Intracellular Osteopontin Promotes the Release of TNFα by Mast Cells to Restrain Neuroendocrine Prostate Cancer

Abstract

Neuroendocrine prostate cancer (NEPC) is an aggressive form of prostate cancer that emerges as tumors become resistant to hormone therapies or, rarely, arises de novo in treatment-naïve patients. The urgent need for effective therapies against NEPC is hampered by the limited knowledge of the biology governing this lethal disease. Based on our prior observations in the transgenic adenocarcinoma of the mouse prostate (TRAMP) spontaneous prostate cancer model, in which the genetic depletion of either mast cells (MC) or the matricellular protein osteopontin (OPN) increases NEPC frequency, we tested the hypothesis that MCs can restrain NEPC through OPN production, using in vitro co-cultures between murine or human tumor cell lines and MCs, and in vivo experiments. We unveiled a role for the intracellular isoform of OPN, so far neglected compared with the secreted isoform. Mechanistically, we unraveled that the intracellular isoform of OPN promotes TNFα production in MCs via the TLR2/TLR4-MyD88 axis, specifically triggered by the encounter with NEPC cells. We found that MC-derived TNFα, in turn, hampered the growth of NEPC. We then identified the protein syndecan-1 (SDC1) as the NEPC-specific TLR2/TLR4 ligand that triggered this pathway. Interrogating published single-cell RNA-sequencing data, we validated this mechanism in a different mouse model. Translational relevance of the results was provided by in silico analyses of available human NEPC datasets and by immunofluorescence on patient-derived adenocarcinoma and NEPC lesions. Overall, our results show that MCs actively inhibit NEPC, paving the way for innovative MC-based therapies for this fatal tumor. We also highlight SDC1 as a potential biomarker for incipient NEPC.

Figure 1.

OPN-proficient MCs are able to restrain NEPC in mice. A, Immunofluorescence for DAPI (cyan), WGA (blue), and OPN (red) on WT and OPN^−/− MCs. Images were acquired with a confocal microscope. B, Quantification of A as a percentage of OPN-positive cells in each field. The histogram depicts the mean ± SD of biological replicates (represented by dots; n = 9 per group). Unpaired t test was used: ****, P < 0.0001. C, Representative immunofluorescence for DAPI (blue), OPN (red), and TRY (green) on prostates of a TRAMP mouse and of Kit^Wsh-TRAMP mice reconstituted either with WT or OPN^−/− MCs. D, Representative immunofluorescence for DAPI (blue), OPN (red), and TRY (green) on a human prostate adenocarcinoma sample. E, Percentage of prostate lesions scored as high-grade prostate intraepithelial neoplasia (HGPIN), adenocarcinoma (ADENO), or NEPC in 25–30-week-old TRAMP (n = 15), Kit^Wsh-TRAMP (n = 15), or OPN^−/−TRAMP mice (n = 17). Where indicated Kit^Wsh-TRAMP mice were reconstituted i.p. with 5 × 10⁶ WT (n = 19) or OPN^−/− (n = 18) MCs at the age of 8 weeks. Numbers within bars represent the percentage of mice with the indicated histology. Fisher exact test was used for the analysis of contingency between different groups. P values are reported as: **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Where the P value is not indicated, the comparison between groups is not statistically significant.

Figure 2.

iOPN mediates the inhibitory function of MCs against NEPC. A, Murine adenocarcinoma (T1525 and T23; n = 6 per group) and NEPC (ST4787, TC566, and TC411K; n = 4 per group) cells (50,000/well) were cultured either alone or with either WT or OPN^−/− MCs (tumor cell:MC ratio 1:1). After 4 days the growth rate of cancer cells was evaluated through trypan blue count. Cancer cells and MCs could be distinguished thanks to their growth in adhesion or suspension, respectively. B, Human adenocarcinoma (22Rv1; n = 6 per group) and NEPC (LASCPC-01, n = 3 per group and NCI-H660, n = 6 per group) cells (100,000/well), were cultured either alone with the human MC cell line HMC1 for 4 days (tumor cell:MC ratio 1:1). The growth rate of cancer cells was evaluated through flow cytometry. Cancer cells and MCs could be distinguished thanks to the selective expression of CD49f and c-Kit respectively, (gating strategy in Supplementary Fig. S4D). C, Secretion of OPN by Supplementary Fig. S4D). C, Secretion of OPN by MCs and prostate cancer cells measured by ELISA in culture supernatants (n = 3 per group). D, Murine ST4787 NEPC cells were cultured either alone or with MCs (tumor cell:MC ratio 1:1; n = 4 per group). Where indicated cells were seeded in a 0.4 µmol/L transwell system (TW). Tumor cell growth was evaluated as in A. E, Immunofluorescence for DAPI (cyan), WGA (blue), and OPN (red) in MC/9, MC/9-CTR, MC9-OPNf, and MC/9-iOPN cells. Images were acquired with a confocal microscope. F, Western blot for OPN in cell lysates from MC/9, MC/9-CTR, MC9-OPNf, and MC/9-iOPN cells. The western blot was validated twice. G, Quantification of F (n = 2 per group). H, Murine ST4787 NEPC cells were cultured either alone or with MC/9-CTR, MC/9-OPNf, or MC/9-iOPN (tumor cell:MC ratio 1:1; n = 6 per group). Tumor cell proliferation was evaluated as in A to D, G, and H, All in vitro experiments were performed at least two times. Pools of all the biological replicates from all the independent experiments were performed and shown in histograms; statistical analysis was run considering all the samples. All histograms depict mean ± SD of biological replicates (represented by dots). One-way ANOVA followed by Tukey’s multiple comparison test was used: *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Where the P value is not indicated, the comparison between groups is not statistically significant.

Figure 3.

iOPN regulates the TLR pathway in MCs. A, Pathways significantly downregulated in NEPC tumors vs. adenocarcinoma of TRAMP mice (dataset that we generated in ref. 6) identified by the MetaCore software. B, RNA-seq normalized data (z-scores) of castration-resistant prostate adenocarcinoma (CRPC-Adeno) and NEPC (CRPC-NEPC) samples from the human Beltran Data set (25) were downloaded from c-BioPortal and filtered for the genes composing a MC signature that we previously generated (7). The boxplot represents the mean expression of the MC signature in CRPC-Adeno and CRPC-NEPC. The adjusted P value between the two groups was calculated using the Wilcoxon-Mann–Whitney test with Benjamini & Hochberg (FDR) correction. C, GSEA performed on the Beltran data set of human CRPC-Adeno and CRPC-NEPC tumor samples. The heatmap shows the significantly enriched upregulated (red) and downregulated (blue) pathways in NEPC, based on normalized enrichment scale. The box specifies the enriched genes in the downregulated hallmark pathway “IL6_JAK_STAT3_signaling.” D, Flow cytometry evaluation of TLR2 and TLR4 in WT and OPN^−/− MCs. E, ST4787 cells (50,000/well) were co-cultured with WT, OPN^−/−, or MyD88^−/− MCs (tumor:MC ratio 1:1; n = 11 per group). Tumor cell growth rate was evaluated by trypan blue count. F, Immunoprecipitation for MyD88 in WT and OPN^−/− MCs after stimulation with LTA (10 μg/mL) and lipopolysaccharide (1 μg/mL) for 30 minutes. Immunoprecipitated samples were then subjected to western blot for OPN. Immunoprecipitation was validated twice. G, Western blot for phosphorylated (p)P65 and pERK1/2 evaluation in WT or OPN^−/− MCs, either unstimulated or after co-culture with ST4787 or T23 (ratio 1:1) for 15 minutes. The western blot was validated twice. H, Western blot for phosphorylated (p) P65 in WT MCs, either unstimulated or co-cultured with ST4787 or T23 cells (ratio 1:1) for 15 minutes. Where indicated, a specific inhibitor of either TLR2 (TL2-C29, 50 μg/mL) or TLR4 (TAK-242, 100 nmol/L) was added. The western blot was validated twice (n = 2). I, Quantification of H. All in vitro experiments were performed at least two times. Pools of all the biological replicates from all the independent experiments were performed and shown in histograms; statistical analysis was run considering all the samples. All histograms depict mean ± SD of biological replicates (represented by dots). One-way ANOVA followed by Tukey’s multiple comparison test was used for the analysis of significance between samples. P values are reported as: *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Where the P value is not indicated, the comparison between groups is not statistically significant.

Figure 4.

MC-derived TNFα specifically inhibits the growth of NEPC cells by inducing apoptosis. A, Murine NEPC (ST4787) or adenocarcinoma (T23) cells (50,000/well) were cultured either alone or in the presence of WT, OPN^−/−, or MyD88^−/− MCs (tumor cell:MC ratio 1:1; n = 3 per group) for 4 days. The heatmap reports the evaluation of IL1β, TNFα, and CCL3 as measured utilizing a multiple immunoassay kit (ProcartaPlex by Thermo Fisher). B, Flow cytometry evaluation of IL1R, and TNFRs (CD120a and CD120b) on ST4787 cells. C, ST4787 NEPC cells were treated with different concentrations (20, 50, or 100 ng/mL) of recombinant (r) IL1β or TNFα (n = 9 per group). Tumor cell proliferation was evaluated through trypan blue count after 4 days. D, T23 (adenocarcinoma) or ST4787 (NEPC) cells were treated with 50 ng/mL of rTNFα and analyzed as in C (n = 4 per group). Ε, Human adenocarcinoma (22Rv1; n = 3 per group) and NEPC (LASCPC-01; n = 6 per group) cells were treated with different concentrations (20, 50, or 100 ng/mL) of rTNFα and analyzed as in C. F, ST4787 cells (50,000/well) were cultured either alone or with WT, OPN^−/−, or TNFα^−/− MCs (tumor cell:MC ratio 1:1). Where indicated, a blocking antibody against TNFα (aTNFα; V1q clone, 10 μg/mL) was added. Tumor cell proliferation was analyzed as in C. N = 9 per group. G, Western blot analysis for cleaved-caspase-3 evaluation in ST4787 cells either unstimulated or co-cultured with WT, OPN^−/−, or TNFα^−/− MCs (ratio 1:1) for 15 minutes. The western blot was validated twice. H, Evaluation of TNFα production by intracellular flow cytometry in WT, OPN^−/−, or MyD88^−/− MCs, either unstimulated or cultured for 16 hours with ST787 or T23 cells (MC:tumor cell ratio 1:1; n = 6 per group). Gating strategy is reported in Supplementary Fig. S7D. I, WT MCs were cultured either alone or in the presence of ST4787 cells (MC:tumor cell ratio 1:1; n = 3 per group). Where indicated, an inhibitor of either NF-κB (BAY-11-7082, 10 µmol/L) or ERK1/2 (PD98059, 10 µmol/L; this compound inhibits MEK, which is upstream ERK1/2) pathway was added. After 16 hours, TNFα production by MC/9 was evaluated by flow cytometry as in H. Histogram reports the relative fluorescence intensity (RFI; ratio of the mean fluorescence intensity of the stained sample and of the fluorescence minus one control). J, MC/9-CTR, MC/9-OPNf, and MC/9-iOPN cells were cultured either alone or with ST4787 cells (MC:tumor cell ratio 1:1; n = 4 per group). After 16 hours, TNFα production by MC/9 was evaluated by flow cytometry as in H. Histogram reports the RFI (ratio of the mean fluorescence intensity of the stained sample and of the fluorescence minus one control), normalized to control condition (MC/9-CTR+ST4787 cells). C to F and H to J, All in vitro experiments were performed at least two times. Pools of all the biological replicates from all the independent experiments were performed and shown in histograms; statistical analysis was run considering all the samples. All histograms depict mean ± SD of biological replicates (represented by dots). One-way ANOVA followed by Tukey’s multiple comparison test (C, E, F, H, I, and J) or Mann–Whitney test (D) was used for the analysis of significance between samples. P values are reported as: *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Where the P value is not indicated, the comparison between groups is not statistically significant.

Figure 5.

MC-infiltrating prostates of TRAMP mice and human patients with focal incipient t-NEPC area express TNFα. A, Representative immunofluorescence for DAPI (blue), TNFα (red), and TRY (green) on tumor lesions from TRAMP mice subjected to surgical castration, showing either a TNFα-negative (top row) or a TNFα-positive MC (bottom row). B, Quantification of TNFα positive MCs in tumor samples from castrated TRAMP mice stained as in A (n = 4 ADENO and n = 3 focal NEPC). C, Representative immunofluorescence for DAPI (blue), TRY (red), and TNFα (green) on tumor samples collected from patients who received neoadjuvant ADT, showing either a TNFα-negative or a TNFα-positive MC. D, Quantification of TNFα positive MCs in human tumor samples, pretreated with neoadjuvant ADT, stained as in C (n = 2 ADENO and n = 4 focal NEPC). In B and D, the violin plots depict the median ± SD of biological replicates (represented by dots). Unpaired t test was used for statistical analysis: *, P < 0.05. Where the P value is not indicated, the comparison between groups is not statistically significant. E, Percentage of prostate lesions, scored as HGPIN, adenocarcinoma (ADENO), or NEPC in 25–30-week-old Kit^Wsh-TRAMP mice reconstituted i.p. with 5 × 10⁶ TNFα^−/− MCs (n = 16) at the age of 8 weeks. Control cohorts of TRAMP (n = 15) and Kit^Wsh- TRAMP (n = 15) mice are the same as shown in Fig. 1E. Numbers within bars represent the percentage of mice with the indicated histology. Fisher exact test was used for the analysis of contingency between different groups. P values are reported as: *, P < 0.05; **, P < 0.01; **, P < 0.001; ****, P < 0.0001.

Figure 6.

SDC1 is the NEPC cell-specific ligand that stimulates TNFα secretion by MCs. A, Flow cytometry evaluation of CD14 and SDC1 in T23 (adenocarcinoma) and ST4787 (NEPC) cells. B, HSPA2 and ANXA2 evaluation by western blot in T23 and ST4787 cells. C, ST4787 were silenced with two different siRNA specific for Cd14 (ST4787-siCD14-1 and ST4787-siCD14-2), Sdc1 (ST4787-siSDC1-1 and ST4787-siSDC1-2), Hspa2 (ST4787-siHSPA2-1 and ST4787-siHSPA2-2), Anxa2 (ST4787-siANXA2-1 and ST4787-siANXA2-2), or for a scramble control (ST4787-scramble). After 72 hours of transfection, ST4787 scramble or silenced cells were cultured either alone or with WT or OPN^−/− MCs (tumor cell:MC ratio 1:1; n = 6 per group). The growth rate of cancer cells was evaluated through trypan blue count. Cancer cells and MCs could be distinguished, thanks to their growth in adhesion or suspension, respectively. D, ST4787 scramble or silenced cells were cultured with WT or OPN^−/− MCs (ratio 1:1; n = 5 per group). After 16 hours the percentage of TNFα positive MCs was evaluated by flow cytometry, as in Fig. 4H. All in vitro experiments were performed at least two times. Pools of all the biological replicates from all the independent experiments were performed and shown in histograms; statistical analysis was run considering all the samples. All histograms depict mean ± SD of biological replicates (represented by dots). One-way ANOVA followed by Tukey’s multiple comparison test (D) or Mann–Whitney test (C) was used for the analysis of significance between samples. P values are reported as: *, P < 0.05; **, P < 0.01; ***, P < 0.001. Where the P value is not indicated, the comparison between groups is not statistically significant. E, Flow cytometry evaluation of SDC1 in adenocarcinoma (22Rv1 human) and NEPC (TC411K murine, NCI-H660 human) cells. F, Analyses on the s.c. RNA seq data set of the TPPRC mouse NEPC model (30). Interrogating the data set, we retrieved 20 different clusters of cells. The dot plot graph shows the expression of genes associated with adenocarcinoma, NEPC, MCs, and other genes of interest (Sdc1, Cd14, TNfa, Tlr2, and Tlr4). Dot size indicates the percentage of cells expressing the gene and dot color indicates the level of gene expression. The UMAPs show the distribution of cells across time of tumor development (top UMAP) and the distribution of the 20 different clusters (bottom UMAP).

Figure 7.

SDC1 expression levels are low in adenocarcinoma and increase in t-NEPC incipient areas in both murine and human prostates. A, Boxplots showing Sdc1 expression in the five phenotypic groups (ARPC, AR⁺/NE⁻; AMPC, AR⁺/NE⁺; ARLPC, AR^low/NE⁻; DNPC, AR⁻/NE⁻; and SNPC, AR⁻/NE⁺) derived from the RNA-seq data of the Labrecque human prostate cancer dataset (29). The RNA-seq data and the metadata containing the association between the samples and the phenotypic groups were downloaded from the GEO repository and the gene expression matrix was filtered for the 98 samples and for our gene of interest (Sdc1). The adjusted P value between the groups was calculated using the One-way ANOVA. B, Immunofluorescence for DAPI (blue), CGA (green), and SDC1 (red) in adenocarcinoma (ADENO) and in focal t-NEPC areas of TRAMP mice, either untreated or subjected to surgical castration, respectively. C, Immunofluorescence for DAPI (blue), CGA (green), and SDC1 (red) in adenocarcinoma (ADENO) and in focal t-NEPC areas from patients either untreated or who received neoadjuvant ADT, respectively.