(A) HT29 cells (1×10⁶) and washed human platelets (100×10⁶) were co-cultured for 20 h using a trans-well (pore size, 0.4 μm) and COX-2 levels were assessed in HT29 cells by Western blot [18]; data are expressed as mean±S.E.M. (n=7), *P<0.05 compared with HT-29. (B and C) Effect of human platelet releasate on HT29 COX-2 expression; HT29 cells were cultured up to 20 h in the absence and in the presence of platelet releasate (obtained by collecting the supernatant of washed human platelets cultured for 20 h, then it was centrifuged at 750 g for 15 min) which contains soluble mediators and vesicles, i.e. MPs and exosomes; at different time-points COX-2 protein levels (B) and mRNA levels (C) were analysed by Western blot (normalized to β-actin levels) and quantitative real-time PCR (qPCR) (normalized to GAPDH) respectively, as previously described [18]; data are expressed as mean±S.E.M. (n=3); *P<0.05 and **P<0.01 compared with HT-29.(D)HuR localization was assessed by confocal microscopy analysis in HT29 cells alone (HT) or incubated with platelet releasate for 20 h; immunostaining of HuR is shown in green, GAPDH (cytoplasmic protein) in red and DAPI (nuclear marker) in blue; merge images are shown (merge); ratios between the pixel sum of HuR staining in cytoplasm and pixel sum of HuR staining in the nucleus were calculated using LAS AF software, 2.2.1, as described in [18]. Data are expressed as mean±S.E.M. (n=14); ns, not significant.