Circulating CD36 is increased in hyperlipidemic mice: cellular sources and triggers of release

Abstract

CD36 is a multifunctional transmembrane glycoprotein abundantly expressed in several cell types. Recent studies have identified CD36 in circulation (cCD36) in several chronic inflammatory diseases, including type 2 diabetes and chronic kidney disease, and proposed cCD36 to be a biomarker of disease activity. Whether cCD36 is present in hyperlipidemia, a condition characterized by oxidative stress and low-grade inflammation, is not known. In addition, the cellular origin of cCD36 and triggers of CD36 release have not been elucidated. We now demonstrate that plasma cCD36 level is increased in hyperlipidemic ApoE^−/− and Ldlr^−/− mice. Using several cell-specific CD36 knockout mice, we showed that multiple cell types contribute to cCD36 generation in hyperlipidemic conditions, with a particularly strong contribution from endothelial cells. In vitro studies have demonstrated that oxidized phospholipids, ligands for CD36 (oxPC_CD36), which are known to accumulate in circulation in hyperlipidemia, induce a robust release of CD36 from several cell types. In vivo studies have demonstrated CD36 release into the circulation of WT mice in response to tail-vein injection of oxPC_CD36. These findings document the presence of cCD36 in hyperlipidemia and identify a link between cCD36 and oxidized phospholipids generated under oxidative stress and low-grade inflammation associated with hyperlipidemia.

Graphical Abstract

Introduction

CD36 is a transmembrane glycoprotein and a member of the class B scavenger receptor family (1). CD36 is highly expressed on microvascular endothelial cells, platelets, macrophages, adipocytes, striated myocytes, mammary epithelial cells, and a lesser expression on several other cells, such as erythrocytes and hepatocytes. CD36 plays a role in lipid and lipoprotein metabolism, atherosclerosis, angiogenesis, thrombosis, and inflammation (2). Recent studies have identified a circulatory form of CD36 (cCD36) (3). cCD36 was first detected in the plasma of diabetes type 2 patients (3), but is also increased in patients with stage 5 chronic kidney disease and several others (3–6). Several studies have suggested that cCD36 might serve as a biomarker of disease activity in chronic inflammatory diseases, including diabetes, non-alcoholic steatohepatitis, and cardiovascular disease (7–11). While these findings suggest the potential importance of cCD36 in these disease conditions, the cellular sources, and triggers of release of cCD36 are still not established. Whether cCD36 is increased in hyperlipidemia, the major cause of cardiovascular disease, is also not known. We have previously demonstrated that hyperlipidemia and oxidative stress lead to accumulation in the circulation of oxidized phospholipids (oxPC_CD36), which are specific ligands for scavenger receptor CD36 (12). Oxidized phospholipids have multiple biological activities to a significant extent, mediated by CD36 (13). In this work, we tested whether oxPC_CD36 can trigger CD36 release from cells, leading to the generation of cCD36 under hyperlipidemic conditions.

In this study, we demonstrated that plasma cCD36 levels are increased in hyperlipidemic ApoE^−/− and Ldlr^−/− mice as compared to relatively normolipidemic control mice. Using several cell-specific CD36 knockout mice, we showed that multiple cell types produce cCD36 in hyperlipidemia with a particularly strong contribution from endothelial cells. In vitro studies demonstrated that specific groups of oxPC_CD36, γ-hydroxyl or γ-oxo alkenal phospholipids, induce a robust release of CD36 from several cell types expressing high levels of CD36. Furthermore, tail-vein injection of oxPC_CD36 into wild type (WT) mice induced CD36 release into the circulation.

Material and methods

Materials

Anti-hCD36 (AF1955) and anti-mCD36 (AF2519) antibodies, and Mouse CD36/SR-B3 DuoSet ELISA kit (DY2519) are from R&D systems (Minneapolis, MN). Amicon Ultra Centrifugal filters were purchase from MilliporeSigma (Burlington, MA). Different members of oxPC_CD36, 9-keto-12-oxo-10-dodecenoic acid ester of 2-lyso-phosphocholine (KODA-PC), 9-hydroxy-12-oxo-10-dodecenoic acid esters of 2-lyso-PC (HODA-PC), and 5-keto-8-oxo-6-octenoic acid esters of 2-lyso-PC (KOOA-PC) were synthesized, purified, and the structures were confirmed by multinuclear NMR and high resolution mass spectrometry as previously described (12,14,15). The purity of synthetic lipids was routinely analyzed by high-performance liquid chromatography (HPLC) with on-line electrospray ionization tandem mass spectrometry (LC/ESI/MS/MS). If lipids were found to be less than 98% pure, they were re-isolated prior to use. 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and palmitoyl-2-linoleoyl-sn-glycero-3-phosphocholine (PLPC) were from Avanti Polar Lipids (Alabaster, AL). All other reagents were purchased from Sigma-Aldrich if not mentioned otherwise.

Experimental animals

Wild type (WT) C57BL/6, ApoE^−/−, Pf4-Cre⁺, Adipoq-Cre⁺, LysM-Cre⁺, and Ldlr^−/− mice were purchased from Jackson Laboratories (Bar Harbor, ME). Tie2eCre mice were generated as mentioned earlier (16). CD36^−/−, ApoE^−/−-CD36^−/−, and CD36^fl/fl mice were generated as described earlier (17–19). To generate lines with cell type-specific deficiency of CD36 on ApoE^−/− background we initially crossed ApoE^−/− mice to CD36^fl/fl and several Cre⁺ mice including macrophage-specific (LysM-Cre⁺), adipocyte-specific (Adipoq-Cre⁺), and platelet-specific (Pf4-Cre⁺) mice. Cell-type-specific CD36 knockout mice were then generated by crossing ApoE^−/−-CD36^fl/fl to the corresponding ApoE^−/−-Cre⁺ mice line. Endothelial specific CD36 knockout mice on Ldlr^−/− background were generated using a similar scheme and Ldlr^−/−, Tie2eCre⁺, and CD36^fl/fl mice. We used sex-, age-, and genetic background-matched mice between 8 and 16 weeks of age in our experiments. We used the Cre⁻ littermates as controls for all genotypes. All strains were on the C57BL/6 background. Animals were housed in ventilated cages with ad libitum access to food and water, on a 14:10 light-dark cycle. To induce hyperlipidemia, ApoE^−/− mice were fed a Western diet (TD.88137, Envigo-Teklad) and Ldlr^−/− mice with high-fat high cholesterol diet (TD.96121, Envigo-Teklad) for 6–8 weeks. All animal procedures were approved by the Institutional Animal Care and Use Committee of the Cleveland Clinic.

Bone marrow transplantation

Bone marrow transplantation was performed using recipient ApoE^−/− mice and bone marrow donor ApoE^−/− and ApoE^−/−-CD36^−/− mice as described (17). Eight-week-old male or female recipient ApoE^−/− mice were lethally irradiated with a single dose of whole-body irradiation (900 rads) on the day of transplantation. Bone marrow cells from the donor 7-week- old male or female ApoE^−/− and ApoE^−/−/CD36^−/− mice were isolated and intravenously injected into each recipient mouse. Four weeks later, the mice were fed with a Western diet for 8 to 10 weeks and used for experiments.

Cells

Murine peritoneal macrophages (MPM) were isolated 48–72 hours after thioglycolate injection as described earlier (20). Platelets were isolated by gel filtration as described (21). Primary Human Dermal Microvascular Endothelial Cells; Normal, Neonatal (HDMVECn) cells were purchased from ATCC and cultured as recommended.

Cell differentiation

3T3L1 pre-adipocyte cells were purchased from ATCC and were cultured as recommended. Confluent cells were cultured in differentiation media containing 1μM dexamethasone (Cayman Chemical), 0.5mM 3-isobutyl-1-methylxanthine (IBMX, Cayman Chemical), and 10μg/ml insulin (Sigma-Aldrich) for 3 days. Then cells were cultured in maintenance media containing 10μg/ml insulin for 5–7 days and then used for experiments.

Platelet isolation

Platelet rich plasma (PRP) were separated and platelets were isolated by gel filtration from blood drawn from healthy human donors or mice as described earlier (17).

Tissue lysate preparation

Mice were perfused with 1X PBS and tissues were collected. Lysate were prepared with 1X RIPA buffer in the presence of protease inhibitors.

Blood collection and plasma separation

Blood was collected into anticoagulant EDTA for endothelial specific CD36 deficient mice and corresponding control mice, (Ldlr^−/−-Tie2eCre-CD36^fl/fl) or rest of the mice in to 3.8% citrate. Plasma was separated by centrifugation at 2000 rpm for 20 minutes.

Sample preparation

To remove cell debris and microparticles, cell culture media, platelet supernatant, and plasma samples were subjected to differential centrifugation using ultra centrifuge as described (22). Samples were centrifuged at 2000 g for 10 minutes at 4°C to remove the cells and cell debris, followed by 10,000 g for 30 minutes at 4°C to remove larger vesicles. Then the samples were centrifuged at 100,000 g for 90 min at 4°C to deplete all the small vesicles including exosomes. For electrophoresis and Western blot analysis the cell culture media and platelet supernatants were further concentrated using a 50kDa centrifugal filter and centrifuging at 10,000 rpm for 20 minutes at 4°C.

Immunoblotting

Immunoblotting was performed as described earlier (17). Platelets and cells were treated with oxPC_CD36. Platelets supernatant and cell culture media were collected and processed for sample preparation following above-mentioned method. Platelet and cell pellets were lysed on ice with RIPA buffer containing protease and phosphatase inhibitors. Lysates were centrifuged at 10000 rpm for 10 minutes at 4°C, and supernatants were collected. Protein concentrations were estimated using BioRad protein assay dye reagent and BSA as standard. Proteins were resolved on SDS-PAGE, transferred to PVDF membrane, and probed with protein-specific primary antibodies, followed by horseradish peroxidase (HRP)-conjugated specific secondary antibodies. Immunodetection was performed by enhanced chemiluminescence (ECL, GE Healthcare/Amersham, Buckinghamshire, UK) according to the manufacturer’s protocol.

ELISA

To measure the concentration of circulating CD36 in hyperlipidemic mice; microparticles free plasma was used to detect CD36 using ELISA kit for mCD36 (R&D systems, Minneapolis, MN) following manufacturer’s protocol. A 96 well plate was coated with capture antibody overnight at room temperature. Next day the wells were washed and blocked with blocking reagent for 1h at room temperature. Then the wells were washed and incubated with samples and different concentrations of standard mCD36 for 2h at room temperature. After that the wells were washed and incubated with detection antibody for 1h at room temperature followed by Streptavidin-HRP for 20 min at room temperature. After washing the wells substrate solution was added and incubated for 20 min at room temperature in dark and added the stop solution and mixed well. Optical density was measured at 450nm using a plate reader. The concentration of CD36 in the samples were determined comparing with the known concentrations of standard mCD36.

In vivo oxPC_CD36 injection

oxPC_CD36 (HODA-PC and KODA-PC) were mixed with PLPC in 1:1 molar ratio or PLPC alone in PBS and 0.1μm vesicles were prepared as described (23). Vesicles were injected into WT mice via tail vein injection to the final concentration of 6μM. After 6 hours the plasma of these mice was collected and analyzed for cCD36.

Statistical analysis

Data are presented as mean ±SD. The statistical significance of differences was evaluated using Student’s t-test. The p values less than 0.05 were considered to be statistically significant.

Results

Circulating CD36 is increased in two models of hyperlipidemic mice

To test whether circulating CD36 (cCD36) is increased in hyperlipidemic conditions, we used the two most commonly used mouse models for hyperlipidemia ApoE^−/− and Ldlr^−/− mice. ApoE^−/− mice were fed either a chow diet or a Western diet, and cCD36 levels in the plasma were assessed by Western blot analysis. cCD36 protein levels were significantly increased in the plasma of hyperlipidemic ApoE^−/− mice fed a Western diet as compared to relatively normolipidemic mice that received a chow diet (Fig 1A). A similar result of increased plasma cCD36 was observed in hyperlipidemic Ldlr^−/− mice fed a proinflammatory high-fat high cholesterol diet as compared to a chow diet-fed normolipidemic mice (Fig 1A). cCD36 band intensity was increased 4–5 folds in hyperlipidemic mice as compared to corresponding normolipidemic mice (Fig 1B). We also confirmed the elevated levels of cCD36 in hyperlipidemic mice using ELISA (Fig 1C) and demonstrated that the absolute levels of cCD36 in hyperlipidemic murine plasma reached 0.1μg/ml.

Figure 1. Hyperlipidemia induced plasma levels of circulating CD36.

(A, B) Blood was collected from Ldlr^−/− and ApoE^−/− mice fed with either a Western diet or high fat diet (hyperlipidemic/hyper) or a chow diet (normolipidemic/norm), plasma was isolated, cells and cell debris were removed by centrifugation as described in experimental procedures and circulating CD36 was detected using Western Blotting (A) or ELISA (n=3) (B). (A) Quantification (mean ± SD) of bands intensity in Western blots is shown in the right side panel (n≥4).

oxPC_CD36 trigger CD36 release from the cells

The triggers of CD36 release into the circulation from cells are not known, and the cellular mechanisms leading to the generation of cCD36 are poorly understood. Biologically active oxidized phospholipids, specific ligands for CD36 (oxPC_CD36), are known to accumulate in tissues and in the circulation of hyperlipidemic mice (21). We next tested whether oxPC_CD36 can induce CD36 release from CD36 expressing murine and human cells in vitro. A representative member of oxPC_CD36, HODA-PC, but not control unoxidized phospholipid (POPC) induced release of CD36 from murine peritoneal macrophages (MPM), cells that interact with oxPC_CD36 in vivo in atherogenesis (Fig 2A). oxPC_CD36 induced CD36 release from MPM in a dose-dependent manner at pathophysiologically relevant concentrations of oxPC_CD36 (Fig 2B) (21). CD36 release was time-dependent, with a significant increase within 1h and a robust increase at 6h of treatment (Fig 2C). The release of CD36 from cells was accompanied by the loss of cellular CD36 (Fig 2A–C). HODA-PC is a γ-hydroxy alkenal phospholipid and representative of oxPC_CD36. We showed that other γ-oxo alkenal phospholipids, members of oxPC_CD36 induced CD36 release from MPM (Fig 2D, data for KODA-PC and KOOA-PC are shown).

Figure 2. Oxidized phospholipids, ligands for CD36 induce CD36 release from cells.

(A-D) Murine Peritoneal Macrophages (MPM) were isolated from wild type mice (WT) and incubated with (A) 20μM of oxPC_CD36 (HODA-PC) or control non-oxidized phospholipid, POPC for 6 hours (h), (B) indicated concentrations (0–20μM) of oxPC_CD36 (HODA-PC) for 6h, (C) 20μM of oxPC_CD36 (HODA-PC) for indicated time (0–6h), and (D) 20μM of different members of the specific subgroup of oxPC_CD36, γ-hydroxyl alkenal phospholipids (HODA-PC) or γ-oxo alkenal phospholipids (KODA-PC and KOOA-PC). Media was depleted of particles by centrifugation as described in experimental procedures. CD36 was detected in the media and cell lysate by Western blot analysis. Actin is used as a loading control for cellular CD36. The quantification data of band intensity shown in bar graph (n≥3). * p<0.05, ** p<0.01, *** p<0.001, and ns p>0.05.

We then tested whether CD36 release in response to oxPC_CD36 is macrophage specific. We found that other major CD36 expressing cells, including primary human dermal microvascular endothelial cells (HDMVECn), platelets, and adipocytes (3T3L1), released CD36 in response to oxPC_CD36 whereas unoxidized phospholipid, POPC had no effect (Fig 3A–C).

Figure 3. Oxidized phospholipids induce CD36 release from various CD36 expressing cells.

(A) Primary human dermal microvascular endothelial cells (HDMVECn) were incubated with 20μM oxPC_CD36 (HODA-PC) or control POPC for 6h and CD36 was detected in the media and cell lysate by Western blot analysis. (B) Human platelets were isolated from platelet rich plasma by gel filtration and incubated with oxPC_CD36 (HODA-PC) and control POPC. Supernatant and lysate were analyzed for CD36 by Western blot analysis. (C) 3T3L1 preadipocytes were differentiated to adipocytes and then incubated with 20μM oxPC_CD36 (HODA-PC) or control POPC for 6h and CD36 was detected in the media and cell lysate by Western blot analysis. Actin is used as a loading control for cellular CD36. Right side panels show the quantification data of band intensity (mean ± SD of at least 3 independent experiments). The quantification data of band intensity shown in bar graph (n≥5). * p<0.05, ** p<0.01, *** p<0.001, and ns p>0.05.

Multiple CD36 expressing cells are involved in cCD36 generation in vivo

To identify the source of the cCD36 in hyperlipidemic mice, we assessed the contribution of several types of CD36 expressing cells. Microvascular endothelium, macrophages, adipocytes, and platelets belong to a small group of cells with high expression of CD36. Therefore, we generated macrophage-specific, adipocyte-specific, and platelet-specific CD36 knockout mice using CD36^fl/fl mice (19) and corresponding Cre^+/− mice (all on ApoE^−/− background). Contribution of endothelial CD36 was assessed using endothelial-specific knockout of CD36 on Ldlr^−/− background generated by crossing Ldlr^−/−-Tie2eCre^+/− mice to Ldlr^−/−-CD36^fl/fl mice. We have validated the cell-type-specific CD36 deficiency by genotyping (data not shown) and Western blot analysis (Fig 4A–D). No changes in the CD36 levels in non-targeted cells were detected (Supplementary Fig S1–S4). Mice were fed hyperlipidemic diet as described in experimental procedures, and the cCD36 levels in the plasma were assessed. We observed a significant reduction in plasma CD36 levels in endothelial-cell specific CD36 deficient mice (Ldlr^−/−-Tie2eCre⁺-CD36^fl/fl) as compared to its littermate control Ldlr^−/−-Tie2eCre⁻-CD36^fl/fl mice (Fig 4E). We also observed a noticeable reduction in plasma CD36 levels in macrophage-specific (ApoE^−/−-LysMCre⁺-CD36^fl/fl), and adipocyte-specific (ApoE^−/−-AdipoqCre⁺-CD36^fl/fl) CD36 deficient mice as compared to their corresponding control Cre-negative mice (Fig 4F, G). While all three groups showed a reduction in cCD36, the contribution of endothelial CD36 was particularly pronounced (~50% decrease) in cCD36 levels in Cre positive mice as compared to Cre negative mice (Fig 4E), whereas macrophages and adipocytes showed ~25% decrease (Fig 4F, G). Noticeable contribution of bone marrow (presumably macrophages) derived CD36 to cCD36 was confirmed using hyperlipidemic ApoE^−/− chimeras with either ApoE^−/−-CD36^−/− or ApoE^−/− bone marrow (Fig 4I). Interestingly, we detected little changes in cCD36 in platelet specific knockout of CD36 (ApoE^−/−-Pf4-Cre⁺-CD36^fl/fl mice, Fig 4H). These data indicate that multiple types of CD36 expressing cells are involved in the production of cCD36 in hyperlipidemic conditions with endothelial cells being the highest contributor.

Figure 4. oxPC_CD36 induce cCD36 generation in vivo involving multiple CD36 expressing cells.

(A-D) Cells and tissues were collected from different tissue-specific-CD36 deficient mice as indicated genotypes and lysates were prepared as mentioned in the experimental procedures and CD36 expression was detected by Western blot analysis showing the validation of cell type specific CD36 deficiencies. Ldlr^−/−/Tie2e-Cre⁺/CD36^fl/fl, ApoE^−/−/LysM-Cre⁺/CD36^fl/fl, ApoE^−/−/Adipoq-Cre⁺/CD36^fl/fl, and ApoE^−/−/Pf4-Cre⁺/CD36^fl/fl mice are labeled as Cre+ and Ldlr^−/−/Tie2e-Cre⁻/CD36^fl/fl, ApoE^−/−/LysM-Cre⁻/CD36^fl/fl, ApoE^−/−/Adipoq-Cre⁻/CD36^fl/fl, and ApoE^−/−/Pf4-Cre⁻/CD36^fl/fl mice are labeled as Cre-, protein ladder (M). (E-H) Plasma from the cell-type-specific mice (Tie2e-Cre+, endothelial cells specific; LysM-Cre+, macrophage specific; Adipoq-Cre+, adipocytes specific; Pf4-Cre+, platelet specific) as mentioned above was collected and assessed for circulatory CD36 by Western blot analysis. (I) ApoE^−/− bone marrow chimeric mice with ApoE^−/− (ApoE^−/−) or ApoE^−/−/CD36^−/− (ApoE^−/−/CD36^−/−) bone marrow fed with a Western diet. Plasma was collected as mentioned earlier and and assessed for circulatory CD36 by Western blot analysis. (J) Wild type (WT) mice were injected with non-oxidized control phospholipid, PLPC or oxPC_CD36 via tail vein injection. After 6 hours we collected the plasma from those mice and analyzed for circulating CD36 by Western blot analysis as mentioned earlier. The quantification data (mean ± SD) of band intensity of Western blots (n ≥ 4). * p<0.05, ** p<0.01, *** p<0.001, and ns p>0.05.

oxPC_CD36 triggers cCD36 generation in vivo

We then tested whether oxPC_CD36 can induce CD36 release in circulation in vivo. Wild type mice received intravenous injections of oxPC_CD36 (6μM) or control, unoxidized phospholipid (PLPC), plasma was collected 6h later, and the cCD36 levels were assessed by Western blot analysis. Injection of oxPC_CD36 induced the accumulation of cCD36 in the plasma, whereas control unoxidized phospholipid PLPC had no effect (Fig 4J). This result strongly suggests the direct role of oxPC_CD36 in the generation of cCD36.

Discussion

In this study, we demonstrated that hyperlipidemia in mice is associated with the accumulation of cCD36 that derives from several cellular sources. Furthermore, we have established that oxidized phospholipids, a lipid peroxidation product present in plasma in hyperlipidemia, triggers cCD36 release from various CD36 expressing cells in vitro and in vivo.

Phospholipids represent a major target for lipid peroxidation, and significant levels of oxidized phospholipids were detected in human atherosclerotic lesions, in the circulation of hyperlipidemic mice and patients with low HDL (12,21). More recently, oxidized phospholipids were also detected in several conditions with chronic inflammation (24–28). Oxidized phospholipids have multiple biological activities and can interact with several receptors on the cellular surface, but they are highly specific ligands for scavenger receptors class B, including CD36 (13,29,30). We hypothesized that oxidized phospholipids could be a link between hyperlipidemia, chronic inflammation, and cCD36. Our in vitro experiments demonstrated that oxPC_CD36 induce CD36 release from various cells, while native non-oxidized phospholipids had no effect. These results were confirmed in vivo using pathophysiologically relevant concentrations of oxidized phospholipids. Thus, our work identifies a novel biological activity of oxidized phospholipids - induction of the release of cellular CD36. We observed that various representatives of γ-hydroxy/oxo alkenal oxidized phospholipids induced cellular release of CD36. Interestingly, γ-hydroxy/oxo alkenal phospholipids can covalently modify CD36 (31). It remains to be established though, whether a covalent modification is required for the induction of CD36 release from cells. Taken together, our results strongly suggest that oxidized phospholipids represent a link between chronic inflammation and cCD36. They further suggest that increase in cCD36 is likely to be found in other cases of chronic inflammation.

Hyperlipidemia is associated with increased oxidative stress, lipid peroxidation, and accumulation of oxidized phospholipids in circulation and in tissues (24,32–35). To test a hypothesis that hyperlipidemic conditions are associated with the increase in CD36 in the circulation, we used ApoE^−/− and Ldlr^−/− mice, two most extensively studied mouse models of hyperlipidemia. Both of these models are characterized by chronic pro-oxidant stress and by high levels of individual oxidized phospholipids in circulation in hyperlipidemia (21,36). We detected increased cCD36 in both models. Western-type diet (ApoE^−/−) or proinflammatory high fat/high cholesterol diet (Ldlr^−/−) induced 4–5 folds increase in cCD36. The connection between hyperlipidemia and increased cCD36 was further confirmed in a limited amount of human samples.

Previous report demonstrated that circulating CD36 is not proteolytically cleaved and associated with a specific subfamily of microparticles instead of being soluble (37). At the same time, other studies showed that CD36 could undergo proteolytic cleavage and be released as a free form of CD36 (38,39). Our findings of cCD36 in microparticles depleted plasma and culture media suggest that oxPC_CD36 induced cCD36 might not be microparticles bound. Further mechanistic studies are warranted to characterize the cCD36, generated in response to oxPC_CD36.

A variety of cell types express CD36. Using several cell lines and primary cells, we have demonstrated that cCD36 release is not cell-specific, oxPC_CD36 induced cCD36 release in vitro from all cells tested. Macrophages, microvascular endothelial cells, adipocytes, and human platelets are known to express CD36 at very high levels and, correspondingly, are potential contributors to cCD36 in vivo. To identify the cellular sources of cCD36 in hyperlipidemia, we generated several conditional knockouts of CD36; we also used bone marrow chimeras to assess the combined role of bone marrow-derived cells. Our data indicate that in vivo endothelial cells likely contribute the most to cCD36, likely due to chronic exposure to circulating oxPC. We observed a more dramatic effect than expected in the bone marrow transplant experiment. While the exact reason for this is not clear, it can not be excluded that other bone marrow-derived CD36 expressing cells besides platelets and macrophages contribute to CD36 generation in hyperlipidemia. The significant contribution of bone marrow-derived cells, and specifically macrophages, as well as adipocytes was also detected indicating that cells outside of circulation still contribute to cCD36. Interestingly, while in vitro experiments showed that human platelets release CD36 in response to oxidized phospholipids, platelet specific CD36 deficient mice showed no significant drop in cCD36 accumulation in hyperlipidemic conditions. This is probably explained by a relatively low expression of CD36 in murine platelets compared to endothelial cells, macrophages, and adipocytes. It should be noted that human platelets express very high levels of CD36; thus, we can not exclude human platelet contribution to the generation of cCD36. cCD36 release from cells in vitro is correlated with a significant decrease in cellular CD36 levels in vitro. While the exact consequences of CD36 loss for the cell function are not clear, it is likely that such loss would alter the lipid metabolism, fatty acids and oxidized lipoproteins uptake, and inflammatory responses to lipid ligands by cells. It should be noted though that in vivo chronic loss of CD36 may be compensated by upregulation of CD36 synthesis stimulated by PPARγ ligands present in oxidized LDL (40). CD36 is critically involved in the sensing and metabolism of native and oxidatively modified lipids and lipoproteins and participates in a number of pathophysiological processes, including inflammatory and immune responses, atherosclerosis, thrombosis, diabetes, and others (2,12,21,41–45). While the role of CD36 as a marker of chronic inflammation was suggested (42,46–49), its role in inflammation is insufficiently understood. The CD36 ectodomain contains binding sites for several ligands, including oxidized phospholipids, and plays a critical role in CD36 function (50–53). Upon ligand binding, CD36 interacts with several co-receptors including toll-like receptors to induce intracellular signaling (17,26,45,54). A recent study showed that free, non-cell-associated CD36 ectodomain in complex with a bacterial ligand, interacted with cell surface TLR and induced pro-inflammatory responses (55). This finding suggests that cCD36 released by cells such as microvascular endothelium in response to oxidized phospholipids may promote an inflammatory response via TLR2 on cells that express no or low levels of CD36 (e.g. circulating immune cells, large vessel endothelial cells) and would otherwise be oblivious to oxidized phospholipids. Such interaction may enhance the pro-inflammatory responses in hyperlipidemia induced by endogenous as well as bacterial stimuli. Future studies are needed to test this hypothesis.

Conclusions

In conclusion, our studies revealed that in hyperlipidemic conditions in mice circulating CD36 in the plasma is highly elevated due to release from multiple cell types with a particularly strong contribution from endothelial cells. We further demonstrated that oxPC_CD36, a family of biologically active oxidized phospholipids ligands for CD36 (17), trigger the CD36 release from cells in vitro and in vivo. Taken together, these findings suggest that oxidized phospholipids serve as a mechanistic link between chronic inflammation and cCD36.

Highlights.

• The circulating CD36 in the plasma is highly elevated during hyperlipidemia.

• OxPC_CD36, ligands for CD36, trigger the CD36 release from cells in vivo and in vitro.

• Multiple cell types contribute to increased plasma levels of CD36 in hyperlipidemia.

• OxPC_CD36 serve as a mechanistic link between chronic inflammation & cCD36.

Abbreviations and nomenclature.

oxPC

Oxidized phospholipids

oxPC_CD36

Oxidized phospholipids, ligands for CD36

cCD36

Circulating CD36

HODA-PC

9-hydroxy-10-dodecenedioic acid esters of 2-lysoPC

KODA-PC

9-keto-12-oxo-10-dodecenoic acid ester of 2-lyso-phosphocholine

KOOA-PC

5-keto-8-oxo-6-octenoic acid ester of 2-lyso-phosphocholine

POPC

1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine

PLPC

1-palmitoyl-2-linoleoyl-sn-glycero-3-phosphocholine

ELISA

Enzyme-linked immunosorbent assay

HDMVECn

Primary Human Dermal Microvascular Endothelial Cells; Normal, Neonatal

MPM

Murine peritoneal macrophages

PBS

Phosphate buffer saline

PRP

Platelet rich plasma

RIPA buffer

Radioimmunoprecipitation assay buffer

ATCC

American Type Culture Collection

LDL

Low Density Lipoprotein

HDL

High Density Lipoprotein

PPARγ

Peroxisome proliferator-activated receptor γ

TLR

Toll-like receptor