The Retinol-binding Protein 4 (rbp4), And Gestational Diabetes

1. Backgrounds and ObjectivesGestational diabetes mellitus (GDM) is defined as glucose intolerance with onset or first recognition during pregnancy. An increased prevalence of GDM all over the world has been long recognized and the prevalence of GDM varies greatly depending on the population and the criteria used. In Asian women especially in Chinese women, the prevalence of GDM substantially higher than for the general population. The imbalance between the capacity of the pancreaticβ-cells and the decreased insulin sensitivity during pregnancy is the main pathogenetic mechanism causing GDM, while our data show maybe the latter one is more important. There is a very close relationship between GDM and Type 2 diabetes; GDM is considered to be a transient unmasking of an underlying predisposition to Type 2 diabetes, induced by the metabolic changes of pregnancy. Women with GDM have a risk of up to 2.6-70% of developing type 2 diabetes after delivery, depending on the duration of follow-up. The genetic basis which is underlying in Type 2 diabetes is likely to be involved in GDM as well. This research consists of two important aims. One is to investigate the clinical feature of GDM, compare the differences between NGT and GDM during pregnancy. The other is to study RBP4, a newly discovered adipokine, whether plays a vital role in pathogenesis of GDM. To investigate the prevalence and metabolic impacts of RBP4 genetic variants, and also to find the RBP4 expression change on mRNA level and protein level in GDM.2. Methods2.1 set up the clinical database2.1.1. Subjects:During September 2005 to March 2008, 1595 Chinese pregnant women who periodically visited the out-patient clinic of Peking union medical college hospital (PUMCH) were included in our study. GCT was undertaken during 24th to 28th pregnant weeks. Women with a Plasma glucose level greater than 7.8mmol/L 1 hour after a 50g GCT underwent 100g OGTT after a week. Diagnosis of GDM was based on ADA criteria. There were 290 cases of GCT negative (GCT plasma glucose level less than 7.8mmol/L), 488 cases of GDM, 235 cases of IGT, and 582 cases of NGT.2.1.2. The anthropometric and clinical featuresThe anthropometric and clinical features of the patients were collected at the first prenatal examination during 13th to 15th pregnant weeks. The information consisted of age, height, progravidic weight, history of gravidity and parity, SBP, DBP, past history, family history of DM, WBC, ALT, ALB, Cr, BUN, etc.2.1.3. OGTTPlasma glucose and specific insulin levels were measured at 0h, 1h, 2h, and 3h while OGTT was performed. Moreover, the levels of CHO, TG, HDL, LDL, CRP and HbAlc at 0h were measured as well.2.2. SNP genotyping:2.2.1. The SNPs were further genotyped in individuals comprising of 505 patients and 687 normal subjects. The patients consisted of 342 cases GDM and 163 cases IGT while the normal subjects consisted of 401 cases NGT and 286 cases GCT negative.2.2.2. The genome DNA was extracted from the blood sample, and the objective fragments were obtained using PCR technique.2.2.3. Four RBP4 SNPs [rs 3758538 (-1265A>C), rs 3758539(-803G>A), rs 12265684(+6969C>G), rs 10882273(+11881T>C)] were genotyped using the LDR technique.2.2.4. Prove the LDR results using directional sequence analysis. The haplotypes within a gene are further analyzed.2.3. Measurement of Serum RBP42.3.1. Sera were extracted from the EDTA blood samples in 0hOGTT of 143 pregnant women (including 74 cases of GDM and 69cases of NGT) and RBP4 levels were measured2.3.2. Serum RBP4 levels were measured using RBP4 (human) ELISA kit.2.4. Analysis of RNA and protein expression levels of adipose tissue. 2.4.1 During November 2007 to January 2008, 41 samples of subcutaneous adipose tissue of the women who underwent selective cesarean in PUMC hospital were collected. There were 13 cases of GCT negative, 17 cases of GDM, 2 cases of IGT, 9 cases of NGT. The GCT negative group and the NGT group were combined as the control control, and the GDM group and the IGT group were combined as the case group.2.4.2 The genome DNA was extracted from the adipose tissue of the GCT negative group and then underwent SNP genotyping using LDR technique.2.4.3 Total RNA was extracted and the level of relative expression of RBP4 mRNA was detected using RT-PCR technique.2.4.4 The level of relative expression of RBP4 protein was detected using Western blotting technique.3. Results3.1 The clinical data analysis3.1.1 Accompanied by the increase of plasma glucose, the age, gravity, parity, pregravidic BMI(GCT negative has no pregravidic BMI but the first prenatal examination BMI), the frequency of positive DM family history, and SBP, DBP, ALT, WBC count in first prenatal examination, together with TG, CRP and specific insulin level in OGTT were gradually increased from GCT negative group to NGT, IGT, GDM group. The plasma BUN, Cr, CHO, HDL and LDL among these groups had no statistically significant difference.3.1.2 NGT→IGT→GDM, the estimated indices of insulin sensitivity were gradually decreased. The estimated indices ofβcell function were significantly lower in GDM than IGT or NGT. There is no difference between IGT and NGT, and the mean value ofβcell function indices were closed. In our opinion, the most appropriate estimated index for evaluatingβcell function was estimated first phase or estimated second phase, and for insulin sensitivity was ISI-OGTT.3.1.3 Pregravidic BMI, TG and DM family history were independent risk factors of GDM in Logistic regression, and the sequence from higher risk to lower risk was DM family history > TG > pregravidic BMI.4.2 SNP genotyping study4.2.1 Hardy-Weinberg equilibriumHardy - Weinberg equilibrium was calculated using the chi-squared test. All SNPs were in Hardy-Weinberg equilibrium either in the cases or in the controls.3.2.2 Single SNP analysisThe P values of two SNP (rs3758539 G, rs12265684 C) were less than 0.05 in the comparison of major homo vs minor homo + hetero and/or major allele vs minor allele frequency in case-control study (rs3758539 G vs A, OR=1.446, 95%CI 1.095-1.911, P=0.009; rs3758539 GG vs AG+AA, OR=1.532, 95%CI 1.131-2.075; rsl2265684 C vs G, OR=1.296, 95%CI 1.014-1.658).The major homo genotype of three SNPs (rs3758539, rsl2265684, rs10882273)were significantly high risk for GDM after Logistic correction for age, gravity, parity, pregravidic BMI, SBP, DBP, CHO, TG and DM family history (rs3758539 GG vs AG+GG, OR=1.832, 95%CI 1.252-2.680, P= 0.002; rsl2265684 CC vs CG +GG, OR= 1.757, 95%CI 1.234-2.458, P=0.001 ; rs10882273, OR=1.678, 95%CI 1.189-2.370, P=0.003).3.2.3 Clinical and metabolic characteristics grouped by RBP4 variantgenotypeThe major homo genotype of three SNPs had significantly higher HbA1c level and/or OGTT plasma glucose level, while among these SNPs rs3758539 G>A appeared to mostly involved in pregnant glucose intolerance: rs12265684 CC vs CG+GG, HbA1c 5.42±0.53 vs 5.31±0.41, P=0.010; rs10882273 TT vs CT+CC, HbA1c 5.42±0.54 vs 5.31±0.42, P=0.021; rs 3758539 GG vs AG+AA, HbA1c 5.44±0.52 vs 5.29±0.42, P=0.006, 2hOGTT glucose level 8.40±1.70 vs 8.11±1.44, P=0.042, 3hOGTT glucose level 7.38±1.47 vs 7.09±1.17, P=0.009. There is no statistical difference in age, pregravidic BMI, positive frequency of DM family history, indices of insulin sensitivity andβcell function among these genetic variants.3.2.4 LD evaluation and haplotype case-control study We estimated linkage disequilibrium among these variants by using haploview 4.0. As a result, rs3758538 was unique but rs3758539, rsl2265684 and rsl0882273 were in a tight LD block which was 12Kb long. We identified 4 common haplotypes among these 3 SNPs (rs3758539, rsl2265684 and 10882273) genotyped in all subjects and two of them were associated with GDM. The GCT haplotype (case vs control 86.0% vs 82.3%, P=0.016, OR=1.322, 95%CI 1.054-1.659) was a high risk of GDM while the AGC (case vs control 8.7% vs 11.3%, P=0.038,OR=0.746, 95%CI 0.566-0.985)was a protection one.3.2.5 Test for sample size or powerUsing the Quanto v.0.4.2, we identified that based on sample size in the present study, we had >85% power(P=0.05) to detect a difference in allele frequency of >10%, corresponding to an OR of >1.5.3.3 Measurement of serum RBP43.3.1 The serum RBP4 level of GDM group was little higher than that of NGT group (21.53±5.96ug/ml vs 20.84±4.31ug/ml, P=0.429), however there was no statistically difference after correction for age, progestational BMI and gestational weeks of OGTT.3.3.2 The serum RBP4 levels in major homo genotypes of rs3758539, rsl2265684 and rs10882273 had a higher tendency in comparison with that in minor homo + hetero genotypes, however, there was no significant difference.3.3.3 Correlation analysis: There was a positive correlation between Serum RBP4 level and age, pregravidic BMI and TG level, and a weakly negative correlation between Serum RBP4 level and ISI-OGTT (correlation= -0.199, P =0.017). After correction of mixed factors with multiple linear regression, we found that the serum RBP4 level was independently correlative with TG (Standardized coefficient beta= 0.331, P = 0.000) and age (Standardized coefficient beta = 0.194, P = 0.018), moreover, the TG level was mostly correlated with the RBP4 level.3.4 RBP4 mRNA and protein expression in subcutaneous adipose tissue3.4.1 RBP4 RT-PCR in adipose tissue3.4.1.1 RBP4 mRNA expression in adipose tissue was significantly increased in GDM compared with the controls (1.438 vs 1.034, P=0.025)after adjustment for age, pregravidic BMI.3.4.1.2 RBP4 mRNA expression and genotype of rs3758539: The GG genotype had higher RBP4/β-actin ratio than AA+AG(1.271 vs 1.007, P=0.185), but the P value did not meet the criteria.3.4.2 RBP4 Western Blotting in adipose tissue3.4.2.1 The results of western blotting showed there was no difference between the case and the control, however the mean value in the case group was little higher(0.779 vs 0.738, P=0.617).3.4.2.2 The GG genotype of rs3758539 had little higher RBP4/β-actin ratio in western blotting than the AG+AA genotypes, although there is no statistical difference.4. conclusionsFour informative SNPs of RBP4 which have been reported to be associated with T2DM or obesity were genotyped in our study. Among these variants, rs3758539, rs12265684 and rs10882273 were in a LD block, and they were all associated with glucose intolerance during pregnancy: The G allele and GG genotype of rs3758539, the C allele and CC genotype of rsl2265684, the TT genotype of rsl0882273, all of them were predisposing factors of GDM. The GCT haplotype which consisted of rs3758539, rs12265684 and 10882273 was a high risk of GDM while the AGC was a protection one. In conclusions, the polymorphisms of the RBP4 gene appear to be involved in the development of GDM in Chinese pregnancy.On mRNA level, we found RBP4 expression in adipose tissue was increased in the case than in the controls. On protein expression level, we found no statistically significant difference between case and control either in serum or in adipose tissue, but there was still a higher expression tendency in case than in control. The GG genotype of rs3758539 has a little higher expression than AA+AG genotypes, but these changes did not meet the statistic criteria. In this research, we first carry out the association analysis between SNPs in RBP4 gene and GDM. Moreover, we also studied the expression of RBP4 on mRNA level and on protein level both in case-control groups and the relevant genotype based on the results of association analysis. We hope that these results could apply a part of theory evidence for the pathogenesis of GDM.