Asprosin inhibits macrophage lipid accumulation and reduces atherosclerotic burden by up-regulating ABCA1 and ABCG1 expression via the p38/Elk-1 pathway

Human monocyte line THP-1 was purchased from American Type Culture Collection (Rockville, MD, USA). Fetal bovine serum (FBS), phorbol-12-myristate acetate (PMA), apoA-I and SB203580 were provided by Sigma-Aldrich (St. Louis, MO, USA). Ox-LDL, acetylated low-density lipoprotein (ac-LDL) and HDL were obtained from Yiyuan Biotechnology (Guangzhou, China). Lentiviral vectors expressing asprosin (LV-Asprosin) or Elk-1^S383A (serine 383 was replaced with alanine) were constructed by Genechem (Shanghai, China). Phosphorylation mutant Elk-1^S383A was generated by using a QuikChange II site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA) according to the manufacturer’s instructions. The mutation of Ser383 to alanine was confirmed by DNA sequencing. Rabbit antibodies against liver X receptor α (LXRα), proprotein convertase subtilisin/kexin type 9 (PCSK9), CD36, Elk-1, phospho-Elk-1 (p-Elk-1, Ser383), p-Elk-1 (Ser389), p-Elk-1 (Thr417), p38, phospho-p38 (p-38), JNK, phospho-JNK (p-JNK), ERK1/2, phospho-ERK1/2 (p-ERK1/2), ABCA1, ABCG1, histone H3 and β-actin were supplied by Abcam (Cambridge, UK). Rabbit antibody against asprosin (FineTest, Wuhan, China), mouse antibody against scavenger receptor class A (SR-A; R&D systems, Minneapolis, MN, USA), and horseradish peroxidase (HRP)-conjugated goat anti-rabbit or mouse IgG (H + L) (Beyotime, Shanghai, China) were obtained as indicated.

THP-1 cells were cultured in RPMI 1640 medium (Solarbio, Beijing, China) containing 2% penicillin–streptomycin and 10% FBS in a humidified atmosphere of 5% CO₂ at 37 °C. The differentiation of monocytes to macrophages was induced by the addition 160 nM PMA for 24 h. After 48 h of incubation with 50 µg/mL ox-LDL, macrophages were converted to foam cells. THP-1 macrophage-derived foam cells were seeded in 24-well plates (2 × 10⁶ cells/well) and then transduced with empty vector (LV-Mock; Genechem) or LV-Asprosin at a multiplicity of infection of 100 in the presence of 8 mg/mL polybrene for 24 h. Subsequently, the medium was changed to fresh 10% FBS/RPMI 1640 medium. After 48 h culture, the transfection efficiency was evaluated by western blot analysis of asprosin protein expression. To interfere with the p38/Elk-1 signaling pathway, THP-1 macrophage-derived foam cells were treated with 10 μM SB203580 for 6 h or transfected with Elk-1^S383A through lentiviral vector for 72 h. Thereafter, cells were transduced with or without LV-Asprosin for an additional 72 h.

The intracellular contents of TC, FC and cholesterol ester (CE) were detected using the HPLC analysis as previously described [23]. Briefly, THP-1 macrophage-derived foam cells were washed three times with phosphate-buffered saline (PBS) and then broken by an ultrasonic processor (Scientz, Zhejiang, China) on the ice. Cell lysates were centrifugated at 12,000 g for 5 min to collect protein samples, followed by detection of protein concentration using the BCA Protein Assay Kit (Beyotime). Cholesterol was extracted from protein samples by n-hexane–isopropanol (3:2, V/V) and then dissolved in isopropanol. Subsequently, 0.4 U cholesterol oxidase were added to each sample for measuring the FC content. The TC content was examined by using 0.4 U cholesterol oxidase in combination with 0.4 U cholesterol esterase. CE was calculated by subtracting FC from TC.

After transfection with LV-Mock or LV-Asprosin, THP-1 macrophage-derived foam cells were washed with PBS and then fixed in 4% paraformaldehyde solution for 10 min. Cells were stained with 0.5% Oil Red O solution (Solarbio) in the dark at 37 °C for 5 min, destained with 60% isopropanol for 15 s, incubated with hematoxylin for counterstaining, and then photographed at× 400 magnification.

The NBD-cholesterol efflux assay was performed according to our previous report [24]. In brief, THP-1 macrophage-derived foam cells receiving the indicated treatments were incubated in phenol red-free RPMI 1640 medium supplemented with 5 µmol/L NBD-cholesterol (Invitrogen, Carlsbad, CA, USA) for 4 h at 37 °C. After washing three times with PBS, cells were incubated with 25 μg/mL apoA-I and 50 μg/mL HDL as lipid acceptors for 4 h to examine ABCA1- and ABCG1-dependent cholesterol efflux, respectively. The medium was collected, and cells were harvested and then lysed by 0.1% Triton X-100 (Gibco, Carlsbad, CA, USA). The fluorescence-labeled cholesterol in the medium and cell lysates was visualized by an inverted fluorescence microscope (Olympus IX53, Tokyo, Japan). The Image-Pro Plus software (Diagnostic Instruments, USA) was employed to analyze fluorescence intensity. NBD-cholesterol efflux capacity was calculated as the percentage of fluorescence density in the medium to total fluorescence density (medium + cell lysates).

After the indicated treatments, cell culture supernatant was collected to detect the levels of IL-1β and IL-6 by using the commercial enzyme-linked immunosorbent assay (ELISA) kits (R&D systems) according to the manufacturer’s instructions. The absorbance at 450 nm was determined by the iMark™ microplate reader (Bio-Rad, Hercules, CA, USA).

The uptake of Dil-ox-LDL by THP-1 macrophage-derived foam cells was performed as previously described [9]. In brief, THP-1 macrophage-derived foam cells were transfected with LV-Mock or LV-Asprosin for 72 h, washed twice with PBS, and then incubated with 10 μg/mL Dil-ox-LDL for 4 h at 37 °C. Cells were washed three times with PBS and then visualized by an inverted fluorescence microscope.

Thirty 8-week-old male apoE^−/− mice with a C57BL/6 background were provided by Cavens lab animal (Jiangsu, China) and maintained under a specific-pathogen free condition. These mice were randomly divided into LV-Mock group and LV-Asprosin group, with 15 animals in each group. All mice were fed a Western diet (0.3% cholesterol, 21% fat; Research Diets) for 12 weeks to establish an atherosclerosis model. During this process, LV-Mock or LV-Asprosin (5 × 10⁷ TU/mouse) were injected into the mice via the tail vein once 3 weeks. At the end of the study, the mice were weighed, anesthetized, and then euthanized to collect the blood, aortas and hearts for further analyses. The animal protocols were reviewed and approved by the Animal Care and Use Committee of the First Affiliated Hospital of University of South China.

The hearts and aortic tissues were immediately isolated from the ascending aorta to the ileal bifurcation and fixed in 4% paraformaldehyde solution after apoE^−/− mice were sacrificed. The adventitia was thoroughly cleaned under a dissecting microscope. The aortas were unfolded along the longitudinal axis, stained with 0.5% Oil Red O solution for 30 min and then photographed using a digital camera. The atherosclerotic burden was calculated as the percentage of Oil Red O-positive area to total aorta area. After washing with PBS, the hearts with upper aortic root were embedded in optimal cutting temperature compound (Sakura Finetek, Torrance, CA, USA) and then serially sectioned (8 μm thickness) throughout three aortic valves. Subsequently, the snap-frozen cross-sections were stained with hematoxylin and eosin (HE), Oil Red O and Masson Trichrome for quantification of lesion area, lipid deposition and collagen content using the Image-Pro Plus 6.0 software.

Blood samples were collected from the retro-orbital plexus of apoE^−/− mice at the end of experiments and then centrifuged at 800 g for 10 min to isolate the plasma. The levels of TC, low density lipoprotein cholesterol (LDL-C), HDL cholesterol (HDL-C) and TG in plasma were measured by using the commercially available enzymatic kits (Biosino, Beijing, China). Additionally, plasma PCSK9 levels were detected by using ELISA (R&D systems) according to the manufacturer’s protocol.

The RCT assay was conducted as we had described previously [8]. J774 macrophages were loaded with 5 μCi/mL [³H]-cholesterol plus 50 μg/mL ac-LDL for 48 h and then suspended in Dulbecco's Modified Eagle Medium (Gibco). Subsequently, J774 macrophages labeled with [³H]-cholesterol (4.5×10⁶ cells/mouse) were intraperitoneally injected into apoE^−/− mice (n = 5 in each group). Plasma samples were taken at 6, 24, and 48 h after injection. The feces were continuously collected until 48 h, and hepatic tissue specimens were obtained at 48 h. The radioactivity of [³H]-cholesterol in the plasma, liver and feces was analyzed by a liquid scintillation counter. The efficiency of in vivo RCT was calculated as the percentage of radioactivity in the plasma, liver or feces to total radioactivity injected at baseline.

Total RNA was extracted from cultured cells and mouse aortas using a TRIzol kit (Invitrogen). The obtained RNA was converted to cDNA using a Superscript First-Strand cDNA Synthesis Kit (Invitrogen). Thereafter, qRT-PCR was performed by using SYBR Green Real-Time PCR Master Mix (Promega) on an ABI 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The primers used in this study were designed and synthesized by Sangon Biotech (Shanghai, China). Their sequences are presented in (Additional file 2: Table S1). The relative expression levels of target genes were determined by the 2^−ΔΔCt method. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal control.

The cultured cells and mouse aortas were lysed by the RIPA buffer containing 0.1 mmol/L phenylmethanesulfonyl fluoride (Beyotime). The cytoplasmic and nuclear proteins were extracted by the corresponding protein extraction kits (Beyotime). The protein concentration was quantified by a BCA Protein Assay Kit. The protein samples were subjected to SDS-PAGE and then transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). After blockade in 5% skim milk for 1 h at room temperature, the membranes were incubated with primary antibodies against asprosin (FNab09797, 1:1000), LXRα (ab176323, 1:2000), PCSK9 (ab95478, 1:1000), CD36 (ab133625, 1:500), SR-A (AF1797, 1:1000), Elk-1 (ab11847, 1:500), p-Elk-1 (Ser383; ab34270, 1:300), p-Elk-1 (Ser389; ab28818, 1:300), p-Elk-1 (Thr417; ab28817, 1:300), p38 (ab223619, 1:500), p-p38 (ab32557, 1:1000), JNK (ab112501, 1:500), p-JNK (ab124956, 1:1000), ERK1/2 (ab17942, 1:1000), p-ERK1/2 (ab223500, 1:400), ABCA1 (ab7360, 1:200), ABCG1 (ab52617, 1:1000), histone H3 (ab176842, 1:3000) or β-actin (ab115777, 1:3000) overnight at 4 °C. Subsequently, the membranes were incubated and reacted with HRP-conjugated secondary antibodies (1:3000) for 2 h at room temperature. The immunoreactive bands were visualized using the BeyoECL Plus kit (Beyotime) on Tanon 5500 (Shanghai, China). β-actin was used as an internal control.

All results are presented as the mean ± standard deviation (SD). An unpaired two-tailed Student’s t-test was applied to compare the differences between two groups. The differences among ≥ three groups were analyzed using one-way ANOVA followed by Tukey’s multiple comparison test. A value of P < 0.05 was thought to be statistically significant. The data were statistically analyzed and graphed using the GraphPad Prism 8.0 software (San Diego, CA, USA).

Lipid accumulation in macrophages is regarded as a major driver of atherosclerosis. To clarify whether asprosin plays a role in this process, we first detected its expression using qRT-PCR and western blot in THP-1 macrophages loaded with or without ox-LDL. As shown in Fig. 1A, the mRNA and protein levels of asprosin were significantly decreased in ox-LDL-treated THP-1 macrophages compared with unstimulated cells, indicating a link between asprosin and lipid metabolism. Subsequently, THP-1 macrophage-derived foam cells were transfected with LV-Mock or LV-Asprosin to overexpress asprosin. The transfection efficiency was verified by a marked elevation in asprosin protein levels after LV-Asprosin treatment (Fig. 1B). At the same time, the HPLC analysis showed that asprosin overexpression dramatically reduced the amounts of intracellular TC, FC, and CE (Fig. 1C). Oil Red O has been widely employed to stain neutral lipids, including those present in lipid-rich foam cells in vitro and in vivo [25]. The Oil Red O staining results demonstrated that intracellular lipid droplets were significantly decreased in response to LV-asprosin transduction (Fig. 1D). These findings suggest that asprosin plays a protective role in controlling lipid accumulation within THP-1 macrophage-derived foam cells.

It is well established that decreased cholesterol efflux mediated by ABCA1 and ABCG1 is a major cause leading to intracellular lipid accumulation [26]. To reveal the underlying mechanism by which asprosin protects against macrophage lipid accumulation, we examined the effect of asprosin on these two transporter expression and cholesterol efflux capacity in THP-1 macrophage-derived foam cells. As expected, asprosin overexpression augmented the mRNA and protein levels of ABCA1 and ABCG1 (Fig. 2A), with an accompanying increase in NBD-cholesterol efflux to apoA-I (Fig. 2B) and HDL (Fig. 2C). CD36 and SR-A belong to the members of scavenger receptor family and play a central role in mediating the internalization of modified lipoproteins by macrophages [27]. In addition to decreased cholesterol efflux, increased cholesterol influx contributes to intracellular lipid accumulation [28]. However, we found no significant difference in CD36 and SR-A expression in THP-1 macrophage-derived foam cells transfected with LV-Asprosin compared with LV-Mock (Fig. 2D). Consistently, no change was detected in Dil-ox-LDL uptake between LV-Asprosin group and LV-Mock group (Fig. 2E), thereby ruling out the possibility that asprosin-induced prevention of macrophage lipid accumulation is caused by decreased cholesterol uptake. Collectively, these observations suggest that asprosin suppresses lipid deposition in THP-1 macrophage-derived foam cells by stimulating ABCA1- and ABCG1-dependent cholesterol export.

In addition to promoting intracellular cholesterol release, ABCA1 and ABCG1 play an important role in inhibiting inflammatory response [29, 30]. The above studies have demonstrated that asprosin functions as a positive regulator of ABCA1 and ABCG1 expression. We next evaluated the impact of asprosin on pro-inflammatory cytokine production and secretion in THP-1 macrophage-derived foam cells. Transfection with LV-Asprosin dramatically down-regulated the mRNA expression of IL-1β and IL-6 (Additional file 1: Fig. S1A). Asprosin overexpression also diminished the levels of IL-1β and IL-6 in the cell culture supernatant (Additional file 1: Fig. S1B). These data suggest that asprosin contributes to alleviation of macrophage inflammatory response.

LXRα, a member of the nuclear receptor superfamily, is thought to be the most important transcription factor to induce ABCA1 and ABCG1 expression [31]. Unexpectedly, we found no difference in LXRα mRNA and protein expression in THP-1 macrophage-derived foam cells treated with LV-Asprosin compared with LV-Mock (Fig. 3A). This finding indicates that LXRα is not implicated in asprosin-enhanced ABCA1 and ABCG1 expression. Interestingly, PCSK9, a subtilisin family-serine protease predominantly expressed in the liver, has been reported to inhibit ABCA1 expression in macrophages [32]. However, asprosin overexpression had no effect on the mRNA and protein levels of PCSK9 in THP-1 macrophage-derived foam cells (Fig. 3B). Thus, like LXRα, PCSK9 is not involved in asprosin-induced up-regulation of ABCA1 and ABCG1 expression. Our previous study has identified Elk-1 as a transcription factor of ABCA1 and ABCG1 [22]. We hypothesized that the stimulatory effect of asprosin on ABCA1 and ABCG1 expression is likely mediated by Elk-1. To test this hypothesis, we measured Elk-1 expression in THP-1 macrophage-derived foam cells and observed similar changes of Elk-1 mRNA and protein levels between LV-Asprosin group and LV-Mock group (Fig. 3C). Given that the transcriptional activity of Elk-1 is dependent on its translocation from the cytoplasm to the nucleus [21], We extracted cytoplasmic and nuclear proteins to determine the distribution of Elk-1 in THP-1 macrophage-derived foam cells. Our results showed that asprosin overexpression increased Elk-1 expression in the nucleus but decreased its expression in the cytoplasm (Fig. 3D), suggesting that asprosin can promote the nuclear translocation of Elk-1. The phosphorylation of Elk-1 at Ser383, Ser389 or Thr417 drives it to enter the nucleus so as to exert a transcriptional activation effect [33‐35]. Next, we examined p-Elk-1 levels at these phosphorylated sites. Interestingly, asprosin overexpression elevated p-Elk-1 levels at Ser383 but not Ser389 and Thr417 (Fig. 3E). To determine the role of Ser383 in asprosin-induced enhancement of ABCA1 and ABCG1 expression and cholesterol efflux, we constructed a phosphorylation mutant Elk-1^S383A, which was transfected into THP-1 macrophage-derived foam cells prior to LV-Asprosin treatment. Mutation to alanine of Ser383 abrogated the effect of LV-Asprosin on ABCA1 and ABCG1 expression (Fig. 3F) as well as NBD-cholesterol efflux to apoA-I (Fig. 3G) and HDL (Fig. 3H). Taken together, these data support the concept that asprosin phosphorylates Elk-1 at Ser383 to induce transcriptional activation of ABCA1 and ABCG1 and promote subsequent cholesterol efflux.

Elk-1 can be activated by three classes of MAPKs: ERK1/2, JNK and p38 [36, 37]. We wondered whether these kinases play a role in the regulation of Elk-1 phosphorylation and ABCA1 and ABCG1 expression by asprosin. To this end, we first treated THP-1 macrophage-derived foam cells with LV-Mock or LV-Asprosin. As illustrated in Fig. 4A, asprosin overexpression markedly promoted the phosphorylation of p38 but not ERK1/2 and JNK, suggesting that asprosin functions an activator of p38. Subsequently, THP-1 macrophage-derived foam cells were incubated with p38 specific inhibitor SB203580, followed by transfection with LV-Asprosin. Pretreatment with SB203580 reduced the effect of LV-asprosin on Elk-1 phosphorylation at Ser383 (Fig. 4B) and nuclear translocation (Fig. 4C). Also, the promotive effect of LV-Asprosin on ABCA1 and ABCG1 expression and NBD-cholesterol efflux was significantly reversed by SB203580 (Fig. 4D-F). These results indicate that asprosin stimulates Elk-1 phosphorylation at Ser383 and promotes ABCA1- and ABCG1-dependent cholesterol release by activating p38 in THP-1 macrophage-derived foam cells.

To observe the influence of asprosin on atherosclerotic lesion formation, apoE^−/− mice were fed a Western diet for 12 weeks and simultaneously received tail vein injection of LV-Mock or LV-Asprosin. At the end of the study, no significant difference was found in weight gain between LV-Mock and LV-Asprosin mice (Fig. 5A). The western blot results showed that LV-Asprosin mice had a 3.84-fold increase in asprosin protein expression in the aortas as compared to controls (Fig. 5B). This finding suggests a higher transfection efficiency of LV-Asprosin in vivo. Importantly, asprosin overexpression significantly diminished the atherosclerotic plaque burden in the aortic arch regions (Fig. 5C) as well as Oil Red O-positive lesion area in whole enface aortas (Fig. 5D). Consistently, LV-Asprosin mice had a marked reduction in cross-sectional lesion area, as evidenced by HE staining (Fig. 5E). The Oil Red O staining of aortic valve cross-sections demonstrated that atherosclerotic plaques in LV-Asprosin mice had a significant decrease in Oil Red O-positive area compared with LV-Mock mice (Fig. 5F), indicating that asprosin overexpression contributes to alleviation of lesional lipid deposition. The Masson Trichrome staining further showed that collagen content within the plaques was higher in LV-Asprosin mice than LV-Mock mice (Fig. 5G), indicating that asprosin overexpression promotes the shift toward a more stable plaque phenotype. These in vivo data suggest that asprosin exerts an atheroprotective effect.

Lipid metabolism disorder is an independent risk factor of atherosclerosis [38]. To gain mechanistical insights into the antiatherogenic action of asprosin, we detected plasma lipid levels and RCT efficiency in apoE^−/− mice injected with LV-Mock or LV-Asprosin. Injection of LV-Asprosin significantly increased plasma HDL-C levels but decreased LDL-C, TC and TG levels (Fig. 6A-D). Meanwhile, the radioactivity of [³H]-cholesterol in the plasma, liver, and feces was higher in LV-Asprosin group than that in LV-Mock group (Fig. 6E), suggesting that asprosin overexpression contributes to macrophage-to-feces RCT in vivo. In addition, PCSK9 plays an important role in regulating lipid metabolism [39]. Similar to the in vitro data, there was no significant difference in plasma PCSK9 levels between LV-Asprosin group and LV-Mock group (Fig. 6F). These observations reveal that asprosin protects against atherosclerosis by improving plasma lipid profiles and promoting RCT.

Finally, we isolated the aortas from apoE^−/− mice to examine the phosphorylation of p38 and Elk-1 and the expression of Elk-1, ABCA1 and ABCG1. In agreement with the in vitro results, the levels of p-p38 were significantly increased in LV-Asprosin group compared with LV-Mock group (Fig. 7A). Elk-1 mRNA and protein expression in LV-Asprosin mice were not different from LV-Mock mice (Fig. 7B). However, injection of LV-Asprosin augmented p-Elk-1 levels at Ser383 (Fig. 7C) and promoted the translocation of Elk-1 from the cytoplasm to the nucleus (Fig. 7D). Also, asprosin overexpression enhanced the mRNA and protein levels of ABCA1 and ABCG1 (Fig. 7E). These findings suggest that asprosin can activate the p38/Elk-1 signaling pathway, leading to up-regulation of ABCA1 and ABCG1 expression in vivo.