Metabolism, Obesity & Diabetes: Recent Studies in Cellular and Animal Models and Human Subjects Highlight Mechanisms and Consequences of Metabolic Dysfunction

Abstract

Obesity and diabetes remain leading causes of reduced health span and life span throughout the world. Hence, it is not surprising that these areas are at the center of highly active areas of research. The identification of novel mechanisms underlying these metabolic disorders sets the stage for uncovering new potential therapeutic strategies. In this issue of Highlights in Arteriosclerosis, Thrombosis and Vascular Biology, we review recently published papers in the journal that add to our understanding of causes and consequences of obesity and diabetes and how these disorders impact metabolic function. Collectively, these studies in cultured cells to in vivo animal models to human subjects add to the growing body of evidence that both cell-intrinsic and cell-cell communication mechanisms collaborate in metabolic disorders to cause obesity, insulin resistance and diabetes and its complications.

Adipocyte Biology: Effects on Vascular and Inflammatory Homeostasis

Adipose tissue is a complex and highly active metabolic organ and adipose dysfunction is linked to cardiovascular disease¹⁸. Adipose tissues are diverse and specific depots have been characterized as “good fat” versus “bad fat.” What accounts for these differences? Indeed, beyond adipocytes, multiple classes of immune cells, such as macrophages, T and B lymphocytes, T regulatory cells and NK cells; nerve tissue; stromovascular cells and endothelial cells (ECs), all housed within a biologically active connective tissue matrix, populate this tissue¹⁹. Recent studies have identified cell-intrinsic and cell-cell communication pathways linked to obesity and diabetes in adipose and other metabolic tissues, such as liver, skeletal muscle, and brain, and how their properties may affect vascular and inflammatory health and homeostasis. For example, cell-intrinsic roles for adipocytes in metabolic dysfunction were illustrated by studies in mice with adipocyte-specific deletion of Nox4 (NADPH oxidase 4), using the Adipoq (Adiponectin) cre recombinase strategy for specific deletion of genes of interest in adipocytes. Mice were fed a high fat/high sucrose diet with added cholesterol. Mice devoid of adipocyte Nox4 exhibited no differences in body weight throughout the study, but experienced a delay in the onset of insulin resistance with an initial attenuation of adipose tissue inflammation that normalized with the sustained feeding with this pro-obesogenic diet²⁰. Early, but not later, in the feeding with this diet, the eWAT (epididymal white adipose tissue) displayed a reduction in 4-HNE (4-hydroxynonenal) staining, a marker of oxidative stress, thereby suggesting that NOX4-derived ROS (reactive oxygen species) may contribute to the development of insulin resistance. Adipose and liver inflammation was also reduced early in the diet feeding in the mice devoid of adipocyte Nox4, implicating adipocyte NOX4 in metabolic tissue inflammation²⁰.

In other studies, the same Adipoq cre recombinase mice were used to test the potential impact of adipocyte-specific deletion of Abca1 (ATP-binding cassette transporter A1) in mice fed a high fat, high cholesterol diet. Loss of adipocyte Abca1 resulted in higher adipocyte plasma membrane content of cholesterol and significantly lower body weight, eWAT fat pad weight and adipocyte size. In the adipocyte Abca1-deleted mice, adipose tissue displayed decreased expression of PPARgamma, CCAAT/enhancer binding protein expression, nuclear SREBP1 (sterol regulatory element binding protein 1) and decreased lipogenesis and triglyceride accretion²¹. However, activation of AKT upon insulin treatment did not differ by adipocyte expression of Abca1, nor did systemic stimulated triglyceride lipolysis or glucose homeostasis²¹.

Beyond the use of the Adipoq cre recombinase strategy, a distinct means to delete adipocyte-expressing genes is the use of the aP2 cre recombinase strategy. Compared to the Adipoq cre recombinase approach, however, the use of the aP2 cre recombinase mice has been reported to delete genes of interest in non-adipocytes, as well²². In a recent publication in ATVB, the aP2 cre recombinase strategy was used to delete Lrp1 (the gene encoding the low-density lipoprotein receptor-related protein 1) to test the effects on the progression of atherosclerosis, as the same line of mice, when fed an obesogenic diet, displayed resistance to diet-induced obesity and reduced hyperglycemia, primarily because of an enhanced thermogenic response in muscle^{23, 24}. However, when mice bearing aP2-mediated deletion of Lrp1 were fed a Western-type diet, they displayed greater adipose tissue inflammation and increased monocyte-macrophage infiltration. When PVAT (perivascular adipose tissue) from normal laboratory diet-fed mice devoid of adipocyte Lrp1 was transplanted into the area surrounding the carotid arteries of mice devoid of the Ldlr (the gene encoding the low-density lipoprotein receptor) prior to Western diet feeding, a three-fold increase in atherosclerosis was observed compared to the mice receiving the wild-type Lrp1-expressing adipose tissue²⁴. In parallel, higher degrees of inflammation in the transplanted fat were observed, suggesting that PVAT expression of Lrp1 in adipocytes, and, possibly, other cells targeted by the aP2 cre recombinase strategy, is important to mitigate PVAT inflammation in Western diet feeding and, thereby, suppress progression of atherosclerosis²⁴. These key studies add to the body of evidence linking inflammatory signals from fat tissue to atherosclerosis and reinforce that the biological sequelae of modulating gene expression in metabolic cells, such as adipocytes, may be highly diet-dependent.

Other recent mechanistic studies queried whether adipocytes from PVAT might store norepinephrine through the vesicular monoamine transporter (VMAT)²⁵. As opposed to retroperitoneal adipocytes, adipocytes from the PVAT of male Sprague Dawley rats expressed VMAT1 and VMAT2 and, functionally, the PVAT adipocytes were able to store norepinephrine²⁵. The results of these experiments prompt future investigation into discerning the consequences of PVAT adipocyte storage of norepinephrine in regulation of hemodynamics and blood pressure, for example.

Distinct studies in isolated human subcutaneous abdominal adipocytes from 1,066 men and women sought to probe regulation of lipolysis and demonstrated that high subcutaneous adipocyte lipolysis and resistance to the anti-lipolytic effects of insulin are associated with elevated levels of TG (triglyceride) and low levels of HDL (high density lipoprotein)-cholesterol, thereby indicating that subcutaneous adipose tissue may causally influence dyslipidemia²⁶. Other recently-published work focused on how distinct cells within adipose and other metabolic tissues may impact metabolic and vascular function. These reports will be considered in the sections to follow.

Vascular Cells and Metabolic Perturbation: Studies in Endothelial Cells and Smooth Muscle Cells

Endothelial Cells

ECs play critical roles in metabolism and metabolic dysfunction^{27, 28}. Considerable attention has focused on the interplay between adipose tissue and ECs and metabolism. In fact, recent studies identified that specialized ECs within adipose tissue play seminal roles in metabolism; Gogg and colleagues reported that while human adipose cells do not secrete ligands for PPARgamma, microvascular ECs within adipose tissue take on this role and have been shown to be responsible for regulation of lipid transport and for secretion of lipids that may activate PPARgamma²⁹. Such findings add to the growing body of evidence that ECs play key roles in metabolism and in modulation of vascular function, processes that may be linked to atherosclerosis.

A recent study published in ATVB probed the vascular effects of selective deletion of endothelial Lep (the gene encoding leptin). When Lep-deficient mice and their respective controls were fed a high fat diet, obesity ensued. Examination of the effects of endothelial deletion of Lep on intimal hyperplasia after chemical injury to the carotid artery revealed that compared to endothelial Lep-expressing mice, those mice devoid of EC Lep displayed significantly greater intimal hyperplasia, neointimal cellularity and proliferation of SMCs (smooth muscle cells), at least in part through upregulation of ET-1 (endothelin-1)³⁰.

In studies in human subjects, the effect of obesity on endothelial function was probed in coronary arterioles retrieved from the right atrial appendage and mediastinal adipose tissue of patients undergoing open heart surgery. In other studies reported in this manuscript, mice (aged 6 or 24 months) were fed a high fat diet and used to test the effect of obesity on endothelial function. It was reported that the presence of obesity impaired endothelium-dependent coronary arteriole dilations only in older patients and in aged mice fed the high fat diet³¹. Interestingly, when adipose tissue from aged obese mice, but not young lean mice, was transplanted into young recipient mice, this resulted in impaired coronary arteriole dilation and increased serum cytokines. Overall, these effects were attributed to age-associated increases in ADAM17 (a disintegrin and metalloprotease domain family 17) in the endothelium. The increased activity of endothelial ADAM17 was traced to diminished inhibitory interactions with caveolin-1³¹. In other studies using small arteries dissected from subcutaneous fat biopsies in human obese versus lean subjects, it was shown that arginase contributes to microvascular endothelial dysfunction in obesity because of higher levels of oxidative stress in aging³².

Distinct work addressed novel mechanisms of protection from altered vascular permeability in diabetes; it was shown that signaling through BMP (bone morphogenetic protein-9)/ALK1 (activin-like kinase) prevents VEGF (vascular endothelial growth factor)-mediated phosphorylation of VE-cadherin and induces the expression of occludin, which serves to prevent retinal vascular permeability, an established harbinger of retinal vascular dysfunction³³.

The interplay between metabolic dysfunction, ECs and inflammation has also been examined. Briot and colleagues studied freshly isolated microvascular ECs from human adipose tissue and found that PPARgamma regulates the expression of FATP1, CD36 and FABP4 and is a major regulator of fatty uptake in these cells³⁴. They showed that endothelial activity of PPARgamma is affected by senescence, which was accompanied by increased inflammation, as evidenced by increased expression of ICAM1, MCP1, IL6 and IL8, all factors that collectively hinder the ability of adipose tissue ECs to respond to and handle lipid fluxes³⁴. In other studies in primary human aortic ECs, treatment with palmitic acid activated STING (stimulator of interferon genes), which resulted in binding to IRF3 (interferon regulatory factor 3) and upregulation of ICAM1³⁵. The authors traced palmitic acid-mediated activation of STING to mitochondrial damage. In mice fed a high fat diet, the STING-IRF3 pathway was activated in adipose tissue and deficiency of STING resulted in partial protection from diet-induced adipose inflammation, obesity, insulin resistance and glucose intolerance³⁵.

Other researchers examined the effect of obesity-associated microvascular endothelial dysfunction with macrophage roles in regulation of vascular levels of H₂S (hydrogen sulfide), a vasorelaxant signaling molecule³⁶. Using mesenteric resistance arterioles from lean and obese mice, Candela and colleagues showed that vasodilation and EC and SMC levels of H₂S were decreased in vessels from obese versus lean mice and that co-culture of macrophages from obese mice, or their conditioned medium, with vessels from lean mice, recapitulated the adverse phenotypes observed in the vessels retrieved from obese mice. The data indicated that the inducible nitric oxide synthase activity of PVAT-resident proinflammatory macrophages promoted vascular dysfunction by reducing the availability of H₂S in the vessel wall³⁶.

The effects of diabetes on impaired wound healing responses and the interplay between vascular and monocyte/macrophage and inflammatory responses have been recently probed. In mice, deletion of the At2 (angiotensin II type 2) receptor in nondiabetic and diabetic mice subjected to unilateral femoral artery ligation after two months of hyperglycemia improved the actions of VEGF, which led to increased blood flow reperfusion, reduced apoptotic ECs and increased small vessel formation after ischemia, especially in diabetic mice³⁷. The underlying mechanisms in the mice devoid of At2 were traced to reduced activity of SHP1 (SH2 domain-containing phosphatase receptor) and restoration of the action of VEGF³⁷.

In other studies, Kimball and colleagues studied mechanisms of impaired wound healing in high fat diet-induced diabetes. These authors identified that Ly6C^HI blood monocytes and particularly the diabetes-specific “second wave” of their influx into 4 mm thick punch biopsy wounds at days 3–4 post-wounding, which corresponded to a spike in the expression of MCP1, drove the impaired wound healing observed in the diabetic condition³⁸. Transcriptomic analysis of the diabetic wound Ly6C^HI cells demonstrated differences in expression of proinflammatory and profibrotic genes when compared with non-diabetic controls³⁸.

López-Díez and colleagues employed a murine model of unilateral hind limb ischemia (femoral artery ligation) in non-diabetic and diabetic mice and showed that mice globally devoid of Ager (the gene encoding RAGE, the receptor for advanced glycation end products), displayed significantly improved return of blood flow and angiogenesis in the affected hind limb compared to the sham controls³⁹. Interestingly, the authors showed that RAGE actions attenuated adaptive inflammation in the ischemic hind limb and suggested microenvironment-specific functions for RAGE signaling in the vasculature³⁹. Hence, these studies highlighted the important interplay between immune cells, such as macrophages, and ECs in the biological response to hind limb ischemia induced by unilateral ligation of the femoral artery.

Smooth Muscle Cells

In addition to ECs, recently published work in ATVB has highlighted new studies in SMCs and their roles in metabolic dysfunction. The AKT signaling pathway plays important roles in vascular functions and in transducing the effects of insulin signaling⁴⁰. Adding to the roles of the AKT signaling pathway in metabolism and vascular biology is a study published in ATVB by Jin and colleagues⁴¹. These authors tested the function of AKT1 versus AKT2 in intimal hyperplasia responses and found that germline or SMC-specific deletion of Akt2 in mice aggravated intimal hyperplasia compared to control mice after arterial denudation injury⁴¹. In contrast, however, while SMC-specific deletion of Akt1 prevented neointimal hyperplasia, germline deletion of Akt1 resulted in severe thrombosis. Administration of rapamycin, which promotes differentiation of SMCs in a manner dependent on AKT2 in vitro, prevented intimal hyperplasia in wild-type mice but exerted no effect in mice devoid of Akt2. To explain these findings, especially in light of the work performed in vitro, the authors showed that there were opposing roles of AKT1 and AKT2 in SMC proliferation, migration, differentiation and responses to rapamycin in vitro. Mechanisms were traced to rapamycin induced expression of MYOCD (myocardin) via phosphorylation of AKT2 and nuclear exclusion of FOXO4, thereby inhibiting its binding to the MYOCD promoter⁴¹. These detailed studies add to the body of evidence linking this key signaling pair, AKT1 and AKT2, to vascular injury and relationships to rapamycin.

Others have studied the AKT pathway in diabetes prompted by the premise that the copper transporter ATP7A (copper transporting/ATPase), which is required for the activation of SOD3 (extracellular superoxide dismutase) and is secreted by SMCs, functions by anchoring to ECs in order to protect these cells by scavenging superoxide in the extracellular space. In type 1 diabetes, ATP7A expression and SOD3 activity are reduced in a manner restored by treatment with insulin. Sudhahar and colleagues recently showed that in vessels from human type 2 diabetic subjects, high fat diet-induced obese mice or obese type 2 diabetic db/db mice, ATP7A expression was downregulated⁴². The downregulation of ATP7A was restored by constitutively active AKT or by deletion of Ptp1b (protein tyrosine phosphatase 1B), consequences of which include enhanced insulin-AKT signaling. It was shown that insulin stimulates binding of AKT2 to ATP7A, which induces phosphorylation of AKT2 at Serine 1424/1463/1466⁴². In Akt2 null vessels, SOD3 activity was reduced and found to be rescued by overexpression of ATP7A. Collectively, these findings link the AKT pathway in SMCs to vascular function and, via connections to insulin, to vascular impairments in both types 1 and 2 diabetes.

Monocytes/Macrophages in Obesity and Diabetes

Monocytes and macrophages play key roles in obesity and in the development and complications of diabetes^43–48. In recent studies, the role of GSK3 (glycogen synthase kinase3) in the pathogenesis of metabolic dysfunction was studied and uncovered interesting insights into the effects on macrophage inflammation. Wang and colleagues fed male C57BL/6 mice a high fat diet for 10 weeks in the presence of vehicle or the GSK3 inhibitors, SB216763 or CHIR99021. Compared to vehicle, the GSK3 inhibitors resulted in reduced inflammation of the visceral adipose tissue with a shift in macrophage polarization in that tissue from a pro- to an anti-inflammatory signature⁴⁹. Of note, the inhibitor-treated animals were also more insulin sensitive. However, there were no differences in body weight between the vehicle- versus the inhibitor-treated groups of mice⁴⁹.

The effects of diabetes on long non-coding (lnc) RNA Dnm3os were studied in macrophages by Das and colleagues. These authors reported that in diabetic human and mouse macrophages, levels of the lnc RNA Dnm3os were elevated through activation of NF-kB⁵⁰. When macrophages were transfected to stably overexpress lncRNA Dnm3os, upregulation of immune response genes and phagocytosis was observed and the knockdown of Dnm3os resulted in reduced inflammatory gene expression⁵⁰. These data identify direct effects of diabetes on mechanisms that regulate macrophage inflammation through lnc RNAs.

In studies performed in diabetic mice undergoing hind limb ischemia mediated by unilateral ligation of the femoral artery, it was shown that deletion of Ager increased circulating Ly6C^HI monocytes and augmented the infiltration of these monocytes into the ischemic diabetic hind limb, in parallel with increased angiogenesis and increased restoration of blood flow³⁹. In vitro studies performed on bone marrow-derived macrophages revealed that exposure of wild-type macrophages to diabetes-relevant high glucose conditions caused increased expression of pro-inflammatory mediators linked to tissue injury and downregulation of anti-inflammatory mediators linked to tissue repair. These effects were mitigated in bone marrow-derived macrophages devoid of Ager. Further, in high glucose conditions, macrophage-EC interactions were attenuated compared to those interactions observed in low glucose conditions. These effects on macrophage inflammation-EC interactions in high glucose conditions were reversed by deletion of macrophage Ager³⁹.

Liver Cells and Metabolic Perturbation

It is well-established that diabetes is associated with dyslipidemia and that unique features are observed in types 1 versus 2 diabetes. Studies in cultured cells and nondiabetic murine livers showed that syndecan1 co-immunoprecipitates with FLOT1 (flotilin1) but not CAV1 (caveolin1)⁵¹; this binding was enhanced in the presence of cholesterol-triglyceride rich lipoproteins (C-TRLs). When FLOT1 was knocked down, there was reduced syndecam1 endocytosis in liver cells. The livers of obese type 2 diabetic mice showed 60–70% less FLOT1 protein versus nondiabetic livers; overexpression of FLOT1 in type 2 diabetic mice normalized plasma TG levels⁵¹. These findings suggested that FLOT1-Syndecam1 interactions are impaired in diabetes, leading to processes that increase circulation of C-TRLs⁵¹.

Other studies tested the hypothesis that the drug metformin, commonly used in type 2 diabetes, affected macrophage reverse cholesterol transport. Incubation of mouse or human primary hepatocytes with metformin resulted in increased expression of Abcg5/8 and the bile salt export pump, Bsep. In vivo, treatment of mice fed a Western diet resulted in upregulation of these molecules with increased initial clearance of ³H-cholesteryl ester HDL from plasma. However, fecal levels of ³H-cholesterol were only marginally affected by metformin treatment⁵². In parallel, metformin treatment increased expression of the LDLR in the mice, thereby increasing uptake of nonradiolabeled cholesterol⁵².

Collectively, these studies report recent advances in our understanding of regulation of cholesterol metabolism and its perturbation in metabolic disorders, such as type 2 diabetes.

Interplay between Metabolic Disorders and Vascular Calcification and Remodeling

The connection between diabetes and vascular calcification and remodeling continues to grow^53–57. Recent studies published in ATVB have addressed these concepts in animal models of metabolic disease. Carmo and colleagues showed that in Vitamin D-fed type 2 diabetic ob/ob mice and C57BL/6 mice, chronic Vitamin D stimulation (14 days treatment) caused expansive hypotrophic vascular remodeling associated with increased vascular calcification, at least in part due to the downregulation of the Vitamin D receptor⁵⁸. The aortas of the Vitamin D-treated mice showed upregulation of MMP (matrix metalloproteinase) activity. In another study, a distinct metalloproteinase, ADAMTS13 (a disintegrin and metalloprotease with thrombospondin type 1 repeats-13) was studied in the context of the diabetic kidney. Although reduced ADAMTS13 has been noted in diabetic kidney, in parallel with thrombotic angiopathy, whether or not there is a connection to the pathogenesis of diabetic nephropathy was not known. Dhanesha and colleagues addressed this question using mice devoid of Adamts13 as well as mice devoid of Adamts13 and Vwf (von Willebrand factor) and showed that after streptozotocin treatment, mice devoid of Adamts13 exhibited increased albuminuria, plasma creatinine and urea, which was accompanied by intrarenal thrombosis⁵⁹. However, mice devoid of both Adamts13 and Vwf demonstrated improvement in these parameters, in parallel with inhibition of intrarenal thrombosis⁵⁹, thereby underscoring the link between these pathways, intrarenal thrombosis and diabetic nephropathic-like changes.

In other studies, Lino and colleagues studied the collagen receptor DDR1 (discoidin domain receptor 1). Because DDR1 is linked to vascular calcification and to signaling via the AKT pathway, the authors tested its role in diabetes. Mice devoid of the Ldlr were made deficient as well in Ddr1 and fed a Western-type high fat diet for 12 weeks to induce both atherosclerosis and diabetes. In the mice devoid of both the Ldlr and Ddr1, reduced vascular calcification in the aortic arches was noted with reduction of RUNX1 staining, thereby suggesting that SMC transdifferentiation into osteoblast like cells was reduced⁶⁰. When primary SMCs were grown in calcifying medium, the cells from Ddr1 null mice showed a significant reduction in phosphorylated AKT and reduced RUNX2 levels and activity. To link the PI3K/AKT pathway to calcification, the authors showed that treatment of the SMCs with Wortmannin, an inhibitor of PI3K, reduced calcification processes in the SMCs grown in the calcification medium⁶⁰. These data, therefore, linked the PI3K/AKT metabolic pathway to DDR1 and vascular calcification in diabetes.

What about calcification in human subjects with diabetes? Raggi and colleagues performed imaging of coronary atherosclerosis using ¹⁸F-sodium fluoride positron emission tomography, which has been suggested to differentiate active, unstable atherosclerosis (i.e., vulnerable plaque) from calcified, stable arterial plaques, and gated chest computed tomography for detection of coronary calcium in 88 subjects with diabetes that were asymptomatic for cardiovascular disease. In that study, the authors reported that despite measurable calcium scores, the prevalence of potentially vulnerable plaques detected by NaF imaging was low⁶¹. It was noted, however, that 45% of the diabetic patients had type 1 diabetes, thus raising the issue of the generalizability of the findings to type 2 diabetes. Overall, this research highlighted that more work would need to be done to establish if NaF imaging might be of utility for cardiovascular risk stratification of diabetic subjects^{61, 62}.

Updates on Lipoproteins and Metabolic Disorders

Although research has uncovered important insights into the mechanisms and consequences of metabolic disorders, especially diabetes, on lipoprotein metabolism, much more remains to be understood⁶³. Research from the Women’s Genome Health Study (WGHS) reported data from 15,813 participants with fasting status, which included 1,453 subjects with incident type 2 diabetes during a mean follow-up period of 18.6 years. In that work, a weighted 40 single-nucleotide polymorphism TG risk score was inversely associated with incident type 2 diabetes, with adjustment for baseline body mass index, HDL cholesterol and TG levels⁶⁴. In the TG-adjusted analysis, large and medium, but not small TG-rich lipoprotein particles, were associated with higher type 2 diabetes incidence, suggesting the importance of these particles subfractions in the clinical profiling of type 2 diabetes risk⁶⁴.

In the case of type 1 diabetes, Frej and colleagues sought to identify mechanisms underlying the increased risk of cardiovascular disease in these patients, despite the presence of higher levels of HDL. As apolipoprotein M (ApoM) and its ligand, sphingosine 1-phosphate (S1P) are responsible for many of the anti-inflammatory effects of HDL, these authors probed if these factors are altered in type 1 diabetes. They reported that ApoM/S1P in light HDL particles was less efficient in inhibiting the expression of VCAM1 (vascular cell adhesion molecule-1) by tumor necrosis factor alpha (TNF) in contrast to the ApoM/S1P located in the denser HDL particles. In type 1 diabetes, however, these patients displayed a higher proportion of the light particles, and, thus, more dysfunctional HDL⁶⁵. They speculated that collectively such findings might explain, at least in part, the higher degree of cardiovascular disease in patients with type 1 diabetes versus controls.

What about the S1P factor in the metabolic syndrome? Even in the absence of diabetes, HDL-cholesterol from patients with metabolic syndrome displays abnormalities, such as TG enrichment and S1P depletion. Denimal and colleagues sought to determine if such abnormalities might impair the ability of HDL to stimulate endothelial NOS (eNOS). These authors reported in studies using human umbilical vein ECs that HDL-mediated activation of eNOS is decreased in non-diabetic patients with metabolic syndrome and that the main factor underlying this finding was the depletion of S1P⁶⁶.

Pharmacological Agents and Therapeutic Approaches – Old and New Actors in the Metabolic Milieu

Insulin

The effects of insulin on atherosclerosis have been a subject of long-standing interest and some controversy, with reports generally suggesting that hyperinsulinemia is a risk factor for cardiovascular disease^67–71. Park and colleagues tested this concept by administration of exogenous insulin (by implantation) to mice devoid of Apoe and fed a high fat diet with the goal of stimulating hyperinsulinemia. These mice, compared to control animals not treated with insulin, displayed insulin resistance, hyperglycemia and hyperinsulinemia⁷². After 8 weeks of high fat feeding, mice were then treated with sham or insulin pellet versus phlorizin (a sodium glucose transporter ½ inhibitor) for an additional eight weeks. Insulin treatment resulted in reduced plasma TG, cholesterol and lipoprotein levels and by en face staining, a significant reduction in atherosclerosis and expression of proinflammatory mediators; endothelial function was improved and a shift from “M1” to more “M2”-like macrophages in the plaques was noted⁷². In contrast, treatment with phlorizin resulted in no differences in atherosclerotic plaques, but phlorizin significantly lowered plasma glucose and glycosylated hemoglobin levels⁷². These intriguing data reignite the interest in the beneficial versus damaging effects on insulin on the vessel wall and suggest that one component of insulin action is to reduce vascular inflammation in high fat-fed animals. In summary, these findings suggested that as compared to simply lowering plasma glucose, insulin treatment reduced endothelial and macrophage dysfunction in mice devoid of Apoe and fed a high fat diet.

As opposed to insulin-loaded pellet implantation in mice for delivery of insulin to the periphery, Xiao and colleagues treated healthy men with intranasal insulin. In rodents and humans, direct administration of insulin into the brain results in suppression of hepatic glucose production and, in rodents, results in reduced secretion of TG-rich lipoproteins. Hence, these authors asked if intranasal treatment in human subjects exerted effects on TG-rich lipoprotein levels. They reported that at a dose of intranasal insulin, which in human subjects suppressed hepatic glucose production, there was no effect noted on TG-rich lipoproteins ApoB100 or ApoB48⁷³.

Liraglutide

The GLP1 (glucagon like peptide-1) agonist, liraglutide, has been shown to reduce postprandial lipidemia but the underlying mechanisms have not been elucidated. Verges and colleagues tested 10 patients with type 2 diabetes and showed that liraglutide induced a significant reduction of the ApoB48 pool due to both a decrease in ApoB48 production and an increase in ApoB48 catabolism and in vitro, liraglutide reduced the expression of genes linked to the synthesis of chylomicrons⁷⁴.

Weight Loss and Exercise

To assess the effects of weight loss and exercise on levels of HDL, 95 subjects with metabolic syndrome and 40 healthy subjects were enrolled to receive 12 weeks of no treatment, weight loss, or weight loss plus exercise. Significant differences in the HDL lipidome were observed at baseline between the metabolic syndrome versus the healthy individuals, with smaller mean particle size and lower cholesterol efflux capacity⁷⁵. After weight loss and especially after weight loss plus exercise, the HDL lipidome and particle size were modulated toward that of the healthy individuals and cholesterol efflux capacity was improved⁷⁵, altogether suggesting that these interventions may be important life style-modification approaches in vulnerable subjects.

In other studies, 32 nonfasting plasma samples from subjects undergoing bariatric surgery at preoperative and the 3 and 12 months postoperative time point were tested using predefined protein panels (Olinks). 19 candidate proteins that were altered among these time points were identified and 8 of them demonstrated the highest expression in the liver (as extracted from other publicly available databases)⁷⁶. Of the 19 candidate proteins, 6 were previously measured in more than 3000 Framingham Heart Study participants and of these, a higher concentration of IGFBP2 (insulin like growth factor binding protein2) at baseline was found to be associated with a lower risk of incident metabolic syndrome and type 2 diabetes after multivariate adjustments⁷⁶. Hence, measures of IGFBP2 may serve as possible biomarkers of insulin resistance; the authors concluded that further studies in well-defined obesity intervention studies should consider the testing of this marker.

Bempedoic Acid

ACLY (ATP citrate lyase) is an enzyme in the cholesterol-biosynthesis pathway, which is upstream of 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR), the target of statins. ACLY catalyzes the ATP-dependent and coenzyme-A-dependent conversion of citrate to oxaloacetate and the high energy biosynthetic precursor, acetyl-coA⁷⁷. Acetyl-coA stimulates the synthesis of fatty acids, cholesterol and acetylcholine and mediates the acetylation of histones and proteins⁷⁷. ACLY is highly expressed in adipocytes, liver and cholinergic neurons, where it links carbohydrate and lipid metabolism. Based on these consideration, if and how ACLY contributed or not to cardiovascular disease is the subject of considerable focus. Human genetic data⁷⁸ and data from subjects receiving bempedoic acid suggest links of this ACLY inhibitor to metabolic improvements with reduction in levels of LDL cholesterol⁷⁹. In a study in mice devoid of the Ldlr and fed a high fat and high cholesterol diet, treatment of the mice for 12 weeks with bempedoic acid resulted in an attenuation of diet-induced hypercholesterolemia, hypertriglyceridemia, hyperglycemia, hyperinsulinemia, fatty liver, obesity and adiposity⁸⁰. In the livers of treated mice, significantly increased expression of genes linked to fatty acid oxidation and decreased expression of inflammatory genes was noted. In the full-length aortas, treatment with this agent resulted in 44% attenuation in atherosclerotic plaques at the aortic sinus versus control treatment⁸⁰. In a large animal model, the Yucatan Miniature pig expressing (+/−) or devoid (−/−) of the LDLR were fed a high fat, cholesterol-containing diet for 160 days. In pigs that were heterozygous or homozygous for deletion of LDLR under these feeding conditions, treatment with bempedoic acid resulted in lower levels of cholesterol and attenuation of atherosclerotic lesion areas⁸¹. Treatment with bempedoic acid had no effect on plasma levels of TG or HDL and had no effect on fasting glucose or insulin or levels of hepatic lipids in both genotypes⁸¹. Hence, studies from human subject genetics and small and large animal models suggest further testing of ACLY and its inhibitor, bempedoic acid, as a means to reduce levels of LDL cholesterol and, importantly, to test its overall safety, particularly in statin-intolerant or statin-refractory patients⁸².

Summary and Perspectives

In summary, these recently-published studies reinforce the power of human genetic analyses with direct hypothesis-testing of mechanisms and consequences related to metabolic diseases in cellular and animal models and in human subjects. The above-noted studies underscore the homeostatic and pathobiological roles played by multiple cell types implicated in vascular biology and metabolism and indicate how their cell-intrinsic and/or intercellular communications may dramatically affect pathologies in such disorders as obesity, diabetes, atherosclerosis, vascular calcification, ischemia, diabetes-associated nephropathy and diabetic wound healing. Collectively, these recent publications add to our understanding of metabolic disorders and highlight new directions in therapeutic approaches for metabolic disease.

Highlights.

Diabetes poses a major risk to health and longevity. Spurred, at least in part, by the rise in obesity, diabetes has become the 7^th leading cause of death in the United States^1–3 and a major risk factor for the development of cardiovascular disease and its complications^4–6. The intimate relationships that entwine obesity, insulin resistance and diabetes with the immune system and regulation of metabolism are being increasingly unraveled. The interplay between tissue resident versus infiltrating immune cells in the metabolic organs is complex and the potential mechanistic influence of immune cell (dys)function and insulin resistance is under intense investigation^7–11. Furthermore, the identification of protective roles for immune cells, such as macrophages, in countering the derangements affecting insulin resistance is adding new insights to these complex phenomena^{12, 13}. In addition, the influence of both cell-intrinsic and intercellular communication mechanisms in metabolic organs is being increasingly recognized as central to the organism’s response to endogenous and exogenous stresses. Mechanisms, such as direct cell-cell contact and intercellular communication instigated by extracellular vesicles, such as exosomes, are being implicated in the development of metabolic dysfunction^{7, 14–17}. In this contribution to the Highlights series in Arteriosclerosis, Thrombosis & Vascular Biology (ATVB), we review recent studies published in the journal in the areas of Obesity, Diabetes and Metabolism that add to the body of data identifying mechanisms and consequences of metabolic dysfunction. Further, these studies point to new therapeutic avenues and modalities for obesity and diabetes. From work in cultured cells, to animal models of metabolic disorders and to studies in human subjects, recent work published in ATVB is providing new insight towards understanding and tackling these complex disorders.

Highlights.

• Obesity, diabetes and metabolic dysfunction are major causes of morbidity and mortality in the United States and throughout the world.

• Recent publications in Arteriosclerosis, Thrombosis and Vascular Biology have shed light on novel mechanisms and consequences of this disorders in studies in cultured cells, animal models and human subjects.

• These recent findings published in Arteriosclerosis, Thrombosis and Vascular Biology shed light on new therapeutic targets and opportunities for the treatment of metabolic diseases.

Abbreviations

ABCA1

ATP binding cassette transporter A1

ACLY

ATP citrate lyase

ADAM17

a disintegrin and metalloprotease domain family 17

ADAMTS13

a disintegrin and metalloprotease with thrombospondin type 1 repeats 13

AGER

advanced glycation endproduct receptor (or RAGE)

ALK1

activin like kinase

APO

apolipoprotein

AT2

angiotensin II type 2

ATVB

Arteriosclerosis, Thrombosis and Vascular Biology

BMP

bone morphogenetic protein

CAV1

caveolin 1

C-TRLs

cholesterol-triglyceride rich lipoproteins

DDR1

discoidin domain receptor 1

EC

endothelial cell

eNOS

endothelial nitric oxide synthase

ET1

endothelin-1

eWAT

epididymal white adipose tissue

FLOT1

flotilin 1

GLP1

glucagon like peptide 1

GSK3

glycogen synthase kinase 3

H₂S

hydrogen sulfide

HDL

high density lipoprotein

HMGCR

3-hydroxy-3-methylglutaryl-coenzyme A reductase

4-HNE

4-hydroxynonenal

IGFBP2

insulin like growth factor binding protein 2

ILF2

interleukin enhancer binding factor 2

IRF3

interferon regulatory factor 3

LEP

leptin

LDLR

low-density lipoprotein receptor

LRP1

low density lipoprotein receptor-related protein 1

MMP

matrix metalloproteinase

MYOCD

myocardin

NOX4

NADPH oxidase 4

PTP1B

protein tyrosine phosphatase 1B

PVAT

perivascular adipose tissue

ROS

reactive oxygen species

S1P

sphingosine 1 phosphate

SHP1

SH2 domain containing phosphatase receptor

SMC

smooth muscle cell

SOD3

extracellular superoxide dismutase

SREBP1

sterol regulatory element binding protein 1

STING

stimulator of interferon genes

TG

triglyceride

TNF

tumor necrosis factor

VCAM1

vascular cell adhesion molecule 1

VEGF

vascular endothelial growth factor

VMAT

vesicular monoamine transporter

vWF

von Willebrand factor

WGHS

Women’s Genome Health Study