Synthetic lethality between VPS4A and VPS4B triggers an inflammatory response in colorectal cancer

Abstract

Somatic copy number alterations play a critical role in oncogenesis. Loss of chromosomal regions containing tumor suppressors can lead to collateral deletion of passenger genes. This can be exploited therapeutically if synthetic lethal partners of such passenger genes are known and represent druggable targets. Here, we report that VPS4B gene, encoding an ATPase involved in ESCRT-dependent membrane remodeling, is such a passenger gene frequently deleted in many cancer types, notably in colorectal cancer (CRC). We observed downregulation of VPS4B mRNA and protein levels from CRC patient samples. We identified VPS4A paralog as a synthetic lethal interactor for VPS4B in vitro and in mouse xenografts. Depleting both proteins profoundly altered the cellular transcriptome and induced cell death accompanied by the release of immunomodulatory molecules that mediate inflammatory and anti-tumor responses. Our results identify a pair of novel druggable targets for personalized oncology and provide a rationale to develop VPS4 inhibitors for precision therapy of VPS4B-deficient cancers.

Keywords: CRC; ESCRT; VPS4B; immunogenic cell death; synthetic lethality.

Figure 1. Expression of VPS4B is downregulated in CRC

1. A

   Left panel, a scheme of chromosome 18 copy number alterations depicting the distal long arm loss across TCGA Pan‐Cancer dataset. Vertical red line indicates the localization of VPS4B. Right panel, enlarged fragment of chromosome 18 showing frequent deletions of VPSB locus in cancer samples. Deletions are marked in blue, and amplified regions are marked in red. Both panels were generated with UCSC Xena browser (https://www.biorxiv.org/content/10.1101/326470v3).

2. B

   Analysis of VPS4B copy number alterations in TCGA Pan‐Cancer dataset. Cancer types were sorted according to the mean VPS4B copy number after removing germline values. The boxes denote the 25^th to 75^th percentile range, and the center lines mark the 50^th percentile (median). The whiskers reflect the largest and smallest observed values. VPS4B copy number alteration data were fetched using UCSC Xena browser.

3. C

   Scatter plot, analysis of VPS4B mRNA expression (number of transcripts per million) plotted against VPS4B copy number from TCGA CRC patient samples (n = 376); plot generated using cBioPortal (Gao et al, 2013). Pie chart, summary of all types of VPS4B copy number alterations based on the analysis of data from 615 CRC samples deposited in the Colorectal Adenocarcinoma (TCGA, Provisional) dataset on cBioPortal (http://www.cbioportal.org/).

4. D

   qPCR analysis of VPS4B mRNA abundance in normal colon, adenoma, and CRC samples. Adenocarcinoma (n = 26); adenoma (n = 42); normal colon (n = 24). Green horizontal bars indicate means, and red whiskers indicate SD. Differences were analyzed using the Kruskal–Wallis test followed by Dunn's multiple comparison post hoc test; ns—non‐significant (P ≥ 0.05), ****P < 0.0001.

5. E

   Examples of immunohistochemical staining of VPS4B in normal colon and matched CRC samples as an illustration of the scoring system used for the evaluation presented in (F). 3+—very intensive staining, 2+—medium‐intensive staining, 1+—weak staining, 0—no staining. Scale bar, 50 μm.

6. F

   Comparative analysis of VPS4B staining performed in 100 pairs of normal colon and matched CRC samples.

Data information: The exact P‐values can be found in the source data for this figure.Source data are available online for this figure.

Figure EV1. Evaluation of VPS4A and VPS4B protein abundance in human tissues and in CRC

1. A

   qPCR analysis of VPS4A mRNA abundance in normal colon, adenoma, and CRC samples. Adenocarcinoma (n = 26); adenoma (n = 42); normal colon (n = 24). Green horizontal bars indicate means, and red whiskers indicate SD. Differences were analyzed using the Kruskal–Wallis test followed by Dunn's multiple comparison post hoc test; ns—non‐significant (P ≥ 0.05).

2. B

   Immunohistochemistry (IHC) evaluation of VPS4A abundance in pairs of normal colon versus matched CRC samples (2 representative pairs out of 100 analyzed). 3+—very intensive staining. Scale bar, 20 μm.

3. C

   Specificity tests of VPS4B IHC staining using human tissues and mouse xenografts. C1—strong granular cytoplasmic staining in the mucosa of the appendix. C2, C3—negative staining in the appendix (C2) and CRC (C3) when primary antibody (anti‐VPS4B) was omitted. C4—granular cytoplasmic staining in the xenograft from wild‐type (VPS4B ^+/+) HCT116 cells. C5—negative staining in the xenograft from HCT116 VPS4B ^−/− cells. Scale bars: C1, C4, and C5—20 μm; C2, C3—50 μm.

4. D

   Specificity tests of VPS4A IHC staining in various human tissues with high and low expression of VPS4A. D1—strong granular cytoplasmic staining in the mucosa of the appendix (arrowhead, and at a higher magnification in D2) and weak staining in the muscle (arrow, and at a higher magnification in D3). Scale bars: D1—200 μm; D2 and D3—20 μm.

Data information: The exact P‐values can be found in Appendix Table S3.

Figure EV2. Analyses of VPS4A and VPS4B mRNA and protein abundance in HCT116 cells and the dependencies between both genes in different cell lines

1. A

   mRNA levels of VPS4A (upper panel) and VPS4B (lower panel) in HCT116 cells 72 h upon siRNA transfection. For VPS4A or VPS4B depletion, two different duplexes (#1 or #2) of siVPS4A or siVPS4B were used. For simultaneous VPS4A+B depletion, various combinations of siVPS4A and siVPS4B duplexes were used. Two different duplexes of siCTRL (#1 or #2) were used as non‐targeting controls. NT—non‐transfected. Values represent normalized counts after including variance normalized transformation performed by the DESeq2 package for RNA‐Seq data analysis. Data are means of four independent experiments ± SEM. Two‐tailed unpaired t‐test; ns—non‐significant (P ≥ 0.05).

2. B

   Upper panel, representative immunoblotting analysis of VPS4A and VPS4B abundance in lysates of HCT116 cells collected 72 h after transfection as in (A). GAPDH was used as a loading control. Lower panel, densitometry analysis of VPS4A and VPS4B abundance based on immunoblotting analysis as shown in the upper panel. NT—non‐transfected. Data are means of four independent experiments ± SEM. Wilcoxon signed rank test; ns—non‐significant (P ≥ 0.05).

3. C

   Correlation between dependency scores and VPS4B copy number for selected cancer cell lines from the DepMap portal dataset (https://depmap.org/portal/). According to the portal, a lower score (below −0.5) means that a gene is more likely to be dependent in a given cell line. A score of 0 is equivalent to a gene that is not essential, whereas a score of −1 corresponds to the median of all common essential genes.

4. D

   Immunoblotting analysis of VPS4A silencing efficiency in HOP62 and SNU410 cell lines. Lysates were prepared 6 days after siRNA transfection with non‐targeting (siCTRL#1) or VPS4A‐targeting (siVPS4A#2 or #5) duplexes. Vinculin was used as a loading control.

Data information: The exact P‐values can be found in Appendix Table S3.

Figure 2. Synthetic lethality between VPS4A and VPS4B inhibits growth of CRC lines in vitro

1. A

   Analysis of viability of HCT116 cells assessed 96 h after transfection with independent non‐targeting siRNA (two different duplexes used, siCTRL#1 or #2) or targeting VPS4A (duplexes #1 or #2), VPS4B (duplexes #1 or #2), or both VPS4 (various combinations of siVPS4A+siVPS4B duplexes). Data are means of three independent experiments ± SEM. All values were normalized, averaged (avrg) viability of siCTRL#1‐ and #2‐transfected cells was set as 100%, the Kruskal–Wallis test followed by Dunn's multiple comparison post hoc test; ns—non‐significant (P ≥ 0.05), **P < 0.01.

2. B

   Left panel, analysis of clonogenic growth of HCT116 cells assessed 15 days after transfection with non‐targeting (siCTRL#1), VPS4A‐ or/and VPS4B‐targeting siRNAs used in various duplex combinations as indicated. Right panel, images of HCT116 clones taken at the day of clonogenic growth assessment. Data are means of three independent experiments ± SEM. NT—non‐transfected. All values were normalized, and clonogenic growth of siCTRL#1‐transfected cells was set as 1. One‐sample t‐test; ns—non‐significant (P ≥ 0.05), ****P < 0.0001.

3. C

   Viability of RKO and SW480 cells assessed as in (A), growth of DLD‐1 cells assessed in BrdU incorporation assay 96 h after siRNA transfection. Various independent non‐targeting (siCTRL#1 or #2) and VPS4A‐ or B‐targeting siRNA duplexes were used (#1, #2, #5 or #1, #2, respectively). All values were normalized, and averaged (avrg) viability of siCTRL#1‐ and #2‐transfected cells was set as 100%. Data are means of three independent experiments ± SEM. The Kruskal–Wallis test followed by Dunn's multiple comparison post hoc; ns—non‐significant (P ≥ 0.05), **P < 0.01, ***P < 0.001.

4. D

   VPS4B copy number and dependency scores of selected cancer cell lines obtained in CRISPR/Cas9 and RNAi screens deposited in the DepMap portal (https://depmap.org/portal/).

5. E

   VPS4B copy number status estimated across different cell lines using TaqMan assay. The error bars represent the minimal and maximal copy number in a given triplicate readout. The BCL2 gene localized on 18q in the close vicinity to VPS4B was analyzed as a control.

6. F

   VPS4B and VPS4A protein abundance in selected normal and cancer cell lines analyzed by immunoblotting. Vinculin and GAPDH were used as loading controls.

7. G

   Analysis of viability of HOP62 lung cancer cells assessed 144 h after transfection with independent non‐targeting siRNA (two different duplexes used, siCTRL#1 or #2) or targeting VPS4A (duplexes #2 or #5), VPS4B (duplexes #1 or #2), or both VPS4 (various combinations of siVPS4A+siVPS4B duplexes). Data are means of four independent experiments ± SEM. All values were normalized, and cell viability of averaged (avrg) siCTRL#1‐ and #2‐transfected cells was set as 100%. Two‐tailed unpaired t‐test; ****P < 0.0001.

8. H

   Top panel, analysis of clonogenic growth of HOP62 cells assessed 14 days after transfection with various siRNA duplexes as indicated. Bottom panel, images of HOP62 clones taken at the day of clonogenic growth assessment. Data are means of four independent experiments ± SEM. All values were normalized, and colony area of averaged (avrg) siCTRL#1‐ and #2‐transfected cells was set as 1. The Mann–Whitney U‐test; **P < 0.01.

9. I

   Analysis of viability of SNU410 pancreatic cancer cells assessed 168 h after transfection with independent non‐targeting siRNA (three different duplexes used, siCTRL#1, #2, or #3) or targeting VPS4A (duplexes #2, #4, or #5), VPS4B (duplexes #1 or #2), or both VPS4 (various combinations of siVPS4A+siVPS4B duplexes). Data are means of three independent experiments ± SEM. All values were normalized, cell viability of averaged (avrg) siCTRL#1‐, #2‐, and #3‐transfected cells was set as 100%. Two‐tailed unpaired t‐test; ****P < 0.0001.

Data information: The exact P‐values can be found in the source data for this figure.Source data are available online for this figure.

Figure EV3. Characterization of HCT116 VPS4B ^−/− cell line and engineering of HCT116 VPS4B^−/− cells with doxycycline (Dox)‐inducible VPS4A‐targeting shRNA expression (HCT116 VPS4B ^−/− shVPS4A) for in vivo studies

1. A

   Left panel, PCR sequencing analysis verifying bi‐allelic VPS4B knockout in two CRISPR/Cas9 engineered clones derived from the HCT116 VPS4B ^+/+ parental line. Right panel, immunoblotting analysis of VPS4B in cell lysates of these clones.

2. B

   Comparison of clonal growth of isogenic HCT116 lines: parental VPS4B ^+/+ and VPS4B ^−/−. Upper panel, analysis of colony area of parental HCT116 VPS4B ^+/+ cells and VPS4B ^−/− clones assessed in the colony formation assay. Data are means of five independent experiments. Values were normalized to the colony area of parental HCT116 VPS4B ^+/+ cells that was set as 1. Error bars are SEM. One‐sample t‐test; ns—non‐significant (P ≥ 0.05). Lower panel, representative images of clonal growth of parental and VPS4B ^−/− HCT116 cells that were used for the quantification presented on the top.

3. C

   Analysis of the tumor growth in mice bearing parental HCT116 VPS4B ^+/+ or VPS4B ^−/− xenografts. n = 3 mice for each group, each mouse bearing two tumors, ± SEM.

4. D

   Confirmation of cell death of HCT116 VPS4B ^−/− cells upon VPS4A depletion. Upper panel, viability of HCT116 VPS4B ^−/− assessed 72 h after siVPS4A transfection (three independent siVPS4A duplexes #2, #4, and #5 were used). Non‐transfected (NT) or siCTRL#1‐transfected cells served as viability controls. Data are means of four independent experiments. Values were normalized to the viability of siCTRL#1‐transfected cells that was set as 100%. Error bars are SEM. One‐sample t‐test; ****P < 0.0001. Lower panel, representative images of HCT116 VPS4B ^−/− clones grown for 15 days after siCTRL#1 or siVPS4A transfection. NT—non‐transfected cells.

5. E

   Immunoblotting analysis of VPS4A abundance in lysates of HCT116 VPS4B ^−/− shVPS4A clones (#1 to #3, each bearing an independent VPS4A‐targeting shRNA construct). To induce shRNA expression, doxycycline (Dox) was administered to the cell culture medium (1 μg/ml) 3 days before cell lysis. Lysates of doxycycline‐treated and non‐treated HCT116 VPS4B ^−/− shCTRL clones (#1 and #2, each bearing an independent non‐targeting shRNA construct) were loaded as controls. Lysates from HCT116 VPS4B ^+/+ cells were used to control VPS4A and VPS4B protein detection. GAPDH served as a loading control.

6. F

   Comparison of the growth rate of HCT116 VPS4B ^−/− line and its derivative clones bearing different shVPS4A (#1 and #2) or shCTRL (#1 and #2) constructs cultured in the presence (+Dox) or absence (−Dox) of doxycycline. Data are expressed as the percentage of the viability of HCT116 VPS4B ^−/− cultured in the absence of doxycycline at day 5 (set as 100%). Values are means of three independent experiments. Error bars are SEM. Two‐tailed unpaired t‐test; ns—non‐significant (P ≥ 0.05), *P < 0.05; **P < 0.01.

7. G

   Monitoring of body weight of mice bearing engrafted HCT116 VPS4B ^−/− shVPS4A#1 cells. Day 1 indicates the first day of doxycycline administration (+Dox) to the drinking water of mice. Data represent the means of body weight of 9 mice used in each experimental condition. Error bars are SEM.

Data information: The exact P‐values can be found in Appendix Table S3.

Figure 3. Synthetic lethality between VPS4A and VPS4B inhibits growth of CRC cells in a mouse xenograft model

1. A

   Schematic illustration of xenograft experiments with HCT116 VPS4B ^−/− cells having doxycycline (Dox)‐inducible expression of shRNA targeting VPS4A.

2. B

   Left panel, growth of HCT116 VPS4B ^−/− shVPS4A cells as xenografts in mice in the presence or absence of doxycycline. Day 1 indicates the first day of doxycycline administration. n = 9 for each group, each mouse bearing one tumor, ± SEM. Two‐tailed unpaired t‐test; ns—non‐significant (P ≥ 0.05), **P < 0.01. Right panel, scatter plot representing end‐point volumes of single xenografts. Bars represent means ± SEM.

3. C

   Immunoblotting analysis of VPS4A abundance in xenograft samples from untreated and doxycycline‐treated mice (6 separate xenograft samples for each group). Lysates of HCT116 VPS4B ^−/− cells transfected with control or VPS4A‐targeting siRNA marked VPS4A protein detection. GAPDH served as a loading control.

4. D

   Growth of HCT116 VPS4B ^−/− shCTRL#1 cells as xenografts in mice in the presence or absence of doxycycline. The arrow indicates the first day of doxycycline (Dox) administration. n = 2 mice for each group, each mouse bearing two tumors, ± SEM.

The exact P‐values can be found in the source data for this figure.Source data are available online for this figure.

Figure EV4. Inhibition of endocytosis and cell cycle progression upon simultaneous depletion of VPS4A and VPS4B

1. A

   Upper panel, microscopy images of HCT116 cells collected 48 h after transfection with control (siCTRL#1) or VPS4A‐ and/or VPS4B‐targeting siRNA (siVPS4A#2 or #5 and siVPS4B#1 or #2). EEA1, Rab7, and LAMP1 were used as markers of early endosomes (EE), late endosomes (LE), and lysosomes, respectively, and were visualized in green. Nuclei were stained with Hoechst 33342 (blue). NT—non‐transfected cells. Scale bar, 10 μm. Lower panel, quantified fluorescence signals from microscopy images. The boxes denote the 25^th to 75^th percentile range, the center lines mark the 50^th percentile (median) and the whiskers reflect the largest and smallest observed values in at least four z‐stacks from three independent experiments. The Welch t‐test (for EEA1 and Rab7) and the Mann–Whitney U‐test (for LAMP1); *P < 0.05; **P < 0.01.

2. B

   Upper panel, example of flow cytometry analysis of AlexaFluor 647‐labeled transferrin (Tf‐A647) uptake by non‐transfected (NT) or siRNA (siCTRL#1 and siVPS4A#2 or #5) transfected HCT116 VPS4B ^−/− cells. Lower panel presents the percentage and mean fluorescence intensity of Tf‐A647‐positive cells. Data are means of four independent experiments ± SEM. Fluorescence intensity of non‐transfected Tf‐A647‐treated cells was set as 1. Statistical significance was assessed using the Mann–Whitney U‐test and t‐test. **P < 0.01; ***P < 0.001.

3. C

   Left panel, example of flow cytometry analysis of cell cycle phase distribution in HCT116 VPS4B ^−/− cells 72 h after transfection with control (siCTRL#1) or VPS4A‐targeting siRNA (siVPS4A#2 and #5). NT—non‐transfected. Graphs were generated using the ModFit program. Right panel, analysis of cell cycle phase distribution in HCT116 VPS4B ^−/− cells transfected with siRNA as indicated. Data are means of four independent experiments ± SEM. Statistical significance was assessed using the Mann–Whitney U‐test. ns—non‐significant (P ≥ 0.05); **P < 0.01.

Data information: The exact P‐values can be found in Appendix Table S3.

Figure 4. Combined knockdown of VPS4 proteins in HCT116 cells induces alterations in the transcriptome

1. A

   Gene ontology (GO) analysis of biological processes for transcriptionally upregulated genes (≥ 1.5‐fold; adjusted P‐value < 0.05) after combined VPS4A+B silencing using the enrichGO function from clusterProfiler.

2. B

   Heatmap visualizing expression of genes related to inflammatory response (left panel) and positive regulation of cell death (right panel) generated from the GO analysis of biological processes across different transfection conditions with at least three biological replicates.

3. C

   Selected pathways related to inflammatory response and programmed cell death among transcriptionally upregulated genes after combined VPS4A+B silencing were identified using the enrichPathway function from ReactomePA.

Figure 5. Combined VPS4A+B depletion induces NF‐κB signaling and caspase‐dependent and caspase‐independent cell death pathways

1. A

   Immunoblotting analysis of the canonical and noncanonical branches of the NF‐κB pathway and mediators of caspase‐dependent cell death. Lysates of HCT116 VPS4B ^−/− cells were collected 66 h after transfection with siRNA (siCTRL#1 or different siVPS4A duplexes: #2, #4, or #5). Lysates of HCT116 VPS4B ^+/+ and non‐transfected HCT116 VPS4B ^−/− cells were used to monitor the basal pathway activity. Representative blot from 10 experiments is shown. NT—non‐transfected; p‐RelA—phospho‐RelA; p‐IκBα—phospho‐IκBα; cl—cleaved caspases or PARP‐1. GAPDH or vinculin served as loading controls.

2. B

   Densitometry analysis of the abundance of the indicated proteins based on immunoblot images as shown in (A). Data are means of 10 (phospho‐Rel and cleaved caspase 7), nine (p52, caspase 8 and cleaved caspase 9), seven (phosphorylated and total IκB), or five (cleaved caspase 3) independent experiments. Error bars are SEM. Statistical significance was assessed using the following tests: one‐sample t‐test (caspase 8, cleaved caspases 9, 7, and 3, total and phospho‐IκBα) or Wilcoxon signed rank test (phospho‐RelA and p52); *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

3. C

   Immunoblot showing induction of the NF‐κB pathway (p52) and apoptosis (cleaved caspase 7) in xenograft samples described in Fig 3C. cl—cleaved caspase. Vinculin was used as a loading control.

4. D

   Analysis of the impact of RIPK1 inhibitor (necrostatin‐1) or pan‐caspase inhibitor (Q‐VD‐Oph) on cell viability of HCT116 VPS4B ^−/− cells transfected with siRNA (non‐targeting siCTRL#1 or different siVPS4A duplexes: #2, #4, or #5). Cell viability was assessed 72 h after siRNA transfection. Necrostatin‐1 (50 μM), Q‐VP‐Oph (20 μM), or vehicle were added to the medium 48 h before viability assessment. Data are means of five independent experiments ± SEM. All values were normalized, and viability of siCTRL‐transfected and vehicle‐treated cells was set as 100%. Two‐tailed unpaired t‐test; **P < 0.01.

Data information: The exact p‐values can be found in the source data for this figure.Source data are available online for this figure.

Figure 6. Synthetic lethality between VPS4A and VPS4B induces release of immunogenic DAMPs and promotes M1 macrophage polarization

1. A, B

   Measurement of ATP (A) and HMGB1 (B) released to the cell medium by HCT116 VPS4B ^−/− cells non‐transfected (NT) or transfected with siRNA (non‐targeting siCTRL#1 or targeting siVPS4A duplexes: #2, #4, or #5). Cell culture media were exchanged 16 h after transfection, and fresh media were conditioned for the next 52‐58 h. For non‐transfected cells (NT), the same treatment protocol was used but without the transfection mixture. Data are means of five (A) or four (B) independent experiments ± SEM. Two‐tailed unpaired t‐test; **P < 0.01, ***P < 0.001.

2. C

   Microscopy images presenting cell surface calreticulin (green) in HCT116 VPS4B ^−/− cells 48 h after transfection with siRNA (non‐targeting siCTRL#1 or targeting siVPS4A#2). As a positive control for detection of cell surface calreticulin, non‐transfected cells (NT) were treated with 2 μM mitoxantrone for 24 h. In blue, DAPI staining. Scale bar, 15 μm.

3. D

   Flow cytometric analysis of calreticulin exposed on the cell surface of VPS4A‐depleted HCT116 VPS4B ^−/− cells 66 h after siRNA transfection (siVPS4A duplexes: #2, #4, or #5 were used). Non‐transfected (NT) or siCTRL#1‐transfected cells served as negative controls. Left panel, percentage of cells positive for calreticulin in the population of live (DAPI‐negative) cells, data are means of four independent experiments ± SEM. The Mann–Whitney U‐test; **P < 0.01. Right panel, representative dot plot diagrams of flow cytometric analysis of cell surface‐exposed calreticulin. Primary rabbit anti‐calreticulin and control isotype IgG antibodies were used for staining, followed by secondary AlexaFluor 647‐conjugated antibody.

4. E

   qPCR analysis of M1 (pro‐inflammatory) and M2 (anti‐inflammatory) macrophage polarization markers in mouse bone marrow‐derived macrophages (BMDMs) incubated for 24 h in conditioned media (CM) collected from control (siCTRL#1), Vps4a‐ and/or Vps4b‐depleted CT‐26 cells. For double Vps4a+b depletion, various combinations of siVps4a#2 or #5 and siVps4B#3 or #4 duplexes were used. Data were normalized and are presented as the fold change of expression of a given M1 or M2 marker in BMDMs treated with CM from Vps4‐depleted cells compared to its expression in BMDMs treated with CM from siCTRL#1‐transfected cells (set as 1). Data are means of four independent experiments ± SEM. One‐sample t‐test; ns—non‐significant (P ≥ 0.05), *P < 0.05, **P < 0.01.

Data information: The exact P‐values can be found in the source data for this figure.Source data are available online for this figure.

Figure EV5. (supporting Fig 6). Combined depletion of Vps4a+b in mouse colon carcinoma CT‐26 cells

1. A

   qRT–PCR analysis of the silencing efficiency of Vps4a (upper panel) and Vps4b (lower panel) in CT‐26 cells 72 h after transfection with siRNA. To deplete Vps4a or Vps4b, two independent siRNA duplexes were used (#2 or #5, and #3 or #4, respectively). To simultaneously deplete Vps4a+b, various combinations of siVps4a and siVps4b were used. All values were normalized; Vps4a or Vps4b expression values in siCTRL#1‐transfected cells were set as 1 and used to normalize mRNA abundance in other conditions. NT—non‐transfected.

2. B

   Immunoblotting detection of Vps4a abundance, inflammatory response (phosphorylated RelA) and apoptosis activation (cleaved caspase 7) in lysates of mouse CT‐26 cells collected 72 h after siRNA transfection as in (A). p‐RelA—phosphorylated RelA; cl—cleaved caspase. Vinculin served as a loading control.

3. C

   Phase contrast microscopy images of CT‐26 cells acquired 3 days after transfection with siVps4a or siVps4b as described in (A). Scale bar, 250 μm.

Figure 7. Model for synthetic lethal interaction between VPS4A and VPS4B

Left panel, in normal cells, both VPS4A and VPS4B act redundantly in several essential intracellular processes. So, a single depletion of any VPS4 paralog (e.g., VPS4A) is tolerated, as unperturbed expression of the other paralog alone (e.g., VPS4B) suffices to substitute for its downregulated counterpart. Right panel, cells that have lost VPS4B expression, e.g., due to oncogenic genome rearrangements, rely exclusively on VPS4A activity. So, inactivation of VPS4A in these cells leads to synthetic lethality that is accompanied by strong induction of an inflammatory response and release of immunogenic DAMPs. Immunomodulatory molecules released by dying VPS4A+B‐deficient cancer cells can elicit paracrine effects on primary immune cells, e.g., reprogramming of macrophages toward the M1 anti‐tumor phenotype.