HNF1B-MODY in the Norwegian MODY Registry and the Norwegian Childhood Diabetes Registry: Clinical insights and prevalence informed by genetic and functional evaluation

This study integrates genetic, functional, and clinical data from Norwegian diabetes registries to characterize the prevalence and features of HNF1B-MODY and 17q12 deletions, demonstrating that impaired transactivation correlates with clinical severity and supporting expanded screening for HNF1B alterations in children with renal anomalies regardless of diabetes status.

The Gist

The Big Picture: A "Master Switch" That's Broken

Imagine your body is a massive, high-tech factory. Inside this factory, there is a Master Control Panel (a gene called HNF1B) that tells different departments how to run. Specifically, this panel controls the Sugar Department (pancreas/diabetes) and the Water Filtration Department (kidneys).

Sometimes, this Master Control Panel gets damaged. When it breaks, the factory doesn't just have a sugar problem; the water filters start leaking too. This specific type of broken panel causes a rare form of diabetes called HNF1B-MODY.

The problem for doctors is that the "damage" to the panel can look very different depending on how you look at it. Sometimes the whole panel is missing (a big chunk of the blueprint is gone), and sometimes just a few screws are loose (a tiny typo in the instructions). Because the symptoms vary so much, it's often hard to tell if a specific typo is actually dangerous or just a harmless quirk.

The Mission: Sorting the "Real" Breaks from the "Fake" Alarms

This study, conducted by researchers in Norway, acted like a team of detectives and mechanics. They wanted to solve two main mysteries:

1. How common is this broken panel? (They checked two huge databases of patients: one for adults with suspected rare diabetes and one for children with diabetes).
2. How do we tell a dangerous break from a harmless one? (They used a "stress test" in the lab to see how well the broken panels actually worked).

The Detective Work: What They Found

1. The "Missing Page" vs. The "Typo"

The researchers found two main ways the HNF1B gene gets messed up:

• The "Missing Page" (17q12 deletion): Imagine the instruction manual for the Master Control Panel has an entire page ripped out. This is a big deletion. They found 28 people with this.
• The "Typo" (Sequence variants): The manual is there, but a letter is misspelled in a crucial sentence. They found 15 people with these specific typos.

The Surprise: Many children with the "Missing Page" didn't even have diabetes when they were found! They were sent to the doctor because their kidneys looked weird on an ultrasound. This taught the doctors a huge lesson: Don't wait for diabetes to appear. If a child has kidney issues, check their HNF1B gene immediately, even if their blood sugar is fine.

2. The "Stress Test" (Functional Analysis)

This is the coolest part of the paper. For the "Typo" cases, the researchers couldn't just guess if the typo was bad. So, they built a miniature factory in a petri dish.

• They took the DNA instructions from the patients.
• They put them into cells (like little test tubes).
• They turned on the lights and asked: "How loud is the factory humming?" (This is called transactivation activity).

The Results:

• The Broken Machines: Some typos made the factory almost silent (very low activity). These patients had severe symptoms: bad kidneys, diabetes, and other issues.
• The Working Machines: Some typos made the factory hum just as loudly as a healthy one. These patients had mild symptoms or no symptoms at all.

The Analogy: Think of it like a car engine.

• Pathogenic (Bad): You take a wrench and smash the engine. The car won't start. (Low activity = Severe disease).
• Benign (Good): You paint a stripe on the bumper. The engine runs perfectly fine. (Normal activity = No disease).

By doing this "engine test," they could reclassify some genetic findings. They said, "We thought this was a broken engine, but it's actually fine!" and "We thought this was a scratch, but it's actually a smashed engine!"

The Numbers: How Common Is It?

• In the Adult Registry: About 7.4% of people with confirmed rare diabetes had this specific broken panel. That's roughly 1 in 13.
• In the Children's Registry: About 0.2% of children with diabetes had the "Missing Page" version.
• The Autoimmune Twist: Some children tested positive for "Type 1 Diabetes" antibodies (which usually means the immune system is attacking the pancreas). But, they also had the broken HNF1B gene. This means a child can have both conditions at the same time. Doctors shouldn't rule out this rare gene just because the child has "Type 1" antibodies.

The Takeaway: Why This Matters

This paper is like a new User Manual for Doctors.

1. Look Beyond the Sugar: If a kid has kidney problems, check for this gene, even if they don't have diabetes yet.
2. Don't Trust the Antibodies Alone: Just because a kid tests positive for Type 1 Diabetes antibodies doesn't mean they only have Type 1. They might have this rare genetic issue too.
3. Test the Engine: If a doctor finds a weird typo in the gene but isn't sure if it's dangerous, they should run a "stress test" (functional analysis) to see if the protein actually works. This stops people from getting misdiagnosed and ensures they get the right treatment.

In short: By combining the patient's story, their family history, and a "lab stress test," the researchers figured out exactly which genetic breaks are dangerous and which are harmless, helping to save families from unnecessary worry and ensuring the right people get the right care.

Technical Summary

1. Problem Statement

Interpreting variants in the HNF1B gene is a significant challenge in clinical genetics due to:

• Phenotypic Heterogeneity: HNF1B-associated disease (MODY5) presents with a wide spectrum of symptoms, including diabetes, renal structural anomalies, genital tract malformations, and neurodevelopmental issues.
• Variable Expressivity & Incomplete Penetrance: Not all carriers develop diabetes, and severity varies widely even within families.
• High Rate of De Novo Mutations: This limits the utility of segregation analysis for variant classification.
• Limited Functional Data: Many missense variants are classified as Variants of Uncertain Significance (VUS) because there is insufficient evidence to determine if they are pathogenic or benign.
• Diagnostic Gaps: Current diagnostic criteria often rely on the presence of diabetes and negative autoantibodies, potentially missing HNF1B cases in children who present primarily with renal anomalies or who have concurrent autoantibodies.

2. Methodology

The study integrated data from two Norwegian registries: the Norwegian MODY Registry (NMR) and the Norwegian Childhood Diabetes Registry (NCDR).

• Cohort Analysis:
  • Analyzed 3,473 participants in the NMR and ~8,200 in the NCDR.
  • Identified individuals with HNF1B sequence variants and 17q12 deletions (17q12del).
  • Performed CNV analysis (MLPA and array genotyping) on subsets of patients to detect 17q12del.
• Functional Assays (In Vitro):
  • Cell Lines: Used HeLa (non-pancreatic) and MIN6 (pancreatic beta-cell) lines.
  • Transactivation Assay: Measured luciferase activity to assess the ability of HNF1B variants to activate target promoters (Rat Albumin and GLUT2). Results were expressed as a percentage of Wild Type (WT) activity.
  • DNA-Binding Assay: Used Electrophoretic Mobility Shift Assays (EMSA) on nuclear fractions to test DNA binding capacity for variants in the DNA-binding domain (DBD).
  • Controls: Included known pathogenic, benign, and synonymous variants.
• Variant Classification:
  • Applied ACMG-AMP-ClinGen guidelines (adapted from HNF1A guidelines).
  • Integrated functional data (Transactivation $\le$ 50% = PS3_Supporting; $\ge$ 85% = BS3_Supporting) with clinical phenotyping using the Faguer score (a metric for clinical burden).
• Statistical Analysis:
  • Calculated registry-level prevalence with 95% confidence intervals.
  • Used Spearman's rank correlation to assess the relationship between transactivation activity and clinical severity (Faguer score).

3. Key Contributions

• Functional-Phenotypic Correlation: Established a strong inverse correlation ($\rho = -0.701, p = 0.002$) between transactivation activity and clinical severity. Variants with lower transactivation activity ($\le$ 50% WT) were associated with significantly higher Faguer scores (more severe multisystem disease).
• Reclassification of VUS: Functional data enabled the reclassification of three variants from VUS to Likely Benign (LB) and provided evidence to reclassify two others from VUS to Likely Pathogenic (LP).
• Prevalence Estimation: Provided precise prevalence estimates for HNF1B alterations in both adult (MODY) and pediatric (Childhood Diabetes) populations, including those with 17q12 deletions.
• Diagnostic Paradigm Shift: Demonstrated that HNF1B screening is crucial in children with renal anomalies even in the absence of diabetes or in the presence of autoantibodies.

4. Key Results

A. Cohort Characteristics and Prevalence

• Total Carriers: Identified 28 individuals with 17q12del and 15 individuals with 14 unique pathogenic/likely pathogenic (LP/P) sequence variants.
• NMR (MODY Registry):
  • HNF1B alterations accounted for 7.4% of genetically confirmed monogenic diabetes cases (36/486 probands).
  • 17q12del prevalence: 1.43% (21/1,462).
  • LP/P sequence variant prevalence: 0.76% (15/1,962).
• NCDR (Childhood Registry):
  • 17q12del prevalence: ~0.2% (7/3,583 screened).
  • Notably, 43% of pediatric 17q12del carriers were positive for pancreatic autoantibodies (GADA and/or IA-2A), challenging the assumption that autoantibodies rule out monogenic diabetes.

B. Clinical Phenotypes

• Renal Involvement: Universal in both LP/P sequence variant carriers and 17q12del carriers (100%).
• Diabetes: Common but not universal. Some carriers presented with isolated renal anomalies without diabetes.
• 17q12del Specifics: Carriers showed higher rates of neurocognitive impairment, electrolyte disturbances (hypomagnesemia), and pancreatic exocrine insufficiency compared to sequence variant carriers.
• Clinical Burden: The median Faguer score was significantly higher for variants with $\le$ 50% transactivation activity (Median 13.0) compared to those with > 50% activity (Median 4.0).

C. Variant Classification Outcomes

• Pathogenic (LP/P): 11 variants confirmed as LP/P. This included Protein Truncating Variants (PTVs) and specific missense variants in the Dimerization (DD) and DNA-binding (DBD) domains (e.g., p.G20R, p.R295C).
• Benign (LB): 3 variants (p.P60R, p.N327K, p.P343S) were reclassified as Likely Benign. These retained normal transactivation activity ($\ge$ 85%) and carriers exhibited phenotypes consistent with Type 1 or Type 2 diabetes rather than HNF1B-specific disease.
• VUS: 2 variants (p.V2L, p.S362F) remained VUS due to normal functional activity and lack of specific HNF1B features, though p.S362F was found in both registries.
• Uncertain: p.Q131H remained a VUS due to modest functional impairment and limited clinical data.

5. Significance and Clinical Implications

• Refined Diagnostic Criteria: The study argues against relying solely on "classical MODY" criteria (young onset, negative autoantibodies, no renal issues). It highlights that HNF1B mutations should be suspected in children with renal structural anomalies regardless of diabetes status or autoantibody positivity.
• Reduced VUS Burden: Integrating functional assays with clinical data successfully reduced the number of VUS, improving diagnostic certainty and enabling better genetic counseling.
• Personalized Management: Early molecular diagnosis allows for proactive surveillance of renal function, electrolyte balance, and pancreatic exocrine function, which are critical for managing HNF1B-associated disease.
• Prevalence Awareness: The findings suggest HNF1B is a more common cause of monogenic diabetes and pediatric diabetes with renal anomalies than previously estimated, particularly when 17q12 deletions are included.

In conclusion, this study provides a robust framework for interpreting HNF1B variants by correlating functional deficits with clinical severity, ultimately advocating for broader genetic screening criteria in pediatric populations with renal anomalies.