2023-The genetic polymorphisms and levels of adipokines and adipocytokines that influence the risk of developing gestational diabetes mellitus in Thai pregnant women

Gene 860 (2023) 147228 Contents lists available at ScienceDirect Gene journal homepage: www.elsevier.com/locate/gene The genetic polymorphisms and levels of adipokines and adipocytokines that influence the risk of developing gestational diabetes mellitus in Thai pregnant women Watip Tangjittipokin a,b,*, Benyapa Thanatummatis c, Fauchil Wardati c, Tassanee Narkdontri a,b, Nipaporn Teerawattanapong a,b,d, Dittakarn Boriboonhirunsarn e a Department of Immunology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand b Siriraj Center of Research Excellence for Diabetes and Obesity, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand c Graduate Program in Immunology, Department of Immunology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand d Research Division, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand e Department of Obstetrics and Gynaecology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand ARTICLE INFO ABSTRACT Edited by: Nancy Hakooz Introduction: Aberrant immune and inflammatory response is thought to be involved in the pathogenesis of gestational diabetes mellitus (GDM). Keywords: Objective: To investigate the genetic polymorphisms and levels of adipokines/adipocytokines that influence the Gestational diabetes risk of developing GDM in Thai women. Adipokines Research design & methods: This case-control recruited 400 pregnant Thai women. A total of 12 gene poly­ Cytokines morphisms at ADIPOQ, adipsin, lipocalin-2, PAI-1, resistin, IL-1β, IL-4, IL-17A, TGF-β, IL-10, IL-6, and TNF-α Gene polymorphisms were analyzed by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) assay and Inflammation RNase H2 enzyme-based amplification (rhAmp) SNP assay. Serum levels of adipokines/adipocytokines were evaluated using Luminex assays. Results: Mean age, weight before and during pregnancy, body mass index before and during pregnancy, blood pressure, gestational age at blood collection, and median 50 g glucose challenge test were significantly higher in GDM women than control. Significantly lower adiponectin and higher IL-4 levels were found in GDM compared to controls (p = 0.001 and p = 0.03, respectively). The genotype frequencies of IL-17A (rs3819025) were significantly different between GDM and controls (p = 0.01). Using additive models, IL-17A (rs3819025) and. TNF-α (rs1800629) were found to be independently associated with increased risk of GDM (odds ratio [OR]: 2.867; 95 % confidence interval [CI]: 1.171–7.017; p = 0.021; and OR: 12.163; 95 %CI: 1.368–108.153; p = 0.025, respectively). In GDM with IL-17A (rs3819025), there was a significant negative correlation with lipocalin-2 and PAI-1 levels (p = 0.038 and p = 0.004, respectively). Conclusion: The results of this study highlight the need for genetic testing to predict/prevent GDM, and the importance of evaluating adipokine/adipocytokine levels in Thai GDM women. Abbreviations: ADIPOQ, Adiponectin; BMI, Body mass index; EtBr, Ethidium bromide; GCT, Glucose challenge test; GDM, Gestational diabetes mellitus; HOMA-IR, Homeostasis model assessment insulin resistance indexl HSP, Henoch-Schonlein Purpura; HWE, Hardy-Weinberg equilibrium; IDF, International Diabetes Federation; IL-10, Interleukin 10; IL-17A, Interleukin-17; IL-1β, Interleukin 1 beta; IL-4, Interleukin 4; IL-6, Interleukin 6; NGT, Normal glucose tolerance; PAI-1, Plasminogen activator inhibitor-1; PCR-RFLP, Polymerase chain reaction-restriction fragment length polymorphism; RA, Rheumatoid arthritis; rhAmp, RNase H2 enzyme-based amplification; T2DM, Type 2 diabetes; TGF-β, Transforming growth factor beta; TNF-α, Tumor necrosis factor alpha. * Corresponding author at: Department of Immunology, Faculty of Medicine Siriraj Hospital, Mahidol University, 2 Prannok Road, Bangkoknoi, Bangkok 10700, Thailand. E-mail address: [email protected] (W. Tangjittipokin). https://doi.org/10.1016/j.gene.2023.147228 Received 25 November 2022; Received in revised form 11 January 2023; Accepted 23 January 2023 Available online 26 January 2023 0378-1119/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/).

W. Tangjittipokin et al. Department of the Department of Obstetrics and Gynecology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand dur­ 1. Introduction ing month year to month year. The inclusion criteria of cases (GDM) were, as follows: 1) pregnant women were diagnosed using a 50 g Diabetes is a chronic disease characterized by high blood glucose glucose challenge test (GCT) screening test with cut off 140 mg/dl; 2) A levels. The International Diabetes Federation (IDF) reported a global 100 g oral glucose tolerance test (100-g OGTT) according to Carpenter/ prevalence of diabetes of 9.3 % or 463 million people in 2019. The Coustan criteria (Mpondo et al., 2015); and 3) pregnant women with a global prevalence of diabetes is estimated to increase to 10.2 % or 578 history of other types of diabetes were excluded. Blood biochemistry million by 2030, and to 10.9 % or 700 million by 2045 (Saeedi et al., parameters, including 50-gram glucose challenge test (GCT) and 100- 2019). During September 2018 to February 2019, the incidence of Thai gram oral glucose tolerance test (100-g OGTT) were examined at the GDM was 18.6 % in 1016 pregnant women enrollment before 20 weeks Department of Clinical Pathology, Faculty of Medicine Siriraj Hospital, of gestation (Phattanachindakun et al., 2022). Gestational diabetes Mahidol University, Bangkok, Thailand. mellitus (GDM) is hyperglycemia with onset recognition during preg­ nancy that leads to complications for both mother and fetus (HAPO The protocol of this study was conformed to the Declaration of Study Cooperative Research Group, 2008). At the beginning of 80 % of Helsinki guidelines and approved by the Siriraj Institutional Review GDM cases, beta (β)-cell dysfunction leads to chronic insulin resistance Board (SIRB) (COA no. Si577/2015), and written informed consent was (Buchanan and Xiang, 2005; Alharbi et al., 2022). Hyperglycemia re­ obtained from all 400 study subjects. sults from pancreatic β-cell destruction that is mediated by immune system T cells (Cnop et al., 2005). Pregnant women with GDM are at 2.2. Sample collection higher risk for subsequent development of type 2 diabetes (American Diabetes Association, 2019), cardiovascular disease, preterm delivery Peripheral venous blood samples were drawn from all 400 study (delivery before 37 weeks of gestation), shoulder dystocia, stillbirth, participants, centrifuged, and maintained at − 80 ◦C until use. Serum hyperbilirubinemia, and hypoglycemia in the newborn during the specimens were prepared for adipokine and adipocytokine measure­ perinatal period (Plows et al., 2018). The development of GDM is ment, and genomic DNA was extracted from white blood cells using a associated with various factors (Chiefari et al., 2017), including, ge­ Flexigene® DNA kit (Qiagen, Valencia, CA, USA) according to manu­ netics, rapid changes in maternal lifestyle, family history, other envi­ facturer’s instructions. ronmental factors during pregnancy, and ethnic group. The meta- analysis of T2DM genetic variants showed a significant relationship with 2.3. SNP genotyping oral glucose tolerance test values in Indian GDM patients (Khan et al., 2019). Lifestyle factors, such as overeating and physical inactivity, can 2.3.1. Polymerase chain reaction-restriction fragment length polymorphism contribute to the development of overweight status or obesity. Increased ADIPOQ (rs182052, rs140531754), adipsin (rs1629038), lipocalin-2 fatty acid and adipokine/adipocytokine production was found in obesity and increased the risk of developing GDM (Chiefari et al., 2017). Al­ (rs73672429 and rs11794980), PAI-1 (rs2070682 and rs6092), resistin terations in adipokine and adipocytokine secretion plays an important (rs3745367 and rs1862513), IL-10 (rs3024490), IL-6 (rs1800796 and role in glucose homeostasis during pregnancy (Ramachandrayya et al., rs2066992) and TNF-α (rs1800629 and rs1799724) were examined via 2020). polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) according to the manufacturer’s instructions and using re­ Adipokines and adipocytokines are proteins that are secreted by striction enzyme digestion. DNA was amplified with a specific primer adipose tissue (Fain et al., 2004). Adipokines have significant potential (Integrated DNA Technologies [IDT], Coralville, IA, USA). In brief, PCR for therapeutic application in clinical settings for conditions that include was performed in 0.2 ml with a total volume of 25 µL in each, which obesity-related metabolic disorders, cardiovascular diseases, and other consisted of 125 ng DNA templates, 10x PCR buffer, 2 mM dNTP, 50 mM diseases (Blüher and Mantzoros, 2015). Adipokines, such as adipo­ MgCl2, 10 µM each of F- and R-primers, immolase Taq DNA polymerase, nectin, adipsin, lipocalin-2, plasminogen activator inhibitor-1 (PAI-1), and sterile distilled water. The PCR cycles were, as follows: (i) dena­ and resistin, are associated with body mass index (BMI), insulin resis­ turation at 94 ◦C for 30 s (10 min for the first cycle), (ii) annealing tance, and features of adiposity (Achari and Jain, 2017). Adipocytokines temperature for 30 s (depending on SNPs), and (iii) extension at 72 ◦C are associated with the regulation of homeostasis, blood pressure, lipid for 30 s (5 min for the last cycle). The PCR products were digested with a and glucose metabolism, insulin resistance and inflammation (Dhasta­ specific restriction enzyme and incubated overnight at the temperature gir, 2017). The analysis of molecular genetic variants can help identify specific to each SNP. The digested products were separated by 12 % the causes of diabetes in pregnant women and infants (Ali Khan, 2021). polyacrylamide gel electrophoresis and visualized by silver staining. The details of genes and SNPs genotyping are listed in Supplementary However, the roles of adipokine and adipocytokine variations and Table S1. levels specific to the development of insulin resistance in GDM have not yet been fully explored or are sufficiently well understood. Accordingly, 2.3.2. RNase H2-dependent PCR genotyping assays the aim of this study was to investigate the genetic polymorphisms and Ten SNPs of adipocytokine genes (IL-1β, IL-4, IL17A, TGF-β, and IFN-g) levels of adipokines and adipocytokines that influence the risk of developing GDM in Thai pregnant women. The genetic variants that were detected by RNase H2 enzyme-based amplification (rhAmp®) SNP were investigated were ADIPOQ (rs182052 and rs140531754), adipsin assay (IDT) using real-time PCR according to the manufacturer’s in­ (rs1629038), lipocalin-2 (rs73672429 and rs11794980), PAI-1 (rs6092, structions. The specific primers were designed using the online rhAmp® rs2070682 and rs1799889), resistin (rs1862513), IL-1β (rs1143643 and Genotyping Design Tool available from IDT at the following link: htt rs16944), IL-4 (rs2243250 and rs2243290), IL-17A (rs2275913 and ps://www.idtdna.com/site/order/designtool/index/GENOTYPING_ rs3819025), TGF-β (rs1800469), IL-10 (rs3024490), IL-6 (rs1800796 PREDESIGN. For each assay, rhAmp utilizes two allele-specific primers and rs2066992), and TNF-α (rs1800629 and rs1799724). and a locus-specific primer. To confirm primer specificity, the determined rhAmp SNP assay primers were subjected to Basic Local Alignment Search 2. Materials and methods Tool (BLAST) (https://blast.ncbi.nlm.nih.gov/Blast.cgi) analysis by comparing them with the reference genomes of the adipocytokine genes 2.1. Subjects (IL-1β, IL-4, IL-17A, TGF-β, and IFN-g). This case-control study recruited 200 pregnant women with GDM PCR amplification reactions were prepared according to the protocol (study group or case), and 200 pregnant women with normal glucose for genotyping with rhAmp SNP assays using rhAmp genotyping master tolerance (NGT) (control group or controls) from the Outpatient mix and rhAmp reporter mix with reference dye (IDT). Ten ng of sample

W. Tangjittipokin et al. Table 1 Comparison of clinical characteristics between cases and control. DNA were used in 5 µL reactions with the following parameters: a pre- reading stage of 30 s at 60˚C; enzyme activation 10 min at 95˚C; 40 cy­ Clinical characteristic Controls Cases p- cles of 10 s at 95˚C, 30 s at 60˚C, 20 s at 68˚C; and a final post- reading stage (n ¼ 200) (n ¼ 200) value of 30 s at 60˚C. The analysis was performed using a LightCycler® 480 PCR system and end-point genotyping was performed using Light­ Age (years) 31.15 ± 5.74 33.99 ± 4.69 <0.001 Cycler® 480 software version 1.5 (Roche Diagnostic, Mannheim, Gestational age at blood 13.23 ± 7.50 18.20 ± 10.14 <0.001 Germany). collection (weeks) 58.53 ± 12.43 61.37 ± 13.62 0.01 2.3.3. Allele-Specific primer PCR Weight before pregnancy 0.02 This technique was used to screen 4G/5G polymorphisms (PAI-1; 62.07 ± 12.67 65.31 ± 13.87 0.01 (kg) 0.002 rs179989) in GDM and controls. Three primers need to be designed for Weight during pregnancy 23.13 ± 4.53 24.33 ± 4.92 0.28 4G/5G polymorphisms (rs179989 insertion/deletion) ASP-PCR geno­ 0.26 typing method and two primers need to be designed for internal control (kg) 24.50 ± 4.66 25.89 ± 5.04 0.04 (Mohammed Suhail et al., 2009). The control upstream primer is used to BMI before pregnancy (kg/ 0.02 confirm the occurrence of DNA amplification in the absence of the allele 38 ± 1.62 37.82 ± 1.67 <0.001 on the genomic DNA. The PCR products were subjected electrophoresis m2) <0.001 on a 2.5 % agarose gel and stained with ethidium bromide (EtBr). The BMI during pregnancy (kg/ 3056.54 ± 3120.19 ± 539.34 <0.001 details of genes and SNPs genotyping are listed in Supplementary 466.99 <0.001 Table S1. m2) 113.94 ± 12.45 116.61 ± 13.02 <0.001 Gestational age at delivery 2.4. Adipokines and adipocytokines measurement 70.67 ± 9.18 72.89 ± 9.96 0.75 (weeks) Serum levels of adipokines including adiponectin, adipsin, lipocalin- Newborn body mass (g) 130.50 (113, 188.50 (165.25, 0.001 2, PAI-1, and resistin in both GDM and controls were evaluated using 145.75) 203.75) MILLIPLEX® MAP Human Adipokine Magnetic Bead Panel 1 (Merck Systolic blood pressure 78.23 ± 8.16 (n 84.77 ± 10.83(n = 0.32 Millipore Corporation, MA, USA) in a 96-well plate. Adipokines were (mmHg) = 62) 103) quantified using Luminex® assays (Luminex Corp., Austin, TX, USA) 143.87 ± 24.23 196.52 ± 20.50(n 0.18 (Hung et al., 2013). Diastolic blood pressure (n = 62) = 102) (mmHg) 116.95 ± 20.00 169.66 ± 25.67 (n 0.31 Serum levels of adipocytokines, including IL-1β, IFN-g, IL-10, IL-17A, (n = 62) = 102) IL-4, IL-6, and TNF-α, in both GDM and controls were evaluated using a 50-gram glucose challenge 106.58 ± 20.22 139.60 ± 24.39(n 0.49 Human Cytokine Magnetic Bead Panel (Merck Millipore Corporation, test (GCT) (md/dl) (n = 62) = 103) MA, USA). Adipocytokines were quantified using Luminex® assays 0.20 (Luminex Corp.) (Benito et al., 2014). OGTT (Baseline) (mg/dl) 72 (50 %) 72 (50 %) 0.91 (n) 47 (54.7 %) 39 (45.3 %) 0.33 2.5. Statistical analysis 66 (47.1 %) 74 (52.9 %) 0.74 OGTT (1st hour) (mg/dl) 15 (50 %) 15 (50 %) 0.03 All statistical analyses were performed using SPSS Statistics version (n) 0.50 22 (SPSS, Inc., Chicago, IL, USA). Kolmogorov-Smirnov test was used to 54315.5 37422.50 0.51 analyze the distribution of continuous patient demographic and clinical OGTT (2nd hour) (mg/dl) (31242.50, (25716.75, characteristic data. Normally distributed continuous data were (n) 93850) 67933.25) compared using Student’s t-test, and non-normally distributed contin­ (n = 190) (n = 190) uous data were compared using Mann-Whitney U test. Comparisons of OGTT (3rd hour) (mg/dl) 4518.00 4443.50 (3555.50, categorical data were performed using either chi-square test or Fisher’s (n) (3650.75, 5190.25) exact test. Categorical, normally distributed continuous, and non- 5331.25) (n = 188) normally distributed data are shown as number (n) and percentage ABO blood groups (n) (n = 192) (%), mean ± standard deviation (SD), and median and interquartile Group O 1001.45 (704.76, 913.57 (588.05, range (IQR), respectively. Genotyping frequencies between GDM and Group A 1368.75) 1334) non-diabetic pregnant women were compared using Pearson’s chi- Group B (n = 190) (n = 190) square test. Michael H. Court’s (2005–2008) online Excel-based HWE Group AB 143.91 (111.79, 135.21 (100.66, Test (https://www.tufts.edu/~mcourt01/Documents/Court%20lab% Adipokines levels (pg/ml) 191.61) 195.37) 20%20HW%20calculator.xls) was used to test for deviation from Adiponectin (n = 193) (n = 188) Hardy-Weinberg equilibrium (HWE) in NGT controls. The observed and 197.80 (160.15, 188.87 (150.16, expected frequencies of the genotypes distributed in HWE were analyzed Adipsin 241.42) 248.30) using Pearson’s chi-square test. The genetic risk for homo minor alleles (n = 200) (n = 192) in GDM subjects was examined by binary logistic regression and referred Lipocalin-2 to the major allele in NGT controls with adjustments for confounding 1.03 (0.96, 2.86) 1.49 (1.11, 1.98) factors that included age, gestational age at blood collection, BMI before Resistin (n = 33) (n = 36) pregnancy, BMI during pregnancy, systolic blood pressure, and diastolic 3.19 (0.54, 9.69) 2.69 (0.68, 9.32) blood pressure. The association between adipokine/adipocytokine PAI-1 (n = 44) (n = 39) levels and clinical characteristics in GDM subjects was analyzed by 1.13 (0.73, 1.99) 1.81 (0.86, 1.85) Spearman’s correlation. Analysis for correlation between adipokine/ Adipocytokines levels (n = 43) (n = 41) adipocytokine gene variations and clinical characteristics in GDM sub­ (pg/ml) 0.88 (0.58, 4.86) 1.45 (0.58, 4.18) jects was analyzed by linear regression analysis with adjustments for the (n = 44) (n = 39) same immediately aforementioned confounding factors. The results of IL-1β 11.73 (2.96, 46.80 (17.89, linear regression analysis are shown as odd ratio (OR) or adjusted odds 42.14) (n = 41) 92.52) (n = 38) ratio (aOR) with their respectively 95 % confidence intervals (95 %CI). IFN-g 3.21 (0.94, 5.97) 3.21 (3.21, 6.76) (n = 43) (n = 40) Il-10 9.19 (6.68, 16.88) 10.62 (6.66, 16.63) (n = 44) (n = 41) IL-17A IL-4 IL-6 TNF-α Values are expressed as number (%), mean ± standard deviation (SD), or median (Q1, Q3). P value < 0.05 indicates statistical significance.

W. Tangjittipokin et al. Gene 860 (2023) 147228 Fig. 1. Serum adipokine levels in gestational diabetes mellitus (GDM) women compared to controls. A p-value<0.05 was considered to reflect statistical significance. pressure and diastolic blood pressure were both significantly elevated in GDM patients compared to control subjects (p = 0.04 and p = 0.02, 3. Results respectively). The median level of the 50-gram glucose challenge test (GCT) was significantly higher in GDM than in controls (p < 0.001). The 3.1. Clinical characteristics of subjects average mean 100-gram oral glucose tolerance test (100-g OGTT) at baseline, the 1st hour, the 2nd hour, and the 3rd hour of GDM were The clinical characteristics of the 200 enrolled pregnant women with significantly higher than NGT (p < 0.001). No significant difference was GDMs and 200 included pregnant women with normal glucose tolerance observed between groups for mean gestational age at delivery, newborn (NGTs; controls) are shown and compared in Table 1. There was a sig­ body mass, and the frequency ABO blood group. nificant difference between GDM cases and controls relative to both mean age (p < 0.001) and mean gestational age at blood collection (p < 3.2. Adipokine/adipocytokine levels and adipokine/adipocytokine genes 0.001). Mean weight before pregnancy and during pregnancy were compared between GDM and controls significantly higher in GDM patients compared to controls (p = 0.01 and p = 0.02, respectively). Mean BMI before pregnancy and during preg­ As shown in Table 1 and Fig. 1A, the level of adiponectin was nancy were also both significantly higher in GDM women than in NGT significantly lower in GDM patients compared to controls (p < 0.001). women (p = 0.01 and p = 0.002, respectively). Mean systolic blood However, there were no statistically significant differences between 4

W. Tangjittipokin et al. Gene 860 (2023) 147228 Fig. 2. Serum adipocytokine levels in gestational diabetes mellitus (GDM) women compared to controls. 5

W. Tangjittipokin et al. Gene 860 (2023) 147228 Fig. 3. Correlation matrix heatmap of Spearman’s Rho correlations between different combinations of adipokine levels, adipocytokine levels, and clinical parameters in gestational diabetes mellitus women. groups for the levels of adipsin, lipocalin-2, resistin, or PAI-1. Serum IL-4 significant differences in the genotypes of IL-17A (rs3819025) between levels were significantly increased in GDM patients compared with GDM and controls (p = 0.01) (Table 3). There were no significant dif­ controls (p = 0.03), as shown in Table 1 and Fig. 2E. There were no ferences between GDM patients and controls for the genotypes statistically significant differences in the levels of IL-1β, IFN-g, IL-10, IL- rs1143643 and rs16944 of the IL-1β gene; rs2243250 and rs2243290 of 17A, IL-6, or TNF-α between GDM patients and NGT controls. the IL-4 gene; rs2275913 of IL-17A gene; rs1800469 and rs3024490 of the TGF-β gene; rs3024490 of IL-10 gene; rs1800796 and rs2066992 of 3.3. Analysis for association between clinical parameters and serum IL-6 gene; rs1800629 and rs1799724 of TNF-α gene; or rs20697272 and adipokine levels or adipocytokine levels in GDM patients rs2430561 of IFN-g gene. Due to the observed lower adiponectin levels and higher IL-4 levels Logistic regression analysis was used to assess for association be­ in GDM patients compared to controls (Table 1), we set forth to deter­ tween each SNP and GDM in 3 different genetic models (dominant, mine if adiponectin and/or IL-4 levels are significantly associated with recessive, and additive). Among the evaluated SNPs, the following SNPs clinical parameters in GDM patients, and that analysis is shown in Fig. 3. were found to be independently associated with increased risk of GDM Spearman’s Rho (rs) correlation analysis revealed negative associations after adjustment for age, gestational age at blood collection, BMI before between serum adiponectin levels with gestational age at blood collec­ pregnancy, BMI during pregnancy, systolic blood pressure, and diastolic tion, weight before pregnancy, weight during pregnancy, BMI before blood pressure. IL-17A (rs3819025) was found to independently predict pregnancy, BMI during pregnancy, newborn body mass, and serum TNF- GDM under both the addictive model and recessive model (adjusted OR α (Spearman’s Rho correlation coefficient [rs] = -0.149, p = 0.041; rs = [aOR]: 2.867; 95 % confidence interval [95 %CI]: 1.171–7.017, p = -0.345, p < 0.001; rs = -0.187, p = 0.010; rs = -0.380, p < 0.001; rs = 0.021 and aOR: 0.302; 95 %CI: 0.125–0.727, p = 0.008, respectively) -0.488, p < 0.001; and, rs = -0.591, p < 0.001, respectively). We also TNF-α (rs1800629) was also found to independently predict GDM under found a significantly positive correlation between IL and 4 levels and both the addictive model and the recessive model (aOR: 12.163, 95 %CI: weight before pregnancy, BMI before pregnancy, serum adipsin, IL-10, 1.368–108.153, p = 0.025, and aOR: 0.081, 95 %CI: 0.009–0.719; p = and IL-6 levels (rs = 0.438, p = 0.006; rs = 0.408, p = 0.011; rs = 0.024, respectively) (Table 3). 0.373, p = 0.021; rs = 0.389, p = 0.016; and rs = 0.508, p = 0.001, respectively). The genotype frequency of the adipokine genes (ADIPOQ: rs182052 and rs140531754; adipsin: rs1629038; lipocalin-2: rs73672429 and 3.4. Genotype frequency of adipokine/adipocytokine genes in GDM rs11794980; PAI-1: rs1799889, rs2070682 and rs6092; resistin: patients and controls rs3745367 and rs1862513) was not significantly different between in GDM patients and control subjects (Table 2). The variants of the adipocytokine genes were investigated by PCR- RFLP and rhAmp SNP assay. The results of those analyses showed 6

Table 2 Genotyping frequencies of adipokine polymorphisms in cases and controls. Gene SNP (major/minor Risk allele Risk allele freq. Subject Genotype frequency, n alleles) freq. (cases) (controls) (%) A/ A/ B/ P A B B ADIPOQ rs182052 (G > A) 0.39 0.39 Cases 70 103 27 0.88 0.08 0.11 Controls 69 107 24 Adipsin rs140531754 (T > 0.32 0.34 Lipocalin- C) 0.35 0.36 Cases 168 31 1 0.21 rs1629038 (G > C) 0.51 0.50 Controls 160 35 5 2 0.47 0.42 PAI-1 rs73672429 (A > T) 0.41 0.42 Cases 91 90 19 0.79 0.13 0.1 Controls 85 93 22 Resistin rs11794980 (C > G) 0.40 0.35 0.45 0.41 Cases 88 84 28 0.99 rs1799889 (4G > Controls 87 84 29 5G) rs2070682 (T > C) Cases 50 97 53 0.94 Controls 53 96 51 7 rs6092 (G > A) Cases rs3745367 (G > A) Controls 56 102 42 0.27 71 91 38 rs1862513 (G > C) Cases Controls 69 97 34 0.78 65 104 31 Cases Controls 148 51 1 0.17 162 36 2 Cases Controls 68 105 27 0.23 78 105 17 Cases Controls 54 114 32 0.36 67 103 30 A/A homo major allele, A/B heterozygote allele, B/B homo minor allele. OR odds ratio, CI confidence interval additive, dominant, and recessive genetic models using logistic regression with/without adjustment for age, g systolic blood pressure (mmHg), diastolic blood pressure (mmHg). The OR and 95% CI of having the risk allele heterozygote allele are shown in the recessive model.

W. Tangjittipokin et al. Addictive model Dominant model Recessive model Pa OR (95 % Pb OR (95 % Pa OR (95 % Pb OR (95 % Pa OR (95 % CI) Pb OR (95 % CI) CI) CI) CI) CI) 8 0.654 0.4511.322 0.9761.006 0.9081.028 0.5850.848 0.3760.742 1.159 (0.640–2.728) (0.664–1.524) (0.645–1.637) (0.471–1.529) (0.383–1.436) (0.608–2.207) 0.206 0.3560.785 0.3670.766 0.1454.974 0.2164.084 1 0.139 0.237 (0.470–1.312) (0.429–1.368) (0.576–42.970) (0.440–37.900) 0.196 (0.025–2.211) (0.023–1.699) 0.836 0.5990.899 0.6360.898 0.5611.215 0.9551.021 0.922 (0.604–1.338) (0.575–1.403) (0.630–2.344) (0.496–2.101) 9 0.504 (0.430–1.977) 0.790 0.893 0.8230.956 0.9570.988 0.9681.011 0.8470.941 (0.396–1.576) 1.046 (0.642–1.422) (0.634–1.540) (0.577–1.773) (0.505–1.751) (0.540–2.028) 9 0.908 0.8571.059 0.7551.074 0.8921.035 0.7050.917 0.8730.959 0.966 (0.569–1.969) (0.684–1.686) (0.628–1.705) (0.586–1.435) (0.572–1.608) (0.531–1.756) 0.439 0.0961.437 0.1821.384 0.5300.854 0.9070.967 4 0.675 1.285 (0.938–2.202) (0.858–2.232) (0.523–1.397) (0.557–1.680) 1.124 (0.682–2.422) (0.651–1.941) 0.357 0.6950.920 0.8831.036 0.6040.869 0.2540.706 1.370 (0.605–1.397) (0.646–1.661) (0.510–1.479) (0.388–1.284) 7 0.194 (0.701–2.679) 1.453 0.572 0.0901.510 0.1261.518 0.5841.960 0.5292.219 (0.827–2.554) 0.490 (0.938–2.431) (0.890–2.590) (0.176–21.788) (0.185–26.580) (0.041–5.840) 8 0.838 0.299 0.3391.221 0.7571.075 0.0940.578 0.2680.666 1.064 1.505 (0.811–1.840) (0.681–1.697) (0.304–1.098) (0.324–1.367) (0.587–1.928) (0.696–3.251) 0.318 0.1731.350 0.1841.386 0.7000.899 0.6360.864 7 0.640 1.418 (0.877–2.077) (0.857–2.241) (0.523–1.546) (0.472–1.582) 0.563 (0.714–2.816) (0.050–6.269) 3 0.080 1.849 (0.928–3.683) 6 0.341 1.348 (0.729–2.493) l, NA not applicable (analytical model not applicable due to low frequency). P-values were calculated for the gestational age at blood collection (weeks), BMI before pregnancy (kg/m2), BMI during pregnancy (kg/m2), are shown in the additive and dominant model. The OR and 95% CI of having the A/A homo major allele, A/B Gene 860 (2023) 147228

Table 3 Genotyping frequencies of adipocytokine polymorphisms in cases and controls. Gene SNP (major/minor Risk allele freq. Risk allele freq. Subject Genotype frequency, Addict alleles) (cases) (controls) n (%) Pa OR A/ A/ B/ P A B B 0.9111 (0.583– IL-1β rs1143643 (G > A) 0.45 0.44 Cases 61 99 40 0.97 0.8230 rs16944 (A/G) 0.48 0.49 Controls 62 100 38 0.96 (0.530– IL-4 rs2243250 (T/C) 0.30 0.30 Cases 53 104 43 0.75 0.7141 rs2243290 (A/C) 0.29 0.29 Controls 51 104 45 0.60 (0.512– IL- rs2275913 (C/T) 0.49 048 Cases 95 90 15 0.97 0.6071 17A rs3819025 (G/A) 0.25 0.23 Controls 92 96 12 0.01 (0.550– Cases 99 85 16 0.7701 Controls 95 93 12 (0.614– Cases 50 105 45 0.033 Controls 51 106 43 2.424 Cases 123 54 23 (1.072– Controls 119 72 9 0.4971 (0.666– TGF-β rs1800469 (A/G) 0.39 0.38 Cases 77 91 32 0.37 0.3270 rs3024490 (A/C) 0.36 0.34 Controls 73 103 24 0.54 (0.365– IL-10 rs3024490 (A/C) 0.36 0.34 Cases 81 93 26 0.54 0.3270 IL-6 rs1800796 (C/G) 0.27 0.24 Controls 84 97 19 0.61 (0.365– rs2066992 (T/G) 0.30 0.26 Cases 81 93 26 0.58 0.6191 TNF-α rs1800629 (G/A) 0.1 0.09 Controls 84 97 19 0.07 (0.496– Cases 102 88 10 0.3851 Controls 112 79 9 (0.647– Cases 98 86 16 0.0687 Controls 108 79 13 (0.867– Cases 167 26 7 Controls 166 33 1 rs1799724 (C/T) 0.08 0.04 Cases 171 28 1 0.11 0.9461 Controls 184 15 1 (0.068– IFN-g rs2069727 (A/G) 0.25 0.30 Cases 119 61 20 0.37 0.4580 rs2430561 (T/A) 0.25 0.30 Controls 105 72 23 0.37 (0.406– Cases 119 61 20 0.4580 Controls 105 72 23 (0.406– A/A homo major allele, A/B heterozygote allele, B/B homo minor allele. OR odds ratio, CI confidence interval additive, dominant, and recessive genetic models using logistic regression with (Pb)/without (Pa) adjustment fo m2), systolic blood pressure (mmHg), diastolic blood pressure (mmHg). The OR and 95% CI of having the risk a A/B heterozygote allele are shown in the recessive model.

W. Tangjittipokin et al. tive model Dominant model Recessive model (95 % CI) Pb OR (95 % CI) Pa OR (95 % CI) Pb OR (95 % Pa OR (95 % CI) Pb OR (95 % CI) CI) 1.033 0.6570.864 0.9601.011 0.8601.044 0.9020.969 0.4521.243 –1.832) (0.452–1.650) (0.660–1.549) (0.647–1.686) (0.588–1.598) (0.705–2.193) 0.9681.013 0.7920.941 0.9490.984 0.9151.026 0.9060.969 0.937 (0.542–1.893) (0.601–1.475) (0.600–1.614) (0.639–1.649) (0.574–1.636) –1.656) 0.4961.374 0.8140.954 0.9271.021 0.6370.825 0.4660.720 (0.550–3.434) (0.642–1.416) (0.655–1.593) (0.372–1.832) (0.297–1.743) 1.167 0.4871.374 0.7310.933 0.9711.008 0.5060.766 0.4450.714 –2.658) (0.560–3.369) (0.629–1.385) (0.646–1.573) (0.349–1.681) (0.300–1.696) 0.5991.186 0.9321.020 0.9160.973 0.7070.913 0.3540.780 1.237 (0.629–2.235) (0.648–1.605) (0.586–1.615) (0.568–1.467) (0.462–1.318) –2.782) 0.021 0.6790.918 0.5150.861 0.015 0.008 2.867 (1.171–7.017) (0.613–1.375) (0.549–1.351) 0.371 0.302 1.089 (0.166–0.827) (0.125–0.727) –1.932) 0.8981.046 0.6120.900 0.4020.823 0.2650.721 0.5460.822 (0.522–2.098) (0.599–1.352) (0.521–1.299) (0.406–1.280) (0.434–1.555) –5.481) 0.2830.662 0.2960.714 0.2630.666 0.7641.063 0.7011.092 1.241 (0.312–1.406) (0.379–1.343) (0.327–1.356) (0.712–1.587) (0.697–1.713) 0.2830.662 0.2960.714 0.2630.666 0.7641.063 0.7011.092 –2.311) (0.312–1.406) (0.379–1.343) (0.327–1.356) (0.712–1.587) (0.697–1.713) 0.715 0.1882.014 0.2531.260 0.2971.267 0.7710.872 0.2330.536 (0.711–5.708) (0.848–1.872) (0.812–1.976) (0.346–2.194) (0.193–1.492) –1.398) 0.2931.587 0.2501.261 0.5961.127 0.5170.778 0.3050.645 0.715 (0.671–3.754) (0.850–1.873) (0.724–1.754) (0.364–1.663) (0.279–1.491) 0.025 0.8750.958 0.5481.198 0.0620.135 0.024 –1.398) 12.163 (0.565–1.627) (0.664–2.161) (0.016–1.107) 0.081 1.270 (1.368–108.153) (0.009–0.719) 0.8311.363 0.047 0.1151.773 0.9860.975 0.8670.784 –3.252) (0.079–23.383) 1.928 (0.870–3.611) (0.061–15.696) (0.045–13.533) 1.414 (1.008–3.690) 0.6730.853 0.1330.737 0.3380.805 0.6921.137 0.8461.074 –3.092) (0.407–1.787) (0.494–1.098) (0.516–1.255) (0.603–2.145) (0.524–2.202) 7.129 0.6730.853 0.1330.737 0.3380.805 0.6921.137 0.8461.074 (0.407–1.787) (0.494–1.098) (0.516–1.255) (0.603–2.145) (0.524–2.202) –58.589) 1.102 –17.755) 0.780 –1.502) 0.780 –1.502) l, NA not applicable (analytical model not applicable due to low frequency). P-values were calculated for the or age, gestational age at blood collection (weeks), BMI before pregnancy (kg/m2), BMI during pregnancy (kg/ allele are shown in the additive and dominant model. The OR and 95% CI of having the A/A homo major allele,

W. Tangjittipokin et al. Table 4 Association between IL and 17A (rs3819025) and TNF-α (rs1800629) genotypes and clinical characteristics among patients with GDM. Clinical characteristics IL-17A (rs3819025) genotypes TNF-α (rs1800629) genotypes G/G G/A A/A P G/G G/A A/A P Number of patients 123 54 23 – 167 26 7 – Gestational age at blood 18.02 ± 10.185 20 ± 10.7 14.87 ± 7.59 0.730 18.50 ± 10.16 17.85 ± 9.89 12.29 ± 10.03 0.361 0.063 collection (weeks) 61.94 ± 13.69 59.01 ± 11.51 63.86 ± 17.21 0.843 61.66 ± 13.82 60.70 ± 13.78 57.01 ± 7.02 0.097 Weight before pregnancy (kg) 65.79 ± 13.95 63.54 ± 12.72 66.96 ± 16.20 0.510 65.84 ± 14.20 63.26 ± 12.73 60.16 ± 8.41 0.965 Weight during pregnancy (kg) 24.42 ± 4.90 23.89 ± 4.61 24.86 ± 5.83 0.918 24.53 ± 5 23.72 ± 4.80 21.80 ± 2.35 0.616 BMI before pregnancy (kg/m2) 25.94 ± 5.04 25.72 ± 4.91 26.04 ± 5.54 0.987 26.17 ± 5.12 24.86 ± 4.70 22.98 ± 2.59 0.397 BMI during pregnancy (kg/m2) 37.90 ± 1.39 37.74 ± 2.13 37.59 ± 1.84 0.486 37.76 ± 1.73 38.18 ± 1.44 37.80 ± 0.84 0.913 Gestational age at delivery 0.517 3128.48 ± 500.18 3018.70 ± 3346.47 ± 0.558 3132.07 ± 3050 ± 571.08 3108 ± 686.13 0.175 (weeks) 583.57 593.71 532.33 0.524 Newborn body mass (g) 115.76 ± 13.73 117.64 ± 12.69 0.680 117.12 ± 13.19 114.44 ± 12.68 112.29 ± 9.71 71.74 ± 11.16 74.96 ± 10.10 0.635 73.50 ± 9.98 70 ± 10.13 69 ± 6.90 0.641 Systolic blood pressure (mmHg) 116.79 ± 12.85 0.547 Diastolic blood pressure 73.03 ± 9.38 184.50 (160, 177 (160, 190) 0.088 188 (165, 202) 197.5 (173.8, 217 (163, 236) 0.080 201) 209) 0.936 (mmHg) 198 (170, 212) 45,947 (25547, 0.429 38,037 (25944.8, 36,506 (26570, 0.405 50-gram glucose challenge test 34,662 (23351, 70329) 0.327 69668) 35,657 (18137, 158033) 37422.5 (26118.8, 59432) 4525 (3517.5, 0.038 4427.5 (3469.3, 62449) 5071 (3687, 0.938 (GCT) (md/dl) 69785.8) 4418 (3394, 4977.5) 0.407 5233.3) 4511 (3304.5, 5443) 0.251 Adipokines levels (pg/ml) 4474 (3671.3, 5238) 726.8 (539.4, 0.004 946.3 (634.18, 5034.5) 584.5 (501.1, 0.873 Adiponectin 5233.3) 869.2 (618.7, 1150.5) 1334) 1060.7 (518.9, 805.4) 0.768 1008.5 (585.7, 1246) 137.8 (88.9, 136.5 (99.9, 1638.5) 140.6 (105.7, 0.765 Adipsin 1512.3) 141.2 (99.3, 206.2) 195.4) 122.4 (97.2, 305) 0.783 133.4 (101.1, 201.2) 157.8 (127.3, 190.8 (149.1, 179.1) 186.3 (175.6, 0.775 Lipocalin-2 183.2) 167.9 (139.8, 193.1) 255.7) 172.7 (146.4, 230.4) 201.5 (166.4, 253.1) 247.7) Resistin 261.7) 1.64 (1.11, 0.829 1.49 (1.11, 1.86) – 1.33 (0.96, 1.92) 1.11 (1.00, PAI-1 1.49 (1.11, 2.20) 2.81) 2.59 (0.76, 2.77) – 3.54 (1.07, 5.50) 6.41 (0.54, Adipocytokines levels (pg/ml) 11.7) 1.42 (0.86, 38.34) – IL-1β 1.07 (0.55, 1.85) 1 (0.80, 1.67) 2.91) 1.39 (0.58, – IFNg 2.50 (0.54, 8.68) 1.96 (0.58, 3.99) 0.346 2.59 (0.68, 7.40) 1.27 (0.58, 5.78) 55.21 (20.78, 14.40) – IL-10 1.49 (0.76, 1.99) 21.67 (15.22, 92.52) 0.951 1.27 (0.86, 1.85) 75.05 (33.89, 54.9800) 3.21 (3.21, 137.50) – IL-17A 0.75 (0.58, 3.89) 3.21 (3.21, 6.35) 0.778 1.39 (0.58, 4.35) 3.21 (3.21, 17.74) 11.29 (6.33, 35.88) – IL-4 75.05 (16.29, 137) 10.62 (8.99, 14.23) 0.597 46.80 (16.54, 9.48 (5.57, 24.74) 0.932 98.87) 22.12) IL-6 3.21 (3.21, 6.11) 3.21 (3.21, 6.49) TNF-α 9.42 (6.61, 14.07) 0.848 10.62 (6.55, 14.77) Data presented as mean ± standard deviation or median and interquartile range. P-values were calculated using linear regression after adjusting for age, gestational age at blood collection (weeks), BMI before pregnancy (kg/m2), BMI during pregnancy (kg/m2), systolic blood pressure (mmHg), diastolic blood pressure (mmHg), as appropriate. 3.5. Association between IL and 17A (rs3819025) and TNF-α gene play important roles in the pathogenesis of obesity-associated diseases (rs1800629) and serum adipokine/adipocytokine levels in GDM patients (Tilg and Moschen, 2006). These biomarkers might be predictors of GDM development, so we set forth to investigate the genetic poly­ The two SNPs that were found to be significantly associated with morphisms and levels of adipokines and adipocytokines that influence GDM patients were then investigated for their relationship with clinical the risk of developing GDM in Thai pregnant women. parameters, including gestational age at blood collection, body weight, BMI, gestational age at delivery, newborn body mass, systolic blood Adiponectin is a protein that is abundantly produced by adipose pressure, diastolic blood pressure, fifty gram glucose challenge test tissue that increases insulin sensitivity, that exerts anti-inflammation level, adipokine levels (adiponectin, adipsin, lipocalin-2, resistin, and effect, and that decrease plasma glucose levels (Greenhill, 2017). A PAI-1), adipocytokines levels (IL-1β, IFN-g, IL-10, IL-17A, IL-4, IL-6, and previous study reported that the deletion of the adiponectin gene in TNF-α) were evaluated using linear regression analysis with adjustment transgenic mice resulted in impaired insulin tolerance (Qiao et al., for age, gestational age at blood collection, BMI before pregnancy, BMI 2017). ADIPOQ gene polymorphisms (rs17846866 and rs1501299) were during pregnancy, systolic blood pressure, diastolic blood pressure. associated in the Saudi population with T2DM patients (Al-Nbaheen, Among GDM patients who carry the IL-17A (rs3819025) genotype, we 2022). Hypoadiponectinemia increased the risk of developing GDM by found a significant decrease in lipocalin-2 and PAI-1 concentrations in more than four times compared to normal pregnancy, so the role of the homo minor allele of this variant (p = 0.038 and p = 0.004, adiponectin in maternal adaptation to pregnancy is crucial (Williams respectively) (Table 4). et al., 2004). In the present study, the maternal serum adiponectin level was significantly decreased in GDM patients compared to controls, and 4. Discussion this is similar to the results of studies conducted in different populations, including South India (Bhograj et al., 2016) and Iran (Mohammadi and Adipokines and adipocytokines are inflammatory mediators that are Paknahad, 2017). We also investigated to see if the observed decreased linked to the endocrine and immune systems, and these molecules are serum adiponectin levels in GDM correlate with various clinical char­ acteristics, including gestational age at blood collection, weight before

W. Tangjittipokin et al. Gene 860 (2023) 147228 Fig. 4. A pictorial representation of the proposed combined mechanism of effect of adipokine/adipocytokine levels, adipokine/adipocytokine genes, and clinical characteristics on the development of gestational diabetes mellitus in Thai women. (Created with Biorender.com). pregnancy, weight during pregnancy, BMI before pregnancy, BMI dur­ with and without diabetes mellitus, or in the hypertension subgroups ing pregnancy, newborn body mass, and serum TNF-α. Maternal serum (Asadikaram et al., 2019). Another study found no significant difference adiponectin levels at 24–28 and 32–35 weeks were reported to be in serum IL-4 levels between GDM and NGT controls (At`egbo et al., negatively associated with pre-pregnancy BMI (Luo et al., 2013). 2006), and IL-4 polymorphisms conferred neither protection against nor Moreover, adiponectin levels in the mother were found to be positively risk for GDM. Another group reported that homozygous wild types (IL-4, correlated with the adiponectin levels in fetal circulation. Infant birth rs2243250, CC) may exert protective effect against type 2 diabetes weight and weight increase during the first year of life are both deter­ mellitus (T2DM), whereas heterozygous genotypes (IL-4, rs2243250, mined by the adiponectin concentration in cord blood. Maternal adi­ CT) may be risk factors for T2DM (Alsaid et al., 2013). ponectin levels during pregnancy may indicate postnatal height growth (Zhang et al., 2016). Low expression of adiponectin (ADP) and high Interleukin-17 (IL-17) is a proinflammatory cytokine that triggers expression of TNF-α were reported to be associated with age, pregesta­ inflammation and the development of obesity-related inflammatory tional BMI, and gestational week (Wei and Zhang, 2020). diseases. The results of an animal model study showed that IL-17A to be highly expressed in adipose tissue, and that it activated the expression of In contrast, interleukin-4 (IL-4) levels were found in GDM patients other proinflammatory cytokine, adipokine, and chemokine genes that compared to controls. Previous study did not find any significant dif­ play a role in insulin resistance (Qu et al., 2016). The results of our study ferences in IL-4 levels in overweight patients compared between those showed independent association between IL and 17A (rs3819025, A/A) 10

W. Tangjittipokin et al. Data availability and GDM. IL-17A rs3819025 is located in the intron region and the Data will be made available on request. downstream pathway of the IL-17A gene, which induces the production of inflammatory molecules and protein remodeling as a response to Acknowledgment innate immunity. Previous study also reported association between IL and 17A rs3819025 and childhood Henoch-Schonlein Purpura (HSP) The authors gratefully acknowledge the women that graciously (Xu et al., 2016). Our study also showed a significant correlation be­ agreed to participate in this study. tween IL and 17A (rs3819025, A/A), and both lipocalin-2 and PAI-1 levels. More specifically, we found the homo minor allele of IL-17A Appendix A. Supplementary data rs3819025 to be associated with significantly lower adipsin and PAI-1 levels. In the present study, serum IL-17 levels and genotype of IL-17A Supplementary data to this article can be found online at https://doi. (rs3819025) were not significantly correlated. No association between org/10.1016/j.gene.2023.147228. serum IL-17 levels and the genotype of IL-17A (rs3819025) has been previously reported (Keramat et al., 2019). References We also found TNF-α rs1800629 to be independently associated with Achari, A.E., Jain, S.K., 2017. Adiponectin, a Therapeutic Target for Obesity, Diabetes, GDM. Similarly, TNF-α rs1800629 genotyping was conducted in China, and Endothelial Dysfunction. Int. J. Mol. Sci. 18, 1321. and the results showed its association with GDM (Wei et al., 2020). Several studies also reported that TNF-α has the potential to be a pre­ Alharbi, K.K., Alsaikhan, A.S., Alshammary, A.F., Al-Hakeem, M.M., Ali Khan, I., 2022. dictor of adverse outcomes of pregnancy since higher levels of TNF-α Screening of mitochondrial mutations in Saudi women diagnosed with gestational were found in GDM patients in China (Wei and Zhang, 2020) and diabetes mellitus: A non-replicative case-control study. Saudi J. Biol. Sci. 29, Malaysia (Salmi et al., 2012). However, the other polymorphisms that 360–365. we studied did not show a significant association between adipokine/ adipocytokine levels and TNF-α rs1800629 in GDM patients. Previous Ali Khan, I., 2021. Do second generation sequencing techniques identify documented study showed that A allele of TNF-α rs1800629 was associated with the genetic markers for neonatal diabetes mellitus? Heliyon 7, e07903. increased of TNF-α levels in asthmatic patients (Yang et al., 2015). Furthermore, there were no differences in the plasma IL-6 and TNF‑α Al-Nbaheen, M.S., 2022. Effect of Genetic Variations in the ADIPOQ Gene on levels with the genotypes of IL‑6‑174G/C (rs1800795) and Susceptibility to Type 2 Diabetes Mellitus. Diabetes, Metabol. Syndrome Obesity: TNF‑α‑308G/A (rs1800629) in Thai ischemic stroke patients (Kamdee Targets Therapy 15, 2753–2761. et al., 2021). Alsaid, A., El-Missiry, M., Hatata, E.-S., Tarabay, M., Settin, A., 2013. Association of IL-4- To the best of our knowledge, this study is the first to examine the 590 C>T and IL-13-1112 C>T gene polymorphisms with the susceptibility to type 2 genotype of all selected adipokines and adipocytokines as the signature diabetes mellitus. Dis. Markers 35, 243–247. for each T cell subset in Thai GDM women. However, a larger sample size may be needed to detect the impact of adipokine/adipocytokine American Diabetes Association, 2019. 2. Classification and Diagnosis of Diabetes: variants and levels on GDM risk in Thai women. Standards of Medical Care in Diabetes—2020. Diabetes Care 43, S14–S31. 5. Conclusion Asadikaram, G., Ram, M., Izadi, A., Sheikh Fathollahi, M., Nematollahi, M.H., Najafipour, H., et al., 2019. The study of the serum level of IL-4, TGF-β, IFN-γ, and IL-17A and TNF-α gene polymorphisms and adiponectin and IL-4 IL-6 in overweight patients with and without diabetes mellitus and hypertension. levels were found to be significantly associated with GDM in pregnant J. Cell. Biochem. 120, 4147–4157. Thai women. The results of this study highlight the need for genetic testing to predict or prevent GDM, and the importance of evaluating At`egbo, J.M., Grissa, O., Yessoufou, A., Hichami, A., Dramane, K.L., Moutairou, K., et al., adipokine and adipocytokine levels in Thai GDM women (Fig. 4). 2006. Modulation of Adipokines and Cytokines in Gestational Diabetes and Macrosomia. J. Clin. Endocrinol. Metab. 91, 4137–4143. Funding Benito, M.J., Gonza´lez-García, M.J., Teso´n, M., García, N., Ferna´ndez, I., Calonge, M., This study was supported by a Graduate Scholarship to BT and FW et al., 2014. Intra- and inter-day variation of cytokines and chemokines in tears of from the Faculty of Medicine Siriraj Hospital, Mahidol University; and healthy subjects. Exp. Eye Res. 120, 43–49. by a Siriraj Research Grant for Research and Development to WT (Grant No. R015934014) from the Faculty of Medicine Siriraj Hospital, Mahidol Bhograj, A., Suryanarayana, K.M., Nayak, A., Murthy, N.S., Dharmalingam, M., Kalra, P., University, Bangkok, Thailand. 2016. Serum adiponectin levels in gestational diabetes mellitus. Indian J. Endocrinol. Metabol. 20, 752–755. Ethics approval. This study involves human participants, and the study protocol was Blüher, M., Mantzoros, C.S., 2015. From leptin to other adipokines in health and disease: approved by the Siriraj Institutional Review Board (SIRB) of the Faculty Facts and expectations at the beginning of the 21st century. Metab. Clin. Exp. 64, of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand 131–145. (COA no. Si577/2015). Buchanan, T.A., Xiang, A.H., 2005. Gestational diabetes mellitus. J. Clin. Invest. 115, Data availability statement 485–491. Data will be made available by the corresponding author upon Chiefari, E., Arcidiacono, B., Foti, D., Brunetti, A., 2017. Gestational diabetes mellitus: an reasonable request from the journal editor. updated overview. J. Endocrinol. Invest. 40, 899–909. Declaration of Competing Interest Cnop, M., Welsh, N., Jonas, J.-C., Jo¨rns, A., Lenzen, S., Eizirik, D.L., 2005. Mechanisms of Pancreatic β-Cell Death in Type 1 and Type 2 Diabetes: Many Differences, Few The authors declare that they have no known competing financial Similarities. Diabetes 54, S97–S107. interests or personal relationships that could have appeared to influence the work reported in this paper. Dhastagir, S., 2017. Adiponectin, Diabetes and Coronary Artery Disease. J. Endocrinol. Thyroid Res. 1, 555553. Fain, J.N., Madan, A.K., Hiler, M.L., Cheema, P., Bahouth, S.W., 2004. Comparison of the Release of Adipokines by Adipose Tissue, Adipose Tissue Matrix, and Adipocytes from Visceral and Subcutaneous Abdominal Adipose Tissues of Obese Humans. Endocrinology 145, 2273–2282. Greenhill, C., 2017. The role of adiponectin in gestational diabetes mellitus. Nat. Rev. Endocrinol. 13, 128. HAPO Study Cooperative Research Group, M.B., Lowe, L.P., Dyer, A.R., Trimble, E.R., Chaovarindr, U., Coustan, D.R., Hadden, D.R., McCance, D.R., Hod, M., McIntyre, H. D., Oats, J.J., Persson, B., Rogers, M.S., Sacks, D.A., 2008. Hyperglycemia and Adverse Pregnancy Outcomes. New Engl. J. Med., 358, 1991–2002. Hung, A.M., Sundell, M.B., Plotnikova, N.E., Bian, A., Shintani, A., Ellis, C.D., et al., 2013. A pilot study of active vitamin D administration and insulin resistance in African American patients undergoing chronic hemodialysis. J. Ren. Nutr. 23, 185–193. Kamdee, K., Panadsako, N., Mueangson, O., Nuinoon, M., Janwan, P., Poonsawat, W., et al., 2021. Promoter polymorphism of TNF-α (rs1800629) is associated with ischemic stroke susceptibility in a southern Thai population. Biomed. Rep. 15, 78. Keramat, F., Kazemi, S., Saidijam, M., Zamani, A., Kohan, H.F., Mamani, M., et al., 2019. Association of interleukin-17 gene polymorphisms and susceptibility to brucellosis in Hamadan, western Iran. Microbiol. Immunol. 63, 139–146. Khan, I.A., Jahan, P., Hasan, Q., Rao, P., 2019. Genetic confirmation of T2DM meta- analysis variants studied in gestational diabetes mellitus in an Indian population. Diabetes Metab. Syndr. 13, 688–694.

W. Tangjittipokin et al. and 2045: Results from the International Diabetes Federation Diabetes Atlas, 9(th) edition. Diabetes Res. Clin. Pract. 157, 107843. Luo, Z.-C., Nuyt, A.-M., Delvin, E., Fraser, W.D., Julien, P., Audibert, F., et al., 2013. Salmi, A., Zaki, M.N., Zakaria, Nor Aliza, G. and Rasool, H.G., 2012. Arterial stiffness, Maternal and fetal leptin, adiponectin levels and associations with fetal insulin inflammatory and pro-atherogenic markers in gestational diabetes mellitus. Vasa 41, sensitivity. Obesity 21, 210–216. 96–104. Tilg, H., Moschen, A.R., 2006. Adipocytokines: mediators linking adipose tissue, Mohammadi, T., Paknahad, Z., 2017. Adiponectin Concentration in Gestational Diabetic inflammation and immunity. Nat. Rev. Immunol. 6, 772–783. Women: a Case-Control Study. Clin. Nutr. Res. 6, 267–276. Wei, Q., Chen, X., Chen, H., 2020. Association of Single Nucleotide Polymorphisms of the IL-6, IL-10, and TNF-α Genes with Susceptibility to Gestational Diabetes Mellitus. Mohammed Suhail, A., Biswas, A., Ranjan, R., Meena, A., Birendra Kumar, Y., Genet. Test Mol. Biomarkers 24, 390–398. Sharma, A., et al., 2009. Plasminogen Activator Inhibitor-1 (PAI-1) Gene 4G/5G Wei, W., Zhang, X., 2020. Expression of ADP and TNF-α in patients with gestational Promoter Polymorphism is Seen in Higher Frequency in the Indian Patients With diabetes mellitus and its relationship with pregnancy outcomes. Exp. Ther. Med. 20, Deep Vein Thrombosis. Clin. Appl. Thrombosis/Hemostasis 16, 184–188. 2184–2190. Williams, M.A., Qiu, C., Muy-Rivera, M., Vadachkoria, S., Song, T., Luthy, D.A., 2004. Mpondo, B.C.T., Ernest, A., Dee, H.E., 2015. Gestational diabetes mellitus: challenges in Plasma adiponectin concentrations in early pregnancy and subsequent risk of diagnosis and management. J. Diabetes Metab. Disord. 14, 42. gestational diabetes mellitus. J. Clin. Endocrinol. Metab. 89, 2306–2311. Xu, H., Pan, Y., Li, W., Fu, H., Zhang, J., Shen, H., et al., 2016. Association between Phattanachindakun, B., Watananirun, K., Boriboonhirunsarn, D., 2022. Early universal IL17A and IL17F polymorphisms and risk of Henoch-Schonlein purpura in Chinese screening of gestational diabetes in a university hospital in Thailand. J. Obstet. children. Rheumatol. Int. 36, 829–835. Gynaecol. 1–7. Yang, Y.H., Liu, Y.Q., Zhang, L., Li, H., Li, X.B., Ouyang, Q., et al., 2015. Genetic polymorphisms of the TNF-α-308G/A are associated with metabolic syndrome in Plows, J.F., Stanley, J.L., Baker, P.N., Reynolds, C.M., Vickers, M.H., 2018. The asthmatic patients from Hebei province, China. Int. J. Clin. Exp. Pathol. 8, Pathophysiology of Gestational Diabetes Mellitus. Int. J. Mol. Sci. 19, 3342. 13739–13746. Zhang, Z.-Q., Lu, Q.-G., Huang, J., Jiao, C.-Y., Huang, S.-M., Mao, L.-M., 2016. Maternal Qiao, L., Wattez, J.-S., Lee, S., Nguyen, A., Schaack, J., Hay Jr, W.W., et al., 2017. and cord blood adiponectin levels in relation to post-natal body size in infants in the Adiponectin Deficiency Impairs Maternal Metabolic Adaptation to Pregnancy in first year of life: a prospective study. BMC Pregnancy Childbirth 16, 189. Mice. Diabetes 66, 1126–1135. Qu, Y., Zhang, Q., Ma, S., Liu, S., Chen, Z., Mo, Z., et al., 2016. Interleukin-17A Differentially Induces Inflammatory and Metabolic Gene Expression in the Adipose Tissues of Lean and Obese Mice. Int. J. Mol. Sci. 17, 522. Ramachandrayya, S.A., D’Cunha, P., Rebeiro, C., 2020. Maternal circulating levels of Adipocytokines and insulin resistance as predictors of gestational diabetes mellitus: preliminary findings of a longitudinal descriptive study. J. Diabetes Metab. Disord. 19, 1447–1452. Saeedi, P., Petersohn, I., Salpea, P., Malanda, B., Karuranga, S., Unwin, N., et al., 2019. Global and regional diabetes prevalence estimates for 2019 and projections for 2030