Research Paper Volume 8, Issue 8 pp 1608—1624

A miR-335/COX-2/PTEN axis regulates the secretory phenotype of senescent cancer-associated fibroblasts


Figure 6. A miRNA/PTEN mediated regulatory SASP exists in senescent CAF of OSCC. In absence of cisplatin treatment SA-β-gal activity is minimal in CAF obtained from genetically stable tumours; CAF/GS-OSCC (BICR69 and BICR70) (n=2). CAF derived from genetically unstable tumours; CAF/GU-OSCC (BICR63 and BICR18) are senescent and are positive for SA-β-gal activity irrespective of passage number (n=3). Cisplatin induces SA-β-gal activity in non-senescent CAF/GS-OSCC and amplifies its activity in the senescent CAF/GU-OSCC, n=3 (A). Senescent CAF/GU-OSCC (n=3) secreted more MCP-1 (B) and IL-6 (C) than non-senescent CAF/GS-OSCC (n=2) and normal (n=4) control fibroblasts. Senescent CAF/GU-OSCC stimulated migration of both D20 (D) and H357 cells (E) in vitro (n=3). Invasion of H357 cells was also increased in presence of senescent-CAF in organotypic models (BICR63, n=3) (F); immunohistochemistry for p16INK4A indicated senescence in fibroblasts incorporated in the organotypic model. qRT-PCR showed miR-335 (G) levels are elevated in CAF of both GS-OSCC (n=2) and GU-OSCC (n=4) tumours. Western blot showed PTEN expression was reduced in senescent CAF/GU-OSCC (n=4) compared to non-senescent CAF/GS-OSCC (n=2) and normal fibroblasts (n=6) (H-I). All experiments were performed independently as indicated by n and with technical repeats. The bars represent mean ± SEM or mean ± STDEV (G). *p<0.05, by one-way ANOVA with post-hoc correction by Holm-Sidak method (C, G, I) and Dunn's method (B), two-way ANOVA with post-hoc corrections by Bonferroni t-test (D-E). (See supplementary Figure S6-S7).