GRAPHICAL ABSTRACT

Highlights

• A secreted protein database is built using adipose tissues under CR intervention.

• CRELD2 is identified as a novel adipokine.

• CRELD2 regulates adipocyte differentiation and adipogenesis in normal and obese states.

INTRODUCTION

Adipose tissue is recognized as an important endocrine organ, secreting bioactive peptides/proteins termed adipokines, which play a crucial role in controlling metabolism, immunity, and other physiological functions [1,2]. For instance, leptin, a protein specifically expressed in adipose tissue, regulates appetite and energy balance, while adiponectin improves insulin sensitivity, promotes fatty acid oxidation, and reduces inflammation [3]. Identifying and characterizing new adipose derived factors and studying their functions could lead to the identification of critical metabolic regulation targets [4].

Caloric restriction (CR) is a well-known intervention that promotes longevity and improves health span in various organisms [5,6]. The most significant benefit of CR is improved metabolic health and reduced risk of age-related metabolic disorders, such as type 2 diabetes mellitus (T2DM) and cardiovascular diseases. Adipose tissue, a critical regulator of energy homeostasis and insulin sensitivity, is one of the most affected tissues by CR. CR reduces fat storage and promotes lipolysis, leading to a decrease in adipose tissue mass, which likely mediates various beneficial effects that are important to metabolic related diseases. Furthermore, CR could influence the adipose tissues transcriptome to impact the secretion of fat-secreting proteins to fulfill their functions [7,8]. And CR intervention can be used as a model to help discover novel adipokines that are closely related to metabolic diseases and develop metabolic regulation targets. Therefore, establishing a secreted protein database under CR intervention, and understanding the effects of CR on adipokines from a systems biology perspective could be of significant importance.

Adipose tissue displays heterogeneity and distinct functions in different regions of the body. Subcutaneous adipose tissue (SAT) has lower metabolic activity, minimal impact on metabolic disorders compared to other types of fat, and may even have beneficial functions. In contrast, visceral fat is strongly associated with metabolic disorders such as insulin resistance, T2DM, hypertension, and cardiovascular disease, and negatively affects metabolism. Brown adipose tissue (BAT) plays a role in thermogenesis and energy expenditure, and has been linked to improved insulin sensitivity and metabolic health [9]. The previous belief that BAT degenerates in adult humans has been challenged by recent studies demonstrating that perivascular adipose tissue (PVAT), with a higher metabolic rate, closely resembles BAT [10]. Therefore, it is essential to study adipose tissue in different body regions separately due to distinct roles in regulating metabolism.

In this study, we used whole-genome expression profiling chips to detect gene expression in adipose tissue of SAT, mesenteric adipose tissue (MAT), and PVAT under access to food ad libitum (AL) and 30% CR conditions. Secretory proteins expressed in adipose tissue were screened for constructing a database of secreted proteins (also called adipokines) using bioinformatics methods. Through the database, we discovered a novel adipokine cysteine rich with EGF like domains 2 (CRELD2), and explored its expression, secretion and involvement in adipose tissue under normal and obese conditions.

METHODS

Methods for animals and treatments, detection of whole-genome transcription in adipose tissue, screening for depot-specific genes expressed in adipose tissue, analysis of differentially expressed genes (DEG) between CR and AL groups, enrichment analysis of DEG, construction of a novel secreted protein database, human samples collection, animal fasting blood glucose and body weight detection, blood lipids assay, enzyme-linked immunosorbent assay (ELISA) for CRELD2 concentration measurement, histological morphological examination of adipose tissue, immunofluorescence staining, the culture and differentiation induction of 3T3–L1 cells, preparation of conditioned medium of adipose tissue and 3T3–L1 differentiated adipocytes, real-time polymerase chain reaction (PCR) analysis and statistical analysis are available in the Supplementary Methods and Supplementary Tables 1-4. Some methods were performed as previously described [4,11-16].

RESULTS

The expression profiles of adipose tissues in different locations exhibit unique features

Distinct depots of rat PVAT surrounding thoracic aorta, inguinal SAT and MAT were collected for detecting whole-genome transcription in AL and CR groups. Rats in AL group were allowed unrestricted consumption of standard rodent chow, and rats in CR group received a daily food allocation equivalent to 70% of the normal chow intake with a duration of 18 weeks. Different adipose tissues exhibited a distinct gene expression profile. Fig. 1 illustrates the expression patterns of tissue-specific genes in AL rats, highlighting the top 15 genes in each tissue. In PVAT, genes associated with cardiac or skeletal muscle, namely Myh6, Mybpc2, Tnni3, Mybphl, Csrp3, Mybpc3, Ckm, Myoz2, Mb, Myl7, and Actn2, were prominently expressed. The preferential activation of these genes in PVAT may be attributed to its developmental origin. PVAT also manifested selective expression of genes linked to BAT, such as Ucp1 and Cox8h, that are involved in oxidative phosphorylation and energy expenditure, suggesting a potential overlapping functionality between PVAT and BAT.

Distinct gene expression profiles characterize various biological processes in SAT. The constitutive photomorphogenesis 9 signalosome associated genes, namely Csn10, Csnd, and Csng, were robustly expressed. Likewise, genes implicated in protein phosphorylation activation, such as Prr27 and Snorc, were also specifically expressed, implying an augmented function of protein ubiquitination or phosphorylation modifications in SAT. Furthermore, there were genes associated with milk or mucus production, tight junction formation and vitamin D metabolism.

MAT exhibited pronounced gene expression patterns associated with immune inflammation, evidenced by the robust expression of genes linked to B-cell development and immunoglobulin expression, including Ighv, Tnfrsf14, Cnr2, Tnfsf11, Reg3g, Vsig4, Tnfrsf17, Glycam1, A2m, Marco, and CD96. Furthermore, MAT demonstrated specific expression of Niban3 and Ripor2, genes implicated in cellular proliferation and apoptosis.

The effects of CR on the transcriptome of PVAT, SAT, and MAT

Compared to AL group, 630 upregulated and 381 downregulated genes were identified in PVAT of CR group (Supplementary Fig. 1A). Gene Ontology (GO) gene function enrichment analysis indicated that these DEG were primarily associated with biological terms such as energy metabolic processes, cellular responses to peptide hormones, and cell differentiation and repair (Supplementary Table 5, Supplementary Fig. 1B). The corresponding Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis suggested that these genes participate in metabolic regulation, cell signal transduction of insulin, glucagon, transforming growth factor β (TGF-β), etc. (Supplementary Table 6, Supplementary Fig. 1C), suggesting significant metabolic regulatory changes in rat PVAT following CR. Gene Set Enrichment Analysis (GSEA) enabled the identification of more comprehensive biological functions and pathways. When compared with GO database, it revealed alterations in perception, signal transduction and cellular structure (Supplementary Fig. 1D). Further comparison with KEGG database showed changes in disease expression profiles related to insulin/glucagon and AMP-activated protein kinase (AMPK) pathways. Additionally, it uncovered changes in new signaling pathways, for instance muscle cell skeleton and prolactin (Supplementary Fig. 1E).

In SAT of CR group, 212 upregulated and 163 downregulated genes were revealed compared to AL group (Supplementary Fig. 2A). These DEG showed significant enrichment in multiple GO biological terms (Supplementary Table 7, Supplementary Fig. 2B), including processes related to miRNA, immune cell differentiation and proliferation, regulation of metabolic processes, and cell signal transduction and responses. GO terms related to cell structure and location were also enriched. KEGG analysis revealed that the DEG are involved in pathways such as nuclear factor-kappa B, tumor necrosis factor, malaria, Chagas disease, and thyroid cancer (Supplementary Table 8, Supplementary Fig. 2C). These pathways are associated with immune inflammation, metabolic regulation and various diseases. GSEA analysis of GO database indicated disease expression profile in SAT is significantly enriched in specific immune processes related to T, B, and other lymphocytes (Supplementary Fig. 2D). GSEA enrichment matched by KEGG also showed some classic immune inflammation pathways (Supplementary Fig. 2E). Overall, the transformation of SAT expression profile primarily indicates significant changes in the immune process under CR.

In MAT, CR group demonstrated expression differences with 140 upregulated and 217 downregulated genes compared to AL group (Supplementary Fig. 3A). GO analysis revealed key categories such as adipocyte differentiation, lipid metabolism, chemokine activity and others (Supplementary Table 9, Supplementary Fig. 3B). The KEGG results include three main categories of cell signal transmission, metabolic pathway and disease regulation. The details can be seen in Supplementary Table 10, Supplementary Fig. 3C. The GSEA results are relate to the organization of extracellular structures in tissues and energy metabolism, including fatty acids, organic acids, and crucial biological processes like angiogenesis and cell migration (Supplementary Fig. 3D). DEG are enriched in signal pathways such as Rap1, cyclic adenosine monophosphate (cAMP), TGF-β, Ras, and peroxisome proliferator-activated receptor (PPAR) (Supplementary Fig. 3E). These pathways also regulate the cytoskeleton, fatty acid degradation, and cancer progression.

Library of potential adipokines

A total of 1,472 putative secreted proteins were identified from our 12 rat chips of SAT, MAT, and PVAT through a comprehensive analysis of gene arrays and secreted protein databases. Among them, 200 genes exhibited an expression level surpassing 2000 in at least one adipose tissue of rats under AL and CR conditions (Table 1, Fig. 2A). Within this high expression set of potential adipokines, 101 genes demonstrated an expression level greater than 2000 across all three adipose tissues. High expression in two adipose tissues included 19 genes in both SAT and MAT, 13 genes in both SAT and PVAT, and three genes in both MAT and PVAT. High expression in one adipose tissue included 34 genes in SAT, 21 genes in MAT, and nine genes in PVAT. From this database, one can obtain adipose depot specific adipokines and CR specifically regulated adipokines. For examples, although nicotinamide phosphoribosyltransferase (NAMPT), also known as VISFATIN, was highly expressed in all three adipose tissues, it was a relatively PVAT specific adipokine with more than 2-fold higher expression in PVAT than SAT and MAT, and upregulated in PVAT of CR group (Fig. 2B). NAMPT has been reported to link the anti-aging process, and shown in our previous research to regulate vascular smooth muscle proliferation and promote neural cell survival during cerebral ischemia [17-20]. Meteorin-like (METRNL, also known as SUBFATIN) was a relatively SAT specific adipokine (Fig. 2B), identified by our lab more than 10 years ago [4], to have a pro-differentiation action in adipose tissue and an improvement in glucose and lipid metabolism against insulin resistance in obesity and diabetes [12], and now has been widely recognized as a protective factor against diabetes and cardiovascular disease [11,21]. In this study, we sought to verify a novel adipokine, especially a relatively MAT specific adipokine that has not been witnessed but appeared significance for metabolism regulation as demonstrated by the above bioinformatics analysis. CRELD2 was selected as the novel adipokine due to no published data on its expression, secretion, and potential function in adipose tissue (Fig. 2B). The Creld2 gene expression in MAT was 5.5 times higher than in PVAT and 3.6 times higher than in SAT under AL condition, and 8.0 times higher than in PVAT and 7.1 times higher than in SAT under CR condition (Fig. 2B). And, its expression was upregulated in MAT by CR (Fig. 2B). Thus, CRELD2 was also named as MESFATIN (refers to mesenteric fat specifically expressed protein).

CRELD2 is extensively expressed in the adipose tissues of mouse, rat, and human

In mice, we examined Creld2 gene expression in BAT of interscapular brown adipose tissue (IBAT) and PVAT and white adipose tissues (WAT) of SAT, periepididymal adipose tissue (PAT), and MAT (Fig. 3A). Cell density in BAT was higher than that in WAT, and the WAT had larger lipid droplet than the BAT, which are in accordance with the reported (Fig. 3B). Compared with BAT (IBAT, PVAT), Creld2 mRNA level in WAT (SAT, PAT, MAT) was higher, especially was the highest in MAT (Fig. 3A). In all examined tissues, Creld2 mRNA expression in liver was the highest (Fig. 3A). At protein level, CRELD2 was extensively expressed in the IBAT, PVAT, SAT, PAT, and MAT of mouse (Fig. 3C). Since adipose tissue contains adipocytes and other types of cells [22], we further examined CRELD2 expression in adipocytes of adipose tissues. Immunostaining for colocalization of CRELD2 (green) and adipocyte marker ADIPONECTIN (red) showed that CRELD2 was expressed in adipocytes of IBAT, PAT, and MAT (Fig. 3D). Tissue expression pattern between CRELD2 and its homologue CRELD1 in mice was compared. Consistently with the higher expression of Creld2 in WAT than that in BAT, Creld1 mRNA expression in WAT of PAT and MAT was also higher than that in BAT of IBAT (Fig. 3E and F). And among examined tissues, Creld1 expression in PAT was highest (Fig. 3F), which is different from the highest expression of Creld2 in liver (Fig. 3A and E).

In rat, Creld2 was also expressed in IBAT, PVAT, SAT, PAT, MAT, liver, spleen, and muscle (Supplementary Fig. 4A). The histomorphology of rat IBAT, PVAT, SAT, PAT, and MAT were presented by HE and Oil Red O staining (Supplementary Fig. 5). And, Creld2 mRNA level in rat MAT was also higher than that in IBAT, PVAT, SAT, and PAT, and was the highest in liver among all examined tissues (Supplementary Fig. 4A). Further immunostaining assay confirmed CRELD2 was extensively expressed in rat adipose tissues (Supplementary Fig. 4B).

In human, Creld2 was also expressed in adipose tissues and other tissues, including SAT, visceral adipose tissue, epiploon, muscle and spleen (Supplementary Fig. 4C). With checking the histomorphology of human adipose tissues, CRELD2 expression at protein level was also confirmed (Supplementary Fig. 4D and E). CRELD2 was extensively expressed in the adipose tissues of human, and co-localized with adipocytes (CRELD2⁺/ADIPONECTIN⁺ cells). Taken together, CRELD2 is extensively expressed in the BAT and WAT of mouse, rat and human, and highly colocalized with adipocytes in adipose tissue. And, Creld2 mRNA expression level in MAT of both mouse and rat is higher than that in IBAT, PVAT, SAT, and PAT, which is consistent with the whole-genome transcription results (Fig. 2B).

CRELD2 is a novel adipokine and involved in adipocyte differentiation and adipogenesis

To answer whether CRELD2 can be secreted, CRELD2 concentration in the serum of mouse and human was examined by ELISA. CRELD2 concentration was about 727±45 pg/mL in the mouse serum (Fig. 4A) and 2.95±0.24 ng/mL in the healthy human serum (Fig. 4B), demonstrating that CRELD2 can be secreted into the circulating blood.

By preparing conditioned medium of mouse PAT and in vitro cultured adipocytes, we also found that CRELD2 existed not only in the conditioned medium of mouse PAT (Fig. 4C), but also in the conditioned medium of adipocytes induced from 3T3–L1 pre-adipocytes (Fig. 4D and E), providing firsthand evidence for CRELD2 as a novel adipokine. Moreover, CRELD2 concentration in the conditioned medium of relatively mature adipocytes (in vitro differentiation of 3T3–L1 cells for 8 to 10 days) was higher than that in less differentiated preadipocytes (in vitro differentiation of 3T3–L1 cells for 0 to 2 days) (Fig. 4E), indicating CRELD2 secretion is upregulated during the differentiation process from preadipocytes to adipocytes. The effect of CRELD2 on the adipocyte differentiation was further examined. At day 2 of 3T3–L1 cells initial differentiation stage, treatment of CRELD2 recombinant protein (3 μg/mL) for 48 hours upregulated the expression of lipogenic genes adiponectin, fatty acid binding protein 4 (Fabp4), and CCAAT/enhancer binding protein alpha (Cebpα) in terminal adipocytes (Fig. 4F). Therefore, CRELD2 is an adipokine which can be secreted from adipose tissue and adipocytes, with increased secretion concentration during adipogenesis process, and promotes adipocyte differentiation and maturation under normal physiological condition.

CRELD2 takes part in regulating adipogenesis process under obesity condition

Leptin-deficient ob/ob mice was introduced to explore the expression pattern of CRELD2 under obesity condition. The body weight, fasting blood glucose, tissues weight of PAT, liver, muscle and heart, the ratio of tissue weight/body weight, and the serum lipid concentration of total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), triglyceride (TG) and high-density lipoprotein cholesterol (HDL-C) in C57BL/6J wild-type (WT) mice and ob/ob mice were examined (Fig. 5A-D), supporting the application of ob/ob mice as a model to mimic the conditions of obesity and T2DM development as reported. We examined CRELD2 concentration in the serum of ob/ob and WT mice. Although the average value of CRELD2 concentration in the serum of ob/ob mice was higher than WT mice, there was no significance between the two groups (Fig. 5E).

High fat diet (HFD) induced obesity mouse model was also introduced. Compared to C57BL/6J mice fed with standard diet, the fasting blood glucose, serum lipid concentration of TC, LDL-C and HDL-C, and PAT weight was higher and TG concentration was lower in C57BL/6J mice fed HFD (C57+HFD), although there was no difference in liver and muscle weight (Supplementary Fig. 6A-C). Similar to the result in ob/ob mice (Fig. 5E), serum CRELD2 concentration in C57+HFD mice had a higher trend, but with no significance compared with that in C57BL/6J mice (Supplementary Fig. 6D). Thus, obesity did not change CRELD2 circulating concentration in the serum of mice.

The expression pattern of Creld2 gene in adipose tissues of ob/ob mice (Fig. 5F) was different from C57BL/6J mice (Fig. 3A). The size of adipocytes in both BAT and WAT of ob/ob mice was larger than that in C57BL/6J mice (Figs. 3B and 5H). Interestingly, compared with WT mice, Creld2 mRNA level was upregulated in BAT (IBAT and PVAT) and downregulated in WAT (SAT, PAT, and MAT) of ob/ob mice (Fig. 5G), suggesting potential role of CRELD2 in regulating adipogenesis process under obesity condition. In other metabolic related tissues of liver, spleen and muscle, Creld2 mRNA in ob/ob mice showed higher expression in spleen (Fig. 5F), rather than higher expression in liver of C57BL/6J mice (Fig. 3A). And Creld2 expression was lower in liver and higher in spleen of ob/ob mice, as compared to WT mice (Fig. 5G). There was also extensive expression of CRELD2 protein in IBAT, PVAT, SAT, PAT, and MAT of ob/ob mice (Fig. 5I). No change of CRELD2 circulating concentration between WT and obese mice does not mean that there is no significant change in local tissues. As shown in Fig. 5J, there is significant change of CRELD2 secretion level in local tissues of MAT and liver of ob/ob mice compared with WT mice, demonstrating different expression of CRELD2 in local tissues under different conditions. Taken together, can affect the expression pattern and local secretion level of CRELD2 in adipose tissues and metabolic related tissues, providing evidence for the potential role of CRELD2 in regulating metabolic related adipogenesis.

DISCUSSION

In this study, we developed a comprehensive database of secreted proteins by integrating transcriptomic data from various adipose tissues using bioinformatics analysis, and identified a total of 200 potential adipokines that exhibited significantly highly expressed in adipose tissues. The transcriptomes of various adipose tissues exhibit similarities with the surrounding tissues, particularly in the case of PVAT and SAT. For instance, PVAT displays high expression levels of genes associated with smooth muscle or myocardial functions, while SAT exhibits high levels of proteins such as extracellular matrix proteins, milk proteins, and tight junction proteins. These observations suggest that adipose tissue in these regions may share a developmental origin with the surrounding tissues and potentially contribute to the maintenance of tissue homeostasis. The common effects of CR on different adipose tissues are predominantly associated with metabolism, insulin signaling pathways, and immune inflammation. These findings suggest that CR exerts similar regulatory effects on multiple aspects of adipose tissue. The genes secreted by all three adipose tissues are highly related to immune inflammation, which may be due to the loose structure of adipose tissue that facilitates infiltration of immune cells. Additionally, each adipose tissue has its own unique characteristics. SAT expresses the most depot specific secreted proteins among the detected adipose tissues, while the proteins secreted specifically by MAT are almost all related to immune inflammation. The PVAT expresses fewer secreted proteins, most of which are related to enhanced metabolism and immune function, consistent with features of BAT.

CRELD family consists of two structurally related members CRELD1 and CRELD2 [23,24]. The major structural difference between these two proteins is the presence of transmembrane domains in CRELD1, but absent in CRELD2, which leads to distinct subcellular localizations and functions. Although there is little report for the secretion of CRELD1, one recent study reported that CRELD1 exists in the plasma proteins of diabetes patients [25]. The secretory biological characteristics of CRELD1 need to be explored in the future. CRELD2 is predominantly localized in the endoplasmic reticulum (ER) and Golgi apparatus, and demonstrated as a secreted glycoprotein with protein disulfide isomerase activity and secreted under normal and ER stress [26,27]. However, there is no report for the expression and secretion of CRELD2 in adipose tissues or adipocytes. In this study, we used whole-genome expression profiling chips to detect the secretome changes of adipose tissues under AL and CR conditions, and identified CRLED2 as novel adipokine. Among 1,472 identified putative secreted proteins, the basic characteristics of adipokines NAMPT and METRNL have been reported by our lab [4,12,17,18]. The new identified adipokine CRELD2 exhibits relatively specific expression in MAT both in chip of rat and in mRNA expression of mice and rat. There are also CRELD2 expression in adipose and other tissues of human. In addition to being secreted in serum of mice and human, CRELD2 can be also secreted from adipose tissues and adipocytes and the secretion level increases with the process of adipogenesis. Moreover, CRELD2 recombinant protein was found to promote the expression of lipogenic genes during adipocyte differentiation process, indicating the contribution of CRELD2 in adipocyte differentiation and maturation.

This study also examined whether CRELD2, as an adipokine, is involved in the regulation of the obesity process. Two obesity models of ob/ob mice and HFD fed mice, representing the endogenous obesity model induced by leptin gene deficiency and the exogenous obesity model induced by HFD, respectively, were introduced [28]. Although there was higher trend of CRELD2 concentration in serum of both ob/ob mice and HFD fed mice, there was no significance compared with that in C57BL/6J mice. No change of CRELD2 circulating concentration between WT and obese mice does not mean that there is no significant change in local tissues, especially in local adipose tissues. Our result demonstrated the different secretion level of CRELD2 in local tissues of MAT and liver under normal and obesity conditions. And also, the expression pattern of CRELD2 in adipose tissue of BAT and WAT and metabolic related tissues of liver and spleen of ob/ob mice was different from that in C57BL/6J mice, suggesting a potential role of CRELD2 in regulating adipogenesis process under obesity condition. These results imply the involvement of CRELD2 in white-to-brown adipose tissue conversion under different metabolic condition. There are many factors that regulate adipocyte differentiation and obesity, including WNT and ER stress [29,30]. CRELD2 has been reported to modulate noncanonical WNT4 signaling [31], and WNT4 takes part in adipocyte differentiation by mediating β-catenin pathway [32]. ER stress has been observed in numerous studies of human adipose tissue, and CRELD2 is a well-known ER inducible factor [30,33]. In obesity, adipose tissue is under hypoxia due to poorly oxygenated, which can interfere with disulfide bonding in the ER lumen with result of initiating ER stress and unfolded protein response (UPR) in adipose tissue. Thus, CRELD2 may regulate adipogenesis and obesity process through WNT4/β-catenin pathway or ER stress and UPR. Future study needs to answer the exact role and potential action of CRELD2 in obesity and metabolic related diseases.

In addition, series studies have explored the secretion property of CRELD2 [33]. It has demonstrated that ER stress, inhibitor of ER-Golgi transport, intervention of glucose-regulated protein 78 (GRP78) expression, mesencephalic astrocyte-derived neurotrophic factor and others can regulate CRELD2 secretion. Here, we proved the secretion of CRELD2 in circulation, local adipose tissue and adipocytes. However, conditions that induce CRELD2 secretion is still not well elucidated. Further studies need to comprehensively understand the mechanism that regulates intracellular transport and secretion of CRELD2. Besides, NAMPT and METRNAL have demonstrated as both angiogenic and adipogenic factors by our lab, both of which can promote angiogenesis as adipokines [4,14,17]. Similarly, CRELD2 has been also reported as angiogenic factor [34], in addition to as adipokine. But it is not clear whether CRELD2 could regulate angiogenesis in vivo due to the limited report. Extensive roles of CRELD2 under physiological and pathological conditions need to be explored.

Taken together, this study analyzed the expression profiles of adipose tissues in different locations of rats between CR and AL groups, and constructed a novel secreted protein database created from multiple adipose depots with beneficial intervention of CR. CRELD2 has been demonstrated as a new adipokine which is involved in adipocyte differentiation and adipogenesis under normal physiological and obese conditions.

SUPPLEMENTARY MATERIALS

NOTES

CONFLICTS OF INTEREST

No potential conflict of interest relevant to this article was reported.

AUTHOR CONTRIBUTIONS

Conception or design: C.Y.M.

Acquisition, analysis, or interpretation of data: all authors.

Drafting the work or revising: S.N.W., Z.Y.L., Q.S.L., P.F.J., C.Y.M.

Final approval of the manuscript: all authors.

FUNDING

This work was supported by grants from the National Natural Science Foundation of China Major Project (No. 82030110, No. 82330117, and No. 81130061), Youth Program of National Natural Science Foundation of China (No. 82003754), and Shanghai Science and Technology Commission Project (No. 20YF1458400 and No. 23ZR1477400). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

ACKNOWLEDGMENTS

None

Fig. 1.

Top 15 genes with specific expression in adipose tissues of different regions from rats of ad libitum (AL) group (four rats per chip, two chips for each adipose depot, and totally six chips for AL group). PVAT, perivascular adipose tissue; MAT, mesenteric adipose tissue; SAT, subcutaneous adipose tissue.

Fig. 2.

Distribution of potential secretory proteins with high expression levels in adipose tissue. (A) A set of 200 genes encoding potential secretory proteins with expression signals exceeding 2,000 in at least one adipose tissue. The distribution of these genes with expression signals more than 2,000 across different adipose tissues is presented. See the Table 1 for details. (B) The expression of nicotinamide phosphoribosyltransferase (Nampt), meteorin-like (Metrnl), and cysteine rich with EGF like domains 2 (Creld2) in various adipose tissues is shown under chow diet ad libitum (AL) and caloric restriction conditions (four rats per chip, two chips for each adipose depot, and totally six chips for AL group and six chips for caloric restriction group). MAT, mesenteric adipose tissue; SAT, subcutaneous adipose tissue; PVAT, perivascular adipose tissue.

Fig. 3.

The cysteine rich with EGF like domains 2 (CRELD2) expression in the adipose tissues and other tissues of mouse. (A) The relative expression of Creld2 mRNA in the interscapular brown adipose tissue (IBAT), perivascular adipose tissue (PVAT), subcutaneous adipose tissue (SAT), periepididymal adipose tissue (PAT), mesenteric adipose tissue (MAT), liver, spleen, muscle, heart, kidney, intestine, colon and rectum of mice (n=3). All data are shown as mean±standard error of the mean (SEM). (B) The histomorphology of IBAT, PVAT, SAT, PAT, and MAT in mice by hematoxylin-eosin (HE) staining and Oil Red O staining (bar=50 μm). (C) The representative images of CRELD2 immunostaining (green) in the IBAT, PVAT, SAT, PAT, and MAT of mice. 4´,6-Diamidino-2-phenylindole (DAPI) labels nuclei (blue) (bar=50 μm). (D) The representative images for immunostaining CRELD2 (green) with adipocyte cell marker ADIPONECTIN (red) in the IBAT, PAT, and MAT of mice. DAPI labels nuclei (blue) (bar=50 μm). (E, F) The relative expression of Creld2 and Creld1 mRNA in the IBAT, PAT, MAT, liver, spleen, muscle, heart, and kidney of mice (n=3). All data are shown as mean±SEM.

Fig. 4.

The secretion of cysteine rich with EGF like domains 2 (CRELD2) in vivo and in vitro, and its effect on adipocyte differentiation. (A) The standard curve of mouse CRELD2 enzyme-linked immunosorbent assay (ELISA) kit (left), and the concentration of CRELD2 in the serum of C57BL/6J mouse (n=5). (B) The standard curve of human CRELD2 ELISA kit (left), and the concentration of CRELD2 in the serum of healthy human (n=10). (C) The concentration of CRELD2 in the conditioned medium (CM) of vehicle group and mouse periepididymal adipose tissue (PAT) group. CRELD2 concentration in the CM of PAT was normalized to the weight of PAT (weight unit: g) (n=3 in vehicle group; n=8 in CM of mouse PAT group). (D) The representative images of adipocytes differentiated from 3T3–L1 preadipocytes (bar=100 μm). (E) The concentration of CRELD2 in the CM of vehicle (no cell), in vitro differentiated preadipocytes for 0 to 2 days and in vitro differentiated adipocytes for 8 to 10 days induced from 3T3–L1 cells (n=3 in each group). (F) Creld2 recombinant protein upregulates the expression of lipogenic genes in adipocytes. The mRNA expression of lipogenic genes of Adiponectin, fatty acid binding protein 4 (Fabp4), and CCAAT/enhancer binding protein alpha (Cebpα) in adipocytes derived from 3T3–L1 cells differentiation. All data are shown as mean±standard error of the mean. O.D., optical density. ^aP<0.01, ^bP<0.001 among the three groups, ^cP<0.001 between day 0–2 CM and day 8–10 CM groups.

Fig. 5.

The cysteine rich with EGF like domains 2 (CRELD2) level in the serum, adipose tissues and other tissues of ob/ob mice and C57BL/6J wild-type (WT) mice. (A) Body weight. (B) Tissue weights of periepididymal adipose tissue (PAT), liver, muscle and heart, and the ratios of PAT weight/body weight, liver weight/body weight, muscle weight/body weight, and heart weight/body weight. (C) Fasting blood glucose level. (D) Serum lipid concentrations of triglyceride (TG), total cholesterol (TC), lowdensity lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C). (E) Serum CRELD2 concentrations (n=5 in WT mice; n=6 in ob/ob mice). (F) The expression of Creld2 mRNA in the interscapular brown adipose tissue (IBAT), perivascular adipose tissue (PVAT), subcutaneous adipose tissue (SAT), PAT, mesenteric adipose tissue (MAT), liver, spleen, and muscle of ob/ob mice (n=3). (G) The relative expression of Creld2 mRNA in the IBAT, PVAT, SAT, PAT, MAT, liver, spleen, and muscle of WT mice and ob/ob mice (n=3). (H) The histomorphology of IBAT, PVAT, SAT, PAT, and MAT in ob/ob mice by hematoxylin- eosin (HE) staining (bar=50 μm). (I) The representative images of CRELD2 immunostaining (green) in the IBAT, PVAT, SAT, PAT, and MAT of ob/ob mice. 4´,6-Diamidino-2-phenylindole (DAPI) labels nuclei (blue) (bar=50 μm). (J) CRELD2 concentration in the conditioned medium of MAT and liver of WT mice and ob/ob mice (n=3 in each group). All data are shown as mean±standard error of the mean. NS, no significance. ^aP<0.001, ^bP<0.01, ^cP<0.05.

REFERENCES

• 1. Cai Z, He B. Adipose tissue aging: an update on mechanisms and therapeutic strategies. Metabolism 2023;138:155328.ArticlePubMed
• 2. Cho DH, Park SM. Epicardial adipose tissue and heart failure, friend or foe? Diabetes Metab J 2024;48:373-84.ArticlePubMedPMCPDF
• 3. Scheja L, Heeren J. The endocrine function of adipose tissues in health and cardiometabolic disease. Nat Rev Endocrinol 2019;15:507-24.ArticlePubMedPDF
• 4. Li ZY, Zheng SL, Wang P, Xu TY, Guan YF, Zhang YJ, et al. Subfatin is a novel adipokine and unlike Meteorin in adipose and brain expression. CNS Neurosci Ther 2014;20:344-54.ArticlePubMedPMC
• 5. Rhoads TW, Anderson RM. Caloric restriction has a new player. Science 2022;375:620-1.ArticlePubMed
• 6. Spadaro O, Youm Y, Shchukina I, Ryu S, Sidorov S, Ravussin A, et al. Caloric restriction in humans reveals immunometabolic regulators of health span. Science 2022;375:671-7.ArticlePubMedPMC
• 7. Barzilai N, Gabriely I. The role of fat depletion in the biological benefits of caloric restriction. J Nutr 2001;131:903S-906S.ArticlePubMed
• 8. Rosendo-Silva D, Gomes PB, Rodrigues T, Viana S, da Costa AN, Scherer PE, et al. Clinical and molecular profiling of human visceral adipose tissue reveals impairment of vascular architecture and remodeling as an early hallmark of dysfunction. Metabolism 2024;153:155788.ArticlePubMed
• 9. Corvera S. Cellular heterogeneity in adipose tissues. Annu Rev Physiol 2021;83:257-78.ArticlePubMedPMC
• 10. Becher T, Palanisamy S, Kramer DJ, Eljalby M, Marx SJ, Wibmer AG, et al. Brown adipose tissue is associated with cardiometabolic health. Nat Med 2021;27:58-65.ArticlePubMedPMCPDF
• 11. Zheng S, Li Z, Song J, Wang P, Xu J, Hu W, et al. Endothelial METRNL determines circulating METRNL level and maintains endothelial function against atherosclerosis. Acta Pharm Sin B 2023;13:1568-87.ArticlePubMedPMC
• 12. Li ZY, Song J, Zheng SL, Fan MB, Guan YF, Qu Y, et al. Adipocyte Metrnl antagonizes insulin resistance through PPARγ signaling. Diabetes 2015;64:4011-22.ArticlePubMedPDF
• 13. Wang SN, Wang Z, Wang XY, Zhang XP, Xu TY, Miao CY. Humanized cerebral organoids-based ischemic stroke model for discovering of potential anti-stroke agents. Acta Pharmacol Sin 2023;44:513-23.ArticlePubMedPMCPDF
• 14. Xu TY, Qing SL, Zhao JX, Song J, Miao ZW, Li JX, et al. Metrnl deficiency retards skin wound healing in mice by inhibiting AKT/eNOS signaling and angiogenesis. Acta Pharmacol Sin 2023;44:1790-800.ArticlePubMedPMCPDF
• 15. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 2014;15:550.ArticlePubMedPMCPDF
• 16. Snyder A, Jedreski K, Fitch J, Wijeratne S, Wetzel A, Hensley J, et al. Transcriptomic profiles in children with septic shock with or without immunoparalysis. Front Immunol 2021;12:733834.ArticlePubMedPMC
• 17. Wang P, Xu TY, Guan YF, Su DF, Fan GR, Miao CY. Perivascular adipose tissue-derived visfatin is a vascular smooth muscle cell growth factor: role of nicotinamide mononucleotide. Cardiovasc Res 2009;81:370-80.ArticlePubMed
• 18. Wang P, Miao CY. NAMPT as a therapeutic target against stroke. Trends Pharmacol Sci 2015;36:891-905.ArticlePubMed
• 19. Wang P, Yang X, Zhang Z, Song J, Guan YF, Zou DJ, et al. Depletion of NAD pool contributes to impairment of endothelial progenitor cell mobilization in diabetes. Metabolism 2016;65:852-62.ArticlePubMed
• 20. Wang P, Guan YF, Li WL, Lu GC, Liu JM, Miao CY. Nicotinamide phosphoribosyltransferase facilitates post-stroke angiogenesis. CNS Neurosci Ther 2015;21:475-7.ArticlePubMedPMC
• 21. Miao ZW, Hu WJ, Li ZY, Miao CY. Involvement of the secreted protein Metrnl in human diseases. Acta Pharmacol Sin 2020;41:1525-30.ArticlePubMedPMCPDF
• 22. Rutkowski JM, Stern JH, Scherer PE. The cell biology of fat expansion. J Cell Biol 2015;208:501-12.ArticlePubMedPMCPDF
• 23. Maslen CL, Babcock D, Redig JK, Kapeli K, Akkari YM, Olson SB. CRELD2: gene mapping, alternate splicing, and comparative genomic identification of the promoter region. Gene 2006;382:111-20.ArticlePubMed
• 24. Robinson SW, Morris CD, Goldmuntz E, Reller MD, Jones MA, Steiner RD, et al. Missense mutations in CRELD1 are associated with cardiac atrioventricular septal defects. Am J Hum Genet 2003;72:1047-52.ArticlePubMedPMC
• 25. Slieker RC, Donnelly LA, Akalestou E, Lopez-Noriega L, Melhem R, Gunes A, et al. Identification of biomarkers for glycaemic deterioration in type 2 diabetes. Nat Commun 2023;14:2533.PubMedPMC
• 26. Oh-hashi K, Kunieda R, Hirata Y, Kiuchi K. Biosynthesis and secretion of mouse cysteine-rich with EGF-like domains 2. FEBS Lett 2011;585:2481-7.ArticlePubMedPDF
• 27. Hartley CL, Edwards S, Mullan L, Bell PA, Fresquet M, Boot-Handford RP, et al. Armet/Manf and Creld2 are components of a specialized ER stress response provoked by inappropriate formation of disulphide bonds: implications for genetic skeletal diseases. Hum Mol Genet 2013;22:5262-75.ArticlePubMedPMC
• 28. Lutz TA. An overview of rodent models of obesity and type 2 diabetes. Methods Mol Biol 2020;2128:11-24.ArticlePubMed
• 29. An SM, Cho SH, Yoon JC. Adipose tissue and metabolic health. Diabetes Metab J 2023;47:595-611.ArticlePubMedPMCPDF
• 30. Khan S, Wang CH. ER stress in adipocytes and insulin resistance: mechanisms and significance (review). Mol Med Rep 2014;10:2234-40.ArticlePubMed
• 31. Dennis EP, Edwards SM, Jackson RM, Hartley CL, Tsompani D, Capulli M, et al. CRELD2 is a novel LRP1 chaperone that regulates noncanonical WNT signaling in skeletal development. J Bone Miner Res 2020;35:1452-69.ArticlePubMedPDF
• 32. Liu R, Li N, Lin Y, Wang M, Peng Y, Lewi K, et al. Glucagon like peptide-1 promotes adipocyte differentiation via the Wnt4 mediated sequestering of beta-catenin. PLoS One 2016;11:e0160212.ArticlePubMedPMC
• 33. Tang Q, Liu Q, Li Y, Mo L, He J. CRELD2, endoplasmic reticulum stress, and human diseases. Front Endocrinol (Lausanne) 2023;14:1117414.ArticlePubMedPMC
• 34. Albig AR, Roy TG, Becenti DJ, Schiemann WP. Transcriptome analysis of endothelial cell gene expression induced by growth on matrigel matrices: identification and characterization of MAGP-2 and lumican as novel regulators of angiogenesis. Angiogenesis 2007;10:197-216.ArticlePubMedPDF

Figure & Data

References

Citations

Citations to this article as recorded by