Association Between Single Nucleotide Polymorphisms in CDKAL1 and HHEX and Type 2 Diabetes in Chinese Population

Introduction

Type 2 diabetes mellitus (T2DM) is a complex disease which is associated with micro- and macrovascular complications. It can bring damage to several organs which is characterized by insulin resistance in peripheral tissues and dysregulated insulin secretion by pancreatic β-cell. The annual prevalence of T2DM in 2019 was 463 million adults globally, and is estimated to increase to over 700 million by 2045.¹ In China, the prevalence of diabetes mellitus showed a rapid increasing and reached 10.9% in 2013.² The Chinese population show a lower insulin secretion level compared with the European population, which implies that the insulin secretion function of pancreatic β-cell is a critical factor for the development of T2DM in the Chinese population.³

Cyclin-dependent kinase 5 regulatory subunit associated protein 1-like 1 (CDKAL1) gene encodes cyclin-dependent kinase 5 regulatory subunit-associated protein 1 (CDK5RAP1)-like 1, which is homologous to the CDK5RAP1, inhibitor of the CDK5 kinase through the inhibition of the CDK5 activator p35.^4,5 CDK5 is a serine/threonine protein kinase which is involved in the glucose-dependent regulation of the insulin secretion; therefore, it plays an important role in the pathophysiology of β-cell dysfunction and predisposition to T2DM.^6,7 Hematopoietically expressed homeobox (HHEX) gene encodes a transcription factor which is involved in Wnt signaling pathway, a fundamental pathway required for cell growth and development.⁸ It plays a vital role in the ventral pancreatic specification.⁹ Decreased somatostatin levels in HHEX-deficient islets causes disrupted paracrine inhibition of insulin release from β-cells.¹⁰ According to the functional mechanism and downstream effects of these two genes, it is reasonable to conclude that CDKAL1 and HHEX are well-accepted pathogenesis-related key genes for T2DM.

The associations of single nucleotide polymorphism (SNP) in both CDKAL1 and HHEX with insulin secretion and/or T2DM had been examined in many studies.^11,12 It is noted that not only are there different association results in different ethnic groups, but also there are different association results in the same population.¹³ For example, the associations of SNPs (rs1111875, rs5015480, and rs7923837) in HHEX with T2DM were observed in Chinese population from Shanghai, but not from Beijing.¹³ Thus, it is necessary to carry out additional association studies to reveal the real roles of SNPs in CDKAL1 and HHEX in susceptibility of T2DM. Here, we conducted a case-control study to investigate whether the SNPs located on CDKAL1 (rs4712524, rs10946398, rs7754840, rs9460546, and rs7756992) and HHEX (rs1111875, rs5015480, and rs7923837) were associated with the T2DM in the Chinese Han population.

Subjects and Methods

Ethics Statement

This study was carried out in accordance with the Helsinki Declaration of 1975, which was revised in 2008. The protocol employed by this investigation was approved by the Institutional Review Board of the Second People’s Hospital of Yunnan Province before investigating. Each recruiter provided informed consent before participating in the study.

Subjects

A total of 1,169 T2DM patients and 1,277 healthy volunteers were enrolled in this study. All participants (T2DM and NDM) self-reported to be of Han ethnicity. The 1,169 patients of T2DM (742 males and 427 females) were diagnosed at the Second People’s Hospital of Yunnan Province between January 2018 and December 2019 in accordance with the diagnostic standard of T2DM that World Health Organization (WHO) criteria published in 1999 and American Diabetes Association (ADA) guidelines in 2020.¹⁴ For the T2DM group, the inclusion criteria are: 1) The T2DM subjects should be diagnosed as diabetes mellitus (DM) by fasting plasma glucose and/or 2 hours postprandial plasma glucose; 2) the T2DM subjects should be with fasting plasma glucose ≥7.0 mmol/L and/or 2 hours postprandial plasma glucose ≥11.1 mmol/L two times. The exclusion criteria are: 1) gGestational diabetes mellitus and other specific types of diabetes were excluded by requiring history illness; 2) the fasting and 2 hours postprandial plasma concentrations of insulin and C peptides were detected to exclude subjects with type 1 diabetes; and 3) subjects younger than 40 were excluded if islet cell autoantibodies and the glutamic acid decarboxylase autoantibodies are positive. For the NDM group, the inclusion criteria are: 1) tThe NDM subjects should be measured by fasting plasma glucose and glycosylated hemoglobin (HbA1C); and 2) the NDM subjects should be with fasting plasma glucose <6.10 mmol/L and HbA1C <5.7%. The exclusion criteria are: 1) sSubjects with a family history of diabetes mellitus were excluded; and 2) subjects with hypertension or coronary heart disease were also excluded.

Laboratory Measurements

Venous blood samples were collected in the morning after the subjects had fasted for 12 hours. Fasting plasma glucose (FPG) levels were measured using the glucose oxidase method. The levels of total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), triglycerides (TG), and low-density lipoprotein cholesterol (LDL-C) were measured by enzymatic methods. HbA1C was measured by immunoturbidimetry. All laboratory measurements were performed using a HITACHI 7600–020 Automatic Analyzer.

Selection and Genotyping of SNPs

We primarily focused on the two genes (CDKAL1 and HHEX) which were shown to have a relationship with pathogenesis of T2DM. Five SNPs located in CDKAL1 gene (rs4712524, rs10946398, rs7754840, rs9460546, and rs7756992) and three SNPs (rs1111875, rs5015480, and rs7923837) located in HHEX gene were selected.

Genomic DNA from peripheral blood was extracted using The QIAamp Blood Mini Kit (Qiagen, Hilden, Germany). MassARRAY Analyzer 4.0 (Agena, Inc) was used to genotype the eight SNPs from the two genes. The PCR primers were designed using the AssayDesigner 3.1 (Sequenom lnc., San Diego, CA, USA). Four microliters of the PCR master mix was mixed with 1 μL template DNA (25 ng/μL) in a 384-well plate. The PCR reaction setting was the same as those in our previous study.¹⁵ Then 2 μL of shrimp alkaline phosphatase (SAP) was added into the PCR products per-well for removing the dNTP; the reaction conditions were set at: 37°C for 20 minutes and 85°C for 5 minutes. Next, 2 μL of the EXTEND Mix was added for single base extension using the following PCR cycle conditions: 1) 94°C for 30 seconds, 2) 94°C for 5 seconds, 3) 52°C for 5 seconds, 4) 80°C for 5 seconds, 5) 72–94°C for 3 minutes, and then steps 2–4 were repeated for 40 cycles with five repetitions of steps 3 and 4 per cycle. Resin purification was performed to 9 μL reaction products then the final products were transferred into a 384-well SpectroCHIP bioarray by MassARRAY Nanodispenser RS1000 machine (Agena, Inc, San Diego, CA, USA). The MALDI-TOF mass spectrometer (Agena, Inc) was used to translate the SpectroCHIP and the raw genotyping data was obtained using the TYPER 4.0 software.

Statistical Analysis

Student’s t-tests were used to compare the ages, glucose, and lipid metabolic parameters (TC, HDL-C, TG, LDL-C, FPG, and HbA1C) between the T2DM and NDM groups while the gender distribution between the T2DM and control groups was tested using the Chi-square test. The Student’s t-test and Chi-square test were performed using the SPSS 21. All polymorphic loci were tested for deviation from the Hardy-Weinberg equilibrium in the control group with a threshold of 0.05. Basic statistical analysis for the association between the alleles, genotypes, haplotypes, and T2DM were performed using the SHEsis software,^16,17 and the T2DM risk was estimated by the odds ratio (OR) with 95% confidence interval (95% CI). The distribution and differences in the haplotypes between the case and control groups were determined using the SHEsis software and the haplotypes with frequencies <0.03 were ignored during the analysis.^16,17 The association between T2DM and genotypes of these SNPs was analyzed using inheritance mode analysis by SNPStats software.¹⁸ The Akaike information criterion (AIC) and the Bayesian information criterion (BIC) were used to determine the best fit model for each SNP.¹⁸ The statistical power was calculated using PS Software.¹⁹ The glucose and lipid metabolic parameters of the different genotypes in each SNP were compared with one-way analysis of variance. Multiple comparisons among the three genotypes were corrected by Tukey’s method using GraphPad Prism software (Version 7.0). The significant threshold after Bonferroni correction for multiple comparisons was indicated by P<0.00625(0.05/8) for each SNP.

Results

Subject Characteristics

The clinical characteristics and the glucose and lipid metabolic parameters of the enrolled subjects are presented in Table 1. The age or gender between the subjects from the T2DM and NDM groups showed no significant difference (P>0.05) as well as age between the male and female subjects from both groups. Otherwise, there were significant differences in the glucose and lipid metabolic parameters (TC, TG, HDL-C, LDL-C, FPG, and HbA1C levels) between subjects from the T2DM and NDM groups, that statistically showed as P<0.001 (Table 1). To be more specific, compared to the subjects from the NDM group, the T2DM subjects exhibited dyslipidemia (higher levels of TC, TG, and LDL-C and lower levels of HDL-C).

Table 1 Clinical Characteristics and Glucose and Lipid Metabolic Parameters of the Subjects Enrolled in the Present Study

Association Between the SNPs and T2DM

The assessment of the genotype frequencies of the eight SNPs in the NDM group had been tested to be in Hardy-Weinberg equilibrium (P>0.05) (Table 2). The allele and genotype frequencies of the eight SNPs in both the T2DM and NDM groups are listed in Table 2. Four SNPs (rs4712524, rs10946398, rs7754840 in CDKAL1, and rs5015480 in HHEX) showed significantly different allele frequencies between the T2DM and NDM groups (P<0.00625) (Table 2). The G allele of rs4712524 (P=0.004, OR=1.184; 95% CI=1.057–1.327), C allele of rs10946398 (P<0.001, OR=1.247; 95% CI=1.112–1.398), and C allele of rs7754840 (P<0.001, OR=1.229; 95% CI=1.096–1.378) in CDKAL1, and the C allele of rs5015480 of in HHEX (P<0.001, OR=1.295; 95% CI=1.124–1.493) functioned as the risk alleles of T2DM. The other four SNPs, rs9460546 and rs7756992 in CDKAL1, and rs1111875 and rs7923837 in HHEX did not show any significant association.

Table 2 Comparison of Genotypic and Allelic Distribution of SNPs (Rs4712524, Rs10946398, Rs7754840, Rs9460546, and Rs7756992) in CDKAL1 and SNPs (Rs1111875, Rs5015480, and Rs7923837) in HHEX

Association of the Haplotypes of the SNPs with T2DM

Linkage disequilibrium (LD) among the SNPs was also estimated, where the LD coefficient “D” was calculated using the SHEsis software. The D’ values are defined in the range [−1, 1], with a value of “1” indicating perfect disequilibrium. A D’ value above 0.8 indicated the existence of different loci in the LD. The LD of the five SNPs (rs4712524, rs10946398, rs7754840, rs9460546, and rs7756992) in CDKAL1 was estimated. The CDKAL1 rs4712524-rs10946398-rs7754840-rs9460546 haplotype was constructed with D’>0.9 (Supplementary Table 1). The haplotype analysis revealed that the frequency distributions of the haplotypes rs4712524-rs10946398-rs7754840-rs9460546 in the T2DM and NDM groups had statistically significant differences (P=0.001) (Table 3). This result indicated that the rs4712524G-rs10946398C-rs7754840C-rs9460546G (OR=1.210; 95% CI=1.076–1.360) was associated with a higher risk of T2DM development, while rs4712524A-rs10946398A-rs7754840G-rs9460546T functioned as a protector against the development of T2DM (OR=0.827; 95% CI=0.735–0.930) (Table 3). The HHEX rs1111875-rs5015480 haplotype analysis with D’>0.9 showed that rs1111875C-rs5015480C was associated with development of T2DM (P<0.001, OR=1.364; 95% CI=1.180–1.576) (Table 4 and Supplementary Table 2).

Table 3 The CDKAL1 Rs4712524-Rs10946398-Rs7754840-Rs9460546 Haplotype Analysis Between the NDM and T2DM Group

Table 4 The HHEX Rs1111875-Rs5015480 Haplotype Analysis Between the NDM and T2DM Group

Mode of Inheritance Analysis

The inheritance model (codominant, dominant, recessive, overdominant, and log-additive) of these SNPs was constructed using SNPStats. The best fit inheritance model for rs4712524, rs10946398, and rs7754840 in CDKAL1 and rs5015480 in HHEX were all log-additive (with the lowest AIC and BIC). For rs4712524, 2G/G+G/A genotype was associated with higher risk of T2DM development (P=0.003, OR=1.19; 95% CI=1.06–1.33) (Table 5). For rs10946398, the 2C/C+C/A genotype was the risk factor compared to AA genotype (P<0.001, OR=1.26; 95% CI=1.12–1.42) (Table 6). For rs7754840, the 2C/C+C/G genotype was associated with higher risk of T2DM development (P<0.001, OR=1.24; 95% CI=1.11–1.40) (Table 7). For rs5015480 2C/C+C/G was the risk genotype (P<0.001, OR=1.30; 95% CI=1.13–1.51) (Table 8). The other SNPs did not exhibit any association with T2DM in the given population (Supplementary Table 3–6).

Table 5 Different Inheritance Models Analysis of the Rs4712524 in CDKAL1 Between the NDM and T2DM Group

Table 6 Different Inheritance Models Analysis of the Rs10946398 in CDKAL1 Between the NDM and T2DM Group

Table 7 Different Inheritance Models Analysis of the Rs7754840 in CDKAL1 Between the NDM and T2DM Group

Table 8 Different Inheritance Models Analysis of the Rs5015480 in HHEX Between the NDM and T2DM Group

Association of Genotypes of the SNPs with Metabolic Phenotypes

In the NDM group, no significant associations of these SNPs with glucose and lipid metabolic parameters, including FPG, TC, HDL-C, TG, LDL-C, and HbA1C, were observed (Supplementary Table 7).

Discussion

CDKAL1 and HHEX are well-accepted pathogenesis-related key genes for T2DM, they play a pivotal role in regulating the insulin secretion function of pancreatic β-cells. This study evaluated eight SNPs on CDKAL1 and HHEX genes to determine the significant association with T2DM. Of those, four SNPs (rs4712524, rs10946398, and rs7754840 in CDKAL1, rs5015480 in HHEX) were associated with T2DM in the Chinese population.

CDKAL1 expression in human pancreatic β-cells participates in the correlation of CDKAL1- and CDK5-mediated pathways.²⁰ It has been shown that CDKAL1 could increase insulin secretion in pancreatic β-cells by inhibiting CDK5.^4,6,7 Recently, many studies have investigated the association between SNPs (rs7754840, rs4712524, rs10946398, rs7756992, and rs10811661) in the CDKAL1 gene and T2DM in different populations around the world.^21–33 However, the association results of some SNPs showed a difference in different populations,^23,34 even in the same population.^26,35 In the current study, we found that rs7754840 was associated with T2DM in a Chinese population. Our results were consistent with other population results,^{27,–29–33,36} especially for the Chinese population (Uyghur and Han).^37–39 In 2014, Klimentidis et al⁴⁰ reported that rs7754840 was significantly associated with HbA1c in a Yup’ik population (P<0.001). And the rs7754840 also has been reported to be associated with decreased β-cell glucose sensitivity and insulin secretion.^11,41 These results indicated that rs7754840 could be associated with the T2DM through influencing the β-cell function and insulin secretion. However, in 2017, Nikitin et al³⁴ found the rs7754840 was not associated with T2DM in Russian. Lee et al²⁶ found rs7754840 was associated with risk of T2DM in Koreans in 2008. However, in 2012, Park et al³⁵ did not find the association between this SNP and T2DM in Koreans. These association results indicated that the different genetic background or other factors could influence the association results in different populations, even in the same population. This phenomenon was also found in the association of rs7756992 in a Chinese population. In 2010, Xu et al⁴² reported an association of this SNP with T2DM in a Chinese population, which is similar with the results in other populations.^25–28,33 However, Dou et al⁴³ found there is no association between this SNP and T2DM in another Chinese population, which is consistent with the results in other populations.^28,39,44 In the current study, we also did not find an association between this SNP and T2DM in the Chinese population. The inconsistency was also found in the association of rs10946398 with T2DM, which was associated with T2DM in the current Chinese population, Chinese She,⁴⁵ and Russian³⁴ population, but not associated with T2DM in a Korean population.²⁵ In summary of the previous studies’ results and our study, we found that these results did not show a clear conclusion of the association between these SNPs and T2DM. In 2015, Locke et al⁴⁶ reported that SNPs in CDKAL1 might show an effect on the expression of CDKAL1. Thus, functional studies are needed to confirm the association and reveal the value of the association in clinical application.

HHEX gene encodes a transcription factor which is involved in a fundamental pathway, the Wnt signaling pathway.¹⁰ In 2014, McKenna et al⁴⁷ found that misregulated HHEX expression within the diabetic islet might contribute to disrupted paracrine control of insulin secretion in T2DM, leading to accelerated β-cell exhaustion and failure. In the current study, we found that rs5015480C in HHEX gene might be associated with high risk of T2DM in the Chinese population. Our result is consistent with the results in German,²³ Korean,²⁵ and other Chinese populations,^13,38,48 including Chinese She population (the frequencies of rs5015480C were T2DM 0.437 vs NGT 0.403).⁴⁵ The rs5015480, which is located upstream of the HHEX gene has been found to be associated with decreased insulin response and β-cell glucose sensitivity.¹¹ In 2014, Klimentidis et al⁴⁰ found that rs5015480 was significantly associated with a combined fasting glucose and HbA1c measure (P<0.001) and with HOMA-B (P<0.001) among the Yup’ik population. These results indicated that rs5015480 could be associated with the T2DM through influencing the β-cell glucose sensitivity.^11,41 Several studies showed that rs1111875 in HHEX gene was related to the risk of T2DM in European^11,49,50 and Chinese^13,48 populations. In the current study, our results showed that rs1111875T was the protective factor of T2DM before Bonferroni correction, which is consistent with results of previous studies.^{11,–13,–48–50} However, there was no significant difference between T2DM and NDM after Bonferroni correction. If the sample size will be increased, the rs1111875 could be associated with the T2DM in the current Chinese population.

The haplotype is more powerful than single SNP in detecting the association between SNPs and complex diseases. In the current study, we observed that the rs4712524G-rs10946398C-rs7754840C-rs9460546G haplotype in CDKAL1 gene was associated with a higher risk of T2DM (P=0.001, OR=1.210; 95% CI=1.076–1.360), which is consistent with the single SNP result. Moreover, the HHEX haplotype rs1111875C–rs5015480C was associated with development of T2DM which is also consistent with the single SNP result. Therefore, the association between these haplotypes and T2DM could provide powerful evidence in assessing the underlying genetic preposition in T2DM.

In the current study, our data just provided the association data between these SNPs and T2DM, however, the mechanisms still need to be verified through functional study, which is one of the limitations of the current study. In addition, some clinical information, such as weight, Body mass index (BMI), the treatment used in patients with T2DM, family history of T2DM, and metabolic-carbohydrate control was important. However, we did not collect these kinds of information in the current study. Thus, the lack of these kinds of information was the other limitation of the current study.

Conclusion

In the current study, we revealed four SNPs (rs4712524, rs10946398, and rs7754840 in CDKAL1, rs5085140 in HHEX) and the haplotype in HHEX and CDKAL1 were associated with T2DM in a Chinese population. The statistical power of the four SNPs was 0.830, 0.965, 0.942, and 0.946, respectively, which indicates that our results are reliable. Our results from an association study would provide reference data for revealing the pathogenesis of T2DM and finding the new biomarkers of T2DM, which could be utilized for the prevention, diagnosis, and treatment of T2DM.