Molecular diagnosis of maturity onset diabetes of the young in Iranian patients: improving management

Abstract

Background

The purpose of this study is to identify the mutations of the most common form of maturity-onset diabetes of the young (MODY), also known as MODY3, in diabetic patients suspected of MODY. This can recommend appropriate medical surveillance of at-risk family members of MODY based on the genetic cause.

Methods

We analyzed the clinical course of 19 patients from 12 unrelated Iranian families with diabetes features. The coding regions and intron–exon boundaries of the hepatocyte nuclear factor 1 alpha (HNF1A) gene were studied by polymerase chain reaction (PCR) and sanger sequencing. Also, the detected mutation was analyzed by bioinformatics tools.

Results

One novel frameshift insertion mutation (p.Glu11Argfs*12) was detected in one of the probands and seven other patients of her family with the heterozygote state. The mutation is located in the exon1 of the dimerization domain of the HNF1A gene. According to the In Silico analysis, the detected mutation is predicted as a pathogenic one.

Conclusions

Differential diagnosis between MODY3 and other forms of diabetes can be considered a necessity in terms of overlapping symptoms of MODY3 with type1 or 2 diabetes. Molecular genetic testing can provide an accurate diagnosis for optimal management.

Introduction

Maturity Onset Diabetes of the Young (MODY) [OMIM #606391] is a monogenic group of diabetes. It is also heterogeneous and defined by hyperglycemia, which is followed by the autosomal dominant inheritance that can usually be detected in at least three generations [1].

Mutation in fourteen various genes including HNF4A, GCK, HNF1A, PDX1, TCF2, NEUROD1, KLF11, CEL, PAX4, INS, BLK, ABCC8, KCNJ11, and APPL1 are determined as MODY 1 to MODY 14 phenotypes respectively [2]. Mutations in HNF1A, HNF4A, and GCK genes are the most frequent cause of MODY, which account for up to 80% of all MODY cases [3]. HNF1A gene (OMIM #142410) is located on chromosome 12q24.2 encoding a transcription factor (HNF1A), which plays a major role in the regulation of gene expression in various metabolic pathways such as insulin secretion, glucose metabolism, and transport [2]. Moreover, it is important in the development and function of beta cells. Accordingly, mutations in HNF1A lead to decreasing glucose-induced insulin secretion and increasing hyperglycemia [4]. MODY3 results from the HNF1A gene mutations are considered the most common form of monogenic diabetes. Affected patients of MODY3 have demonstrated a high risk of microvascular and macrovascular complications besides being in the severe condition of hyperglycemia [5]. HNF1A gene mutations have high penetrance, the average age of 65% of patients are reported before 25 years. Nonetheless, approximately all subjects develop diabetes by the age of 50 years old [6]. According to published data more than 400 different mutations have been identified in the HNF1A gene [3, 7]. Although MODY subjects are accounted for approximately 1–2% of patients with diabetes, they are frequently misdiagnosed as type 1 or type 2 diabetes because of overlapping clinical symptoms [4]. These patients are treated with insulin, whereas they may be well controlled on oral hypoglycemic agent sulphonylurea [8].

This study aimed to identify MODY cases by molecular diagnosis to develop the correct diagnosis that subsequently leads to improve glycaemic control. A further benefit is to evaluate the prevalence of HNF1A mutations in diabetic patients suspected of MODY according to the most likely clinical criteria.

Methods

Subject selection and clinical examination

In the current study, 19 diabetic patients who were suspected of MODY from 12 unrelated families with the Arab community were analyzed. Informed consent was obtained from all participants in the study. All patients were diagnosed upon by the International Society for Pediatric and Adolescent Diabetes and confirmed at the outpatient’s pediatric endocrine clinic of Abuzar Children’s Medical Center and diabetes clinic of Golestan Hospital [Craig, 2006 #296]. Both hospitals were affiliated with Ahvaz Jundishapur University of Medical Sciences in Ahvaz during our three years of assessment. The criteria are as follows: positive family history of diabetes in at least three consecutive generations with an autosomal dominant inheritance pattern and presence of at least one proband with early-onset diabetes (before 30 years of age). Moreover, some clinical parameters to exclude type 1 or 2 diabetes from MODY were assessed which included body mass index (BMI), serum insulin level, C-peptide level and history of ketoacidosis. The study was approved by the Ethics Committee of the Ahvaz Jundishapur University of Medical Sciences (IR.AJUMS.REC.1396.822).

DNA extraction and amplification of HNF1A gene

All patients underwent HNF1A gene mutation analysis. Genomic DNA was extracted from the whole peripheral blood (containing EDTA) using YEKTA TAJHIZ (cat. No: FABGK001) DNA blood kit. The 10 exons and their intron–exon boundaries of the HNF1A gene were amplified by PCR with 8 pairs of primers for 9 exons of the HNF1A gene [9]. PCR reactions were performed by using the 100–200 ng of template DNA, 10 pmol of each primer 1.5 mMol MgCl2, 200 µmol dNTPs, 4U prime Taq DNA polymerase, and 10X buffer in 50 µl reaction tubes. For visualizing, separating fragments of DNA by size, and purification, Agarose Gel Electrophoresis alongside a DNA ladder was assayed.

Sequencing and mutational analysis of the HNF1A gene

Sanger sequencing and data analysis were performed for all PCR fragments for each patient using an ABI Prism 3700 apparatus (Big Dye Terminator sequencing kit, Applied Biosystems). The familial co-segregation analysis was performed for the detected mutation in patients, as well. We used ENSEMBL (https://www.ensembl.org/index.html), HGMD (basic) (http://www.hgmd.cf.ac.uk/ac/index.php), and dbSNP (https://www.ncbi.nlm.nih.gov/snp/) as the databases to verify the identified variants to determine whether the variants were previously reported as the pathogenic mutation or polymorphism, or they should be considered as a novel mutation. plus, In Silico analysis was used to predict the pathogenic role of the novel detected mutation, using Mutation Taster (http://www.mutationtaster.org). The genomic alignment was also performed for the new detected mutation to compare the region between different species via ENSEMBL databases. Moreover, we used the Protein Homology/analogy Recognition Engine (Phyre2) website (http://www.sbg.bio.ic.ac.uk/~phyre2/html/page.cgi?id=index), for protein modelling prediction of HNF1A polypeptide structure compared with mutated protein one. The new mutation was named according to the guideline of the Human Genome Variation Society, using the human HNF1A transcribe sequence as reference (ENST00000541395.5).

Statistical analysis

All variables are presented as the Mean ± SEM. An independent t-test was used to compare the Age and BMI values between two distinct groups of MODY and diabetes (non-HNF1A mutation). A two-tailed test with a type error level (α) set at 5% was applied in all statistical analyses. Statistical analyses were performed using SPSS for Windows software (version 26; SPSS Statistics, IBM Corporation, Armonk, NY, USA).

Results

All subjects were selected from the Arab community in the southwest of Iran. The parents were consanguineous in 7 out of the 12 unrelated patients’ families. 7 Out of the 19 diabetic patients were male (36.8%) and 12 (63.2%) were female; 8 (42.1%) of them had MODY mutation, and 11 (57.9%) were diabetes. The mean age at the time of the study was 25.4 ± 2.5 years and the mean BMI was 22.7 ± 0.7 kg/m². In Comparison between HNF1A mutation carriers and non-HNF1A mutation, no significant statistical differences in BMI were observed (P = 0.179), however, it was significant in age (P = 0.009).

At the beginning, we sequenced 12 patients as proband with MODY clinical characteristics for the HNF1A gene. Approximately all patients were glycemic controlled with insulin. We identified a novel frameshift mutation c.29-30insAC (p.Glu11Argfs*12), as a possible pathogenic cause with the heterozygote state in exon 1 in the patient PIII-1 (Fig. 1A). This mutation has created a premature stop codon 12 residues downstream at codon position 23 which is located in the dimerization domain of HNF1A protein. Interestingly, this particular mutation has not been reported in the 1000 genomes database. In order to have further exploration, the disease-causing role was predicted for this variant by applying Mutation Taster (Table 1). Many pieces of evidence indicate that this region has been conserved across species. Utilizing In Silico protein structure analysis by the Phyre2 program, showed that c.29-30insAC mutation in the dimerization domain of the HNF1A gene is produced a truncated protein structure (Fig. 2). The patient PIII-I is a female from a consanguineous family. The patient was diagnosed with type 1 diabetes and had been treated with insulin by the age of 11 years old. She was suffering from insulin overdose such as confusion, dizziness, and irritability after taking insulin. Meanwhile, the proband’s mother had gestational diabetes during pregnancy with microvascular complications, including renal parenchymal echogenicity that may be due to the primary stage of renal impairment in diabetes. Also, proband’s grandfather was suffering from nephropathy which is one of the microvascular complications of diabetes. The patient’s family history was strongly positive for diabetes. All affected members including the patient's mother (II-2), three aunts (II-3, II-4, and II-9) and two uncles (II-6 and II-8), grandfather on her mother's side (I-1) (Fig. 1B) were similarly controlled by insulin as a type 1 patients. Subsequently, all members of the family were screened for the detected mutation. The results revealed that affected members, in the same way, were heterozygous for p.Glu11Argfs*12 mutations. while, the unaffected ones (I-2, II-1, II-5, II-7, II-10, III-2, and III-3) were wild type. (Fig. 1B).

Fig. 1.

A Sequence chromatography of a novel frameshift mutation c.29-30insAC with heterozygote state compared to wild type form of the HNF1A gene. The arrow is indicating the position of the mutation. B The pedigree of genetic inheritance of the MODY3 affected family. The genotypes are shown with “WT” = wild type; M = mutation

Table 1.

Predicted data of c.29-30insAC mutation by mutation taster tool

Summary	Prediction disease causing NMDf Amino acid sequence changed Frameshift Protein features (might be) affected Splice site changes
The analysed issue	Analysis result
HGNC Symbole	HNF1A
Alteration (phys. location)	chr12:121416600_121416601insAC
DNA changes	c.29_30insAC/cDNA.52_53insAC/g.255_256insAC
Alteration type/region	Inse/CDS
AA changes	E11Rfs*12
Position(s) of altered AAa	11 (frameshift or PTC—further changes downstream)
known variant	Variant was neither found in ExAC nor 1000G
Frameshift	Yes
phyloP/phastCons	(flanking) 0.726/(flanking) 0.554
Length of protein	NMD
Position (AA) of stop codon in WTb/Mc AA sequence	663/22
Theoretical NMD boundary in CDSd	1811
Wild type AA sequence	MVSKLSQLQT ELLAALLESG LSKEALIQAL GEPGPYLLAG EGPLDKGESC GGGRGELAEL PNGLGETRGS EDETDDDGED FTPPILKELE NLSPEEAAHQ KAVVETLLQE DPWRVAKMVK SYLQQHNIPQ REVVDTTGLN QSHLSQHLNK GTPMKTQKRA ALYTWYVRKQ REVAQQFTHA GQGGLIEEPT GDELPTKKGR RNRFKWGPAS QQILFQAYER QKNPSKEERE TLVEECNRAE CIQRGVSPSQ AQGLGSNLVT EVRVYNWFAN RRKEEAFRHK LAMDTYSGPP PGPGPGPALP AHSSPGLPPP ALSPSKVHGV RYGQPATSET AEVPSSSGGP LVTVSTPLHQ VSPTGLEPSH SLLSTEAKLV SAAGGPLPPV STLTALHSLE QTSPGLNQQP QNLIMASLPG VMTIGPGEPA SLGPTFTNTG ASTLVIGLAS TQAQSVPVIN SMGSSLTTLQ PVQFSQPLHP SYQQPLMPPV QSHVTQSPFM ATMAQLQSPH ALYSHKPEVA QYTHTGLLPQ TMLITDTTNL SALASLTPTK QVRSRPAGPP LACDRAPHPH IPRAQEAALL PQVFTSDTEA SSESGLHTPA SQATTLHVPS QDPASIQHLQ PAHRLSASPT VSSSSLVLYQ SSDSSNGQSH LLPSNHSVIE TFISTQMASS SQ*
Mutated AA sequence	MVSKLSQLQT RSSWRPCSSQ G*

All positions are in base pairs (bp) if not explicitly stated differently

^aAA/aa = amino acid; ^bWT = wild type; ^cM = mutated; ^dCDS = coding sequence; ^eins = insertion; ^fNMD = nonsense-mediated mRNA decay

Fig. 2.

Structural comparison of the molecular models of HNF1A Glu11Argfs*12 and HNF1A polypeptides determined by the phyre2 program. A HNF1A polypeptide wild type at the 3D structure. B The c.29-30insAC mutation in the dimerization domain is produced a truncated protein

Discussion

This study has been executed for the first time in the southwest of Iran in the patients suspected of MODY and their families. We detected a novel frameshift insertion mutation (p.Glu11Argfs*12) in the dimerization domain of the HNF1A protein. Furthermore, analysis of the Phyre2 website predicted that the most essential parts of HNF1A have been deleted in the mutant model and the protein structure has changed in comparison with the normal model. The molecular characteristics of the HNF1A gene mutations might possess a pivotal role in the severity and the onset age of the disease as well. The HNF1A gene produces three isoforms (HNF1A(A) to HNF1A(C)) through alternative splicing and polyadenylation. Consequently, the HNF1A(A) isoform is produced by full-length transcript contains the 10 exons, whereas HNF1A(B) and HNF1A(C) isoforms are produced by a transcript that contain 7 and 6 exons, respectively [10]. HNF1A protein comprises three functional domains; NH2-terminal dimerization domain (amino acids 1–32), DNA-binding domain (amino acids 91–281), and COOH-terminal transactivation domain (282–631) [11, 12]. It is noteworthy that the average age of incidence in patients with the mutations located in exons 1–6 which affect all three HNF1A isoforms is younger than those who have mutations in exons 8–10 that affect only isoform HNF1A(A) [7]. Therefore, the mutations affecting the dimerization/DNA-binding domains are associated with a lower age of onset compared to those affecting the transactivation domain. Meanwhile, it has been shown that the age of diagnosis is lower in patients with truncating mutations than in those who have missense mutations [7]. In our studied proband (PIII-I, Fig. 1B), the age of onset was 11 years old and the identified mutation (p.Glu11Argfs*12) was a truncating mutation that affected the dimerization domain. Accordingly, this mutation is the more likely cause of diabetes manifestation for MODY3 at the early age of the proband, the same as some other affected patients in the family. Variable clinical manifestation related to age at onset between family members might be defined by different lifestyles and environmental factors. However, it is yet to be elucidated. According to the previous studies, the occurrence of the HNF1A gene mutations caused almost 20–50% of the MODY cases in various ethnicities, which demonstrates the prevalence of diversity among the various populations. [13]. Based on clinical investigation, the prevalence of MODY in Asia was estimated at 5% to 20% of diabetes [14]. Moghbeli et al. in 2017 screened 34 unrelated families with the clinical features of MODY. One novel missense mutation (c.C49G) in the exon1 HNF1A gene was identified in one of the probands and his family [9]. Following the same study, Eskandari et al. in 2019 detected a missense mutation (c.758G>A) in HNF1A coding regions of two unrelated families [15], which was reported for the first time by Colclough et al. [14]. Nevertheless, definite results have yet to be achieved to be able to estimate the genetic mutation rate of the HNF1A gene. Notably, genetic studiesneed to be conducted in a widespread approach in the Iranian population. On the other hand, the prevalence of MODY is rare but is often misclassified as type 1 or type 2 diabetes. A previous observational study transferred the way in which patients were treated with insulin to sulphonylurea. As a result, most of the patients did respond well to low-dose of sulphonylurea as the first line of treatment [16]. Therefore, wider investigations based on molecular genetic testing by integrating clinical features of the diabetic patients suspected of MODY have an important implication. Identifying the majority of cases in terms of establishing a diagnostic pathway to provide genetic counselling to patient relatives, as well as reaching the clinical application would be a beneficial accomplishment.

Conclusion

In conclusion, molecular genetic analysis is the authentic way to distinguish MODY from other types of diabetes especially type 1 diabetes due to the young age at diagnosis in both groups. Due to the autosomal dominant inheritance of MODY and the regarding the possible risk of 50% of recurrence to offspring, families deserve to be quite aware of the disorder and the genetic approach to make better decisions to have a further child. So that, this becomes an important issue that MODY diagnostic tests with genetic counseling should be included in diabetes diagnostic experiments. Some advantages are considered with the accurate molecular diagnosis; it provides prognosis in the high-risk familial, also, better monitoring of the affected individual causes reduce the possible side effect and the costs of treatment by oral hypoglycemic agent compared to daily insulin injections. As MODY cases frequency and subsequent investigations are not adequated, further investigations are required to obtain meaningful outcomes about MODY patients. Also, it would increase our knowledge of the genetic databases in Iranian MODY patients. Combined with these, we hope to raise awareness to help identify MODY patients.

Funding

This work was supported by Ahvaz Jundishapur University Medical Sciences (Research Project No: D-9610).

Declarations

Conflict of interest

Nothing to declare.

Ethics approval

The study was approved by the Ethics Committee of the Ahvaz Jundishapur University of Medical Sciences (IR.AJUMS.REC.1396.822).

Consent to participate

Informed consent to participate was obtained from all individual participants included in the study.

Consent for publication

Informed consent for publication was obtained from all individual participants included in the study.