PDCD5 says no to NO

Blood vessels are more than mere conduits, and their dynamic function is regulated to a large extent by their inner layer, the endothelium. Endothelial cell-derived nitric oxide (NO), a product of endothelial NO synthase (eNOS), is important for normal endothelial function and vascular health. Programmed cell death 5 (PDCD5) is a regulator of multiple cellular processes through its interactions with other proteins. In PNAS, Lee et al. (1) describe a novel mechanism by which PDCD5 alters vascular function by reducing NO production.

Integrated function of the cardiovascular system is essential for the maintenance of homeostasis in complex multicellular organisms. Mutations that affect components of this system frequently result in embryonic lethality (2, 3), and maladaptive changes during adulthood lead to cardiovascular disease, the number one cause of death worldwide (4, 5). The endothelium functions as a selective barrier and controls vascular tone, hemostasis and thrombosis, inflammation, and redox balance (6). Endothelial cells regulate these functions in part by producing NO, which can diffuse across cell membranes to work as a paracrine factor on neighboring smooth muscle cells and circulating leukocytes and platelets. Impaired NO synthesis or availability contributes significantly to vascular diseases, including hypertension and atherosclerosis (6, 7).

Endothelial cells produce NO from l-arginine mainly through the activity of eNOS, although other isoforms may be involved (7). Regulation of eNOS activity is complex and occurs at transcriptional, posttranscriptional, and posttranslational levels. Posttranslational modifications such as phosphorylation, acylation, and nitrosylation control eNOS subcellular localization and activity (8). For instance, phosphorylation of human eNOS at Ser-617, Ser-635, and Ser-1177 is stimulatory, while phosphorylation at Ser-116 or Thr-495 is inhibitory (9). A myriad of extracellular signals including shear stress, acetylcholine, bradykinin, histamine, and vascular endothelial growth factor (VEGF) stimulate NO production through distinct kinases that converge on eNOS (7, 8), but the regulatory mechanisms and differential contributions of these pathways have not been fully elucidated. One important kinase is Akt, which phosphorylates eNOS at Ser-1177 to enhance its activity in a calcium-independent manner (10).

PDCD5 is a ubiquitously expressed cytosolic and nuclear protein. It facilitates p53-dependent apoptosis through its interaction with mouse double minute 2 homolog (MDM2) (11) and may function as a tumor suppressor. PDCD5 has also been reported to regulate other cellular processes [e.g., autophagy (12)], although the relevant mechanisms are less defined. PDCD5 binds multiple cellular proteins, including histone deacetylase 3 (HDAC3) (13). One function of HDAC3 in leukemia cells is to bind to and deacetylate Akt, thereby facilitating Akt phosphorylation and activation (14). Lee and coworkers (13) have previously shown that binding of PDCD5 to HDAC3 displaces p53 and promotes HDAC3 degradation in cancer cells. However, previous work defining a role for PDCD5 in the vascular system has been limited to a study showing that it is induced by cerebral ischemia/reperfusion and contributes to infarction and vascular permeability (15).

Lee et al. (1) explore the role of PDCD5 in regulating the HDAC3–Akt interaction in the context of endothelial function. In vitro studies demonstrated that loss of PDCD5 in endothelial cells increases the interaction of HDAC3 with Akt—tested by coimmunoprecipitation and proximity ligation assays—augmenting levels of phosphorylated forms of Akt and eNOS, and NO production. Functionally, loss of PDCD5 in endothelial cells was shown to increase microvessel sprouting in an in vitro aortic ring assay (1). Conversely, overexpression of PDCD5 in VEGF-stimulated endothelial cells decreased the HDAC3–Akt interaction, levels of phosphorylated eNOS and Akt, NO production, and endothelial tube formation in vitro (1). None of these effects were observed when a mutant form of PDCD5, defective for binding HDAC3, was overexpressed (1), demonstrating that PDCD5 binding to HDAC3 is critical. These observations suggest that PDCD5 interacts with HDAC3 and disrupts HDAC3–Akt binding, which, in turn, decreases Akt phosphorylation and Akt-dependent activation of eNOS in endothelial cells (Fig. 1). Lee et al. further evaluate the importance of HDAC3 in this pathway using genetic knockdown and pharmacologic inhibition. Decreases in HDAC3 abundance or function in VEGF-stimulated endothelial cells impaired Akt and eNOS phosphorylation, NO production, and microvessel sprouting from aortic rings (1), consistent with the proposed model. These mechanistic studies focus on presumed cytoplasmic events but do not exclude contributions from changes that may stem from loss of PDCD5 or HDAC3 in the nucleus (11).

Fig. 1.

Programmed cell death 5 (PDCD5)-dependent inhibition of endothelial nitric oxide synthase (eNOS). In endothelial cells in culture, loss of PDCD5 increases both the interaction of the kinase Akt with histone deacetylase 3 (HDAC3) and phosphorylation of Akt. This in turn enhances Akt-dependent phosphorylation and activation of eNOS, and subsequent nitric oxide (NO) production from l-arginine. Conversely, increased expression of PDCD5 disrupts the interaction of HDAC3 with Akt and decreases the levels of phosphorylated Akt and eNOS, reducing NO production. Red dotted arrows indicate processes that are restricted.

To explore this mechanism in vivo, Lee et al. generated mice with endothelial-specific deletion of PDCD5 and used partial carotid artery ligation and administration of the Paigen diet to induce vascular remodeling. Deletion of PDCD5 in endothelial cells markedly reduced neointimal growth compared with wild-type mice studied in the same protocol (1). Moreover, mice with endothelial cell-specific deletion of PDCD5 exhibited increased endothelial expression of phosphorylated Akt (1), consistent with their model. Caveats include that the PDCD5–HDAC3 and HDAC3–Akt interactions and levels of HDAC3 and phosphorylated eNOS were not assessed in the in vivo context; and that loss of PDCD5 in the hematopoietic lineage may also have contributed to the observed effects because the Tie2-Cre deletor used in these studies may also be active in those cells (16). Despite these limitations, the in vivo findings are interesting and constitute a rationale to study PDCD5-dependent inhibition of Akt and eNOS in endothelial cells in clinically relevant models of vascular disease, including atherosclerosis and hypertension.

PDCD5 can be detected in serum and other fluids (17, 18), and Lee et al. studied human serum samples from a high-risk cohort to test whether serum levels correlate with adverse cardiovascular conditions. They found an inverse relationship between levels of PDCD5 and those of nitrite/nitrate (stable end products of NO) (1). Moreover, circulating PDCD5 levels were higher in males and in patients with hypertension, diabetes mellitus, left ventricular hypertrophy, higher coronary calcium scores, and lower HDL levels (1). These results are provocative, although the predictive power of serum PDCD5 levels will require further study. These correlations also raise interesting biological questions: Given its ubiquitous expression (19), what are the sources of circulating PDCD5? Specifically, do levels of PDCD5 in serum reflect endothelial dysfunction? Conversely, is PDCD5 in the blood able to act on and change the phenotype of endothelial cells? Some clues to this last question are provided by pertinent in vitro studies in Lee et al. showing that PDCD5 is detected in media conditioned by human endothelial cells, and addition of recombinant human PDCD5 directly to cell culture media reduces the levels of phosphorylated forms of Akt and eNOS in endothelial cells. Finally, is PDCD5—in the blood and/or in cells—a potential target for therapeutic intervention? This possibility is exciting, although the fact that PDCD5 impacts other cellular processes (11–13, 15) also needs to be considered.

This study adds a new and likely complex PDCD5-based angle to our understanding of mechanisms that affect eNOS activity, endothelial dysfunction, and cardiovascular disease. Assessing the significance and specificity of this regulatory pathway in biology and disease is an important goal for future studies.