HATs in Smooth Muscle Cell Phenotype Switching: Redundant No Longer

Despite an identical genome in every cell, differential gene expression determines call fate: cardiomyocyte versus leukocyte, vascular smooth muscle versus endothelial cell. For some cells, fate is permanent, whereas others, such as smooth muscle cells (SMC), retain the ability to dedifferentiate and take on different phenotypes. In vascular smooth muscle, the switch from a contractile phenotype, that controls blood vessel diameter and tone, to proliferative, inflammatory or other phenotypes, is necessary for vascular repair. However, when dysregulated, SMC phenotype switching contributes to cardiovascular disorders including atherosclerosis, in-stent restenosis and coronary allograft vasculopathy. Whereas substantial advances have been made in understanding SMC phenotype switching, we have yet to successfully target SMCs phenotype for therapeutic benefit. In this issue of Circulation, Chakraborty and colleagues¹ report opposing roles for two proteins that control vascular smooth muscle cell phenotype, with important therapeutic implications.

Understanding gene transcriptional regulation to determine cell fate and function has been a long and continuously evolving journey. Almost 40 years ago, transcription factors were discovered that bind to specific sites on DNA to regulate target gene transcription.² Some were termed master regulators of specific cell fates, like MyoD for skeletal muscle or serum response factor (SRF) for vascular smooth muscle. However, how a transcription factor can regulate distinct genes under different circumstances in response to physiologic and pathologic stimuli remains unclear. In the early 1990’s, transcriptional co-activators were identified that are recruited to DNA by transcription factors to mediate gene activation. Many coactivators were later found to function by inducing post-translational modifications of DNA or histone proteins, thereby modulating accessibility of genes for transcription and contributing to differential gene expression based on cell type or conditions.

CREB-binding protein (CBP) is one such coactivator with histone acetyltransferase (HAT) activity that was identified in 1994.³ The highly homologous protein p300 was identified soon after⁴. Over the next decade, CBP and p300 were found to coactivate many other TFs and to function as HATs.⁵ Hence CBP and p300 were considered as integrators of external signals to modulate gene transcription by inducing histone modifications. For over 2 decades, p300 and CBP have been presumed to be redundant, given their high degree of homology and common functions. Over 3500 published manuscripts refer to them collectively as “CBP/p300” (or vice versa). Yet, rarely in nature are two distinct proteins maintained evolutionarily for identical function. Thus, understanding potential differences is critical if one hopes to target CBP/p300 therapeutically for cardiovascular disease.

Chakraborty et al now report opposing roles for CBP and p300 in vascular SMC phenotype regulation and determine a novel mechanism for their distinct impacts on SMC fate with potentially important therapeutic implications.¹ Chakraborty et al first show that in settings in which SMCs dedifferentiate to the proliferative phenotype, as in the murine vascular injury model and in human hearts with coronary allograft vasculopathy, CBP and the acetylation of lysine 27 on histone H3 (H3K27ac) are both increased, while p300 and H3K9ac are decreased. Using traditional drivers of SMC phenotype switching and loss-of-function approaches in primary cultured human SMCs in vitro, they show that p300 is necessary to promote the differentiated, contractile phenotype, and CBP promotes dedifferentiation into the proliferative and migratory phenotype seen in disease states. This is confirmed in vivo using novel mouse models with CBP or p300 specifically deleted from SMC: loss of p300 promotes neointima formation and loss of CBP attenuates the SMC response to injury. Mechanistically, Chakraborty et al show that p300 promotes the differentiated SMC phenotype by enhancing histone acetylation of SRF binding sites in SMC contractile genes and that CBP does the opposite. Importantly, they go on to show that this difference is mediated by distinct interactions with TET methylcytosine dioxygenase 2 (TET2), a DNA-modifying enzyme that the Martin lab has previously shown promotes SMC differentiation in allograft vasculopathy.⁶ Here they demonstrate that p300 (but not CBP) interacts with TET2 by Co-IP, upregulates TET2 expression, and that p300 is necessary for DNA methylation by TET2 of contractile gene loci. Conversely, CBP inhibits TET2 expression, and hence p300 activity, while recruiting histone deacetylases to contractile genes.

Whereas CBP and p300 have generally been held to function interchangeably, this is not the first demonstration of unique roles for these highly homologous histone acetyltransferases. In the late 1990s, CBP and p300 were found to have distinct functions in differentiation of an embryonal carcinoma cell line, with p300 specifically promoting the differentiated phenotype.⁷ More recently, RNA profiling and chromatin immunoprecipitation (ChIP)-sequencing revealed that CBP and p300 regulate distinct genes in differentiating skeletal myoblasts⁸ and modulate histone acetylation at distinct sites across the genome of mouse embryonic stem cells ⁹. However, these prior studies showing different roles of CPB and p300 were all done in vitro, since total body knockout of either CBP or p300 is embryonic lethal. Thus, the development by Chakroborty et al of mice with SMC-specific deletion of each coactivator, and the finding that each has a distinct phenotype, is the first demonstration of distinct roles for CBP and p300 in vivo. This new data, combined with the previously published studies distinguishing CBP and p300 function in vitro, support a new paradigm in which p300 promotes differentiation of cells while CBP drives distinct cellular responses in response to stress or in the setting of disease.

This overall concept that SMC phenotype switching is controlled at least in part at the level of epigenetic modifications of DNA and histones is also not new.¹⁰ In addition to this study showing a role for histone acetylation in human allograft vasculopathy, histone modifications have been shown to contribute to SMC function other vascular disorders including atherosclerosis and vascular aging. A prior study showed that human carotid endarterectomy specimens have increased SMC H3K27ac compared to carotid tissue from trauma patients without atherosclerosis.¹¹ In a recent publication examining vascular aging, both CBP and H3K27ac were found to increase with age in mouse aorta and in human primary SMCs in association with increased promoter histone acetylation and expression of genes involved in vascular fibrosis and stiffness.¹² The current study builds on these prior studies by suggesting therapeutic SMC targets to attenuate vascular disease.

Some limitations should be considered. This manuscript only used loss-of-function approaches both in vitro and in vivo. P300 knockout in vivo enhanced neointima formation and knockdown in vitro decreases SMC contractile gene expression. Whether activation or overexpression of p300 might be protective was not tested. Here only wire injury and allograft vasculopathy are examined so it will be important to explore the differential role of CBP and p300 in other, more common, disorders in which vascular remodeling is implicated, including atherosclerosis, aging, aneurysm or hypertension. The authors conclude that drugs specifically targeting CBP may have benefits in cardiovascular diseases in which SMC phenotype switching plays a role. Since CBP transactivates many transcription factors and total body KO of CBP is lethal, this strategy will likely have additional effects that may be unwanted. Also, the HAT domains of CBP and p300 are highly homologous thus it may be difficult to identify selective inhibitors. Furthermore, as p300 protects from neointima and CBP promotes it, a non-specific inhibitor may not be beneficial. However, the finding that only p300 interacts with TET2 and that this interaction is necessary for the pro-differentiation function of p300 in SMCs supports targeting the TET2 interaction as a potential novel strategy.

Despite these limitations, this study has important clinical implications. The data in this manuscript support that identifying ways to promote p300 function and inhibit CBP could have cardiovascular benefits. There are not currently specific p300 inhibitors but the finding that p300 alone interacts with TET2 to methylate DNA may be a targetable mechanism by which these otherwise very homologous HATs have different impact. The impact of inhibiting both CBP and p300 on vascular outcomes is unknown but should be tested and the results considered as CBP/p300 inhibitors are being developed for cancer treatment.

Finally, as always, this study raises multiple additional questions that remain to be answered. How are p300 and CBP regulated in SMC; in particular, what makes p300 decrease with injury and CBP increase with injury or aging? This regulatory mechanism might be a therapeutic target. How are p300 and CBP differentially targeted to the genome in SMC to perform these distinct functions? Many other studies show that CBP/p300 coactivate the same transcription factors and regulate similar genes. In addition to determining whether p300 might be protective in other disorders in which vascular SMC phenotypic modulation is important (atherosclerosis, vascular aging, aneurysm formation, pulmonary hypertension), might this mechanism also contribute to SMC proliferative phenotypes outside the vasculature such as uterine fibroids or lymphagioleiomyomatosis? This information could expand the potential impact. And finally, one might test the concept that p300 has a more universal role in promoting cell differentiation fates while the switch to CBP occurs when cells respond to stress or disease. If so, the novel mechanism might be relevant in other tissues and other diseases outside the cardiovascular system.