Endothelial cells play an active role in maintaining vascular health. Their response to different flow patterns and circulating molecules initiate signaling cascades that determine endothelial phenotypes and ultimately vascular health or disease. One key molecule that initiates endothelial pro-health cascades is the AMP-activated protein kinase (AMPK). Responding to the onset of cellular stress, AMPK not only maintains the cellular energy status by activating catabolic pathways, while shutting down anabolic ones, it also promotes endothelial differentiation and quiescence while attenuating inflammation. However, the precise signaling events that result in AMPK's beneficial effects remain elusive.

Here, we show that both Poly (ADP-ribose) polymerase-1 (PARP-1) and nucleolin (NCL) are phosphorylated by AMPK. PARP-1 is an abundant nuclear protein that once activated, uses NAD+ to "PARylate" itself as well as other protein targets. It serves multiple functions in the cell including DNA repair, epigenetics, mitosis, and transcription by binding to TTGATATAAAT sequences within the genome. The B-cell lymphoma 6 (Bcl-6) intron 1 contains a PARP-1 binding sequence. PARP-1 binds to this sequence it the intron in its inactive state, but upon activation by AMPK phosphorylation, it dissociates allowing transactivation of Bcl-6 protein. Bcl-6 acts as a co-repressor for VCAM-1, MCP-1, and MCP-3, which inhibits the recruitment of macrophages to the endothelium thus attenuating the local inflammatory response. In addition to the phosphorylation of PARP-1, AMPK regulates the transcriptome through the phosphorylation of nucleolin.

We also demonstrate that AMPK regulates endothelial health phosphorylation of nucleolin. Nucleolin is a ubiquitous, multifunctional protein that traffics to multiple cellular locations. It facilitates ribosome biogenesis in the nuceolus, transcription in the nucleus, mRNA stability in the cytoplasm, and angiogenesis at the cell surface. Upon phosphorylation of nucleolin by AMPK, nucleolin transiently translocates to the nucleus where it transactivates KLF2 expression. KLF2 is a zinc finger containing shear stress responsive transcription factor that regulates the majority of the transcriptome in response to pulsatile flow. Additionally, KLF2 is involved in the differentiation of endothelial cells and maintains endothelial quiescence.

Taken together, AMPK regulates the PARP-1 anti-inflammatory and nucleolin/KLF2 pathways which promote endothelial health.

Mitochondria serve to regulate energy and redox homeostasis, ultimately determining cell fate. While mitochondrial dysfunction is characteristic of many metabolic, cardiovascular and neurodegenerative diseases, AMP-activated protein kinase (AMPK), an energy and stress sensor, is known to regulate mitochondrial biogenesis and function. This work is to study a novel mechanism by which AMPK regulates both energy and redox homeostasis through regulation of NAD+ synthetase 1 (NADSYN) and two epigenetic modulators, DNA methyltransferase 1 (DNMT1) and retinoblastoma binding protein 7 (RBBP7). This study combines bioinformatics screening and experimental validation to discover novel substrates of AMPK that are involved in energy and mitochondrial homeostasis in the cell.

.KLF2, originally named lung kruppel-like factor (LKLF), is a zinc finger transcription factor that is heavily involved in vascular development. Recent studies have identified this shear stress responsive gene as a key regulator of endothelial function by modulating endothelial quiescence and vascular tone. Key molecules involved in vascular function such as endothelial nitric oxide synthetase (eNOS) and thrombomodulin have been demonstrated to be direct downstream targets of KLF2. Furthermore, KLF2 upregulation is also associated with downregulation of adhesion molecules and inflammatory cytokines providing an atheroprotective effect. Like other members of the KLF family, KLF2 binds the DNA motif `CACCC' to regulate expression of downstream targets. In vivo, KLF2 is upregulated by pulsatile shear stress conferring an atheroprotective effect. The HMG-CoA reductase inhibitor, statins, have also been shown to upregulate KLF2 via an AMPK dependent mechanism.

We hypothesize that the KLF2 can directly regulate miRNA expression through binding of regulatory regions to promote or repress transcription. The change in miRNA expression can be one means by which KLF2 promotes an atheroprotective effect. Towards this end, we performed a genome wide search for putative KLF2 binding sequences within the regulatory regions of shear stress responsive miRNAs. Preliminary experiments, have demonstrated that miR-483 is upregulated by pulsatile shear stress, statin treatment, as well as KLF2 overexpression. We hypothesize that the upregulation of miR-483 can result in transcriptional repression of a subset of KLF2 responsive genes. Angiotensin Converting Enzyme (ACE) is key enzyme in the rennin-angiotensin system responsible for regulation of blood pressure via the cleavage of the Angiotensin I to the active peptide, Angiotensin II. Akt2 is a serine/threonine kinase that is involved in cell functions including proliferation, growth, and inflammation. Bioinformatics predictions as well as preliminary data have demonstrated that these genes could be a potential target of miR-483.

Despite significant advances in medicine, atherosclerosis and its complications remain the leading causes of morbidity and mortality in most developed countries. Endothelium, lining the first layer of blood vessels, serves as a dynamic interface in maintaining vascular homeostasis. Endothelial dysfunction is a hallmark of atherosclerosis development, signified by impaired nitric oxide (NO) bioavailability derived from endothelial nitric oxide synthase (eNOS). To investigated the molecular mechanisms by which eNOS-NO is regulated under physiological, pathophysiological and pharmacological contexts, I examined the role of two metabolic master molecules, AMP-activated protein kinase (AMPK) and sirtuin 1 (silent information regulator 2 homolog, SIRT1), in eNOS regulation and NO bioavailability.

Primarily, I demonstrated that AMPK directly phosphorylates eNOS Ser-633, in addition to Ser-1177, leading to endothelial NO production, in response to hormonal (adiponectin), physical (shear stress), and pharmacological (atorvastatin) stimuli. Further, I found reduced eNOS phosphorylation in aortas from mice deficient in AMPKa2 (AMPKa2^-/-), confirming the AMPK-mediated eNOS regulation in vivo. In addition, with the application of Nano-liquid chromatography/tandem mass spectrometry (LC/MS/MS), I proved that eNOS Ser-633 was able to compete with acetyl-CoA carboxylase (ACC) Ser-79 for AMPK phosphorylation. These findings suggest that AMPK phosphorylation of eNOS Ser-633 is a functional signaling event for NO bioavailability in endothelial cells (ECs).

In the second part, I focused on SIRT1, a NAD⁺-dependent deacetylase and its modulation on eNOS. First, I found that SIRT1 responds to laminar shear stress with increased protein level and activity in ECs. In parallel to the observation that SIRT1 was significantly higher in ECs exposed to physiological pulsatile flow than pathophysiological oscillatory flow, SIRT1 was shown to be higher in the mouse thoracic aorta exposed to athero-protective flow than in the aortic arch under athero-prone flow. Next, I revealed the functional consequences of the shear stress-enhanced SIRT1 in ECs, including mitochondrial biogenesis and eNOS deacetylation. Subsequently, I studied the interplay of AMPK phosphorylation and SIRT1 deacetylation on eNOS. By using AMPK inhibitor and eNOS Ser-633 and -1177 mutants, I demonstrated that AMPK phosphorylation of eNOS may prime SIRT1 deacetylation of eNOS with attendant NO production. To verify this finding in vivo, I compared eNOS acetylation status in thoracic aortas from AMPKa2^-/- mice and their wild-type littermates and found that AMPKa2^-/- mice had a higher eNOS acetylation. Collectively, my study suggests that AMPK and SIRT1 could synergistically enhance eNOS-derived NO under physiological cues, such as athero-protective flow.

There are numerous deleterious effects of repeated bouts of acute high glucose (HG) or chronic HG. At the molecular level, HG causes oxidative stress, increases advanced glycation end products which activate their cognate receptors (AGE/RAGE), and activate pro-inflammatory pathways. At the physiological level, tooth decay (lactic acid demineralization of enamel), gum disease, obesity (pro-inflammatory cytokines, adipokines), erectile dysfunction, vaginal infections result from HG. Important to this project, HG damages blood vessels leading to the cardiovascular complications in diabetes. HG accelerates macrovascular disease increase risk of strokes, myocardial infarction and renal failure. Further, HG promotes an increased susceptibility to dermal breakdown, forming a portal of entry for microbial infections. This is accompanied by inadequate wound healing of decubitus ulcers, trauma/accident-induced injury or cutaneous infections.

An integral component of wound healing is the process of neovascularization or reperfusion of the injured or ischemic area to supply necessary nutrients and for removal of necrotic or damaged cells and debris. A primary component of neovascularization is endothelial cell (EC) migration to sites of injury. It is well established that HG conditions result in disturbed endothelial migration and wound healing. My specific aim is to delineate the molecular mechanisms of the HG impaired EC migration. The working hypothesis of this project it that high glucose inhibits the pro-angiogenic AMP-activated kinase, thereby inhibiting matrix metaloproteinase-2 collagen cleavage and matrix degradation, which under physiological glucose levels, facilitates wound healing through neovascularization. This pathway is mediated by the phosphorylation and activation of AKT2 (protein kinase B-â), which subsequently increases activating transcription factor 2 (ATF-2) binding to the MMP-2 promoter to increases its expression and activity.

Lipid synthesis and storage in adipose tissues contribute to energy homeostasis, which is inversely regulated by activated AMP-activated protein kinase (AMPK). Here, I report that Ca2+/calmodulin-dependent protein kinase kinase beta (CaMKKbeta) negatively regulates lipogenesis in adipocytes through its phosphorylation of AMPK. Specifically, the glucagon-activated CaMKKbeta/AMPK cascade inhibits acetyl-CoA carboxylase (ACC) activity through the phosphorylation of ACC both in vitro and in vivo. Fasting increases, whereas refeeding decreases, the phosphorylation of AMPK and ACC in CaMKKbeta+/+, but not in CaMKKbeta-/- mice. Under ad libitum feeding, the increased expression of sterol regulatory element binding protein-1c (SREBP-1c)-mediated lipogenic gene is concurrent with decreased phosphorylation of AMPK-regulated ACC in white adipose tissue (WAT) of CaMKKbeta-/- mice. In line with the inhibitory role of CaMKKbeta in lipid synthesis and storage, CaMKKbeta-/- mice exhibit increased body weight, adiposity, and adipocyte hypertrophy, although their appetite is comparable with the CaMKKbeta+/+ littermates. Additionally, metformin, an AMPK activator, does not increase the phosphorylation of AMPK and ACC in WAT of CaMKKbeta-/- mice. Thus, the CaMKKbeta/AMPK signaling is an important molecular event in physiological and pharmacological (e.g., metformin) regulation of adiposity.

The vascular endothelium is essential to maintain normal vascular homostasis and its dysfunction is a hallmark of atherosclerosis development. Up-regulated by laminar flow, the transcription factor Kruppel-like factor 2 (KLF2) is a crucial integrator for maintaining multiple endothelial functions, including anti-inflammation, anti-thrombosis, vasodilatation, and anti-angiogenesis. To investigate the molecular mechanism by which KLF2 is regulated by different flow pattern, I examined the regulation of KLF2 expression at both transcriptional level and post-transcriptional level response to different flow pattern.

In the first part of the study, I demonstrated that AMP-activated protein kinase (AMPK) is necessary and sufficient to regulate the expression of PS-induced KLF2 and its downstream (eNOS and ET-1). In addition, I found PS induced the phosphorylation of ERK5 and MEF2, which regulates the KLF2 expression, is AMPK-dependent in ECs. Furthermore, the phosphorylation levels of ERK5 and MEF2, as well as the expression of KLF2, were significantly reduced in the aorta of AMPK2 knockout mice when compared with wild-type control mice. These findings suggest that AMPK/ERK5/MEF2 is a functional signaling for the regulation of KLF2 transcription.

MicroRNAs (miRNAs) are non-coding small RNAs that regulate gene expression at the post-transcriptional level. In the second part, I examined the role of miRNAs, particularly miR-92a, in the atheroprotective flow-regulated KLF2. Fistly, I found KLF2 is regulated by miRNA by knockdown Dicer with siRNA. miR-92a directly binds to KLF2 3'UTR and inhibit its mRNA level and protein level. Subsequent studies revealed that, atheroprotective laminar flow downregulated the level of miR-92a to induce KLF2, and the level of this flow-induced KLF2 was reduced by pre-92a. Consistent with these findings, miR-92a level was lower in the endothelium of atheroprotective than atheroprone areas of the mouse aorta. Furthermore, mouse carotid arteries receiving pre-92a exhibited impaired vasodilatory response to flow. Collectively, my study suggests that atheroprotective flow patterns decrease the level of miR-92a, which in turn increases KLF2 expression to maintain endothelial homeostasis.

Maintaining vascular endothelial cell (EC) homeostasis is integral to preventing cardiovascular disease. Here, we demonstrate that EC function is tightly regulated with cholesterol homeostasis. The reciprocal cholesterol efflux and cholesterol biosynthesis pathways also contribute to EC function or dysfunction, respectively. This study focuses on a two-pronged approach: 1) How a key regulator of EC function, namely Krüppel-like factor 4 (KLF4), relates to cholesterol efflux and oxidation to promote an anti-inflammatory and anti-oxidative effect in ECs to prevent cardiovascular disease; and 2) How key mediators of cholesterol biosynthesis, i.e., sterol regulatory element-binding protein 2 (SREBP2), relate to the regulation of EC dysfunction contributing to cardiovascular disease. To investigate both pathways, I began with a systems biology approach by using RNA-sequencing of ECs in which KLF4 or SREBP2 is overexpressed. From this global view, novel pathways regulated by KLF4 or SREBP2 that exert opposing functional outcome in ECs were identified. Furthermore, KLF4 was demonstrated to promote endothelial homeostasis in part through the activation of cholesterol oxidation and efflux via the induction of cholesterol-25-hydroxylase (Ch25h) and liver X receptor (LXR). Additionally, KLF4-regulated suppression of the pro-inflammatory SREBP2 and NLRP3 inflammasome pathway contributed to an athero-protective phenotype in mice, further defining that KLF4 regulates a counterbalance between LXR and SREBP2. In contrast, SREBP2 was determined to be both necessary and sufficient to induce EC dysfunction through the upregulation of endothelial-to-mesenchymal transition (EndoMT). This is further implicated in the pro-fibrotic cardiovascular disease, idiopathic pulmonary fibrosis. This two-sided and reciprocal approach provided evidence that cholesterol homeostasis must be tightly regulated by KLF4 in order to maintain EC function. Moreover, when cholesterol regulation is shifted towards SREBP2, ECs become dysfunctional leading to exacerbated cardiovascular disease. The opposing functions of these two key transcription factors provide insight into the complex regulatory network between EC function and cholesterol balance.