Rare Variants in NEUROD1 and PDX1 Are Low-Penetrance Causes of MODY, Whereas Those in APPL1 and WFS1 Are Not Associated With MODY

Abstract

An accurate genetic diagnosis of maturity-onset diabetes of the young (MODY) is critical for personalized treatment. To avoid misdiagnosis, only genes with strong evidence of causality must be tested. Heterozygous variants in NEUROD1, PDX1, APPL1, and WFS1 have been implicated in MODY, but strong genetic evidence supporting causality is lacking. We therefore assessed their existing genetic evidence and performed gene-level burden tests in a large MODY cohort, alongside two established MODY genes as positive controls (HNF1A- high penetrance, RFX6 -low penetrance). The first reported MODY-associated variants in NEUROD1, PDX1, APPL1, and WFS1 were <1:20,000 frequency. Based on the small number of large published pedigrees per gene (n < 3), MODY-associated variants showed only modest cosegregation in these genes. Crucially, ultra-rare (minor allele frequency <1:10,000) protein-truncating and predicted-damaging missense variants in APPL1 and WFS1 were not enriched in a MODY cohort (n = 2,571) compared with population control individuals (n = 155,501; all P > 0.05). In contrast, variants in NEUROD1 and PDX1 were enriched, albeit at levels comparable to RFX6. Multiple sensitivity analyses corroborated these findings. In summary, rare heterozygous variants in NEUROD1 and PDX1 are low-penetrance causes of MODY, while those in APPL1 and WFS1 lack robust genetic evidence for causality and should not be included in MODY testing panels.

Article highlights: Testing genes with limited evidence of causality risks misdiagnosis of maturity-onset diabetes of the young (MODY). There is limited evidence as to whether rare variants in PDX1, NEUROD1, APPL1, and WFS1 cause MODY. Rare damaging variants in APPL1 and WFS1 are not enriched in a MODY cohort, but those in PDX1 and NEUROD1 are, at a level similar to low-penetrance MODY genes. PDX1 and NEUROD1 should be included in MODY gene panels, while heterozygous APPL1 and WFS1 variants should not be reported as causes of MODY.

Figure 1. Gene-level burden analysis of ultra-rare variants in the MODY cohort

The figure shows gene-level burden analyses comparing ultra-rare variants (MAF <1×10⁻⁴) between a European-ancestry MODY cohort (n=2,471) and UK Biobank controls (n=155,501). Analyses include: (A) synonymous variants (B) protein-truncating variants; and (C) damaging missense variants (REVEL >0.7). HNF1A and RFX6 served as high- and low-penetrance positive controls, respectively. Asterisks (*) indicate significance after multiple testing corrections (p<0.002). We provide an odds ratio and a 95% confidence interval for each association.

Figure 2. Functional domain-specific burden tests

The figure shows gene-level burden analyses comparing ultra-rare variants (MAF <1×10⁻⁴) in a European-ancestry MODY cohort (n=2,471) against UK Biobank controls (n=155,501). We show data for A) protein-truncating variants, (B) damaging missense variants (REVEL >0.7), and (C) synonymous variants. We show data for all ultra-rare variants and split them by their location within or outside the functional domains. We provide an odds ratio and a 95% confidence interval for each association. Asterisks (*) indicate significance after multiple testing corrections (p<0.002).