Shabnam Montazeri, Faculty of Medicine and Health, International Medical University (IMU), 126, Jalan 19/155B, Bukit Jalil, 57000 Kuala Lumpur, Malaysia. E-mail: firstname.lastname@example.org
Gestational diabetes mellitus (GDM), defined as carbohydrate intolerance diagnosed for the first time during pregnancy, affects both maternal and fetal health. Possession of a specific genetic polymorphism can be a predisposing factor for susceptibility to some diseases. The aim of this study was to investigate the association between single nucleotide polymorphisms (SNP) in the promoter gene of interleukin-10 (IL-10) as well as tumor necrosis factor-alpha (TNF α) with the development of GDM.
Two hundred and twelve consecutive series of eligible normal pregnant women (controls) and gestational diabetes mellitus women were selected based on the study's inclusion and exclusion criteria.
DNA was extracted from blood and genotyped for IL-10 at three positions and TNF α for gene polymorphism using the polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) method. Plasma levels of IL-10 and TNF α at different gestational periods as well as postpartum were quantified using enzyme linked immunosorbent assay (ELISA).
The results of the study showed that the difference in the frequency of SNP at position −597 in the promoter of the human IL-10 gene between the control and GDM groups was statistically significant (p < 0.05). In contrast, there was no significant difference in the frequency of SNP at the other two sites in the promoter region of the human IL-10 gene (−824 and −1082) as well as position −308 in the promoter of the human TNF-α (p > 0.05). In addition, there was no significant difference between the two groups in terms of plasma levels of IL-10 as well as TNF α in different stages of pregnancy.
SNP at position −597 was significantly associated with the development of GDM and shows potential for use as a predictive marker for GDM.
Gestational diabetes mellitus (GDM) is defined as carbohydrate intolerance of varying severity with onset or first recognition during pregnancy (Metzger 1991). It is associated with maternal and fetal complications. In most cases, the diagnosis of GDM is made in late second or early third trimester of pregnancy using conventional screening tests (Luesley and Baker 2004). However, there is neither worldwide agreement on the standard cut-off point in blood sugars nor a standard glucose tolerance test.
Although a significant amount of information regarding the pathogenesis of GDM is available, it is still not known why GDM develops in some pregnant women. Additionally, the extent of occurrence of GDM associated complications in both mother and fetus is still high regardless of the available methods of management and control of GDM.
The presence of a specific genetic polymorphism seems to predispose individuals to some diseases and hence the detection of such genetic polymorphism might be useful as a marker enabling early diagnosis. Polymorphism of a single nucleotide can be detected by restriction digestion of the PCR product with digestive enzymes.
Tumor necrosis factor-alpha (TNF α), on the other hand, is a pro-inflammatory cytokine. It is one of the key cytokines that have been implicated in mediating insulin resistance (Dandona et al. 1998). Some studies have shown that single nucleotide polymorphisms (SNP) in the TNF α promoter gene can regulate plasma levels of TNF α levels and the action of insulin (Fernandez-Real et al. 1997; Dalziel et al. 2002). However, these findings are not supported by other studies (Koch et al. 2000; Rusmusse et al. 2000).
These data have not been reflected in GDM. Considering the important role of these cytokines (i.e. IL-10 and TNF α) in pregnancy, presence of polymorphisms in their genes is potentially important. The aim of this study was to investigate the association between SNP in the human promoter region of the IL-10 gene and TNF α gene with the development of GDM. Since these SNPs are in the regulatory region, they could affect the transcription of the gene which in turn may affect cytokine secretion. If a strong relationship exists between these markers and pregnancy, the establishment of such a relationship, could lead to their use as a predictor of GDM.
MATERIAL AND METHODS
This is a case-control prospective study carried out in women attending the Obstetric Care Centre at the Department of Obstetrics and Gynecology, Hospital Tuanku Jaafar (HTJ), Seremban, Malaysia. Both GDM and normal subjects were recruited randomly based on consent to participate in the study and fulfilling the specified inclusion and exclusion criteria. Inclusion criteria for normal subjects included singleton pregnancy, aged 18–35 years old, parity no more than five, no known risk factor for GDM (such as diabetes in first degree relative, previous history of GDM, and history of macrosomia or stillbirth), no glucose intolerance in pregnancy following a 50 g glucose challenge screening test (GCT) and the absence of any other medical disorders.
Singleton pregnant women, aged 18–35 years old and parity no more than five, who exhibited glucose intolerance and/or had risk factors for GDM were screened with 75 g glucose tolerance test (MGTT). All those who were found to have glucose intolerance according to the WHO criteria (WHO 1999) were recruited as subjects for the GDM group. All subjects attended regular antenatal care until delivery at the Obstetric Care Centre. In addition, GDM mothers were appropriately monitored and managed to regulate the blood sugar levels. One hundred and two control women and 110 GDM women completed the study. Serum HbAlc was measured at 36 weeks of pregnancy for all GDM subjects as an indication of diabetic (glycemic) control. This test was performed at an accredited pathology laboratory (PathLab (M) Sdn. Bhd.).
Informed consent was obtained after the purpose of the study was explained to all the participating women by the attending obstetrician.
Collection of blood samples
Five ml of venous blood was drawn from all subjects in both groups three times during pregnancy (i.e. second trimester, 32 and 36 weeks) and once after delivery (six weeks post partum) in the course of routine investigations done at medical institute. Blood for the study was collected in heparinized tubes. Plasma and blood cells were separated by centrifugation (1500 rpm, 10 min, 4°C) and stored at 20°C until analysis.
Detection of SNP using PCR and RFLP
Genomic DNA was extracted from peripheral blood leukocytes using a genomic isolation DNA commercial kit as recommended by the manufacturer (QIAGEN, Germany). The extracted DNA was amplified by PCR using published forward and reverse primers (Padyukov et al. 2004; Hee et al. 2007) (LNTRON Biotechnology, Korea). PCR was performed using commercially available PCR premix according to the manufacturer's recommendation protocol (DMTRON Biotechnology, Korea) and a PTC-100 Pltier thermal cycle (MJ Research). Cycling condition for PCR was: 30 cycles of 95°C for 30 s, 45°C for 30 s and 72°C for 1 min.
The PCR product was purified using a PCR purification kit as described by the manufacturer (LNTRON Biotechnology, Korea). The purified PCR product was incubated in a thermal bath (DUAL Thermal Bath ALB 128, Korea) at 37°C overnight (for EcoNI or RsaI) and at 55°C overnight (for MaeIII). For RFLP, the PCR digested products were separated by electrophoresis and viewed using a gel documentation system (BioDoc-It Gel documentation System) (UVP, USA). The promoter region of the human IL-10 gene containing the three SNPs, i.e. −1082 G>A, −824 Orand −597 C>A was successfully genotyped for all subjects.
Tumor necrosis factor-alpha (TNF α)
The extracted DNA was amplified by PCR using published primers (Um and Km 2003, Casano-Sancho et al. 2006, Stanulla et al. 2001)(INTRON Biotechnology, Korea). The PCR consisted of 36 cycles, which included the following conditions: one cycle of 94°C for 3 min, 35 cycles of 94°C for 1 min, 60°C for 1 min and 72°C for 1 min. The 107 bp PCR product was purified using a PCR purification kit as described by the manufacturer (INTRON Biotechnology, Korea). The purified product was digested with NcoI restriction enzyme (Vivantis, Malaysia) at 37°C for 2 h. The digested product was heat-inactivated at 65°C for 20 min. For RFLP, the PCR digested products were separated by electrophoresis and viewed using a gel documentation system (BioDoc-It Gel documentation System) (UVP, USA). SNP at position −308 G>A was successfully genotyped for all subjects (Table 1).
Table 1. Technical data for the genotyping of single nucleotide polymorphism in the human IL-10 and TNF α.
Restriction pattern (bp)
Forward: 5′ 3′ AAG ACA ACACTACTA AGGCTT CCTT
G:306 + 278
Reverse:5′ 3′ TAA ATA TCC TCA AAG TTCC
A: 306 + 252 + 26
Forward: 5′ 3′ ATCCAAGAC AAC ACT ACT AA
C: 292 + 217 + 79
Reverse:5′ 3′ TAA ATA TCC TCA AAG TTC C
T: 509 + 79
Forward: 5′ 3′ ATCCAAGACAAC ACT ACTAA
C: 306 + 232 + 42
Reverse:5′ 3′ TAA ATA TCC TCA AAG TTC C
A: 240 + 232 + 66 + 42
Forward: 5′ 3′ AGG CAA TAG GTT TTG AGG GCC AT
G:87 + 20
Reverse:5′ 3′ TCC TCC CTG CTC CGA TTC CG
Quantification of plasma levels of IL-10 and TNF-α by ELISA
Plasma levels of IL-10 and TNF α were determined at different stages of pregnancy i.e. second trimester, 32 and 36 weeks and after delivery (six weeks post partum) for both control and GDM subjects using ELISA according to the manufacturer recommended protocol (eBioscience company, USA); The concentration of IL-10 in the plasma of each subject was calculated based on a standard curve.
SPSS (ver. 16) was used for statistical analysis, p < 0.05 was considered statistically significant. The Hardy–Weinberg equilibrium test was determined by using a χ2-test based on expected frequencies. Haplotype frequencies in both groups were calculated using EH (estimating haplotype). The χ2-test was used to compare genotype and allele frequencies. Generalized estimating equitation (GEE) was used to compare IL-10 levels at different stages of pregnancy and in two groups. Logistic regression with odds ratio and 95% confidence interval (CI) were used to determine the risk and predictive value of the SNP in the promoter of the human IL-10 gene as well as TNF α for the development of GDM as well as sensitivity and specificity for GDM prediction.
Demographic data of the GDM and control subjects are shown in Table 2. GDM subjects were older, of higher parity and had higher BMI than the normal controls. Ethnic distribution did not differ in the two study groups (p > 0.05). It was noted that the HbAlc level was less than 6.5% in about 75% of the GDM subjects.
Table 2. Demographic data in GDM and control subjects. *p-value < 0.05 was set as statistically significant.
Control (n = 102) (%)
GDM (n = 110) (%)
p < 0.05*
p < 0.05*
p > 0.05
p < 0.05*
The genotype and allele frequencies of the polymorphisms at three positions in the promoter of the human IL-10 gene and one position in the promoter of the TNF α are shown in Table 3. The difference between control and GDM groups in terms of SNP genotype in site −597 of IL-10 promoter gene was significantly different (p = 0.03) with an odds ratio of 2.2 (95% CI 1.21–4). At this position, GDM subjects displayed significantly higher frequencies of the mutant–type allele whereas the control subjects exhibited more wild-type allele. The p-value was 0.0001 with an odds ratio of 2.17 (95% CI 1.462–3.216).
Table 3. The relationship between genotype and allele frequencies of IL-10 and TNF α SNPs in control and GDM groups. *p-value < 0.05 was set as statistically significant.
Control (n = 102) (%)
GDM (n = 110) (%)
p < 0.05*
p > 0.05
p > 0.05
p > 0.05
Logistic regression with odds ratio and 95% CI were also used to determine the risk and predictive value of SNP in IL-10. Based on the logistic regression test SNP in the promoter of the IL-10 gene at position −597 had significant association with GDM (p = 0.01). Subjects who were carrying mutant-type of IL-10 at position −597 were 2.2 times more likely to have GDM compared to those who did not have. The sensitivity of the test is 60% and the specificity is 57%. This alludes to the polymorphism at position −597 of the human IL-10 gene to be associated with the development/incidence of GDM. However, with respect to the other SNP studied in the human IL-10 gene, when IL-10 promoter gene at positions −824 and −1082 were compared, our results show that there was no significant (p > 0.05) difference in the genotype and allele frequencies between the GDM and control groups.
The difference in the haplotype frequencies amongst the three SNP (−597, −824 and −1082) in the promoter region of the human IL-10 gene was found to be statistically significant (p < 0.05) between control and GDM groups. The CCG haplotype, which consisted of all wild-type alleles, was detected in higher frequency in the control group. In contrast, there was a higher frequency of the ATA haplotype, which consisted of all mutant alleles, in GDM subjects. We also detected two rare haplotype i.e. ATG and ACA, which consisted of a mixture of wild-type and mutant alleles at the three sites studied.
The concentrations of plasma IL-10 at different stages of pregnancy was quantified using ELISA. Figure 1 show the IL-10 levels in both groups. The differences in the plasma IL-10 levels between both study groups as well as in different stages of pregnancy were compared using generalized estimated equitation (GEE) test. There were no significant (p > 0.05) differences observed in plasma levels of IL-10 in terms of different stages of pregnancy including at post partum between control and GDM subjects. In our study, the trend of IL-10 levels tended to increase up to the third trimester and then decrease. The IL-10 levels were lower in GDM subjects than control subjects but did not reach significant levels.
The IL-10 level in each stage of pregnancy was compared with different SNP at three positions in each study group respectively. At p-value more than critical value of 0.05, no significant difference was seen. However, the GDM mothers who carried the homozygous mutant alleles (AA) at position −597 in the promoter region of the human IL-10 gene appeared to have lower levels of plasma IL-10 at 32 and 36 weeks of pregnancy compared to the control group though it did not reach the significant level (Table 4, 5).
Table 4. Relationship between plasma IL-10 levels in different stages of pregnancy and SNPs at three position in the promoter region of the human IL-10 gene in two study groups.
Plasma IL-10 levels (mean rank)
Six weeks post partum
Table 5. Relationship between plasma TNF α levels in different stages of pregnancy and SNP at position −308 in the promoter region of the human TNF α gene in two study groups.
Plasma TNF-α levels (mean rank)
Six weeks post partum
Our results indicate that there is no significant (p > 0.05) difference in the genotype as well as allele frequencies of SNP in the TNF α promoter gene at this position i.e. −308 amongst the GDM and control groups. Comparison between wild-type genotype and combination of hetero and mutant-type (GG vs GA+AA) of SNP in the promoter region of TNF α in both study groups also showed no statistically significant difference (Table 6). The concentrations of plasma TNF-α at different stages of pregnancy were quantified using ELISA. Figure 2 shows the TNF α levels in both groups. The differences in the plasma TNF α levels between both study groups as well as in different stages of pregnancy were compared using generalized estimated equitation (GEE) test. There were no significant (p > 0.05) differences observed in plasma levels of TNF α in terms of different stages of pregnancy including at post partum between control and GDM subjects. Considering the differences in ethnicity and BMI in controls and cases, the multivariate analysis was done to assess the possibility of these variables as confounding factors with three SNPs in the IL-10 promoter gene as well as TNF α individually, no significant difference was seen (data not shown).
Table 6. Comparison between wild-type genotype vs. combination of hetero and mutant-type of SNP in the promoter region of TNF α in both study groups.
Control (n = 102)
GDM (n= 110)
There has been relatively little research in the area of GDM genetics per se. Genetic predisposition to GDM has been suggested since GDM clusters in families. Identification of the underlying genetic causes of GDM will eventually give a better view of the mechanisms that contribute to the pathophysiology of the disease. Furthermore, it may improve options to possibly prevent GDM and complications for the mother and her child (Shaat and Groop 2007).
On the other hand, the relationship between the presence of SNP in the IL-10 as well as TNF α and their plasma levels are of the clinical interest because of the pivotal role of these cytokines in the stimulation or regulation of inflammatory and immune responses (Howard et al. 1992). The aim of this study was to investigate the association between SNP in the human promoter of the IL-10 gene as well as TNF α gene with the development of GDM. One of the major anti-inflammatory cytokines is IL-10. This cytokine has been reported to be able to counteract negative effects of pro-inflammatory cytokines leading to insulin resistance (Choy and Panayi 2001), which is what initiated this study.
The results obtained show that the frequency of SNP in position −597 in the promoter region of the human IL-10 between control and GDM subjects was significantly different. In addition, the difference between control and GDM groups in terms of genotype at this site (−597) was also significantly different. It is interesting to note that at this position, the GDM subjects displayed significantly higher frequencies of the mutant-type allele whereas the majority of the control subjects exhibited the wild-type allele. It can then be concluded that SNP at position −597 in the promoter region of the human IL-10 gene is associated with the development/incidence of GDM. In contrast, there were no significant differences in the other two SNP located in the promoter of the human IL-10 gene at positions −824 and −1082 between GDM and control groups.
Using the haplotype analysis, with the three SNPs located in the promoter of the human IL-10 gene (−507, −824, and −1082), we established three major haplotype i.e. CCG, CCA and ATA (Peng et al. 2006). The ATA haplotype, which consisted of mutant-type alleles in all three SNPs, was detected in higher frequency in the in GDM subjects compared to control subjects (0.035 vs 0.029). To date, in a study involving non-pregnant-women, there has not been reported any significant difference in terms of genotype of SNPs at positions −597 and −824 in the promoter region of the human IL-10 gene between diabetic mellitus (DM) and control subjects. However, there was a significantly higher frequency of the T-allele (−824T) and A-allele (−597A) in patients with type 2 DM (Chang et al. 2005). Likewise, the results of study on type 2 diabetes in Caucasian Italian subjects showed that variation of the IL-10 in the promoter region at positions −597, −824 and −1082 was associated with insulin resistance and obesity (two risk factors for type 2 diabetes) in non-diabetic subjects but not in type 2 diabetes (Scarpelli et al. 2006).
Interleukin-10 plays an important role during pregnancy to prevent any unwanted activation of the mother's immune system that could be detrimental to the fetus (Marzi et al. 1996; Raghupathy et al. 1999; Karthukorpi et al. 2001; Holmea et al. 2003). So, it was reassuring that we observed an upward trend in the levels of plasma IL-10 during pregnancy. The levels of IL-10 increased up to the third trimester, after which this level decreased. This observation is expected as pregnancy is proposed to induce a T-helper-2 (TH2) type of immune response. However, there was no significant difference in the plasma IL-10 levels between the two study groups in the different stages of pregnancy. A previous study conducted to assess the physiological effects of pregnancy on plasma IL-10 concentration, had shown that plasma levels of IL-10 were significantly higher in healthy pregnant women compared to non-pregnant women (Holmea et al. 2003).
In our study, the GDM mothers who carried the homozygous mutant alleles (AA) at position −597 in the promoter region of the human IL-10 gene appeared to have lower levels of plasma IL-10 at 32 and 36 weeks of pregnancy compared to the control group but the difference did not reach the significant level. The possible explanation could be that the position −597 is not the only factor to determine IL-10 secretion. Besides, IL-10 transcriptional regulation is currently not well described. It seems like genetic and environmental factors also have effects on IL-10 secretion (Grondal et al. 1999).
There is evidence to suggest that TNF α can inhibit insulin signaling (Hotamisligil et al. 1993) and therefore impair insulin secretion (Tsiotra et al. 2001). As a result, SNP in the human TNF α may contribute to insulin resistance and thus lead to type 2 diabetes mellitus. We detected the frequencies of SNP at position −308 in the promoter region of the human TNF α gene. In our population, the frequency of this allele amongst the control and GDM subjects were 4.9% and 4.5% respectively, while the genotype frequency amongst the control and GDM subjects were 2.0% and 2.7% respectively. In fact, some investigations have found that the frequency of the mutant allele i.e. the A-allele at position −308 in the promoter of the human TNF-α gene, is very rare in Asians (Dandona et al. 1998) and this, to a large extent, reflects one of the genetic characteristics of Eastern Asians. However, the reason for this rarity is unclear (Um and Kim 2003).
Based on the statistical analysis of this study, we could not detect any significant difference in the genotype and allele frequencies at this SNP between the control and GDM subjects. A previous study has shown that −308 A allele in the promoter of the human TNF α gene was associated with higher risk for type 2 diabetes mellitus (Kubaszek et al. 2003). However the correlation between SNP in TNF α gene promoter and type 2 DM is still controversial, because of discrepancies among different studies (Shiau et al. 2003). While some reported the association of the polymorphism and insulin resistance (Fernandez-Real et al. 1997), no correlation was observed between TNF α−308A genotype with insulin sensitivity or insulin secretion in studies of healthy young relatives of type 2 DM (Koch et al. 2000; Rusmusse et al. 2000) or with prevalence of diabetes (Shiau et al. 2003) or even with obesity in obese populations (Walston et al. 1999; Lee et al. 2000).
In our study, TNF α levels showed a downward trend during pregnancy and increased during late third trimester of pregnancy. This observation is consistent with previous studies showing an increase in plasma TNF α in late pregnancy (Kirwan et al. 2002). The suppression of TNF α production during pregnancy seems to favor the sustenance of a normal pregnancy (Vince and Johnson 1996). There were no significant differences observed in the plasma levels of TNF α at different stages of pregnancy (second trimester, 32 weeks, 36 weeks and six weeks post partum) between control and GDM subjects. In studies on serum TNF α in pregnancy, fasting serum TNF α levels at 39 weeks of pregnancy was found to be higher in GDM cornpared to non-diabetic pregnant mothers (Winkler et al. 2002; Wang et al. 2004). However reports of changes in TNF α levels during normal pregnancy and GDM are equivocal and there are controversies between studies largely related to discrepancies in methodology used to analyze TNF α.
Mclachlan et al. (2006) did not observe an association between TNF-α and insulin sensitivity in either GDM or control subjects and concluded that TNF α does not seem to contribute greatly to insulin resistance. Winkler et al. (2002) proposed that circulating TNF α should not be considered as endocrine cytokine as it may not alter the action of insulin, in fact, it may act in a paracrine or autocrine manner.
The aim of current study was to explore the possible association of both IL-10 and TNF α SNP in GDM in a local population in Malaysia. We could not establish any cause–effect relationship. At the most, this study shows variations at the sites studied in IL-10 and TNF α associated with GDM. SNP at position −597 promoter gene of IL-10 was significantly associated with development of GDM. This finding is interesting and may used as a predictive factor. Although based on our results, the polymorphisms in IL-10 and TNF α were largely not associated with GDM, our data do not exclude the possibility of an association with SNP and GDM.
Whilst exploring the need for a more discriminate tool for the early detection of GDM remains elusive, we have tried to draw some valuable correlated data for these two cytokines which may prove valuable in future studies. To our best knowledge, there have not been any other studies associating SNPs in IL-10 and TNF α to GDM to date.
The authors would like to thank Dr Ravindran Jegasothy, Dr Krishna Kumar, medical officers, nurses and staff of the Department of Obstetrics and Gynecology of Hospital Tuanku Jaafar, Seremban, Malaysia for facilitating the recruitment of patients and use of clinical data. This study was made possible through a Grant from the International Medical University (IMU 125/2006). The study protocol was approved by Research and Ethics Committee of the International Medical University, Malaysia (IMU).