Mutant p53 cancers reprogram macrophages to tumor supporting macrophages via exosomal miR-1246

Abstract

TP53 mutants (mutp53) are involved in the pathogenesis of most human cancers. Specific mutp53 proteins gain oncogenic functions (GOFs) distinct from the tumor suppressor activity of the wild-type protein. Tumor-associated macrophages (TAMs), a hallmark of solid tumors, are typically correlated with poor prognosis. Here, we report a non-cell-autonomous mechanism, whereby human mutp53 cancer cells reprogram macrophages to a tumor supportive and anti-inflammatory state. The colon cancer cells harboring GOF mutp53 selectively shed miR-1246-enriched exosomes. Uptake of these exosomes by neighboring macrophages triggers their miR-1246-dependent reprogramming into a cancer-promoting state. Mutp53-reprogammed TAMs favor anti-inflammatory immunosuppression with increased activity of TGF-β. These findings, associated with poor survival in colon cancer patients, strongly support a microenvironmental GOF role for mutp53 in actively engaging the immune system to promote cancer progression and metastasis.

Fig. 1

Carcinoma cells harboring mutp53 exert a non-cell-autonomous effect over macrophages. a Primary human monocytes were grown and differentiated towards three different lineages of macrophages (M0, M1, and M2) and co-cultured with an isogenic set of HCT116 cells differing by their p53 status (+/+ = WT p53, −/− = p53 null, mut = mutp53, p.R248W). RNA was extracted and subjected to qPCR analysis with primers specific to TNF-α and IL-10. Values were normalized for GAPDH mRNA in the same sample. b Primary monocytes were grown as in a and co-cultured with HT29 (mutp53- R273H) cells that underwent stable shRNA knock-down using scrambled oligo (ShCon) or specific to p53 (Shp53). TNF-α and IL-10 were analyzed as in a. c Co-cultured macrophages were seeded onto a cy-3-gelatin covered glass slide for 48 h and gelatin degradation rates were measured. d Co-cultured macrophages were harvested, stained with fluorescent antibodies against CD163 and CD206 and analyzed by flow cytometry. Relative intensities were compared with isotype controls. e Co-cultured macrophages were seeded in an eight-well chamber slide and incubated with fluorescent zymosan particles for 24 h, after which zymosan phagocytosis was evaluated. f,g Co-cultured macrophages were harvested and reseeded in an electrical-impedance monitoring chamber for 5 days to measure either migration (f) or invasion (g) properties. All experiments in this figure were repeated three times, error bars represent standard errors

Fig. 2

Exosomes shed from tumor cells are taken up by neighboring macrophages. a Exosomes were isolated from HCT116 cells harboring either WT, mutant (R248W), or no p53 as described in the Methods section. Isolations were either filtered (0.22 µm) or kept unfiltered during the procedure. Subsequently, isolations were lysed and subjected to western blot analysis with the indicated antibodies for exosomal markers. Calnexin served as a cellular contaminants marker. b,c Exosomes isolated from HCT116 cells underwent a nanoparticle tracking analysis (NTA) to determine exosomal size distribution (b) and concentration (c). The exosome samples were compared with cell-free medium that underwent similar isolation procedure (NEG). d Exosomes isolated from HCT116 cells were labeled using a Syto RNAselect dye before being incubated with macrophages for periods of 24 or 48 h. Accumulation of exosome uptake was captured in time-lapse movies. Macrophage nuclei are labeled with DAPI, plasma membranes are labeled with CellMask far-red. Bars = 25 µm. e RNA was extracted from HCT116-derived exosomes and its integrity and quality were tested using a bioanalyzer system before it was subjected to miRNA expression assay and normalized to the most abundant 100 miRs. f A representative comparison displaying greater than twofold changes in miRs between HT29 cells either knocked down for mutp53 (Shp53) or not (ShCon). g Changes in expression for four prominent miRs that were found to be significantly more abundant in mutp53 HCT116 and HT29 cells

Fig. 3

Exosomes shed from mutp53 tumor cells carry specific microRNA cargo. a Macrophages were co-cultured with HCT116 cells. RNA was extracted from the macrophages and subjected to qPCR analysis with primers specific to miR-1246. Values were normalized for RNU48 in the same sample. b Macrophages were grown in the presence of 10 µg exosomes isolated from HCT116 cells differing by their p53 status. miR-1246, miR-21, and miR-4454 levels were measured as in a. c M1 and M2 macrophages were transfected with LNA-miR-1246 mimic (mim) and compared with an equivalent control vector (con). RNA was extracted from the macrophages and subjected to qPCR analysis with primers specific to TNF-α and IL-10. d HCT116 mutp53 (R248W) cells were transfected with LNA-miR-1246 mimic (mim) or miR-1246 inhibitor (inhib.) and compared with an equivalent control vector (con). Forty-eight hours later, cells were co-cultured with M2 macrophages for an additional 3 days after which RNA was extracted from the macrophages and subjected to qPCR analysis with primers specific to TNF-α, CCL2, and IL-10. e Exosomes were isolated from HCT116 cells harboring mutp53 and separated from free proteins by a size-exclusion column. Macrophages were incubated with either pbs negative control (Con), the exosomes fraction (Exos) or the free proteins fraction (Prot.). RNA was extracted from the macrophages and subjected to qPCR analysis with primers specific to the indicated genes. Data are presented as fold changes compared with a PBS-treated control. f M2 macrophages were grown in the presence of 10 µg exosomes isolated from HCT116 cells differing by their p53 status for different time periods. RNA was extracted from the macrophages and subjected to qPCR analysis with primers specific to pre-miR-1246. All experiments in this figure were repeated three times, error bars represent standard errors

Fig. 4

miR-1246 is associated with mutp53 and plays a role in reprogramming TAMs. a,b M2 macrophages were co-cultured with either mutp53 or WT p53 HCT116 cells for 6 days. Subsequently, 10⁵ reprogrammed macrophages were mixed with 5 × 10⁵ fresh HCT116 WT cells (carrying a luciferase vector) and co-injected subcutaneously into NOD-SCID mice. Each group consisted of 10 or 11 animals. Tumor development was monitored weekly. c,d On day 56 of the experiment described in a, b, mice were killed and their liver and lungs were monitored for metastatic foci (c), and the number of organs observed with metastases were compared with a group of mice injected with HCT116 cells alone (d). e,f M2 macrophages were transfected with LNA-miR-1246 mimic and compared with an equivalent control vector. Three days later, the transfected macrophages were mixed with luciferase expressing HCT116 cells and co-injected subcutaneously to the back of NOD-SCID mice (HCT116 + M2 with control mimic, n = 5, HCT116 + M2 with miR1246 mimic, n = 5). Mice were monitored weekly for tumor growth using an IVIS imager (e) and luminescent fluxes were quantified (f). Error bars represent standard errors

Fig. 5

Mutp53 positively correlates with TAMs in CRC patients. a Strong epithelial immunostaining of p53 in the cancerous glands is correlated with diffuse and strong immunostaining of CD163 and CD206 in the surrounding stroma, in a representative human CRC case carrying mutp53 compared with a WT p53 case. Bars represent 100 µm. c Diffuse and strong CD206 immunopositivity in the invasive front of a mutp53 human CRC case as compared with a WT p53 case. The left panels display tumors at low magnification where the invasive front is marked with a dashed line. The boxed area is displayed in the right panels at a higher magnification. Bars represent 200 µm in the left panels and 100 µm in the right panels. b,d Staining abundance and intensity were calculated for all 42 specimens as described in the Methods section. Cases were grouped by p53 typing outcome—missense p53 mutations were defined as “mutp53 group”, while insertions, deletions and no mutations were defined as “WT+ indels group”. The analysis was conducted either for the entire cancerous tissue (tumor, b) or focused in tumor boundaries (invasive front, d). e RNA from frozen tissue samples of 29 CRC cases was extracted and subjected to mRNA expression microarray analysis. The heatmap depicts the expression patterns of the genes whose abundance was significantly upregulated or downregulated between the “WT + indels” (n = 13) group and the “mutp53” group (n = 16). A specific molecular signature of genes upregulated in mutp53 CRC cases is detailed. f Survival curve of CRC patients divided by TP53 status comparing between the WT+ indels combined with non-GOF p53 mutants (WT + indels + other mut) with patients carrying tumors with GOF mutations (GOF mut). Error bars represent standard errors

Fig. 6

Mutp53 correlates with miR-1246 in CRC patients. a MicroRNA microarrays were used to profile microRNA expression levels in 27 WT p53 cases and 28 mutp53 cases. Complete table with p values is presented in the Supplementary Information (Supplementary Table 4). b Strong epithelial (middle panel) and stromal (lower panel) miR-1246 hybridization signal in mutp53 tumor compared with WT p53 tumor. Bars = 100 µm. c Co-detection of miR-1246 and CD206 in the stroma of sporadic colorectal carcinomas harboring mutp53 and WT p53. White arrowheads demonstrate miR-1246+/CD206+ cells. Scale bars: 200 μm (low magnification); 25 μm (high magnification). DAPI was used for nuclear stain (blue). d Exosomes were isolated from plasma of CRC patients categorized as “WT + indels” (n = 11) or “mutp53” (n = 25). RNA was extracted from the exosomes and miR-1246, miR-21, and miR-4454 levels were measured using qPCR. Quantification and normalization were performed using NTA and miR-454. Error bars represent standard errors

Fig. 7

Increased TGF-β signaling and immunosuppression in mutp53 CRC patients. a Strong epithelial immunostaining of p53 is correlated with strong non-epithelial immunostaining of FOXP3 in a representative human CRC case carrying mutp53 and compared with a WT p53 case. b FOXP3 staining abundance and intensity was calculated for all specimens as described in the Methods section. c,d A general inflammatory gene set (c) and hallmark TGF-β signaling gene set (d) produced by a GSEA analysis of mRNA expression levels of the GOF p53 mutants group of CRC tumors compared with the rest of the missense p53 mutants group (other mut). Specific genes significantly upregulated in the GOF group are displayed in red while inflammatory genes downregulated in the GOF group are in blue. NES normalized enrichment score, FDR false discovery rate. GOF mutants refers to the following p53 positions: R245, R248, R175, R273, and R282. e The proposed molecular model by which mutp53 facilitates the tumor microenvironment. Colon tumor cell acquires a mutation in TP53 yielding an increased release of exosomes containing miR-1246. Such exosomes are received by neighboring macrophages that undergo a phenotypic shift resulting with enhanced secretion of anti-inflammatoy cytokines (which also recruit immunosuppressive T-regulatory cells) and epithelial−mesenchymal transition (EMT) promoting factors contributing to tumorigenesis and eventually to poor prognosis. Error bars represent standard errors