Nada A. Abumrad

Long-chain fatty acids in the blood are prevented from unfettered movement into nonfenestrated tissues or the arterial wall. During fasting, nonesterified FAs are released from adipose tissue into the circulation and bind to albumin, forming a complex >65 kDa, with limited ability to efficiently cross endothelial cell (EC) barriers without a specific receptor. For this reason, nonhepatic tissue distribution of circulating FA parallels EC expression of the FA-binding protein CD36 (cluster of differentiation 36). The deletion of CD36 in ECs reduces nonesterified FA uptake by the heart, muscle, and brown adipose tissue. The other major transport system for FAs is via lipoproteins. Circulating FAs are contained within TRLs (triglyceride-rich lipoproteins), chylomicrons during the postprandial period, and VLDL (very low-density lipoprotein) both postprandially and during fasting. LPL (lipoprotein lipase) on capillary ECs releases FAs from TRLs and likely allows their passage into tissues, in part, via a CD36-independent process. ECs can also internalize lipoprotein particles, followed by the transendothelial movement of lipids. In this review, we will discuss the pathways of EC uptake of FAs from circulation, how this process affects both EC and tissue biology, and the importance of these processes for systemic metabolism and vascular health. We will conclude with speculations on methods to modulate EC FA uptake and their implications for human health.

BACKGROUND: Movement of circulating lipids into tissues and arteries requires transfer across the endothelial cell (EC) barrier. This process allows the heart to obtain fatty acids, its chief source of energy, and apoB-containing lipoproteins to cross the arterial endothelial barrier, leading to cholesterol accumulation in the subendothelial space. Multiple studies have established elevated postprandial TRLs (triglyceride-rich lipoproteins) as an independent risk factor for cardiovascular disease. We explored how chylomicrons affect ECs and transfer their fatty acids across the EC barrier. METHODS: We had reported that media from chylomicron-treated ECs lead to lipid droplet formation in macrophages. To determine the responsible component of this media, we assessed whether removing the extracellular vesicles (EVs) would obviate this effect. EVs from control and treated cells were then characterized by protein, lipid, and microRNA content. We also studied the EV-induced transcription changes in macrophages and ECs and whether knockdown of SR-BI (scavenger receptor-BI) altered these responses. In addition, using chylomicrons labeled with [ ¹³ C]oleate, we studied the uptake and release of this labeled by ECs. RESULTS: Chylomicron treatment of ECs led to an inflammatory response that included production of EVs that drove macrophage lipid droplet accumulation. The EVs contained little free fatty acids and triglycerides, but abundant phospholipids and diacylglycerols. In concert with this, [ ¹³ ]C labeled chylomicron triglycerides exited ECs primarily in phospholipids. EVs from chylomicron-treated versus untreated ECs were larger, more abundant, and contained specific microRNAs. Treatment of macrophages and naive ECs with media from chylomicron-treated ECs increased expression of inflammatory genes. CONCLUSIONS: EC chylomicron metabolism produces EVs that increase macrophage inflammation and create LDs. Media containing these EVs also increases EC inflammation, illustrating an autocrine inflammatory process. Fatty acids within chylomicron triglycerides are converted to phospholipids within EVs. Thus, EC uptake of chylomicrons constitutes an important pathway for vascular inflammation and tissue lipid acquisition.

Lipoprotein lipase (LPL) is critical for clearance of circulating triglycerides and for tissue fatty acid supply. LPL is primarily synthesized and secreted by adipocytes into the interstitium and must traffic from there to the abluminal/basolateral side of capillary endothelial cells. There, LPL binds glycosylphosphatidylinositol-anchored protein 1, GPIHBP1, which stabilizes the protein and facilitates its movement across the endothelial cells to the luminal side where it functions in hydrolysis of lipoprotein triglycerides. Importance of LPL traffic is supported by findings that rare mutations in GPIHBP1 cause hypertriglyceridemia. However our understanding of how LPL is secreted by adipocytes and traffics to endothelial cells is incomplete. Here we examined the possibility that secretion and traffic of adipocyte LPL might involve generation of small extracellular vesicles (sEVs/exosomes) which often mediate cell-cell communication. Proteomic analysis of sEVs secreted by adipocytes showed them enriched in LPL. To study LPL secretion and transfer we generated human derived pre-adipocytes (HPA) that stably express tagged LPL (FLAG and His epitopes). LPL pulldown and sEV isolation from HPA conditioned media documented that greater than 70% of secreted LPL is present in sEVs. The mechanism for LPL secretion in sEVs was found to involve the ESCRT-independent neutral sphingomyelinase 2 (nSMase2) pathway, as treatment with the nSMase2 inhibitor GW4869 reduced secretion by 80%. The above observations were reproduced using highly sensitive nanoparticle flow cytometry. The sEV associated LPL has lipolytic activity and it is released by heparin addition indicating it is on the sEV surface. In addition, using human derived microvascular endothelial cells with stable lentiviral expression of GPIHBP1 we show that LPL positive sEVs transfer LPL to these cells, but not to control cells without GPIHBP1. Our findings suggest that sEV formation by nSMase2 controls adipocyte LPL secretion and traffic, that sEVs protect LPL activity and facilitate LPL transfer to GPIHBP1 on endothelial cells.

Background: Movement of circulating lipids into tissues and arteries requires transfer across the endothelial cell barrier. This process allows the heart to obtain fatty acids (FAs), its chief source of energy and apolipoprotein B (apoB)-containing lipoproteins to cross the arterial endothelial barrier leading to cholesterol accumulation in the subendothelial space. Multiple studies have established elevated postprandial triglyceride-rich lipoproteins (TRLs) as an independent risk factor for cardiovascular disease (CVD). We explored how chylomicrons affect ECs and transfer their FAs across the EC barrier. Methods: We had reported that media from chylomicron-treated ECs leads to lipid droplet (LD) formation in macrophages. To determine the responsible component of this media, we assessed whether removing the extracellular vesicles (EVs) would obviate this effect. EVs from control and treated cells were then characterized by protein, lipid and microRNA (miR) content. We also studied the EV-induced transcription changes in macrophages and ECs and whether knockdown of scavenger receptor-BI (SR-BI) altered these responses. In addition, using chylomicrons labeled with [ ¹³ C]oleate, we studied the uptake and release of this labed by ECs. Results: Chylomicron treatment of ECs led to an inflammatory response that included production of EVs that drove macrophage LD accumulation. The EVs contained little free fatty acids and triglyceride, but abundant phospholipids and diacylglycerols. In concert with this, [ ¹³ ]C labeled chylomicron triglycerides exited ECs primarily in phospholipids. EVs from chylomicron treated versus untreated ECs were larger, more abundant, and contained specific miRs. Treatment of macrophages and naïve ECs with media from chylomicron-treated ECs increased expression of inflammatory genes. Conclusions: EC chylomicron metabolism produces EVs that increase macrophage inflammation and create LDs. Media containing these EVs also increases EC inflammation, illustrating an autocrine inflammatory process. FAs within chylomicron triglycerides are converted to phospholipids within EVs. Thus, EC uptake of chylomicrons constitutes an important pathway for vascular inflammation and tissue lipid acquisition.

Aims/hypothesis: Microvascular dysfunction contributes to insulin resistance. CD36, a fatty acid transporter and modulator of insulin signalling, is abundant in microvascular endothelial cells. Humans carrying the minor allele (G) of CD36 coding variant rs3211938 have 50% reduced CD36 expression and show endothelial dysfunction. We aimed to determine whether G allele carriers have microvascular resistance to insulin and, if so, how this affects glucose disposal. Methods: Our multi-disciplinary approach included hyperinsulinaemic–euglycaemic clamps in Cd36 ^−/− and wild-type mice, and in individuals with 50% CD36 deficiency, together with control counterparts, in addition to primary human-derived microvascular endothelial cells with/without CD36 depletion. Results: Insulin clamps showed that Cd36 ^−/− mice have enhanced insulin-stimulated glucose disposal but reduced vascular compliance and capillary perfusion. Intravital microscopy of the gastrocnemius showed unaltered transcapillary insulin flux. CD36-deficient humans had better insulin-stimulated glucose disposal but insulin-unresponsive microvascular blood volume (MBV). Human microvascular cells depleted of CD36 showed impaired insulin activation of Akt, endothelial NO synthase and NO generation. Thus, in CD36 deficiency, microvascular insulin resistance paradoxically associated with enhanced insulin sensitivity of glucose disposal. Conclusions/interpretation: CD36 deficiency was previously shown to reduce muscle/heart fatty acid uptake, whereas here we showed that it reduced vascular compliance and the ability of insulin to increase MBV for optimising glucose and oxygen delivery. The muscle and heart respond to these energy challenges by transcriptional remodelling priming the tissue for insulin-stimulated glycolytic flux. Reduced oxygen delivery activating hypoxia-induced factors, endothelial release of growth factors or small intracellular vesicles might mediate this adaptation. Targeting NO bioavailability in CD36 deficiency could benefit the microvasculature and muscle/heart metabolism. Trial registration: Clinicaltrials.gov NCT03012386 Data availability: The RNAseq data generated in this study have been deposited in the NCBI Gene Expression Omnibus (www.ncbi.nlm.nih.gov/geo/) under accession code GSE235988 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE235988). Graphical

Aim: To examine factors underlying why most, but not all, adults with obesity exhibit impaired insulin‐mediated glucose uptake, we compared: (1) adipose tissue fatty acid (FA) release, (2) skeletal muscle lipid droplet (LD) characteristics, and (3) insulin signalling events, in skeletal muscle of adults with obesity with relatively high versus low insulin‐mediated glucose uptake.Methods: Seventeen adults with obesity (BMI: 36 ± 3 kg/m²) completed a 2 h hyperinsulinemic–euglycemic clamp with stable isotope tracer infusions to measure glucose rate of disappearance (glucose Rd) and FA rate of appearance (FA Ra). Skeletal muscle biopsies were collected at baseline and 30 min into the insulin infusion. Participants were stratified into HIGH (n = 7) and LOW (n = 10) insulin sensitivity cohorts by their glucose Rd during the hyperinsulinemic clamp (LOW< 400; HIGH >550 nmol/kgFFM/min/[μU/mL]).Results: Insulin‐mediated suppression of FA Ra was lower in LOW compared with HIGH (p < 0.01). In skeletal muscle, total intramyocellular lipid content did not differ between cohorts. However, the size of LDs in the subsarcolemmal region (SS) of type II muscle fibres was larger in LOW compared with HIGH (p = 0.01). Additionally, insulin receptor‐β (IRβ) interactions with regulatory proteins CD36 and Fyn were lower in LOW versus HIGH (p < 0.01), which aligned with attenuated insulin‐mediated Tyr phosphorylation of IRβ and downstream insulin‐signalling proteins in LOW.Conclusion: Collectively, reduced ability for insulin to suppress FA mobilization, with accompanying modifications in intramyocellular LD size and distribution, and diminished IRβ interaction with key regulatory proteins may be key contributors to impaired insulin‐mediated glucose uptake commonly found in adults with obesity.

Lipid metabolism plays a critical role in lymphatic endothelial cell (LEC) development and maintenance. Altered lipid metabolism is associated with loss of lymphatic vessel integrity, which compromises organ function, protective immunity, and metabolic health. However, the role of lipid metabolism in LEC function is not well understood. Insulin is a key regulator of lipid metabolism and protein palmitoylation, the reversible post-translational protein modification by palmitate that affects protein stability, trafficking, protein-protein, and protein-membrane interactions. Human LECs are highly sensitive to insulin and can develop insulin resistance in vitro , but whether insulin regulates LEC protein palmitoylation and function is unknown. To examine the role of palmitoylation in LEC function, we generated the first palmitoylation proteomics profile in human LECs, validated insulin regulated targets and profiled differences in palmitoylation between lymphatic and blood endothelial cells. Palmitoylation occurred primarily in proteins involved in LEC vesicular or membrane trafficking, translation initiation, and in those found in membrane rafts. Insulin enriched palmitoylation of LEC proteins involved in GTPase signaling, ubiquitination, and junctional anchoring. We also determined that the long-chain fatty acid receptor CD36 mediates optimal lymphatic palmitoylation. CD36 silencing in LECs doubled palmitoylation targets involving proteins related to inflammation and neutrophil degranulation contributing to an activated inflamed endothelium. These results suggest that the coordination of the process of palmitoylation is critical for normal lymphatic endothelial function.

To examine factors underlying why most, but not all adults with obesity exhibit impaired insulin-mediated glucose uptake, we compared: 1) rates of fatty acid (FA) release from adipose tissue, 2) skeletal muscle lipid droplet (LD) characteristics, and 3) insulin signaling events in skeletal muscle collected from cohorts of adults with obesity with “HIGH” versus “LOW” insulin sensitivity for glucose uptake. Seventeen adults with obesity (BMI: 36±3kg/m ² ) completed a 2h hyperinsulinemic-euglycemic clamp with stable isotope tracer infusions to measure glucose rate of disappearance (glucose Rd) and FA rate of appearance (FA Ra). Skeletal muscle biopsies were collected at baseline and 30min into the insulin infusion. Participants were stratified into HIGH (n=7) and LOW (n=10) insulin sensitivity cohorts by their glucose Rd during the hyperinsulinemic clamp (LOW550 nmol/kgFFM/min/[µU/mL]). Insulin-mediated suppression of FA Ra was lower in LOW compared with HIGH (p<0.01). In skeletal muscle, total intramyocellular lipid content did not differ between cohorts. However, the size of LDs in the subsarcolemmal region (SS) of type II muscle fibers was larger in LOW compared with HIGH (p=0.01). Additionally, insulin receptor (IR) interactions with regulatory proteins CD36 and Fyn were lower in LOW versus HIGH (p<0.01), which aligned with attenuated insulin-mediated Tyr phosphorylation of IRβ and downstream insulin-signaling proteins in LOW. Collectively, reduced ability for insulin to suppress FA mobilization, with accompanying modifications in intramyocellular LD size and distribution, and diminished IR interaction with key regulatory proteins may be key contributors to impaired insulin-mediated glucose uptake commonly found in adults with obesity. KEY POINTS: Although most adults with obesity exhibit impaired insulin-mediated glucose uptake (insulin resistance), some remain sensitive to insulin. Factors that “protect” adults with obesity from developing resistance to insulin-mediated glucose uptake are poorly understood. Potent suppression of fatty acid (FA) mobilization from adipose tissue by insulin is a strong predictor of whole-body insulin-mediated glucose uptake. Participants with higher sensitivity for insulin-mediated glucose uptake had smaller intramyocellular lipid droplets (LDs) within the subsarcolemmal region of type II skeletal muscle fibers. Novel findings revealed that insulin receptor (IR) interaction with the long-chain fatty acid transport protein, CD36, and the Src-family kinase, Fyn, directly associated with higher rates of glucose uptake under basal and hyperinsulinemic conditions. Together, the findings suggest impaired suppression of FA release from adipose tissue associates with reduced glucose uptake in skeletal muscle due in part to a defect in IR activation by CD36/Fyn and altered subcellular LD characteristics.