p53 further extends its reach

This article refers to:

p53 is a central sensor of severe cellular insults, including genotoxic, oxidative and oncogenic stress. Depending on the context, p53 cell-autonomously orchestrates responses of temporary cell cycle arrest coupled to DNA repair or permanent senescence, or it induces apoptotic, autophagic and necrotic forms of cell death.^1-3 However, little is known about the non-cell autonomous effects of p53. Recent findings from Lujambio et al. uncover a new paracrine oncosuppressor function of p53 in chronic liver damage, a cytokine-dependent process that promotes an anti-tumorigenic microenvironment (at least in part) by modulating immunosurveillance.⁴

Upon liver injury, stromal hepatic stellate cells (HSC) become activated and drive the fibrotic scarring of the liver parenchyma. Chronic liver injury coupled to smoldering fibrosis can result in cirrhosis, a main risk factor for hepatocellular carcinoma (HCC). The same authors had previously shown that p53 mediates a senescence program in HSCs that limits fibrosis and maintains tissue homeostasis. Such a p53-dependent senescence program is executed in cooperation with NFκB and induces the so-called senescence-associated secretory phenotype (SASP). The SASP is marked by the downregulation of extracellular matrix (ECM) proteins including collagen, by the upregulation of ECM-degrading enzymes and by production of immunomodulatory factors and inflammatory cytokines.^5,6

The fibrosis-inhibitory effect of damage-activated p53 in HSCs was confirmed by Lujambio et al. in different stages of liver damage, including extended liver cirrhosis. Moreover, the authors demonstrated that the HSC-specific ablation of Trp53 (the murine gene coding for p53) exacerbates the malignant transformation of hepatocytes as induced by carbon tetrachloride (CCl₄) and diethylnitrosamine (DEN). This strongly suggests that a non-cell autonomous mechanism of tumor suppression is at work in p53-competent livers, operating via a p53-mediated senescence/SASP response. Lujambio et al. also showed that genes associated with cell proliferation and ECM are downregulated by p53, while genes involved in cell cycle arrest, exocytosis/vesicle transport and activation of immune signaling are upregulated. By analyzing the medium conditioned by p53-activated (senescent) or p53-ablated (proliferative) HCSs, the authors could determine the factors preferentially secreted in these two conditions, namely, interleukin (IL)-6, intercellular adhesion molecule 1 (ICAM1) and interferon γ (IFNγ) by senescent HSCs, and IL-3, IL-4, and IL-5 by proliferative HSCs. Interestingly, the most consistent changes upon p53 depletion were observed in the levels of cytokines known to promote the activation and differentiation of macrophages (i.e., IL-4, IFNγ).

Macrophages are known for their ability to modulate liver fibrosis. Tumor-associated macrophages (TAMs), which are the most abundant immune cells within the tumor microenvironment, exert a major influence on tumorigenesis.⁷ TAMs can exist in two distinct phenotypes, M1 and M2, which are quite plastic and are mainly defined by the associated gene expression profile. Besides exerting a broad anti-pathogen activity, M1 macrophages are activated by lipopolysaccharide (LPS) and IFNγ and exert three main antitumor functions: (1) they secrete immunostimulatory cytokines including tumor necrosis factor α (TNFα), IL-1, IL-12 and IL-23; (2) they exert prominent antigen-presenting functions; and (3) they can direct kill malignant cells, at least in selected circumstances. In contrast, pro-tumorigenic M2 macrophages, upon activation by IL-4, IL-10 and IL-13, downregulate MHC expression while secreting immunosuppressive cytokines such as IL-10, arginase, growth factors including transforming growth factor β (TGFβ) and platelet-derived growth factor (PDGF) as well as angiogenic mediators like vascular endothelial growth factor (VEGF). Taken together, these bioactive molecules promote tumor growth, angiogenesis, tissue remodeling and metastasis.^8-10 Since the M2 phenotype often predominates among TAMs, elevated TAM levels generally correlate with poor prognosis.¹¹

Intriguingly, Lujambio and colleagues showed that liver-derived macrophages cultured in the medium conditioned by proliferating HSCs predominantly exhibit an M2 phenotype. Conversely, when cultured in the medium conditioned by senescent HSCs, liver-derived macrophages are skewed toward an M1 phenotype. Moreover, in co-culture experiments, senescent HSCs were destroyed by (presumably M1) macrophages (generating an anti-fibrogenic environment), while proliferating HSCs were not. Even more interesting, the incubation of (presumably M2) macrophages with the medium conditioned by proliferative HSCs significantly increased the proliferation of co-cultured premalignant hepatocytes. This suggests that the ablation of Trp53 stimulates a proliferative and supposedly tumorigenic microenvironment (at least in part) as it skews macrophage differentiation toward an M2 phenotype. The switch to an M1 macrophage phenotype upon p53 activation was also observed in CCl₄-damaged livers in vivo. In contrast, Trp53-ablated livers showed a severe overall reduction in the abundance of macrophages, with a relative increase in M2 polarization. As for the molecular mechanism underpinning the p53-induced SASP of HSCs, the authors confirm previous data pointing to a cooperativity between p53 and NFκB. Indeed, the knockdown of NFκB phenocopies Trp53 ablation in this setting.

It is known that hepatic carcinogenesis is intimately linked to inflammation¹² and that macrophage polarization, especially in relationship with TAM-derived cytokines, chemokines and growth factors, plays an important role in HCC progression.¹³ However, the molecular links that connect inflammation and HCC tumorigenesis are not fully understood. For example, IL-6 and TNFα favor tumor proliferation. TGFβ, a critical growth factor for hepatic carcinogenesis, is secreted by M2 macrophages. Increased signaling via the TGFβ-SMAD3 axis promotes HCC progression in mouse models by inducing the epithelial-to-mesenchymal transition (EMT), which directly promotes invasion and metastasis.^14-17 Likewise, high levels of tumor-infiltrating M2 macrophages are associated with poor prognosis in HCC patients.¹⁸ On the other hand, IFNγ (which favors the accumulation of M1 macrophage and was shown to be downregulated by p53 depletion by Lujambio et al.)⁴ has previously been involved in HCC progression. Nevertheless, IFNγ upregulation has also been found to correlate with improved prognosis among HCC patients.¹⁹ One of the critical components in the inflammation/tumorigenesis connection that eventually drives HCC is NFκB. NFκB can exert both pro- and anti-tumorigenic effects, depending on the tumor type and stage at which it is activated. Thus, NFκB activation in liver parenchymal cells mediates oncosuppressive functions (analogous to the findings of Lujambio et al. in HSCs),⁴ while it exerts a pro-tumorigenic effect in TAMs.^20-22

Overall, the novel insights into the regulation of macrophage polarization by the p53-SASP axis in HSCs provided by Lujambio et al. represent a significant step forward toward understanding of HCC initiation and progression. These findings extend our knowledge of the well-established cell-autonomous p53 oncosuppressive network to a minor, yet crucial, organ-specific mesenchymal cell type that acts non-cell autonomously to establish an anticancer immune microenvironment and restrain fibrogenesis. The results by Lujambio and colleagues also generated several interesting questions. For example, how does exactly p53 control NF-κB activation in HSCs? How the p53-dependent polarization of macrophages affects the tumor environment and thus influence tumor progression also warrants further exploration. Most importantly, does this mechanism apply to hepatic carcinogenesis in humans? How frequent are mutational or regulatory disruptions of the p53 program in viral, chemical or alcoholic hepatitis-driven human HCCs? If confirmed, this novel role of p53 in the remodeling of the tissue microenvironment may pave novel avenues for the development of anticancer drugs.