FAM3A Inhibits Plasma Triglyceride Levels via Microvasculature‐Involved Enhancement of VLDL‐Triglyceride Hydrolysis and Lipid Uptake

ABSTRACT

FAM3 metabolism regulating signaling molecule A (FAM3A) is ubiquitously expressed. However, its biological role remains largely unknown. This study investigated the regulatory roles of FAM3A in plasma triglyceride (TG) levels. Proteomics revealed upregulated lipid metabolism‐related proteins, including CD36 and TCF15, in skeletal muscle of FAM3A‐transgenic mice. Clinically, plasma FAM3A was reduced in patients with metabolic disorders. FAM3A‐transgenic mice and recombinant FAM3A (rcFAM3A) injection in mice significantly reduced plasma TG levels, upregulated tissue lipid uptake and lipid transporter expression in skeletal muscle. Microvascular endothelial cells (MVECs) from soleus and heart of FAM3A‐transgenic mice or rcFAM3A‐infused mice exhibited increased fatty acid (FA) uptake and trans‐endothelial transport. Accordingly, FAM3A increased CD36 and TCF15 expression in MVECs derived from soleus, and increased CD36 expression was also observed in both C2C12 and MVECs treated by FAM3A in vitro. The siRNA‐mediated knockdown of TCF15 or CD36 effectively inhibited FA uptake. Notably, FAM3A increased the luminal localization and activity of lipoprotein lipase (LPL) in endothelium, enhancing plasma VLDL‐TG hydrolysis both in vivo and in vitro. However, FAM3A showed no effect on hepatic VLDL production or intestinal lipid absorption. Importantly, FAM3A exerted these effects primarily by activating PPARα signaling, as pharmacological blockage of PPARα with GW6471 inhibited FAM3A‐induced upregulation of CD36 and TCF15 expression, endothelial lumen LPL localization and activity, FA uptake and trans‐endothelial transport, and plasma TG clearance. Collectively, these findings reveal a pivotal role of FAM3A in plasma TG catabolism, identifying that the FAM3A/PPARα axis drives LPL‐mediated TG lipolysis, lipid‐transendothelial delivery, and tissue FA uptake.

FAM3A enhances the PPARα signaling, leading to upregulated expression of lipid transporters including CD36 and TCF15. The upregulated CD36 and TCF15 enhance FA uptake and trans‐endothelial delivery in MVECs. Meanwhile, FAM3A‐enhanced PPARα signaling also increases LPL expression in myocytes and LPL position in lumen side of microvasculature to promote the lipolysis of triglyceride (TG) and liberating fatty acids (FA) in plasma. The increased lipolysis, endothelial FA uptake, and trans‐endothelial FA transport ultimately result in a reduction of plasma TG level.

1. Introduction

Hypertriglyceridemia is an independent risk factor for cardiovascular disease apart from the well‐established risk factors such as elevated low‐density lipoprotein (LDL) and reduced high‐density lipoprotein (HDL) cholesterol levels. In plasma lipid metabolism, the triglyceride (TG)‐rich lipoproteins are prominently hydrolyzed by lipoprotein lipase (LPL). The hydrolysis of lipoprotein‐TG occurs mainly in the adipose, heart, and skeletal muscle tissues, generating fatty acids that are either utilized for fuel to produce ATP in striated muscle or stored in the form of TG in adipose tissue [1, 2]. LPL is abundant in the adipose, heart, and skeletal muscle tissues, but its transcripts and protein can be found in many other tissues [3, 4]. LPL is expressed in parenchymal cells, such as myocytes and adipocytes, assembled into head‐to‐tail homodimers, and then secreted into the surrounding interstitium. Here, LPL is further transported to the luminal face of capillary endothelial cells (wherein it is accessible to TG‐rich lipoproteins) and then processes TG‐rich lipoproteins. This critical step primarily relies on a series of transporters specifically expressed on the surface of endothelial cells, among which the scavenger receptor CD36 has been studied most extensively. It has been demonstrated that this molecule is pivotal in the uptake of long‐chain fatty acids [5, 6]. Peroxisome proliferator‐activated receptor α (PPARα), as a nuclear transcription factor, plays a central role in the regulation of gene expression, particularly genes involved in lipid metabolism, and has been closely linked to the pathogenesis of dyslipidemia [7]. Recent studies indicate that the activation of PPARα improves lipid metabolic disorders (MD) by upregulating genes associated with hepatic fatty acid oxidation (including LPL) and suppressing pathways related to lipogenesis [8]. However, despite the discovery of PPARα involved in LPL biology, there are many new questions to be resolved regarding the regulation of TG catabolism by microvasculature in the TG metabolism field.

Family with sequence similarity 3 (FAM3) is a novel cytokine‐like gene family identified in 2002 and includes four members designated FAM3 metabolism regulating signaling molecule A (FAM3A), FAM3B, FAM3C, and FAM3D [9]. Increasing evidence indicates that FAM3 proteins are important regulators of glucose and lipid metabolism and that they play a significant role in the pathogenesis of many human diseases such as diabetes and cancer [10, 11]. Previous studies have shown that FAM3B is involved in the pathogenesis of Type 2 diabetes via inhibition of pancreatic β cell function and hepatic insulin sensitivity [11]. The rodent FAM3A and FAM3B have similar expression patterns. However, the biological role of FAM3A remains largely unknown. FAM3A is composed of 230‐amino‐acid residues and is expressed in almost all tissues. The universal distribution of FAM3A in different tissues has indicated that the biological function of FAM3A is not restricted or tissue‐specific. Most specifically, FAM3A is expressed predominantly in the vascular endothelium, particularly capillaries. Recent research has found a link between FAM3A and lipid homeostasis, demonstrating that FAM3A suppresses hepatic lipogenesis by inducing pannexin 1 expression and subsequent ATP release [12]. FAM3A can also promote ATP synthesis and release by binding directly to the F1 component of F1Fo‐ATP synthase to maintain metabolic homeostasis [13, 14]. Additionally, activation of FAM3A by doxepin or imipramine treatment markedly ameliorated hyperglycemia, liver steatosis, and obesity in obese mice [15, 16]. Notably, these metabolic benefits are absent in FAM3A‐deficient mice, suggesting that the lipid and glucose metabolism regulatory effects of these drugs depend on FAM3A. Taken together, these findings suggest that FAM3A is a promising therapeutic target for lipid and glucose metabolism disorders. However, current research on FAM3A and lipid metabolism has predominantly focused on its function in the liver, while its impact on systemic lipid metabolism, particularly in plasma and capillaries, remains largely unexplored.

In the present study, we provided evidence that FAM3A regulated plasma TG concentrations. FAM3A significantly enhanced plasma VLDL‐TG hydrolysis by increasing the microvascular endothelial luminal localization and activity of LPL. Moreover, FAM3A altered lipid delivery across microvascular endothelial cells and subsequent tissue FA uptake in a PPARα‐dependent manner. These findings reveal a pivotal role of FAM3A in regulating plasma TG catabolism through coordinately upregulating the expression of lipid transporters and endothelial luminal LPL localization and activity.

2. Materials and Methods

2.1. Ethics Statement

All procedures involving experimental animals were approved by the Institutional Animal Care and Use Committee of the Ethics Review Board of Peking Union Medical College Hospital (Approval No. [XHDW‐2023‐083‐2]; Beijing, China).

2.2. Human Specimen Collection and Clinical Characteristics

A total of 100 patients with varicosity were included in this study (the basic clinical characteristics are provided in Table S1); 50 patients with varicosity and concomitant metabolic disorder (MD), including diabetes, hyperlipidaemia, or obesity, and 50 patients with varicosity without MD were included. Plasma samples were collected and separated from 5 mL peripheral blood samples by centrifugation at 2000 rpm for 10 min at 4°C within 2 h. The separated plasma was stored at −80°C until the detection of FAM3A levels using commercially available ELISA assays (Lifespan, #LS‐F35367).

The use of human specimens was approved by the Ethics Review Board of Peking Union Medical College Hospital (Beijing, China). Written informed consent for the use of plasma samples was obtained from all patients included in the study.

2.3. Generation of FAM3A‐Transgenic (FAM3A_Tg) Mice

The generation of systematic transgene mice of FAM3A (NM_001379181.1) was described in detail previously [17]. In brief, pronuclear microinjection of the CAG‐m‐FAM3A construct was performed in C57BL/6J fertilized oocytes. Transgene‐positive founder mice (F0) were screened by PCR (primers: FAM3A‐F: 5′GCTGGTTGTTGTGCTGTCTCATC and FAM3A‐R: 5′GCTGTCACGGAAGGCTAGATC) and outcrossed to wild‐type C57BL/6J mice to generate germline‐transmitted F1 offspring.

2.4. Animal Experiments

The C57BL/6J mice (Vital River, China) and FAM3A‐transgenic mice were kept in a SPF environment with a light‐, temperature‐, and humidity‐controlled environment (12‐h light/dark cycle at 20°C–22°C, humidity 45% ± 5%), and had free access to standard laboratory diet (Vital River, China) and water. Twelve‐ to fourteen‐week‐old mice were fed a synthetic high‐fat diet (HFD) containing 45% energy from lard (Researchdiet, Canada, #D12451) for 16 weeks.

To infuse FAM3A in vivo, the recombinant FAM3A protein (NM_021806; Origene, #TP303495) was slowly dissolved in sterile PBS at 4°C before use to a final concentration of 1 μg/mL. The recombinant protein solution was injected into C57BL/6J mice via the tail vein with 200 μL (containing 0.2 μg FAM3A protein) each time and once every 7 days throughout the HFD feeding period. For PPARα inhibition, mice received GW6471 (15 mg/kg/day; Topscience, #T8486) by oral gavage throughout the 16‐week HFD feeding period. GW6471 was initially dissolved in DMSO and then diluted with a vehicle containing 10% 1‐methyl‐2‐pyrrolidone and 5% Cremophor EL. Control mice received the corresponding vehicle according to the same dosing schedule. Mice were allocated to experimental groups in a balanced manner to minimize cage, batch, and handling biases; formal randomization and blinding were not performed.

To investigate the role of TCF15 in FAM3A‐mediated regulation in vivo, recombinant adeno‐associated virus serotype 9 (AAV9) vectors expressing an shRNA targeting mouse Tcf15 (AAV9‐sh_Tcf15) or a scrambled non‐targeting shRNA (AAV9‐sh_NC) were utilized. The vectors were constructed, packaged, and purified by Genomeditech (Shanghai, China). Wild type (WT) and FAM3A_Tg mice were first fed a HFD for 12 weeks. At week 12, mice were assigned to three experimental groups: (1) WT mice receiving AAV9‐sh_NC; (2) FAM3A_Tg mice receiving AAV9‐sh_NC; (3) FAM3A_Tg mice receiving AAV9‐sh_Tcf15. For all AAV‐based experiments, mice received a single intravenous injection of the respective AAV9 vector via the tail vein at a dose of 5 × 10¹¹ viral genomes (vg) in a final volume of 200 μL of sterile PBS, and were subsequently maintained on HFD for an additional 4 weeks. At the end of the 16‐week HFD intervention, soleus skeletal muscles were collected for subsequent analyses, and TCF15 knockdown efficiency was verified by Western blot.

For the analysis of the plasma metabolic parameters, mice were fasted for 8 h and blood samples were collected for analysis of the levels of TG (Abcam, #ab65336) using a colorimetric method. The LPL concentration in plasma or coronary perfusate was determined using mice LPL ELISA Kit (UscnLife, USA, #E0386m) according to the manufacturer's protocol. The values in the coronary perfusate were normalized to the heart weight.

2.5. KEGG Pathway Enrichment Analysis

Differentially expressed proteins between FAM3A_Tg and WT mice were identified from the normalized soleus muscle proteomic data using the limma package (version 3.56.2) in R. Proteins with a fold change (FC) ≥ 1.5 and p < 0.05 were considered differentially expressed. Gene symbols corresponding to the differentially expressed proteins were converted to Entrez Gene IDs using the org.Mm.eg.db (version 3.17.0) annotation package. KEGG pathway enrichment analysis was performed using the clusterProfiler package (version 4.8.3) with Mus musculus (mmu) as the reference organism. Enriched pathways were visualized using a dot plot, in which GeneRatio represents the proportion of differentially expressed proteins mapped to each pathway, dot size indicates the number of differentially expressed proteins (Count), and dot color represents the adjusted p value (p.adjust).

2.6. Gene Set Enrichment Analysis (GSEA)

Gene Set Enrichment Analysis (GSEA) was performed using the entire proteomic dataset. Gene symbols were converted to Entrez Gene IDs using the org.Mm.eg.db (version 3.17.0) annotation package. All quantified proteins were ranked according to the log₂ fold change of FAM3A_Tg versus WT mice, from the most upregulated to the most downregulated proteins. KEGG‐based GSEA was then performed using the gseKEGG function in the clusterProfiler package (version 4.8.3) with Mus musculus (mmu) as the reference organism. Enrichment profiles of representative pathways, including Fatty acid metabolism and PPAR signaling pathway, were visualized using the enrichplot package (version 1.20.1).

2.7. Isolated Heart Perfusion

To determine the change of LPL amount at the vascular lumen, an isolated heart perfusion was performed with Langendorff technique as described previously [2]. Briefly, mice were anesthetized and perfused with 15 mL of modified Tyrode's buffer followed by removal of the heart and cannulation of the aorta with the innominate artery tied below. Hearts were perfused retrogradely with 5 mL of modified Tyrode's buffer containing 1% FA‐free bovine serum albumin (BSA) and heparin (5 U/mL, Sigma) to measure coronary endothelium‐bound LPL. The coronary effluent was collected in timed fractions over 10 min and assayed for LPL mass using Western blot. LPL activity at the indicated times was measured using its activity assay kit.

2.8. Histological Immunofluorescence

The soleus tissue was fixed with 4% paraformaldehyde and then dehydrated and embedded in paraffin. Paraffin‐embedded tissues were cut into 4‐μm sections and prestained, and then incubated at room temperature with 5% normal goat serum for 20 min. After the sealed serum incubation, the tissue section was further incubated with primary antibody (anti‐CD36: Abcam, Cat#ab252923, 1:1000; anti‐TCF15: Novus, Cat#NBP2‐55818, 1:1000; anti‐CD31: Santa cruz, Cat#sc‐376 764, 1:500; anti‐LPL: Proteintech, Cat#28602–1‐AP, 1:500), overnight at 4°C in a wet box. Sections were then incubated with the appropriate fluorophore‐conjugated secondary antibodies for 60 min at room temperature in the dark. After the slices were slightly dried, the antifade mounting medium was added with DAPI (Origene, Cat#ZLI‐9600). The fluorescence‐stained sections were then visualized using Nikon ECLIPSE Ti2‐E.

2.9. Immunohistochemistry

The soleus tissue was cut to 4 μm sections and then sections were incubated with 3% hydrogen peroxide to block endogenous peroxidase activity. Tissue sections were blocked at room temperature with 5% normal goat serum for 20 min and incubated with primary antibodies anti‐CD36 (Abcam, Cat#ab252923, 1:1000), anti‐TCF15 (Novus, Cat#NBP2‐55818, 1:1000), at 4°C overnight. Then the HRP‐conjugated secondary antibody was added and incubated. The DAB substrate was added, and the sections were examined with light microscopy.

2.10. Cell Culture and Treatment

2.10.1. Isolation and Culture of Microvascular Endothelial Cells (MVECs)

The soleus skeletal muscle and heart tissues from WT or FAM3A transgenic mice were dissected and rinsed with ice‐cold PBS and dissected into very small pieces using sterile scalpels. All pieces were transferred into a gentle MACS C tube (Miltenyi Biotec, #130–096‐334) containing digestion medium (Thermo Fisher Scientific, #10829018); 1× penicillin/streptomycin (Thermo Fisher Scientific, #15140122); 2× Antibiotic‐Antimycotic (Thermo Fisher Scientific, #15240062); 1 mM sodium pyruvate (Thermo Fisher Scientific, #1360070); 1× MEM Non‐Essential Amino Acids Solution (Thermo Fisher Scientific, #11140035) supplemented with 0.2% collagenase IV (Worthington, #LS004188), 1.25 U/mL dispase (Thermo Fisher Scientific, #171055–041), 7.5 μg/mL DNase I (Sigma‐Aldrich, #D4527‐10KU) and 2 mM CaCl2. Samples were incubated in a 37°C water bath for 40 min, shaking the suspension every 10 min. At the end of the incubation time, the sample was further dissociated using the gentle MACS dissociator system (program C; pre‐programmed by manufacturer). The digested tissue was filtered first through a 70 μm cell strainer (Sigma‐Aldrich, #CLS431751‐50EA) and then through a 40 μm cell strainer (Sigma‐Aldrich, #CLS431750‐50EA). Samples were centrifuged at 300g for 15 min. Next, the cell suspension was enriched for endothelial cells using CD31 MicroBeads (Miltenyi Biotec, #130‐097‐418) according to the manufacturer's instructions. The CD31 enriched single cell suspension was washed with WB and stained with CD45‐PE‐Cy7 (clone 30‐F11 Thermo Fisher Scientific, #25‐0451‐82); CD31‐FITC (clone 390; Thermo Fisher Scientific, #11‐0311‐85) and Viable dye eFluor 450 (Thermo Fisher Scientific, #65‐0863‐18). FACS sorting was based on the CD31+CD45− fraction. Viable, CD31+, CD45‐ mouse soleus MVECs (msMVECs) and heart MVECs (mcMVECs) were sorted into collecting medium (10% FBS, Thermo Fisher Scientific, #A38401).

The commercial msMVECs or human cardiac MVECs (hcMVECs, Sciencell, USA, #6000) were cultured in ECM (Sciencell, USA, #1001) supplemented with ECGS (Sciencell, USA, #1052), 5% FBS (Sciencell, USA, #0025), and P/S solution (Sciencell, USA, #0503) in standard conditions (95% air/5% CO₂/37°C). For lipid‐loading experiments, recombinant FAM3A (NM_021806; Origene, #TP303495) or GW6471 (Topscience, #T8486) treatment experiments, cells of passages 2–3 were grown until confluence.

2.10.2. Culture of Myocytes

The commercially acquired mouse skeletal myocytes C2C12 (Fenghui, China, #CL0058) and MVECs from mouse skeletal muscle tissue (Procell, #CP‐M336) were cultured in DMEM supplemented with 10% FBS (Sciencell, USA, #0025) and P/S solution. For experiments, C2C12 cells and MVECs were treated with recombinant FAM3A (rcFAM3A) at a concentration of 200 ng/mL for 2 h, 6 h, and 12 h. For PPARα inhibition experiments, cells were pre‐treated with PPARα chemical inhibitor GW6471 (4 μM) for 30 min and subsequently stimulated with rcFAM3A (200 ng/mL) for 6 h.

2.11. Transfection of Small‐Interfering RNA

To investigate the roles of TCF15 and CD36 in FAM3A‐mediated endothelial FA uptake, small‐interfering RNA (siRNA) targeting mouse or human TCF15 or CD36, together with a non‐targeting control siRNA (si‐control), was transfected according to the manufacturer's instructions. Briefly, isolated mouse soleus or cardiac MVECs (msMVECs and mcMVECs) and commercial MVECs (msMVECs and hcMVECs) were plated at 40%–60% confluence, then transfected with siRNA duplexes at a final concentration of 20 nM using Lipofectamine RNAiMAX (Invitrogen, #13778150). After 6–8 h, the transfection medium was replaced with fresh culture medium. Cells were transfected with siRNAs for 24 h and subsequently subjected to high lipid (HL) conditions for 8 h, followed by the corresponding functional or molecular analyses.

2.12. In Vivo FA Uptake Assay

For the assessment of FA uptake in vivo, mice were injected intravenously with a 2 μCi dose of [1‐¹⁴C]‐Oleic Acid (¹⁴C‐OA; PerkinElmer) dissolved in saline (1:4). For the analysis of ¹⁴C‐OA in tissues, 30 min after intravenous injection, mice were anesthetized with an i.p. injection of tribromoethanol (0.2 mL/10 g) and perfused with a 0.9% saline solution. Soleus and heart were dissected, weighed, and solubilized by the addition of Solvable (PerkinElmer, 1 mL/100 mg tissue) in a glass scintillation vial and incubated at 60°C overnight. After cooling to room temperature, 30% (w/w) hydrogen peroxide was added, followed by heating at 60°C for another hour to minimize color quenching of samples. Finally, scintillation fluid (Normascint, Scharlab) was added to each sample, followed by vigorous shaking. The vials were then allowed to equilibrate in the dark for at least 60 min before scintillation counting using an LKB‐Wallac‐Rackbeta‐1214 Counter (PerkinElmer). Final data are expressed as % injected dose per gram of tissue.

2.13. In Vitro FA Uptake Assay

The MVECs were incubated with QBT reagent (BODIPY‐labeled dodecanoic acid coupled to a quenching element; Molecular Devices) in HBSS buffer +0.1% FA‐free BSA for 30 min to determine free FA uptake. Before FACS analysis, cultured cells were washed and harvested. Resuspended cells were washed with PBS several times. Then, the cells were resuspended in ice‐cold PBS and analyzed immediately by flow cytometry. The BODIPY‐FA‐uptake in cells was quantified by the corresponding mean fluorescence intensity (MFI).

2.14. In Vitro FA Transport Transwell Assay

FA transport was analyzed in cultured MVECs. The cells were seeded on 24‐well transwell inserts (0.4 μm pore size; Corning). The day after, medium was changed to ECM supplemented with 5% fetal bovine serum (FBS), EGCS, P/S, and 10 units/mL heparin. Then, 50 μCi/well ¹⁴C‐Palmitic acid (PerkinElmer) and 0.75 μg/mL Texas‐Red Dextran (MW 70000; Invitrogen) were added to the apical chamber. Eight minutes later, the entire lower liquid compartment was collected and used as sample for time zero. Thereafter, 50 μL samples were collected from the lower compartment at 30 min, 1 h, 2, 4, and 6 h and replaced with 50 μL of medium. For quantification of Dextran fluorescence, samples were measured with a fluorescence reader. For measurement of the ¹⁴C signal, samples were suspended in Emulsifier Safe Liquid Scintillation Cocktail (PerkinElmer) and total radioactivity was measured by liquid scintillation counting. Each experiment was performed in quadruplicate.

2.15. Endothelial Monolayer Permeability Assay

Endothelial monolayer permeability was measured as described previously [18]. Briefly, endothelial cells were treated with rcFAM3A and cultured in transwell plates. Fluorescein‐isothiocyanate‐labeled BSA (FITC‐BSA, 10 mg/mL, Sigma) was given to the cells. Cell culture medium from the lower compartment was removed after 60 min. FITC‐BSA was quantified in a fluorescence spectrofluorophotometer. BSA flux was valued as a ratio between fluorescence intensities in the lower compartment (60 min) and the upper compartment (0 min). The values were expressed as a percentage of control.

2.16. Real‐Time PCR

Total RNA from tissues was extracted with Trizol reagent (Sigma‐Aldrich, USA) and transcribed into cDNA using a PrimeScript RT Reagent Kit (Takara, Japan) following the manufacturer's instructions. The reaction mixture was taken as the template to perform real‐time PCR. The primers used were listed in Table S4.

2.17. Western Blot Analysis

Anti‐CD36, TCF15, LPL, PPARα, RXRa, β‐actin, GAPDH and HRP‐conjugated secondary antibodies were commercially obtained, and the detailed information for antibodies was listed in Table S5. Western blot was performed according to previous protocols widely described. To quantify protein signals, we subtracted background and normalized the value to β‐actin and GAPDH.

2.18. Postprandial TG Response

To determine the effect of FAM3A on the postprandial TG response, mice were injected by tail vein with rcFAM3A for 7 days. Mice were fasted for 5 h, and the blood samples were drawn (time = 0). Then, the mice received an intragastric load of 200 μL of olive oil. Additional blood samples were drawn at the time points of 1, 2, 4, and 6 h after the gavage, and the sera were assayed for TG. All the TG values are corrected for that at time = 0.

2.19. Intestinal TG Absorption

To measure intestinal lipid absorption, C57BL/6 mice were injected by tail vein with rcFAM3A for 7 days. FAM3A_Tg mice and rcFAM3A‐treated C57BL/6 mice were fasted overnight, and then 12.5 mg of Triton WR1339 (10% solution in saline) was injected via the tail vein to block lipolysis. One hour later, mice received an intragastric load of [³H]triolein (5 μCi) (GE Healthcare, Little Chalfont, UK) in 200 μL of olive oil. Blood samples were drawn from the tail before gavage (t = 0) and at 1 and 2 h after gavage, and plasma radioactivity was measured as described previously [19].

2.20. Hepatic VLDL‐TG Production

To determine the effect of FAM3A on the hepatic VLDL‐TG production rate, C57BL/6 mice were injected by tail vein with rcFAM3A for 7 days. FAM3A‐transgenic mice and rcFAM3A‐infused mice were fasted for 5 h, and then were anesthetized with isoflurane gas and injected via the tail vein with Trans³⁵S label (150 μCi/mouse, MP Biomedicals, USA, #51006). Half an hour later, 12.5 mg of Triton WR1339 (10% solution in saline) was injected to block lipolysis [20], and blood samples were drawn by tail bleeding (time = 0). Subsequent blood samples were taken at 10, 30, 60, and 90 min after injection of Triton WR1339 by tail bleeding. At the 120 min time point, mice were exsanguinated and the VLDL fraction of each mouse was isolated quantitatively from 400 μL of serum after density gradient ultracentrifugation [21]. VLDL‐TG was measured as described previously [22]. VLDL‐apoB was precipitated selectively by isopropanol [23] and counted for the incorporated ³⁵S.

2.21. LPL Activity Measurement

Mice were injected by tail vein with rcFAM3A for 7 days, FAM3A‐transgenic mice and rcFAM3A‐infused mice were fasted for 5 h, and plasma samples were collected 5 min after the tail vein injection of 300 U/kg heparin (Sigma). Postheparin LPL activity was measured using a LPL activity assay kit (Fluorometric) following the manufacturer's protocol (Cell Biolabs, CA, #STA‐610). Fluorescent TG, which is quenched by a trinitrophenyl group intramolecularly, was used as substrate. Hydrolysis by lipase produces pyrene fluorescence. Hepatic lipase activity can be distinguished in the presence of 1 mol/L NaCl, which totally abolishes LPL activity.

2.22. Preparation of VLDL‐Like Emulsions

VLDL‐like TG‐rich emulsion particles were prepared as previously described [24]. Briefly, 100 mg of total lipid (weight ratio of triolein: egg yolk phosphatidylcholine: lysophosphatidylcholine: cholesteryl oleate: cholesterol (70:22.7:2.3:3.0:2.0)) supplemented with [³H]triolein (50–200 μCi) (Perkin Elmer, USA, #NET431005MC) was sonicated with a Soniprep 150 at an output of 10 μm. An emulsion fraction containing 80‐nm sized emulsion particles was obtained by consecutive density gradient ultracentrifugation steps as described previously [25]. The TG content of emulsions was determined as described above.

2.23. Clearance of VLDL‐Like Emulsions In Vivo

WT and FAM3A‐transgenic mice were fed an HFD for 16 weeks. In parallel, C57BL/6J mice were fed an HFD for 16 weeks and received tail‐vein injections of rcFAM3A (0.2 μg in 200 μL PBS) or saline once every 7 days throughout the HFD feeding period. Tissue distribution of VLDL‐like emulsion particles was determined as detailed elsewhere [25]. The fed mice were anesthetized as described above, and their abdomens were opened. 200 μL of VLDL‐like [³H]triolein‐labeled emulsion particles were administered via the vena cava inferior at a dose (1.0 mg of TG) that exceeded the endogenous plasma TG content in both experimental groups. Fifteen min later, perirenal white adipose tissue, heart, liver, hind limb muscle tissue, and spleen were harvested and solubilized in Solvable (overnight at 60°C), and ³H radioactivity was determined [22].

2.24. Statistical Analysis

Data are presented as mean ± SEM. Statistical analyses were performed using IBM SPSS Statistics version 25.0. Data distribution was assessed using the Kolmogorov–Smirnov test. Comparisons between two independent groups were performed using a two‐tailed independent Student's t‐test. For comparisons among more than two groups, one‐way ANOVA followed by Tukey's post hoc test was used. Time‐course data with repeated measurements were analyzed using two‐way ANOVA, followed by Bonferroni's post hoc test where appropriate. All statistical tests were two‐sided, and p < 0.05 was considered statistically significant. Sample sizes were determined based on previous studies and the variability observed in preliminary experiments.

3. Results

3.1. FAM3A Is Associated With Lipid‐Metabolism and Reduces Plasma TG Levels

FAM3 family members have been reported to be active in metabolic processes. Our previous study has established that HFD led to decreased plasma FAM3A levels in mice [17]. In the present study, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis and Gene Set Enrichment Analysis (GSEA) of mass spectrometry‐based proteomic data from the soleus muscle of HFD‐fed C57BL/6 wild type (WT) and FAM3A‐transgenic mice further revealed that FAM3A may play crucial roles in regulating lipid metabolism, such as fatty acid metabolism and PPAR signaling pathway (Figure 1a,b). To determine the functional effect of FAM3A on systemic lipid metabolism in vivo, we measured plasma TG levels in FAM3A‐transgenic mice and recombinant FAM3A (rcFAM3A)‐infused mice for 7 days. The results showed that both FAM3A_Tg and rcFAM3A injection in mice exhibited significantly lower plasma TG concentrations (Figure 1c,d).

FIGURE 1.

FAM3A is associated with lipid metabolism. (a) KEGG pathway enrichment analysis. The bubble plot displays the top enriched biological pathways identified from the differentially expressed proteins based on proteomic data from mass spectrometry, which were determined in soleus muscle from the mice fed an HFD for ten weeks (biologically independent animals/group). The x‐axis represents the GeneRatio. (b) Gene Set Enrichment Analysis (GSEA) profiles. Enrichment plots are shown for the “Fatty acid metabolism” and “PPAR signaling pathway” gene sets. (c, d) The plasma TG levels were measured and graphed from the wild type (WT) mice and FAM3A_Tg mice fed with HFD for 16 weeks (c) and the C57BL/6 mice fed with HFD for 16 weeks and meanwhile treated with or without recombinant FAM3A (rcFAM3A, (c, d); n = 14 biologically independent animals/group). Data are presented as mean ± SEM. Statistical significance was calculated with a two‐tailed independent t test.

3.2. FAM3A Enhances In Vivo Lipid Transport Protein Expression and Lipid/FA Uptake

It has been generally acknowledged that plasma lipids such as triglyceride (TG) need to be broken down/lipolysis (by lipoprotein lipase (LPL)) into fatty acid (FA) and then pass through capillary endothelium (by carriers of lipid transporter) to enter parenchymal tissues to provide energy demands. Specifically, these lipid transport carriers are crucial in regulating plasma TG levels.

Based on the above high‐throughput screening we identified a series of lipid transport‐associated proteins among the term ‘fatty acid metabolism’, including cluster of differentiation 36 (cd36), fatty acid transport protein 4 (fatp4), and glycosylphosphatidylinositol‐anchored high‐density lipoprotein‐binding protein 1 (gpihbp1) (Figure 2a). Furthermore, we also included transcription factor 15 (TCF15) and meox2 as proteins associated with lipid transport, as the previous study indicated that TCF15 and meox2 facilitated cardiac FA transport by driving endothelial CD36 and LPL expression [26] (Figure 2a). The crucial lipid transport‐associated proteins were further assessed by qRT‐PCR. Significantly, cd36 and tcf15 were up‐regulated in FAM3A‐transgenic mice compared with WT controls (Figure 2b).

FIGURE 2.

FAM3A promotes tissue FA uptake. (a) A heatmap shows the differentially expressed proteins associated with fatty acid transport. (b) These crucial proteins were confirmed by qRT‐PCR by using soleus from C57BL/6 WT mice and FAM3A_Tg mice which were fed with HFD for 16 weeks (n = 5 biologically independent animals/group). (c, d) Western blot images and quantifications to evaluate the expression of CD36 and TCF15 in gross soleus tissues from the WT mice and FAM3A_Tg mice fed with HFD for 16 weeks (c) and the C57BL/6 mice fed with HFD for 16 weeks and meanwhile treated with or without rcFAM3A (d; n = 6 biologically independent animals/group; quantitative comparisons between samples were run on the same gel). (e, f) Representative images and quantifications of immunofluorescence (e) and immunohistochemical (f) staining identified CD36 and TCF15 protein in soleus skeletal muscles from WT mice and FAM3A_Tg mice fed with HFD for 16 weeks (n = 6 biologically independent animals/group). Scale bar: 100 μm in (e); and Scale bar: 500 μm, insets: 40 μm in (f). (g) The amount of ¹⁴C radioactivity denoted FA uptake in soleus and heart tissues from the WT mice and FAM3A_Tg mice which were fed with HFD for 16 weeks or the saline‐ and rcFAM3A‐treated C57BL/6 mice which were fed with HFD for 16 weeks. The value was expressed as % of injected ¹⁴C‐Oleic Acid (OA) dose (unit: ID) per gram (n = 6 biologically independent animals/group). Data are presented as mean ± SEM. Statistical significance was calculated with two‐tailed independent t test and p values are indicated (^ns p ≥ 0.05).

We further analyzed the effects of FAM3A on CD36 and TCF15. As illustrated in Figure 2, both FAM3A_Tg and rcFAM3A injection in mice promoted protein levels of CD36 and TCF15 in gross soleus tissues (Figure 2c,d). This trend was further proved by in situ immunofluorescent staining of localization of CD36 or TCF15 (Figure 2e; Figure S1) and by in situ immunohistochemical staining of CD36 or TCF15 in soleus from FAM3A‐transgenic mice (Figure 2f).

We further explored whether FAM3A could impact FA uptake in vivo. As expected, FA uptake was significantly increased in soleus and heart tissues in both FAM3A‐transgenic mice and rcFAM3A‐infused mice compared with their respective controls (Figure 2g). Together, these results suggest that FAM3A plays a crucial role in promoting FA uptake in vivo.

3.3. FAM3A Enhances Lipid Uptake in Microvascular Endothelial Cells and Lipid‐Transendothelial Delivery

In tissues, the lipid transport–related proteins are enriched in endothelium, which facilitate the lipid entry of parenchymal tissues to provide energy demands. We therefore assessed the role of FAM3A in lipid transport in microvasculature.

As expected, both FAM3A_Tg and rcFAM3A injection in mice increased the protein levels of CD36 and TCF15 in freshly isolated MVECs from soleus (Figure 3a).

FIGURE 3.

FAM3A induces lipid‐transendothelial delivery and FA uptake in microvascular endothelial cells. (a) Western blot images and quantifications to evaluate the expression of CD36 and TCF15 in isolated mouse soleus MVECs (msMVECs) from FAM3A_Tg mice and rcFAM3A‐treated C57BL/6 mice (a, n = 6 biologically independent group; quantitative comparisons between samples were run on the same gel). (b) Western blot images and quantifications to evaluate the expression of CD36 in skeletal muscle cell C2C12 and MVECs from mouse skeletal muscle tissue treated with or without rcFAM3A (b, n = 6 biologically independent group; quantitative comparisons between samples were run on the same gel). (c) Representative images of immunofluorescence staining of co‐localization of CD36 or TCF15 with CD31 protein in soleus skeletal muscles from WT mice and FAM3A_Tg mice fed with HFD for 16 weeks. Scale bar: 100 μm. (d, e) The msMVECs were isolated from soleus tissues in WT mice and FAM3A_Tg mice (d) and the commercial msMVECs were treated with rcFAM3A (200 ng/mL) or knockdown of TCF15 and CD36 by siRNA for 8 h under a high lipid (HL) condition (e). Quantification of mean fluorescence intensity of BODIPY labeled DA denoted the FA uptake by cultured msMVECs (d and e, n = 4 biologically independent samples/group). (f) Time course of ¹⁴C‐palmitic acid amount in the lower part of a transwell chamber was presented, and ¹⁴C‐palmitic acid value (relative to the signal at time zero) denoted the amount of FA transported through the monolayer of msMVECs from FAM3A_Tg mice and commercial msMVECs (n = 3 biologically independent samples/group). Data are presented as mean ± SEM. Statistical significance was calculated with two‐tailed independent t test (a, b, d), one‐way ANOVA followed by Tukey's post hoc test (e), or two‐way ANOVA followed by Bonferroni's post hoc test (f). p values are indicated (^ns p ≥ 0.05).

We further explored whether FAM3A could impact in vitro FA uptake in and transport across MVECs. Significant increases in CD36 expression were also confirmed by in vitro experiments both in skeletal muscle cell C2C12 and MVECs from mouse skeletal muscle tissue after stimulation of rcFAM3A (Figure 3b). This observation was further demonstrated by in situ immunofluorescent staining of co‐localization of CD36 or TCF15 with the endothelial‐specific marker CD31 (Figure 3c; Figure S2a).

As measured by fluorescence detecting, uptake of fluorescently labeled dodecanoic acid was significantly increased in cultured MVECs from soleus (msMVECs) and heart (mcMVECs) of FAM3A‐transgenic mice (Figure 3d; Figure S2b) and in cultured commercial msMVECs and human cardiac MVECs (hcMVECs) after stimulation of rcFAM3A (Figure 3e; Figure S2c) in contrast with their parallel controls. Notably, knockdown of TCF15 or CD36 by small interfering RNA (siRNA) significantly inhibited FA uptake in vitro (Figure 3e; Figure S2c). Moreover, AAV9‐mediated knockdown of TCF15 repressed FAM3A‐induced upregulation of CD36 expression and FA uptake in vivo (Figure S3). Correspondingly, an in vitro transwell assay revealed that both FAM3A transgenic and rcFAM3A treatment (Figure 3f; Figure S2d) facilitated FA transport across an endothelial monolayer, suggesting that CD36 and TCF15 may be involved in FAM3A's function in FA uptake and lipid‐transendothelial delivery.

Moreover, we observed that rcFAM3A showed no effect on endothelial monolayer permeability (Figure S2e), excluding the possibility that these effects are caused by altered monolayer leakiness. Together, these findings suggest that FAM3A may induce FA trans‐endothelial delivery via the promotion of CD36 and TCF15.

3.4. FAM3A Promotes LPL Localization in Microvascular Lumen and Plasma VLDL‐TG Lipolysis In Vivo

In plasma lipid metabolism, the triglyceride (TG)‐rich lipoproteins are prominently hydrolyzed by lipoprotein lipase (LPL). In order to be accessible to TG‐rich lipoproteins, it is necessary that LPL is transported to the luminal face of capillary endothelial cells. At the lumen, LPL actively metabolizes the TG core of lipoproteins to FA which is used as fuel within tissues. To determine the position of LPL at the vascular lumen, heart tissues from FAM3A‐transgenic mice were isolated and perfused in the Langendorff mode with heparin to release coronary luminal LPL. As shown, the LPL amount at the vascular lumen was markedly increased in FAM3A‐transgenic mice compared with control ones (Figure 4a,b), Furthermore, LPL activity in the perfusate also increased accordingly in FAM3A‐transgenic mice, significantly at 1–2 min of the early perfusion time (Figure 4c), indicating a stimulating effect of FAM3A on microvascular lumen LPL localization.

FIGURE 4.

Effect of FAM3A on coronary luminal LPL amount and localization and VLDL‐TG lipolysis. (a, b) FAM3A_Tg mice were fed with HFD for 16 weeks, then hearts were removed and perfused in the Langendorff mode with heparin (5 U/mL) to release coronary luminal LPL. LPL amount (a) and concentration (b) in the perfusate were determined using Western blot and ELISA, the LPL amount was quantified as fold change above WT control (n = 5 biologically independent samples/group). (c) LPL activity was determined at the indicated times (n = 5 biologically independent samples/group). (d, e) Postheparin plasma LPL concentration and activity were measured (n = 8 biologically independent samples/group). (f) C57BL/6 WT mice and FAM3A_Tg mice were fed with HFD for 16 weeks and the lpl gene expression levels which were evaluated by qRT‐PCR are shown (n = 5 biologically independent animals/group). (g–i) Western blot images and quantifications to evaluate the expression of LPL in isolated mouse soleus and heart from FAM3A_Tg mice (g) and rcFAM3A‐treated C57BL/6 mice (h), and in C2C12 treated with or without rcFAM3A (n = 6 biologically independent group; quantitative comparisons between samples were run on the same gel). (j) Representative images and quantifications of immunofluorescence staining identified LPL protein in soleus skeletal muscles from WT mice and FAM3A_Tg mice fed with HFD for 16 weeks (n = 6 biologically independent animals/group). Scale bar: 100 μm. (k, l) FAM3A_Tg mice and rcFAM3A‐treated C57BL/6 mice received a large intravenous bolus of [³H]triolein‐labeled emulsion particles (1.0 mg of TG). Fifteen min later, mice were killed and the tissue distribution of ³H radioactivity was determined (n = 5 biologically independent animals/group). pWAT denotes perirenal white adipose tissue. Data are presented as mean ± SEM. Statistical significance was calculated with two‐tailed independent t test (a, b, d–j) or two‐way ANOVA followed by Bonferroni's post hoc test (c). p values are indicated (^ns p ≥ 0.05).

We further studied the effect of FAM3A on the postheparin plasma LPL amount and activity by analyzing the fasted plasma from FAM3A‐transgenic mice and rcFAM3A‐infused mice. Both FAM3A_Tg and rcFAM3A injection in mice promoted the plasma LPL amount and activity compared with their parallel control groups (Figure 4d,e). Meanwhile, qRT‐PCR results showed that lpl was up‐regulated in FAM3A‐transgenic mice compared with WT controls (Figure 4f). Expectedly, the protein level of LPL in soleus and heart was significantly increased in both FAM3A‐transgenic mice and rcFAM3A‐infused mice (Figure 4g,h). Significant increases in LPL expression were also confirmed by in vitro experiments in skeletal muscle cell C2C12 after stimulation of rcFAM3A (Figure 4i). This result was further confirmed by the in situ immunofluorescent staining of LPL (Figure 4j).

We further questioned whether FAM3A enhanced LPL‐mediated lipolysis. We therefore assessed whether FAM3A influenced the amount of FA dissociated from the plasma VLDL‐TG in vivo by detecting the tissue distribution of [³H]oleate derived from intravenously injected [³H]triolein‐labeled VLDL‐like emulsion particles. In line with the results that FAM3A enhanced postheparin plasma LPL amount and activity, both FAM3A_Tg and rcFAM3A injection in mice increased uptake of ³H radioactivity in tissues including heart, liver, skeletal muscle, and pWAT (Figure 4k,l). In contrast, the uptake by the spleen, which mainly reflects whole particle uptake due to the absence of LPL activity in the spleen, remained unchanged. Collectively, these findings indicate a potency of the stimulating effect of FAM3A on LPL lipolytic function and plasma VLDL‐TG clearance in vivo.

3.5. FAM3A Promotes Postprandial TG Clearance Without Effect on Intestinal TG Absorption

We further explored the mechanisms underlying the FAM3A‐induced reduction in plasma TG levels. In general, the balance between TG input and output determines its final amount in the plasma pool. There are two main sources of TG in plasma, namely exogenous: the fat ingested from food is absorbed by intestinal mucosa, and endogenous: TG is synthesized by the body (mainly in the liver). Since the plasma TG clearance process was evaluated above, we next explored the impact of FAM3A on exogenous and endogenous sources of plasma TG production.

We questioned whether the effect of FAM3A on TG reduction was caused by enhanced lipolytic conversion of TG‐rich lipoproteins. Therefore, we detected whether FAM3A affected postprandial lipemia, anticipating that FAM3A would stimulate LPL activity and therefore enhance catabolism of TG‐rich lipoproteins. For this purpose, mice were injected with recombinant FAM3A and then were given an intragastric gavage of olive oil to force a rapid and extensive chylomicron production. Recombinant FAM3A‐infused mice showed a significant lowering of the postprandial TG response compared with control (Figure 5a). These findings indicated that FAM3A in mice resulted in a rapid clearance of chylomicron‐TG. Because FAM3A is also present in the intestine, it is likely that the reduced postprandial TG response is induced by a direct negative effect of FAM3A on intestinal lipid absorption. To exclude this possibility, we measured intestinal lipid absorption levels. However, it seemed that FAM3A showed no effect on intestinal lipid absorption (Figure 5b). These findings suggest that the decreased postprandial TG response may be attributed to a stimulating effect of FAM3A on LPL‐mediated lipolysis.

FIGURE 5.

Effect of FAM3A on postprandial TG clearance, intestinal TG absorption and hepatic VLDL production. (a) Mice were fed a HFD for 16 weeks, and then were injected consecutively with rcFAM3A for 7 days. The mice were fasted and received an intragastric olive oil load. Plasma samples were drawn just before the gavage (time = 0) and at the indicated time points after the gavage and assayed for TG. TG values are corrected for that at time = 0 (n = 5 biologically independent animals/group). (b) C57BL/6 mice were injected by tail vein with rcFAM3A for 7 days. FAM3A_Tg mice and rcFAM3A‐infused mice were fasted overnight and injected with Triton WR1339, and then gavaged with 200 μL of olive oil containing 5 μCi of [³H]triolein. Plasma samples (10 μL) were collected to measure radioactivity to determine intestinal lipid absorption (n = 5 biologically independent animals/group). (c–h) C57BL/6 mice were injected by tail vein with rcFAM3A for 7 days. FAM3A_Tg mice and rcFAM3A‐infused mice were fasted overnight. The fasted animals received intravenous injections of Trans³⁵S label and Triton WR1339 to block lipolysis. Plasma samples were taken within 90 min after injection and analyzed for TG (c and f, n = 5 biologically independent animals/group). After 120 min, mice were exsanguinated and plasma VLDL was isolated and assayed for TG (d and g, n = 5 biologically independent animals/group) and [³⁵S]apoB (e and h, n = 5 biologically independent animals/group). Data are presented as mean ± SEM. Statistical significance was calculated with two‐way ANOVA followed by Bonferroni's post hoc test (a–c, f) or two‐tailed independent t test (d, e, g, h). p values are indicated (^ns p ≥ 0.05).

3.6. FAM3A Shows no Effect on VLDL Production in Liver

Next, we explore the effect of FAM3A on hepatic VLDL‐TG production from endogenous viewpoints. We blocked lipolysis and the clearance of VLDL with Triton WR1339 and assessed hepatic VLDL production rate to determine whether hampered hepatic VLDL production contributes to the reduced TG level in plasma. However, neither FAM3A‐transgenic mice nor rcFAM3A‐infused mice significantly affected the VLDL‐TG secretion rate in plasma (Figure 5d,g). Moreover, the VLDL‐[³⁵S]apoB production rate was also unchanged (Figure 5e,h). Apparently, FAM3A had no effects on the intrahepatic lipidation of apoB and VLDL particle production. Additionally, an assessment of VLDL isolated 2 h after Triton WR1339 injection indicated no significant changes in VLDL lipid content. Neither FAM3A_Tg nor rcFAM3A injection in mice affected plasma VLDL‐TG content (Figure 5c,f), when lipolysis and the clearance of VLDL were blocked. Therefore, these results revealed that FAM3A has no effects on lipidation of apoB and VLDL‐TG content within hepatic VLDL production.

3.7. FAM3A/PPARα Axis Drives Lipid‐Transendothelial Delivery and FA Uptake in Microvascular Endothelial Cells

We further explored the underlying mechanisms involved in the biological function of FAM3A. PPARα is a main PPAR isoform in skeletal muscle tissues. Both FAM3A overexpression in mice and stimulation in cells enhanced PPARα protein levels (Figure 6d–f). Moreover, blockade of PPARα by GW6471 abolished FAM3A‐induced promotion of plasma LPL amount and activity in both FAM3A‐transgenic mice and rcFAM3A‐infused mice (Figure 6a,b). Meanwhile, significant decreases in lpl expression were also confirmed by qRT‐PCR experiments in FAM3A‐transgenic mice after inhibition by GW6471 (Figure 6c). As detected by Western blot, the protein levels of PPARα were up‐regulated in FAM3A transgenic mice compared with WT controls (Figure 6d). The in vitro study also confirmed the role of FAM3A in promotion of PPARα protein level both in skeletal muscle cell C2C12 and MVECs from mouse skeletal muscle tissue (Figure 6e,f). We therefore questioned whether those metabolic actions of FAM3A were due to enhanced PPARα signaling. As expected, both the in vivo and in vitro results showed that blockade of PPARα by GW6471 abolished FAM3A‐induced promotion of CD36 and LPL. Correspondingly, this trend was further proved by in situ immunofluorescent staining of localization of CD36, TCF15 and LPL (Figure 6g and Figure S4a).

FIGURE 6.

FAM3A affects lipid‐transendothelial delivery and FA uptake through enhancing PPARα signaling. (a, b) Postheparin plasma LPL concentration and activity were measured in FAM3A_Tg mice (a) and rcFAM3A‐infused C57BL/6 mice (b) which were injected with GW6471 and fed with HFD for 16 weeks (n = 8 biologically independent samples/group). (c) The lpl gene expression levels were evaluated in FAM3A_Tg mice injected with GW6471 and fed with HFD for 16 weeks (n = 5 biologically independent samples/group). (d) Western blot images and quantifications to evaluate the expression of PPARα and CD36 in isolated mouse soleus from FAM3A_Tg mice and rcFAM3A‐infused C57BL/6 mice which were injected with GW6471 and fed with HFD for 16 weeks. (e) Western blot images and quantifications to evaluate the expression of PPARα and CD36 in MVECs from mouse skeletal muscle tissue treated with GW6471. (f) Western blot images and quantifications to evaluate the expression of PPARα and LPL in C2C12 treated with GW6471 (d–f, n = 6 biologically independent group; quantitative comparisons between samples were run on the same gel). (g) Representative images and quantifications of immunofluorescence staining identified localization and expression of TCF15 and LPL in soleus skeletal muscles from FAM3A_Tg mice injected with GW6471 and fed with HFD for 16 weeks (n = 6 biologically independent animals/group). Scale bar: 100 μm. h The amount of ¹⁴C radioactivity denoted FA uptake in soleus and heart tissues from FAM3A_Tg mice and rcFAM3A‐infused C57BL/6 mice which were injected with GW6471 and fed with HFD for 16 weeks. The value was expressed as % of injected ¹⁴C‐oleic acid (OA) dose (unit: ID) per gram (n = 6 biologically independent animals/group). (i, j) The msMVECs were isolated from soleus tissues in FAM3A_Tg mice injected with GW6471 and fed with HFD for 16 weeks (i) and the commercial msMVECs were pre‐treated with GW6471, and then stimulated with rcFAM3A (j, 200 ng/mL). Quantification of mean fluorescence intensity of BODIPY labeled DA denoted the FA uptake by cultured MVECs (n = 4 biologically independent samples/group). (k) Time course of ¹⁴C‐palmitic acid amount in the lower part of a transwell chamber was presented, and ¹⁴C‐palmitic acid value (relative to the signal at time zero) denoted the amount of FA transported through the monolayer of msMVECs from FAM3A_Tg mice and commercial msMVECs treated with GW6471 (n = 3 biologically independent samples/group). Data are presented as mean ± SEM. Statistical significance was calculated with two‐tailed independent t test (a–j) or two‐way ANOVA followed by Bonferroni's post hoc test (k). p values are indicated (^ns p ≥ 0.05).

We further investigated the regulatory roles of the FAM3A‐PPARα axis in FA uptake and in FA transport across endothelial monolayer in vivo and in vitro. As shown, the improved FA uptake effects of FAM3A were repressed significantly by GW6471 treatment in soleus and heart tissues in both FAM3A‐transgenic mice and rcFAM3A‐infused mice (Figure 6h). Moreover, the increased uptake of fluorescently labeled dodecanoic acid induced by FAM3A was abolished by GW6471 treatment in cultured msMVECs and mcMVECs of FAM3A‐transgenic mice (Figure 6i; Figure S4b) and in cultured commercial msMVECs and hcMVECs after stimulation of rcFAM3A (Figure 6j; Figure S4c). Furthermore, GW6471 treatment also abolished FAM3A‐induced lipid‐transendothelial delivery in cultured commercial msMVECs and hcMVECs (Figure 6k; Figure S4d). Additionally, PPARα and RXR usually form heterodimers to regulate gene expression. Interestingly, we observed that FAM3A_Tg mice had a higher RXRa protein level in the muscles than WT mice (Figure S5).

Taken together, these findings suggested that FAM3A promotes lipid‐transendothelial delivery and FA uptake may be through enhanced PPARα signaling.

3.8. Inhibition of PPARα Abolishes the TG‐Lowering Effect of FAM3A

To determine whether the lipid‐lowering effect of FAM3A depends on PPARα signaling, we evaluated plasma TG levels after pharmacological inhibition of PPARα. The FAM3A‐induced decrease in plasma TG was abolished in both FAM3A‐transgenic and rcFAM3A‐infused mice treated with GW6471 (Figure 7a,b). Furthermore, clinical data revealed that patients with varicosity and concomitant MD exhibited significantly lower plasma FAM3A levels compared to varicosity controls without MD (Figure 7c). Together, these results demonstrate that FAM3A decreased plasma TG levels through a PPARα‐dependent mechanism, which coordinately upregulates lipid uptake, lipid transendothelial delivery, and LPL‐mediated lipolysis via activation of PPARα signaling, ultimately leading to enhanced plasma TG clearance.

FIGURE 7.

TG‐Lowering effect of FAM3A depends on PPARα signaling. (a) The plasma TG levels were measured and graphed in the WT mice and FAM3A_Tg mice which were treated with GW6471 and fed with HFD for 16 weeks. (b) The plasma TG levels were measured and graphed in the rcFAM3A‐infused C57BL/6 mice which were treated with GW6471 and fed with HFD for 16 weeks (a and b, n = 10 biologically independent animals/group). (c) The plasma FAM3A levels were measured and graphed in patients with varicosity with metabolic disorder (MD), including diabetes, hyperlipidaemia, or obesity (patient_MD, n = 50), or without MD (patient_ctl, n = 50). Data are presented as mean ± SEM. Statistical significance was calculated with two‐tailed independent t test. p values are indicated (^ns p ≥ 0.05).

4. Discussion

In this study, we have demonstrated that FAM3A decreased plasma TG levels. The reduced plasma TG is due to enhanced plasma VLDL‐TG hydrolysis by LPL. Meanwhile, FAM3A facilitated lipid‐transendothelial delivery via CD36 and TCF15 and enhanced subsequent tissue uptake of liberated fatty acids (FA). We further propose that the position of LPL in the microvasculature lumen and the increase of FA uptake are mediated in a PPARα‐dependent manner. These findings reveal a pivotal role of the FAM3A/PPARα axis in regulating VLDL‐TG hydrolysis and FA uptake.

Based on the known metabolic functions of FAM3B and their similar expression patterns, we speculated that FAM3A may also function in metabolic regulation just like FAM3B. However, few studies have addressed FAM3A properties involved in metabolism‐relevant events recently. Wang et al. reported that FAM3A attenuated hyperglycemia, insulin resistance, and fatty liver in diabetic mice [13], suggesting that FAM3A is protective in the liver with MD. In the present study, we have provided experimental proof for the plasma TG‐lowering effect of FAM3A. Both FAM3A_Tg and recombinant FAM3A (rcFAM3A) injection in mice significantly reduced plasma TG levels (Figure 1). In addition, we found that both FAM3A_Tg and rcFAM3A injection in mice significantly enhanced tissue lipid uptake in skeletal muscle and heart tissues and upregulated the expression of lipid transporters in skeletal muscle (Figure 2). Consistent with these in vivo observations, MVECs isolated from the soleus of FAM3A‐transgenic mice and rcFAM3A‐infused mice exhibited increased FA uptake (Figure 3d,e) and upregulated the expression of lipid transporters including CD36 and TCF15. Meanwhile, FAM3A increased the expression of endothelial lipid transporter in both C2C12 myocytes and MVECs. Collectively, these results demonstrate that FAM3A enhances lipid transport and FA uptake in vivo and in vitro.

Although the proteins involved in FA uptake in and transport across endothelial cells may play crucial roles in pathological events [27, 28], currently little is known about the factors or the mechanisms that mediate in FA shuttling from the lumen to the basal side of endothelial cells [26, 29, 30], with the exception of peroxisome proliferator‐activated receptor (PPAR)‐γ [31], vascular endothelial growth factor (VEGF)‐B [32, 33], CD36 [34], Meox2, and TCF15 [26]. VEGF‐B has been identified as a paracrine factor in charge of FA uptake through endothelial cells in adipose tissue, skeletal muscle, and heart [32, 33]. CD36 was the FA transporter expressed at the highest level in heart endothelial cells [34]. Meox2 and TCF15 were exclusively expressed in endothelial cells, and they are critical regulators of FA transport across cardiac endothelial cells [26]. In our study, the observed upregulation of CD36 and TCF15 strongly supports the view that FAM3A promotes the expression of these key lipid transporters. Meanwhile, both FAM3A transgenic and rcFAM3A treatment facilitated FA transport across an endothelial monolayer (Figure 3f; Figure S2d). Taken together, CD36 and TCF15 play an important role in mediating FAM3A's function in FA uptake and lipid‐transendothelial delivery.

In the present study, we explored the mechanisms underlying the plasma TG‐lowering effect of FAM3A. In plasma, endogenous TG is transported by VLDL, which is assembled in the liver from triglycerides, cholesterol, and apolipoproteins. However, neither FAM3A‐transgenic mice nor rcFAM3A‐infused mice showed significant effects on the VLDL‐TG secretion rate in plasma, lipidation of apoB, and TG content in VLDL (Figure 5c–h), indicating that the reduced TG level in plasma is not attributed to hampered hepatic VLDL production. These findings suggest that altered hepatic VLDL production is unlikely to be the primary contributor to the TG‐lowering effect of FAM3A. However, Wang et al. proposed that hepatic overexpression of FAM3A repressed FA synthesis and increased VLDL‐TG secretion, therefore attenuating hepatic steatosis in Type 2 diabetic mice [13]. It seems that these findings are controversial. However, in their study, increased VLDL‐TG secretion was observed under the specific conditions of FAM3A overexpression restricted to the liver and Type 2 diabetes. We speculate that FAM3A functions in different cells, tissues, and organs with crosstalk may lead to local or whole consequences. Additionally, maybe hormones, especially metabolic ones, such as insulin, are crucial for FAM3A efficacy.

Undoubtedly, we also questioned whether the TG‐lowering effect of FAM3A was caused by enhanced lipolytic conversion of TG‐rich lipoproteins. Following an intragastric fat load, a diminished postprandial hypertriglyceridemia was observed in rcFAM3A‐infused mice (Figure 5a), indicating that FAM3A may result in a rapid clearance of chylomicron‐TG. Moreover, we excluded a negative effect of FAM3A on intestinal lipid absorption (Figure 5b), also supporting the potency of the stimulating effect of FAM3A on TG‐rich lipoprotein catabolism. LPL is the main enzyme responsible for the lipolytic processing of TG‐rich lipoproteins in plasma, the proper regulation of which is important for controlling the delivery of lipid nutrients to tissues. We therefore explored the effect of FAM3A on LPL lipolytic function. Firstly, using the langendorff heart perfusion mode, we observed that hearts from FAM3A‐transgenic mice on an HFD exhibited increased postheparin LPL amount at the vascular lumen, and increased LPL activity accordingly (Figure 4a,b). Both FAM3A‐transgenic mice and rcFAM3A‐infused mice resulted in an increased postheparin LPL amount and activity in plasma (Figure 4d,e). We further explored the direct effect of FAM3A on LPL expression (Figure 4g–i), significant increases of LPL were also confirmed in vivo and in vitro. In addition, both FAM3A transgenic and rcFAM3A treatment exhibited an increased accumulation of ³H activity into LPL‐expressing tissues representing increased [³H]oleate deposition (Figure 4k,l). From these results, we thus conclude that FAM3A has the potency to enhance plasma LPL amount and function, contributing to the stimulation of LPL lipolytic function and FA uptake in vivo.

LPL is synthesized by parenchymal cells. To process TG‐rich lipoproteins, however, LPL needs to be transported from the interstitial spaces to the luminal face of capillary endothelial cells (wherein it is accessible to TG‐rich lipoproteins) [35]. Currently, the regulators responsible for LPL position in the capillary lumen have not been fully identified. Most importantly, it should be noted that FAM3A is expressed predominantly in the vascular endothelium, particularly capillaries [9]. Therefore, we further explored the direct effect of FAM3A on LPL microvascular endothelial apices position. As expected, we found that LPL amount and activity at the vascular lumen of FAM3A‐transgenic mice significantly increased at 1–2 min of the early perfusion time, after which the activity of LPL in the vascular lumen returned to normal levels (Figure 4c). Meanwhile, immunofluorescence analysis revealed that the level of LPL co‐expressed with endothelial cells was significantly increased (Figure 4j). These results, to some extent, suggested that FAM3A changes LPL position in coronary vessel lumen.

We further explored the mechanisms underlying the stimulatory effect of FAM3A on LPL lipolytic function and FA uptake in. PPARα, as a ligand‐activated nuclear hormone receptor, plays important roles in regulating tissue‐specific and systemic lipid metabolism. Under fasting conditions, ligand‐bound PPARα drives the transcription program of key metabolic pathways: including peroxisomal and mitochondrial β‐oxidation, FA transport, lipoprotein catabolism, and ketogenesis. Previous human observations have demonstrated that PPARα coding variants are implicated in altered lipid metabolism [36]. Previous studies have indicated that PPARα facilitates triglyceride clearance in the heart by upregulating LPL expression and enhancing triglyceride hydrolase activity [37], which is consistent with the human findings. We infer that PPARɑ may be involved in the regulation of lipid‐transendothelial delivery and FA uptake in soleus. In the current study, we confirmed that FAM3A promoted the expression of PPARɑ in vivo and in vitro (Figure 6d–f). Most importantly, we found that lipid‐transendothelial delivery elicited by FAM3A could be alleviated by a PPARα antagonist in vitro (Figure 6k). Furthermore, FAM3A‐induced decrease in plasma TG levels and promotion of FA trans‐endothelial delivery and uptake in soleus were significantly abolished by the inhibition of PPARα (Figures 6 and 7). Therefore, our study identified FAM3A/PPARα axis as a novel signaling pathway that regulates lipid‐transendothelial delivery and FA uptake in skeletal muscles. Despite the finding of elevated PPARα induced by FAM3A, the screen of the proteomic data in soleus revealed few lipid oxidation‐related proteins with significant change in the comparison between FAM3A_Tg mice and WT mice. However, previous studies reported the effects of FAM3A on the FA β‐oxidation in liver. FAM3A promoted the entry of CaM into the nucleus which as a co‐activator of FOXA2 to enhance carnitine palmitoyl transferase 2 (CPT2) transcription and thereby promote FA oxidation [15]. FAM3A bound to various acyl‐L‐carnitines, and knockdown of FAM3A significantly reduced [³H]‐palmate oxidation in mitochondria [38]. However, PPAR signaling was not assessed in these studies. It is likely that PPARɑ is implicated in the regulatory role of FAM3A in FA oxidation. Moreover, it should be acknowledged that previous studies revealed some signaling pathways were implicated in the biological function of FAM3A, including AKT, ERK1/2, STAT3, etc. [17, 39, 40]. However, these signaling pathways were not further investigated together with PPARɑ or TCF15 in this present study and require further exploration.

In conclusion, this study demonstrates that the FAM3A/PPARα axis plays a pivotal role in regulating lipid metabolism by enhancing LPL‐mediated VLDL‐TG hydrolysis at the vascular lumen and facilitating subsequent FA transendothelial transport and tissue uptake. However, several limitations should be acknowledged: (1) The lack of knowledge of the FAM3A receptor limits the in‐depth explorations of the mechanisms by which FAM3A affects PPARα, governs LPL translocation, and regulates the endothelial expression of TCF15 and CD36; (2) Additionally, as one partner of the PPARα‐RXR heterodimer, RXRa was also upregulated by FAM3A. However, we did not make a further exploration to decipher the underlying mechanism and biological outcomes; (3) Moreover, the findings primarily derived from animal models necessitate future validation in human physiological and pathological contexts. Future investigations should focus on delineating the role of the FAM3A/PPARα‐RXR signaling in metabolic diseases and exploring its potential as a therapeutic target for modulating lipid metabolic disorders.

Author Contributions

D.Y., X.S., conducted animal experiments. D.Y., X.S., W.X., and L.Y. conducted molecular biology experiments. X.S. conducted histological experiments. D.Y., X.S. collected and analyzed the experiment data. J.W., B.Q. conducted proteomics experiment, database data investigation and analysis. X.S., D.Y., and Y.Z. drafted the manuscript. Y.Z. supervised the project and obtained funding.

Funding

This work was supported by grants from the Natural Science Foundation of China (Grant No. 82470504), Beijing Natural Science Foundation (Grant No. L246053), and Beijing Natural Science Foundation (Grant No. L256071).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: FAM3A affects lipid transport protein expression.

Figure S2: FAM3A affects lipid transport across microvasculature.

Figure S3: The silencing of TCF15 repressed the effects of FAM3A on CD36 and LPL.

Figure S4: FAM3A affects lipid‐transendothelial delivery and FA uptake.

Figure S5: FAM3A upregulates RXRa protein level.

Table S1: Clinical characteristics of varicosity patients without (Control) or with metabolic disorders (MD).

Table S2: Experimental animals.

Table S3: Cultured cells and related reagents.

Table S4: List of Primers for qRT‐PCR.

Table S5: Antibodies.

Table S6: Other important reagents and resources.

Data S1: Source Data.

Data S2: Uncropped Blots.

Data Availability Statement

The raw mass spectrometry proteomic data are deposited in the iProX database (accession number: IPX0012785000). This paper does not report the original code. Supporting Information figures and tables are available as a word document “Supporting Information” in the Supporting Information. Unprocessed blots are available as a PDF file “Uncropped Blots” in the Related files. Correspondence and reasonable requests for materials should be addressed to DY or YZ.