Sirt1 and cell migration

Abstract

Sirt1 is a type III histone deacetylase implicated in a wide range of physiological and pathophysiological roles. Acting though a myriad of non-histone substrates, Sirt1 modulates transcriptional regulation of energy metabolism and stress response, with important consequences on cell survival and a myriad of human pathologies. Sirt1 has an apparent (albeit context- and tissue type-dependent) role in tumorigenesis, acting particularly through its deacetylation of tumor suppressor gene products such as p53 and Rb. Recent works have now revealed that cortactin, an F-actin binding protein with established roles in protrusive actin dynamics, is a Sirt1 substrate. Cortactin could be acetylated by the acetyltransferase p300, and its deacetylation by Sirt1, either directly or indirectly, retards cell migration. In conjunction with deacetylation of other oncogenic targets, Sirt1’s modulation of cell migration and invasion may be an important additional aspect of its tumorigenic activity.

Introduction

The sirtuin family proteins are nicotinamide adenine dinucleotide (NAD)-dependent class III histone deacetylases conserved in eukaryotes.^1,2 The mammalian genome harbors seven sirtuin paralogues, and Sirt1 has the highest homology to Saccharomyces cerevisiae Sir2, a gene famous for its putative role in lifespan extension in several model organisms.³ Sirt1 has been linked to multiple physiological functions and pathological roles. It deacetylate histones, but it also has a wide range of non-histone sustrates. Sirt1’s activity on peroxisome proliferator-activated receptor-γ (PPARγ)⁴ and its transcriptional co-activator PPARγ coactivator-1α (PGC-1α)^5,6 regulates metabolic homeostasis in energy intensive tissues, providing adaptive transcriptional changes to nutrient availability and environmental stimuli. Sirt1’s deacetylation of members of the forkhead box class O (FOXO) family⁷ and nuclear factor κB (NFκB)⁸ regulates key aspects of cellular stress response and survival.

Another prominent activity of Sirt1 is one associated with cancer. Sirt1 may either be oncogenic or tumor-suppressive, depending on the type of malignancy and the context of analysis.^9,10 Earlier studies have shown that Sirt1 deacetylates Lys382 of p53, thereby repressing its transcriptional activity and attenuating p53-mediated cell cycle arrest and apoptosis.¹¹ Sirt1 expression is in fact negatively regulated by several transcription factors with tumor suppressor activities, including p53, Hypermethylated in cancer 1 (Hic1),¹² and Deleted in breast cancer 1 (Dbc1).^13,14 Human cancers manifesting an inactivation of any of these tumor suppressor genes could therefore enhance Sirt1 expression. Elevated Sirt1 expression is indeed found in several human cancers,¹⁰ and Sirt1 silencing could often sensitize cells to apoptosis.^15,16 Sirt1 also binds to the cell cycle regulator E2F1 and could inhibit its apoptotic activity during DNA damage response.¹⁷ Furthermore, Sirt1 activity may enhance cancer cell resistance to chemotherapy via its induction of the multidrug resistance gene Mdr1.¹⁸

On the other hand, emerging evidence also suggests that Sirt1 could be tumor suppressive. Sirt1’s deacetylation of lys310 of the RelA/p65 subunit of NFκB could sensitized cells to tumor necrosis factor α (TNFα)-induced apoptosis,¹⁹ and its deacetylation of β-catenin keeps the latter oncogenic factor in the cytoplasm, suppressing its transcriptional activities that drive cell proliferation in colon cancer.²⁰ In fact, Sirt1 appears to exert pleiotropic effects in colon cancer.²¹ While it's silencing accelerates colon carcinoma HCT116 xenograft formation in rodents (and its overexpression inhibits tumor formation), Sirt1 inhibition also sensitizes cells to apoptosis by chemotherapeutic drugs. Sirt1 levels are in fact lower in some human cancers, such as the tumor supressor breast cancer 1(Brca1)-associated breast cancers.^22,23 Sirt1 heterozygosity appears to predispose p53 heterozygous mice to tomorigenesis,²² and it negatively regulates the expression of an anti-apoptotic protein, survivin.²³ Most studies on the association of Sirt1 with cancer had focused on cell proliferation and apoptosis. It is unclear if Sirt1 activity may modulate cancer cell motility and migration.

Cancer cells often had enhanced migratory and invasive phenotypes, with membrane protrusions forming more frequently and readily in response to external cues. Migration and invasion require dynamic changes in the polymerization of actin filaments. In cancer cells, several molecules that regulate actin dynamics could be upregulated.²⁴ Cortactin is an F-actin binding protein with prominent roles in actin cytoskeletal network assembly, particularly in membrane protrusions associated with cell invasion activities.^25,26 Cortactin plays critical roles in the assembly and maturation of the invadopodia, which are invasive membrane protrusion of certain cancer cells. In these structures, cortactin’s function is known to be regulated by phosphorylation by Src kinase and extracellular signal regulated kinase (ERK). Cortactin’s phosphorylation releases its inhibitory interactions with the actin severing protein cofilin, and the latter could then sever actin filaments to create barbed ends at the invadopodia to support actin-related proteins (Arp) 2/3 complex-dependent actin polymerization.²⁷ Zhang and colleagues have earlier demonstrated that cortactin is also subjected to another type of post-translational modification, namely acetylation.²⁸ The authors showed that deacetylatio of cortactin by histone deacetylase 6 (HDAC6) influences actin-dependent cell motility.²⁸ As discussed below, Zhang and colleagues have now showed that cortactin could also be deacetylated by Sirt1,²⁹ and this hypoacetylation of cortactin is associated with an increase in cell motility.

Sirt1 Interacts with and Deacetylates Cortactin

The decetylation reaction catalyzed by sirtuins is coupled to cleavage of NAD into 1-O-acetyl-ADP ribose and nicotinamide, and the latter is a rather specific inhibitor of sirtuins’ deacetylase activity (but not other HDACs). Zhang and colleagues have previously observed that nicotinamide treatment increased the level of cortactin deacetylation. The authors surmised that other than HDAC6, cortactin may also be deacetylated by a member of the sirtuin family. To address this possibility, the authors performed an affinity pull-down with recombinant cortactin fused to glutathione S-trasnferase (GST) with lysates of 293T cells transfected with the Flag-tagged forms of all seven mammalian sirtuins. Of all the mammalian sirtuins, only Sirt1 detectably associates with GST-cortactin. Overexpressed cortactin co-imunoprecipitated Sirt1, but not with Sirt2. Importantly, reciprocal co-immunoprecipitation of endogenous cortactin and Sirt1 could be demonstrated using lysates of S13, an ovarian cancer line expressing high levels of both proteins. Molecular dissection of the interactions domains indicated that Sirt1, like HDAC6, binds to cortactin’s repeat region.

The authors have previously shown that the p300/CBP-associated factor (P/CAF) co-activator complex, which is an acetyltransferase, is responsible for the acetylation of lysine residues in cortactin’s repeat region. P/CAF is the major, if not the sole acetyl transferase responsible for cortactin acetylation in vivo. Transfected p300, but not other acetyltransferases such as Tip60 and HBO, efficiently acetylates co-transfected cortactin. The acetylation level of endogenous cortactin is also considerably lower in p300^−/− mouse embryonic fibroblasts compared to wild type cells. On the other hand, co-transfected Sirt1 (but not the deacetylase-dead H363Y mutant) significantly reduced acetylation of cortactin. Furthermore, recombinant Sirt1 was also able to deacetylate cortactin in cell lysates in the presence of NAD⁺. The highly specific Sirt1 inhibitor EX-527 effectively abrogated cortactin decaetylation in a dose-dependent manner.

These results suggest that Sirt1 could bind to cortactin, and that cytoplasmic cortactin is likely a direct substrate of Sirt1. Although Sirt1 was first described as a nuclear protein, its presence and activity in the cytoplasm has been demonstrated in multiple cell types, including cancerous tissues. The authors’ immunohistochemical analysis indicated that Sirt1 is present in both nuclei and cytoplasm of ovarian cancer tissues. As Sirt1 also deacetylates p300 (which inhibits the latter’s acetyltransferase activity), Sirt1 likely modulates cortactin acetylation status both directly (by deacetylating it) and indirectly (by reducing its acetylation by p300).

Cortactin Acetylation Status and Cancer Cell Migration

What are the physiological significance of cortactin acetylation/deacetylation, and the role of Sirt1 in this regard The authors’ previous work indicated that hyperacetylation of cortactin inhibited its interaction with F-actin, prevented its translocation to the cell periphery, and impaired cell motility. The sirtuin inhibitor nicotinamide slowed the migration of ovarian cancer cells (OV2008) in a wound-healing assay. Mouse embryonic fibroblasts from Sirt1-deleted mouse have hyperacetylated cortactin, and are less motile than wild type fibroblasts—a phenotype that could be reversed by expression of exogenous Sirt1 or a decetylation mimetic cortactin mutant. Expression of wild type or deacetylation mimetic mutant of cortactin also increased the mobility of ovarian and breast cancer cell lines. These results are all in support of the notion that decaetylation of cortactin increases cell migration.

In a survey of the acetylation status of cortactin in breast cancer tissues compared to non-cancerous benign counterparts, the authors noted that cortactin expression is increased in many of the cancerous tissues. However, corresponding increases in the acetylated pool of cortactin were not apparent. When normalized against total cortactin expressed, cortaction in five of eight of the cancerous tissues were hypoacetylated compared to benign controls. Of these five samples, Sirt1 is overexpressed in four of them, and its expression is therefore inversely correlated with cortactin acetylation. These results, taken together, suggest that elevation of Sirt1 levels in cancer tissues could result in decreased cortactin acetylation, which in turn promotes cancer cell motility.

Sirt1’s Expanded Rolein Tumorigenesis

The work of Zhang and colleagues points to a novel aspect of Sirt1’s role in human malignancy, namely cancer cell migration and invasion. Many important and interesting issues remained unresolved. Cortactin deacetylation by both HDAC6 and Sirt1 likely affects its interaction with F-actin, as well as actin-modifying factors such as cofilin.²⁷ Whether the acetylation of cortactin affects its other important post-translational modification, namely phosphorylation, remains to be investigated. Another question is how Sirt1 may act in conjunction with HDAC6 in terms of cortactin deacetylation. Both deacetylases appear to bind to and act on residues within the repeat region of cortactin, and presumably the extent to which each deacetylase is involved in modulating cortactin deacetylation in a given cell or tissue would depend on their relative levels and activity. Both HDAC6 and Sirt1 are fairly ubiquitously expressed, but their relative activities in different types of tissue malignancies are likely to be different.

Another point that is of interest and needs to be further clarified is the role of another sirtuin, Sirt2, in cortactin deacetylation. The author have previously shown that Sirt2 silencing also elevated cortactin acetylation.²⁸ Interestingly, both HDAC6 and Sirt2 are also tubulin deacetylases that could modulate microtubule-dependent cell motility.^30,31 These two proteins could also interact with each other in vivo.³¹ Although whether Sirt1 has deacetylation activity on tubulin is unknown, the HDACs could potentially enhance cell migration and invasion via coordinated modulation of actin filaments and microtubule dynamics. This coordination may have physiological or developmental roles, and its aberrant regulation would contribute to cancer invasion and metastasis.

Finally, it is worth exploring to what extent does Sirt1-mediated deacetylation promote migration and invasion of the myriad of human cancers. In other words, would Sirt1 be an effective clinical target in attenuating cancer cell invasion and metastasis Sirt1 inhibitors (and activators) are already being tested as cancer therapeutics, and their effects on cancer invasion and metastasis would be an additional important clinical efficacy parameter to watch out for.