P300, A New Player in Mechanosensitivity and Activation of Cancer-Associated Fibroblasts

The physical properties of the extracellular environment such as matrix composition and mechanics are key factors that determine cancer cell behavior.^1,2 Stiffening matrix is a result of a complex multidirectional interplay between stromal cells such as cancer- associated fibroblasts (CAF), infiltrating immune cells, and cancer cells.

Fiber realignment and strain stiffening enable cancer cells to contract and break basement membranes, withdraw adhesions, and form invadosomes.³ Cells can “sense” their environment, and respond by complex mechanotransductive pathways.⁴ Extracellular physical signals are known to affect integrin subunit assembly and the recruitment of several cellular constituents such as focal adhesion kinase and associated Src, eventually leading to an increasing number and activity of invadosomes. Downstream activation of the RhoA GTPase ultimately controls actin cytoskeletal dynamics and invadopodia formation. Its effector, ROCK, is able to further increase matrix stiffness via a β-catenin–mediated transcriptional program. Matrix cross-linking also has a significant role in rigidity, and it was shown that the combined inhibition of lysyl oxidase 2 (LOXL2) and the transforming growth factor β receptor I had a potent inhibitory effect on collagen accumulation, softening the matrix.⁵

The mechanical environment of tumor cells can strongly influence the response to chemotherapy or radiation therapy, and clinical outcomes. Integrin β 1/focal adhesion kinase activation was central to hepatocyte growth factor- dependent ERK, AKT, STAT3, and cyclin D1 signals linking matrix stiffness to increased proliferative responses and chemoresistance.⁶ Changes in stiffness were observed to induce β 1-integrin–dependent transcriptional programs that activate migration and angiogenesis-related genes. These signatures were correlated with worse survival in 9 different tumor types.⁷ The matrix architecture and the mode of fiber alignment could also have an impact on out- comes. In a recent study, perpendicular fiber alignment (tumor-associated collagen signature-3) was a predictor of overall survival in breast cancer.⁸

CAFs are key to controlling matrix deposition and dynamics. CAFs can elicit signals that align the fibronectin matrix, allowing the directional migration of cancer cells, and aligned fibronectin is a prominent feature of invasion sites in prostate and pancreatic carcinomas.⁹ CAFs also produce an aberrant matrix, enabling invasion, with an increase in collagens IV, VI, VII, X, decorin, and laminin subunits.^10,11

Thus, the question as to how CAFs become activated in the tumor environment is very relevant. In the liver, a common site of hematogenous metastases, hepatic stellate cells (HSCs) are thought to give rise to smooth muscle a-actin–positive CAFs. How does the stiffening matrix during metastatic invasion in the liver enable trans- differentiation of quiescent HSC into CAF/MF phenotypes?

The study by Dou et al¹² highlights the novel role of the p300 acetyltransferase in durosensitivity and the epigenetic regulation of HSC plasticity. Increasing matrix rigidity induced the RhoA-Akt-p300 pathway, resulting in p300 nuclear translocation and the expression of smooth muscle a-actin and a panel of tumor-promoting factors, including CXCL12, CTGF, and PDGF A and B. P300 is a ubiquitous transcriptional regulator that plays a role in histone acetylation and chromatin remodeling, and an increase in p300 activity has been linked to a more aggressive phenotype in several cancers. By atomic force microscopy, the authors demonstrated that colorectal metastases in the liver were “stiffer” than the surrounding parenchyma, and that the expression of p300 was increased in both patient specimens and in mice. In quiescent HSC on soft matrix, p300 levels remained low owing to proteasomal degradation, whereas on stiff hydrogel matrix, both the protein levels and phosphorylation were induced, and p300 translocation to the nucleus was observed. RhoA, as described, is a critical “sensor” for increased matrix stiffness targeting Akt, leading to its phosphorylation. Because a stiff extracellular matrix could cause increased CAF migratory activity, it would be interesting to see whether the RhoA/p300 pathways could indeed be part of invadosome activation, and play a role in migration/invasiveness.

The authors confirmed their findings in elegant in vivo studies. They used a tumor cell/conditioned medium co-implantation model where the deletion of p300 from HSC inhibited tumor growth, whereas seeding HSC grown on a hard matrix, enhanced it. Using a conditional knockout model deleting p300 from collagen-producing cells, both fibrosis (CCl₄) and metastatic growth were decreased. Furthermore, C646, a selective small molecule inhibitor of p300, decreased colorectal metastases in mice. Intriguingly, p300 was induced not only in cancer-associated myofibroblasts, but also in metastatic tumor cells. This finding lends further credence to prior studies where metastatic progression was associated with increasing p300 activity, tumor cell survival, and proliferation.¹³

Recent studies demonstrated the role of Yes-associated protein in communicating mechanosensitive signals.¹⁴ Because p300 has been involved in Yes-associated protein acetylation, it would be worth exploring whether these pathways converge and p300 could indeed be a master regulator of mechanosensitive signals.

Clinically, increasing liver stiffness is associated with a higher risk of hepatocellular carcinoma, and an increased proliferative response and chemoresistance at a molecular level. Thus, p300 activation could be a molecular “hot spot” in durosensitivity as it regulates a positive feedback loop between cancer cells and CAFs. Targeting p300 may open novel treatment approaches that could be used in combination with conventional chemotherapies.

See “P300 acetyltransferase mediates stiffness-induced activation of hepatic stellate cells into tumor-promoting myofibroblasts,” by Dou C, Liu Z, Tu K, et al, on page 2209.