Human samples.

All women underwent a 75-g oral glucose tolerance test at 24–28 wk of gestation. GDM was diagnosed according to the International Association of Diabetes and Pregnancy Study Group criteria [plasma glucose: 0 h (fasting) ≥ 5.1 mmol/l; 1 h ≥ 10.0 mmol/l; 2 h ≥ 8.5 mmol/l]. Full-term placentas were collected from eight normal women and eight GDM patients at Shenzhen Maternity and Child Healthcare Hospital. Placental tissues were taken within 15 min of operative delivery, immediately snap-frozen in liquid nitrogen, and then stored at −80°C. Informed, written consent was obtained from the participants. All research protocols associated with clinical samples were approved by the Committee on the Use of Humans for Teaching and Research, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, and by Shenzhen Maternity and Child Healthcare Hospital, Southern Medical University.

Animal experiments.

Virginal female 7- to 8-wk-old C57BL/6 mice obtained from Vitalriver (Keaoxieli, China) were acclimated to housing conditions for 1 wk, including fresh food and water ad libitum, temperature of 22.0 ± 1°C, humidity of 40–60%, and a 12:12-h light-dark cycle. Timed mating of animals was conducted by placing females with fertile males to induce pregnancy (D0 is the day of vaginal plug). Thirty-two pregnant mice were divided into two groups: chow and HFD (45% kcal fat, 35% kcal carbohydrates, and 20% kcal proteins; Research Diets). Eight pregnant mice were transfected with small interfering RNA (siRNA) with a terminal-dTT modification for GPR1 or with a negative control (NC) in the tail vein (8, 40) from gestational day 9 (D9) to D18, respectively. Each siRNA (NC) sense, UUCUCCGAACGUGUCACGU-dTdT, and anti-sense, ACGUGACACGUUCGGAGAA-dTdT, and GPR1 sense, CCACUCUUUGGUUCCUCAA-dTdT, and anti-sense, UUGAGGAACCAAAGAGUGG-dTdT) (Genepharma), was administered at 0.1 nmol/g per injection once every other day. All research protocols associated with animal experiments were approved by the Committee on the Use of Live Animals for Teaching and Research, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences (approval no. SIAT-IRB-160226-YYS-ZHANGJ-A0198.). In addition, all animal experimental methods were conducted in accordance with the approved guidelines and regulations.

IPGTT, fasting blood glucose, body weight, and serum parameters.

Intraperitoneal glucose tolerance tests (IPGTTs) were conducted in unrestrained mice that were fasted for 16 h and injected intraperitoneally with 2 g of d-(+)-glucose/kg body wt [anhydrous d-(+)-glucose; Sibopharm Chemical Regent], and blood glucose was sampled from the tail at 0, 15, 30, 60, and 120 min after the glucose injection on D16 by a glucose meter (Roche). Body weight was measured after siRNA injection. Blood was collected via retroorbital bleeding on D18. Blood was centrifuged at 3,500 rpm and 4°C for 5 min, and serum was collected from the supernatant. Serum insulin levels were measured using insulin kits (Millipore). Serum triglyceride (TG) levels were measured using kits (Dongou). Testosterone, progesterone, estradiol, and prolactin (PRL) levels in conditioned media were measured using commercial iodine-125 radioimmunoassay kits (Beijing North Biotechnology Research Institute). The sensitivity of the progesterone and estradiol radioimmunoassays was 20 ng/ml. The intra-assay and interassay errors were < 10% and < 15%, respectively.

Placenta weight and TG.

Placentas were immediately collected and weighed, and placental wet weights were averaged per litter. Placenta TG was measured in placenta tissue homogenate that was strictly prepared according to the TG kit (Dongou) instructions.

Histopathology, immunohistochemistry, and confocal microscopy.

Tissues, including pancreas and placenta samples, were then rapidly removed from the animals. Mouse tissues were fixed in Bonn’s liquid, dehydrated in alcohol, embedded in paraffin, and sectioned. Morphological examination was performed on 5-μm sections of paraffin-embedded tissue with hematoxylin and eosin (H&E) stain, according to standard histological protocols. Islet areas were calculated at the tail of the pancreas, and quantification of immunohistochemistry was calculated using ImageJ software, as previously described (1). Antigens were retrieved in citrate buffer using microwave heat. Immunohistochemistry was carried out. The expression of GPR1 (Wuhan Booute Biotechnology) was determined by confocal microscopy, as described previously. JEG3 cells were probed with rabbit anti-GPR1 primary antibody to GPR1 (Wuhan Booute Biotechnology) and then incubated with goat anti-rabbit IgG Alexa 568 (Invitrogen).

Cell culture and treatment.

JEG3 cell line was obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Cells were cultured as previously described (35). Stable knockdown and overexpression of GPR1 in JEG3 were established using Lipofectamine 2000 transfection reagent (Invitrogen). The targeted GPR1 short hairpin RNA (shRNA) vector (sense, CACTCTCTGATTGTCATTATAT; anti-sense, ATATAATGACAATCAGAGAGTG; Open Biosystems) contained a puromycin resistance gene (Invitrogen), and the human GPR1 (NM_001098199.1) expression vector (Stanford University) contained a G418 resistance gene. Transfections with pSM2C-GPR1-shRNA or an empty vector, and with pcDNA3.0-GPR1 or an empty vector, were carried out according to the manufacturer’s protocol, and cells were selected in medium containing 2 µg/ml of puromycin and 500 µg/ml of G418 (Sigma), respectively, for 2–3 wk. The colonized cells were isolated, amplified, and used for subsequent experiments. Cell lipid accumulation was assessed in stably transfected JEG3 cells treated with 5% bovine serum albumin (Sigma) and free fatty acids (400 µM palmitic acid and 800 µM oleic acid; Sigma) for 24 h after 1 h of incubation with a phosphorylated AKT (pAKT) inhibitor (wortmannin; Selleck) or with DMSO.

Oil Red O staining.

Oil Red O staining of placenta was performed in 10-μm frozen sections that were fixed in 4% paraformaldehyde and dehydrated in a gradient of sucrose. Tissue and cell staining with the Oil Red O (Sigma) method was performed as previously described (32).

Western blotting and real-time RT-PCR analysis.

Protein extraction from tissues and cells and Western blotting were performed as previously described (27). Antibodies for GPR1, glucose transport protein-3 (GLUT-3; Abcam), pAKT/AKT (CST), fatty acid binding protein 4 (FABP4), and β-actin (Sigma) were used. Real-time qPCR was performed as previously described (15). Total RNA was isolated from mouse tissues or cells using RNA plus liquid kits (Takara) following the manufacturer’s protocol. RNA (0.5–2 μg) was reverse transcribed into cDNA using reverse transcription kits (Toyobo). Expression levels of several genes were determined by real-time RT-PCR using a Roche LightCycler instrument (Roche) with SYBRgreen (Toyobo) detection, according to the manufacturer’s protocol. The RT-PCR primers are as follows: mouse primer sequence of 18srRNA (forward, GATCCCGGCTCTTAATATTCGAAT; reverse, GCCAGAGTCTCGTTCGTTATC); chemerin (forward, TACAGGTGGCTCTGGAGGAGTTC; reverse, CTTCTCCCGTTTGGTTTGATTG); Gpr1 (forward, GGAGCTCAGCATTCATCACA; reverse, GACAGGCTCTTGGTTTCAGC); Glut1 (forward, GAGCATCTTCGAGAAGGCAGGTGT; reverse, GGCCACAATGAACCATGGAATA); Glut2 (forward, TCCTGGTCTTCACCCTGTTTAC; reverse, GGTCATCCTCACACACTCTCTG); Glut3 (forward, CTCTTCAGGTCACCCAACTACGT; reverse, CCGCGTCCTTGAAGATTCC); C/ebpα (forward, AGCAACGAGTACCGGGTACG; reverse, TGTTTGGCTTTATCTCGGCTC); Pparγ (forward, TCTGGGAGATTCTCCTGTTGA; reverse, GGTGGGCCAGAATGGCATCT); human primer sequence of 18srRNA (forward, CGAAAGCATTTGCCAAGAAT; reverse, AGTCGGCATCGTTTATGGTC); chemerin (forward, TACAGGTGGCTCTGGAGGAGTTC; reverse, CTTCTCCCGTTTGGTTTGATTG); Gpr1 (forward, GGAGCTCAGCATTCATCACA; reverse, GACAGGCTCTTGGTTTCAGC).

Statistical analysis.

All data are shown as means ± SE. Statistically significant differences were confirmed using Student’s t-test or two-way ANOVA using the software GraphPad Prism trial version 7 (GraphPad Software, La Jolla, CA). P < 0.05 was determined to be statistically significant.

Chemerin and GPR1 expression in GDM patient placenta.

For clinical application of these data, such as utilizing GPR1 signaling for diagnosis and therapy, we measured chemerin/GPR1 signaling expression on GDM patient placentas. Real-time RT-PCR analysis showed that chemerin (Fig. 1A) was significantly increased and that GPR1 (Fig. 1B) was significantly decreased in GDM patient full-term placentas compared with the levels in normal placentas. Next, we further confirmed that GPR1 protein (Fig. 1C) was also significantly decreased in GDM patient placentas (Fig. 1E). GLUT-3 also showed a significantly decreased trend (Fig. 1, C and D). These results from clinical data suggest that GPR1 may plays an important role in development of GDM.

Upregulation of chemerin/GPR1 signaling in the mice placenta during pregnancy.

To understand the pathophysiological significance of GPR1 in GDM, we first analyzed the mouse placental expression patterns of chemerin (Fig. 2A) and GPR1 (Fig. 2B), which displayed an increase via RT-PCR test during pregnancy. Circulating chemerin (Fig. 2C) also showed increase via ELISA during pregnancy. In addition, to explore the effect of GPR1 signaling on GDM, a pregnant mouse model with an induced metabolism disorder was generated by feeding a HFD after mating. Glucose tolerance was remarkably impaired in the HFD-fed pregnant mice at D16 (Fig. 2D). Meanwhile, the circulating chemerin levels were significantly increased (Fig. 2E), but GPR1 mRNA expression showed a significant decrease compared with that in the chow-fed pregnant mice (Fig. 2F) at D18. Taken together, these results reveal that chemerin/GPR1 signaling was upregulated during pregnancy; however, GPR1 expression levels were downregulated in HFD-fed pregnant mice, which raised the question of whether chemerin/GPR1 signaling plays a role in metabolism homeostasis during pregnancy.

GPR1 knockdown impairs glucose homeostasis in HFD-fed pregnant mice.

To further study the function of GPR1 signaling in metabolism during pregnancy, GPR1 signaling was knocked down in chow- and HFD-fed pregnant mice via injecting siRNA-GPR1 into the tail veins. The GPR1 knockdown pregnant mice displayed normal body weights (Fig. 3, A and B). Then IPGTTs were performed to further identify the effect of GPR1 on metabolism homeostasis. Glucose tolerance was not dramatically different in the chow-fed GPR1 knockdown pregnant mice at D16 (Fig. 3C), whereas prominent glucose intolerance was observed in the HFD-fed GPR1 knockdown pregnant mice at D16 (Fig. 3D). Similar results were obtained via quantification of the IPGTT area under the curve with one-way ANOVA, showing significantly impaired glucose tolerance compared with that in the chow-fed pregnant mice injected with the NC (Fig. 3E). As we know, circulating insulin plays a role in regulating blood glucose homeostasis. We also measured serum insulin levels via an ELISA kit. Serum insulin was significantly increased in the HFD-fed pregnant mice compared with that in the chow-fed pregnant mice after injection of NC, and there was a dramatic decrease in serum insulin in the GPR1 knockdown group compared with that in the NC group of HFD-fed pregnant mice at D18 (Fig. 3F). These data indicated that the HFD induced the change of glucose intolerance, and GPR1 signaling knockdown exacerbated glucose intolerance and reduced insulin secretion, suggesting that GPR1 signaling knockdown may impact glucose metabolism during pregnancy.

Blood glucose could be regulated by the β-cell producing insulin. It was previously reported that PRL is required for maternal β-cell proliferation and insulin secretion during pregnancy (16). Since HFD-fed GPR1 knockdown pregnant mice displayed decreased β-cell proliferation at D18, we measured the serum PRL levels, which showed a significant decrease in the GPR1 knockdown group compared with the NC group of HFD-fed pregnant mice (Supplemental Fig. S1A; Supplemental data are available in the data supplement online at the journal website). In addition, previous clinical results have revealed that total testosterone levels were higher among women with GDM than among women without GDM in the first trimester (13). Steroid hormones, including testosterone, progesterone, and estrogen, were measured in the serum of all of the groups of pregnant mice at D18. Serum testosterone displayed a significant increase in the GPR1 knockdown mice compared with the NC mice in the respective HFD-fed and chow-fed groups. However, there was no difference in serum testosterone levels in NC mice between the HFD- and chow-fed pregnant groups (Supplemental Fig. S1B). Interestingly, progesterone and estrogen levels were somewhat altered. Serum progesterone was significantly increased in the GPR1 knockdown mice compared with the NC mice in the chow-fed group, and it was dramatically decreased in the HFD-fed mice compared with the chow-fed mice after GPR1 knockdown (Supplemental Fig. S1C). Contrary to the testosterone data, serum estrogen displayed a significant decrease in the GPR1 knockdown mice compared with the NC mice in the respective HFD-fed and chow-fed groups (Supplemental Fig. S1D). PRL receptor expression levels were significantly decreased in HFD-feeding mice compared with GPR1 knock down (Supplemental Fig. S1F). Estrogen receptor-α expression levels were significantly decreased in chow-feeding mice treatment with small interfering RNA for GPR1 (siGPR1) and HFD-feeding mice treatment with small interfering RNA for NC (siNC) compared with chow-feeding mice treatment with siNC (Supplemental Fig. S1G). Estrogen receptor-β expression levels were only observed significantly decreased in HFD-feeding mice compared with chow-feeding mice treatment with siNC (Supplemental Fig. S1H). PRL could promote insulin secretion, but PRL could not recover insulin secretion in GPR1-knockdown MIN6 cells. Insulin secretion has been shown to have an increased tendency and was not significantly different (Supplemental Fig. S1, I and J).

To address whether GPR1 was important for maternal β-cell proliferation during pregnancy, we analyzed islet areas and localization of related proteins, including insulin, GPR1, and proliferating cell nuclear antigen (PCNA). According to the representative photomicrographs of pancreatic H&E-stained sections (Fig. 3G) that were quantified via ImageJ software (Fig. 3H), the islet areas were significantly increased in the HFD-fed NC pregnant mice compared with the chow-fed NC pregnant mice, and there was a dramatic decrease in the islet areas in the GPR1 knockdown group compared with the NC group of HFD-fed pregnant mice at D18. From the immunohistochemical staining (Fig. 3I), we can observe GPR1 protein localization on β-cells according to insulin signaling; meanwhile, siGPR1 prominently induced GPR1 protein expression compared with that in the NC group of HFD-fed pregnant mice. PCNA (Fig. 3, J and K) are proliferation markers that were decreased in the GPR1 knockdown group compared with the levels in the NC group of HFD-fed pregnant mice at D18.

These results indicate that GPR1 might contribute to GDM development. GPR1 knockdown altered insulin release and impacted β-cell proliferation in the HFD-fed mice during pregnancy.

GPR1 knockdown impairs GLUT-3 expression and lipid metabolism in HFD-fed pregnant mouse placentas.

To further understand the functions and roles of the placenta, which is an important organ that regulates physiology through secreting hormones and cytokines during pregnancy, it was reported that HFD could dramatically influence placenta development (21). First, we measured expression levels via RT-PCR and Western blotting, which indicated a significant decrease in GPR1 signaling in the placenta via injection of siGPR1 (Fig. 4, A–C). We also analyzed GPR1 protein localization on D18 in the mouse placentas. GPR1 was mainly expressed in the labyrinth and junction zone, and GPR1 signaling was also reduced in the placenta via injection of siGPR1 compared with that in the NC-injected mice (Fig. 4E). Meanwhile, GPR1 signaling was also reduced in the fetus via injection of siGPR1 compared with that in the NC-injected mice (Supplemental Fig. S2). In addition, pAKT levels were decreased relative to the total AKT expression in the placenta in the HFD-fed GPR1 knockdown mice compared with the NC mice (Fig. 4, B and D). Nevertheless, the placenta weights were dramatically decreased in the GPR1 knockdown group compared with the NC group of the HFD-fed pregnant mice at D18 (Fig. 4F). The fetal weights were prominently increased in the HFD-fed group compared with the chow-fed group after injection of NC, and there was a significant decrease in the fetal weights in the GPR1 knockdown group compared with the NC group of HFD-fed pregnant mice at D18 (Fig. 4G). These results suggested that GPR1 may play an important role in GDM through the placenta.

The placenta carries out the function of transportation to provide nutrients and oxygen to the fetus and to eliminate waste matter into the mother. In this process, it is always accompanied with energy metabolism (27). In this study, we thoroughly explored the processes of glucose transport and lipid metabolism. First, we measured GLUT expression, both Glut1 and Glut2 mRNA showed no difference (Supplemental Fig. S3), and GLUT-3 showed significant decrease in the GPR1 knockdown mice compared with the NC mice via Western blotting and RT-PCR (Fig. 5, A and B, and Supplemental Fig. S2). To characterize the effect of GPR1 signaling on lipid metabolism, we also measured metabolic parameters in the serum and placenta. Maternal serum TG levels in the HFD-fed GPR1 knockdown mice were significantly decreased compared with those in the NC mice (Fig. 5C). Placenta TG levels in the HFD-fed GPR1 knockdown mice were remarkably lower than those in the NC mice (Fig. 5D). Of note, lipid accumulation in the placenta tissue from the HFD-fed GPR1 knockdown mice was decreased compared with that from the NC mice by Oil Red O staining (Fig. 5E). Next, we measured related markers of lipid metabolism. FABP4 protein (Fig. 5F) was also dramatically decreased via quantification of Western blotting (Fig. 5G) in the HFD-fed GPR1 knockdown mice compared with the NC mice. Meanwhile, RT-PCR analysis of peroxisome proliferator-activated receptor-γ (PPAR-γ) (Fig. 5H) and CCAAT-enhancer binding protein-α (C/EBP-α) (Fig. 5I) showed significantly decreased mRNA expression levels in the HFD-fed GPR1 knockdown mice compared with the NC mice. Taken together, decreased GLUT-3 protein, C/EBP-α and PPAR-γ mRNA levels, low serum and placenta content, and reduced placenta lipid accumulation support the hypothesis that GPR1 knockdown may impair nutrient transport and metabolism, leading to a disruption in metabolism homeostasis and to the development of GDM.

GPR1 promotes GLUT-3 and lipid accumulation in trophoblasts.

To understand the mechanism of GPR1 in maternal metabolism homeostasis, we explored the impact of GPR1 signaling on trophoblasts. First, we constructed stable transfection cell lines with GPR1 knockdown and overexpression via an shRNA vector and an expression vector, respectively. Immunostaining analysis revealed that GPR1 expression in the stably transfected GPR1 knockdown cell line was barely detectable compared with that of the NC. Meanwhile, GPR1 localized on the surface and nucleus of the cell line, which may be associated with GPR1 internalization (Fig. 6A). In the stably transfected GPR1 knockdown cell line (Fig. 6B), GPR1 protein expression was lower than that in the cells transfected with an empty vector. Interestingly, GLUT-3 and FABP4 protein levels were decreased when GPR1 signaling was reduced. We assessed lipid accumulation (Fig. 6C) and showed that the GPR1 knockdown significantly decreased lipid accumulation via quantification of Oil Red O staining (Fig. 6D). In the GPR1-overexpressed cell line, GPR1 protein expression was higher than that in the cells transfected with an empty vector (Fig. 6E). In contrast to their expression in the GPR1 knockdown trophoblasts, GLUT-3 and FABP4 protein levels were increased. Lipid accumulation (Fig. 6F) in the GPR1-overexpressed cell line was significantly decreased via Oil Red O staining (Fig. 6G). It was noted that pAKT was also associated with GPR1 signaling expression (Fig. 6, B and E). These results indicate that GPR1 may regulate glucose transport through GLUT-3 and regulate lipid metabolism in trophoblasts. Therefore, we hypothesize that decreased GPR1 signaling through inhibiting the pAKT signaling pathway mediates the placental expression of GLUT-3 in HFD-fed mice, leading to glucose homeostasis, and decreased GPR1 signaling disrupts lipid metabolism by reducing FABP4 expression.

GPR1 regulates glucose and lipid homeostasis in trophoblasts through pAKT.

To further explore the GPR1 signaling pathway in energy homeostasis in trophoblasts, we studied the effect of AKT phosphorylation on this process by utilizing a pAKT inhibitor. We confirmed the efficiency of the inhibitor, and the AKT phosphorylation level was decreased with inhibitor treatment for 30 and 60 min in the GPR1-overexpressed cell line (Fig. 7A). GLUT-3 protein was decreased after the GPR1-overexpressed trophoblasts were supplemented with the inhibitor for 24 h compared with in the cells supplemented with DMSO (Fig. 7B). Lipid accumulation (Fig. 7C), similar to GLUT-3 protein expression, showed a significant decrease with inhibitor treatment compared with DMSO treatment via quantification of Oil Red O staining (Fig. 7D). These results suggested that GPR1 signaling may stimulate the pAKT pathway to regulate glucose and lipid homeostasis in trophoblast cells. Together, these results also supported our hypothesis that GPR1 signaling knockdown decreases AKT phosphorylation-controlled glucose homeostasis by reducing GLUT-3 expression and disrupts lipid metabolism by reducing FABP4 expression in the placenta.