values were derived where p > 0.05 ns, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. Source data are provided as a Source Data file.

Fig. 5. Palbociclib sensitises cells to UPR activation/ER stress and in combination with Rt3D induces an ER stress signature.

A FACS analysis of A375 cells treated with Rt3D ± palbociclib (1 μM) stained for PI at 48 h. Data presented are mean values ± SEM, n = 3 biologically independent experiments. P values were determined by two-way ANOVA of the sub-G1 compartment corrected for multiple comparisons. B Western blots showing apoptotic markers in A375 cells treated with Rt3D (0.01–1) ± palbociclib (1 μM) at 48 h, representative for 3 biological replicates. Quantification of cleaved caspase-3 bands are shown for the mean values ± SEM, n = 3 for the biologically independent experiments. A one-way ANOVA was used with the p value corrected for multiple comparisons. C Cell viability in A375 cells treated with Rt3D ± palbociclib (1 μM) ± caspase-4 inhibitor Z-YVAD-FMK, by MTT (72 h). Data represented are mean values ± SEM, n = 2 independent biological experiments. D siRNA silencing of caspase-4 or scrambled control (SCR), in A375 cells treated with Rt3D (0.1) plus palbociclib (1 μM) by crystal violet. Quantification is shown for the mean values ± SEM, n = 3 biological replicates. P values were determined by a two-tailed t-test. E Proteomic analysis of A375 cells treated with Rt3D (0.1–1) in combination with palbociclib (1 μM) showing upregulated (red) and downregulated (blue) proteins by combination treatment categorised under the gene ontology (GO) term ‘response to endoplasmic reticulum stress’ (48 h, data are from 1 independent experiment). F Proteomic analysis of A375 cells treated with Rt3D (0.1) in combination with palbociclib (1 μM) showing upregulated (red) and downregulated (blue) ER chaperones (24 h, data are from 1 independent experiment). G siRNA knockdown of UPR components or scrambled control (SCR), in A375 cells treated with Rt3D (0.1) plus palbociclib (1 μM) by crystal violet. Data show mean values ± SEM, n = 3 for the biologically independent experiments. A one-way ANOVA was used with the p value corrected for multiple comparisons. H RT-qPCR of CHOP at 48 h following Rt3D + palbociclib (1 μM) in A375. Data presented are mean values ± SEM, n = 3 for the biologically independent experiments. A one-way ANOVA was used with the p value corrected for multiple comparisons. I Cell viability in A375 and Mel624 cells treated with thapsigargin (TG, 0.04 μM) ± palbociclib (1 μM) by MTT assay at 72 h. Data presented are mean values ± SEM, n = 3 biologically independent experiments. A one-way ANOVA was used with the p value corrected for multiple comparisons. J RT-qPCR of the spliced form of XBP1 at 48 h following Rt3D + palbociclib (1 μM) in A375. Data presented are mean values ± SEM, n = 3 biologically independent experiments. A two-way ANOVA was used with the p value corrected for multiple comparisons. P values were derived where p > 0.05 ns, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. Source data are provided as a Source Data file.

Fig. 6. Crosstalk between RNA sensors and ER stress/UPR activation enhances RNA sensing and IFN response.

A A375 cells treated with transfected or B untransfected poly I:C (10 μg/mL), combined with thapsigargin (TG, 0.01, 0.05 μM) and cell viability shown by crystal violet assay. C RT-qPCR of RIG-I (DDX58), MDA-5 (IFIH1), IFNα IFNβ in A375 cells treated with transfected poly I:C (0.1, 2, 10 μg/mL), ±thapsigargin (TG, 0.05 μM) at 48 h. Data presented are mean values ± SEM, n = 3 biologically independent experiments. A two-way ANOVA was used with the p value corrected for multiple comparisons. D RT-qPCR of indicated gene expression in A375 cells treated with RIG-I agonist 3p-hpRNA (1 μg/mL) ± thapsigargin (0.05 μM) at 48 h. Data presented are mean values ± SEM, n = 3 for the biologically independent experiments. A one-way ANOVA was used with the p value corrected for multiple comparisons. E RT-qPCR of the ERV MLT1C49 in A375 cells treated with 3p-hpRNA ± thapsigargin at 48 h. Data presented are mean values ± SEM, n = 3 for the biologically independent experiments. A one-way ANOVA was used with the p value corrected for multiple comparisons. F IFNβ measured by ELISA in cell-free supernatant from A375 cells treated with Rt3D (0.1 MOI) ∓ palbociclib (1 μM) ± IRE1a inhibitor (STF-083010). Data shown are from 1 independent biological experiment, but is representative of 2 independent experiments. G RT-qPCR of RIG-I (DDX58) and IFNb in A375 cells treated with 1 μM palbociclib (+) ± poly I:C (2, 10 μg/mL) at 48 h. Data presented are mean values ± SEM, n = 3 for the biologically independent experiments. A two-way ANOVA was used with the p value corrected for multiple comparisons. H Tumour volumes for C57BL/6 mice bearing 4434 tumours treated with either sham injection with vehicle (n = 6), a single intra-tumoural injection of poly I:C (50 μg, n = 12), palbociclib (90 mg/kg, n = 6), or the combination (n = 11). Data are presented as mean values ± SEM, and p values were derived from one-way ANOVA adjusted for multiple comparisons from area under curve values from individual mice, where p > 0.05 ns, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. Source data are provided as a Source Data file.