A novel RFX6 heterozygous mutation (p.R652X) in maturity-onset diabetes mellitus: A case report.

Heterozygous RFX6 mutation has emerged as a potential cause of maturity-onset diabetes mellitus of the young (MODY). A 16-year-old female was diagnosed with diabetes by her family doctor and was referred to our institution for genetic examination. Genetic testing revealed a novel RFX6 heterozygous mutation (NM_173560: exon17: c.1954C>T: p.R652X) in the patient and in her mother and brother. She had no islet-specific autoantibodies and showed a reduced meal-induced response of insulin, glucose-dependent insulinotropic polypeptide, and glucagon-like peptide-1, which is consistent with the phenotype of MODY due to heterozygous RFX6 mutation. In conclusion, we report a case of MODY due to a novel heterozygous mutation, p.R652X.

INTRODUCTION

Maturity onset diabetes mellitus of the young (MODY) is a monogenic form of diabetes mainly due to gene mutation related to pancreatic β‐cell dysfunction . Mutations in HNF1A, HNF1B, and GCK are the most common causes of MODY and are responsible for ~40% of its etiology, but unidentified MODY genes are involved .

Regulatory factor X6 (RFX6) is a member of the regulatory factor X (RFX) family of transcription factors , . RFX6 regulates the differentiation and function of insulin‐producing cells , . A homozygous RFX6 gene defect proves the Mitchell‐Riley syndrome, which is characterized by neonatal diabetes with pancreatic hypoplasia, duodenal and jejunal atresia, and gall bladder agenesis . Recently, heterozygous RFX6 gene mutations have been noted in MODY cases . RFX6‐related MODY cases have characteristics such as lack of islet autoantibodies and reduced secretion of insulin and glucose‐dependent insulinotropic polypeptide (GIP) in response to glucose ingestion . Here we report a patient with a novel heterozygous RFX6 mutation (p.R652X).

CASE REPORT

A 16‐year‐old female visited her family doctor after being found positive for urinary glucose in her school medical checkup. She had a family history of diabetes (Figure 1). Her plasma glucose and HbA1c levels were 467 mg/dL and 10.8%, respectively; she was diagnosed with diabetes and referred to our institution for genetic examination. Upon admission, her height, body weight, and body mass index were 142 cm, 42 kg, and 20.8, respectively. Autoantibodies against glutamic acid decarboxylase and insulin and islet antibody‐2 were negative, and basal levels of pituitary, thyroid, and adrenal hormones were normal (Table 1). She had no diabetic neuropathy or retinopathy, while her urinary albumin level was 43.1 mg/g Cre. Her abdominal ultrasound and computer tomography (CT) scan revealed no pancreas, intestine, or gall bladder anomalies except for fatty liver. After 7 days of intensive insulin therapy (a total of 10 units per day), 500 mg metformin was started instead of insulin therapy. Her fasting glucose levels fell to 90–100 mg/dL and she was discharged 16 days after admission. While her HbA1c remained >7% for 4 months after discharge, it improved substantially after initiation of the glucagon‐like peptide‐1 (GLP‐1) receptor agonist liraglutide (Figure 2a).

Figure 1

A family tree of the patient. Squares, circles, and arrows indicate males, females, and proband, respectively. Family members, including her brother, mother, and maternal grandfather had a history of diabetes. d. Deceased. M and N, wild type and p.R652X mutation alleles, respectively. Roman numerals on the left of the pedigrees indicate generation number, and the numbers below the symbols indicate the subject’s number within each pedigree. Arrow shows the proband. The proband’s family members generally have short stature regardless of the presence (M/N) or absence (N/N) of the identified heterozygous RFX6 mutation (p.R652X): III‐2 (M/N) 142 cm, II‐1 (N/N) 155 cm, II‐3 (M/N) 152 cm, III‐1 (M/N) 156 cm, I‐1 148 cm, I‐2 155.5 cm and II‐2 156 cm, respectively. Thus, it is likely that the short stature of the proband is unrelated to the heterozygous RFX6 mutation, p.R652X.

Table 1

Laboratory data of the patient

Urinalysis		AST (U/L)	23 (13–30)	Hormones
Specific gravity	>1.040	ALT (U/L)	29 (7–23)	IGF1 (ng/mL)	210
Protein	‐	LDH (U/L)	131 (124–222)	GH (ng/mL)	1.51
Glucose	3+	ALP (U/L)	258 (106–322)	FT3 (pg/mL)	3.03 (2.39–4.06)
RBC	‐	γGTP (U/L)	20 (9–32)	FT4 (ng/dL)	0.97 (0.76–1.65)
Ketone	‐	AMY (U/L)	40 (44–132)	TSH (μIU/mL)	0.24 (0.54–4.26)
Blood counts		CRE (mg/dL)	0.45 (0.46–0.79)	Cortisol (μg/dL)	15.5 (6.24–18.0)
WBC (×103/μL)	7.32 (3.3–8.6)	UA (mg/dL)	4.7 (2.6–7.0)	ACTH (pg/mL)	31.9 (7.2–63.3)
RBC (×106/μL)	4.35 (3.86–4.92)	Na+ (mEq/L)	136 (138–145)	Insulin (mU/L)	5.10 (1.84–12.2)
Hb (g/dL)	12.8 (13.7–16.8)	K+ (mEq/L)	4.0 (3.6–4.8)	CPR (ng/mL)	1.49 (1.84–12.2)
Ht (%)	37.5 (40.7–50.1)	Cl− (mEq/L)	106 (101–108)	Others
MCV (fL)	86.2 (83.6–98.2)	Ca2+ (mg/dL)	9.6 (8.8–10.1)	Anti‐GAD Ab (U/mL)	<5.0 (<5.0)
MCHC (%)	29.4 (31.7–35.3)	Pi2− (mg/dL)	3.1 (2.7–4.6)	IA2 (U/mL)	0.7 (<0.6)
Plt (×104/μL)	30.4 (15.8–34.8)	T‐cho (mg/dL)	187 (142–220)	Anti‐Insulin Ab (U/mL)	<0.4 (<0.4)
Biochemistry		TG (mg/dL)	127 (30–150)
TP (g/dL)	6.8 (6.6–8.1)	HDL‐c (mg/dL)	51 (30–150)
Albumin (g/dL)	4.5 (4.1–5.1)	Glucose (mg/dL)	329

ACTH, adrenocorticotropic hormone; ALP, alkaline phosphatase; ALT, alanine aminotransferase; AMY, amylase; AST, aspartate aminotransferase; BUN, blood urea nitrogen; ChE, choline esterase; CPR, C peptide immunoreactivity; CRE, creatinine; FT3, free tri‐iodo‐thyronine; FT4, free thyroxine; GAD, glutamate decarboxylase; GH, growth hormone; Hb, hemoglobin; Ht, hematocrit; IA‐2, islet antigen 2; IGF‐1, insulin‐like growth factor; LDH, lactate dehydrogenase; MCHC, mean corpuscular hemoglobin concentration; MCV, mean corpuscular volume; Plt, platelet; RBC, red blood cell; T‐Bil, total bilirubin; T‐cho, total cholesterol; TG, triglyceride; TP, total protein; UA, uric acid; WBC, white blood cell; γGTP, γ‐glutamyltransferase.

Figure 2

Clinical course and the response of insulin and incretins to meal ingestion in the patient. (a) HbA1c (closed square) and body weight (open square). (b) Meal tolerance test (480 kcal; carbohydrate 58.4%, protein 15.9% and fat 25.7%) was performed 9 months after her discharge. Glucagon‐like peptide‐1 (GLP‐1) liraglutide was stopped 7 days before the test. We compared the proband’s data with our previous results of 18 drug‐naïve individuals with type 2 diabetes (T2DM) and 17 individuals with normal glucose tolerance (NGT) . Values of glucose, insulin, GLP‐1, and glucose‐dependent insulinotropic polypeptide (GIP) at the indicated times are represented as mean ± SEM for NGT and type 2 diabetes.

A meal tolerance test was performed 9 months after her discharge (Figure 2b). Compared with individuals having normal glucose tolerance (NGT) , the proband showed higher glucose levels before and after meal ingestion. Insulin and C‐peptide were lower in the proband than in those with NGT despite the higher glucose levels before and 30 min after meal ingestion. The insulinogenic index [(insulin30 min − insulin0 min)/(glucose30 min − glucose0 min)] was lower in the proband (0.44) than that of NGT (0.83). Fasting and postprandial levels of GIP and GLP‐1 were lower than those of NGT. Due to her family history of diabetes (Figure 1) and the early onset of her disease, we sequenced MODY‐related genes of the proband, her father, mother, and brother. We found a heterozygous RFX6 mutation (RFX6: NM_173560: exon17: c.1954C>T: p.R652X) in the proband (III‐2) and in her mother (II‐3) and brother (III‐1).

We obtained approval from the ethics committee of Gifu University Graduate School of Medicine (Approval number 29‐191). Genetic testing was carried out after counseling by a clinical genetic specialist. Written informed consent was obtained.

DISCUSSION

We report a case of MODY due to the novel heterozygous RFX6 mutation p.R652X. Our patient had no islet autoantibodies and reduced insulin and GIP response, which is consistent with the characteristics of RFX6‐related MODY cases . Our patient also had a reduced GLP‐1 response. These findings are consistent with the reduced levels of GIP and GLP‐1 found in Rfx6‐deficient mice and may underlie the substantial improvement in glycemic control by the GLP‐1 receptor agonist liraglutide in this case (Figure 2a). Consistently, it was reported that patients with the RFX6 mutation show a good therapeutic response to dipeptidyl peptidase‐4 (DPP‐4) inhibitors . Further clinical studies are warranted to determine the efficacy of GLP‐1 receptor agonists and DPP‐4 inhibitors in RFX6‐related MODY cases.

Insulin and C‐peptide levels were reduced in the proband before and 30 min after meal ingestion compared with those of NGT, and the insulinogenic index was lower in the proband than that in NGT. These results clearly indicate impaired insulin secretion in the proband, which may be related to the reduced GLP‐1 and GIP response (Figure 2b). RFX6 upregulates the expression of the insulin gene and other genes involved in insulin secretion (e.g., glucokinase and voltage‐dependent calcium channel) in human β‐cell line EndoC‐βH2 cells . Thus, it is likely that the novel heterozygous RFX6 mutation p.R652X directly impairs β‐cell function in humans.

RFX6 directly binds to an X‐box motif located at −288 to −269 bp from the transcription initiation site of the human insulin gene and activates insulin gene transcription . We found that RFX6(R652X) not only failed to activate the human insulin promoter but also suppressed RFX6‐induced activation of the human insulin gene dose‐dependently (Figure S1). Subcellular localization of RFX6(R652X) revealed that the p.R652X mutation had little effect on nuclear localization of the protein, except that some fractions of the protein caused perinuclear aggregation (Figure S2). These results together strongly suggest that RFX6(R652X) interacts with RFX6 in the nucleus, thereby suppressing RFX6‐induced activation of the target genes involved in insulin secretion. However, it is also possible that perinuclear aggregation of RFX6(R652X) disturbs β‐cells, thereby impairing insulin secretion.

In conclusion, we report a case of MODY due to the novel heterozygous mutation p.R652X.

DISCLOSURE

The authors declare no conflict of interest.

Figure S1 | Transcriptional activities of RFX6 and RFX6(R652X) in the human embryonic kidney (HEK) 293 cells. (a) Schematics of RFX6 and RFX6(R652X). (b) Luciferase activities of human insulin promoter in HEK293 cells transfected with indicated amounts of pCMV6b (Empty), pCMV6b‐RFX6 (RFX6) or pCMV6b‐RFX6(R652X) (RFX6(R652X)). Data represent the mean ± SD of 6 wells per condition. *Indicates P < 0.05 (Tukey’s test) versus 0.4 μg pCMV6b‐RFX6. Data were analyzed using Graphpad Prism versus 9 (Graphpad Software, CA, USA).

Click here for additional data file.

Figure S2 | Subcellular localization of RFX6 and RFX6(R652X) tagged with 3xFLAG epitope at the N‐terminus in indicated cell lines. (a) Schematics of RFX6 and RFX6(R652X) tagged with 3xFLAG epitope at N‐terminus. (b) Representative images of HEK293, INS‐1 832/12 cells and STC‐1 cells expressing RFX6 and RFX6(R652X) tagged with 3xFLAG epitope. Magnification, x40. Red arrows indicate perinuclear aggregates of 3xFLAG‐RFX6 (R652X).

Click here for additional data file.

Supplementary Material | Supplementary methods.

Click here for additional data file.