387 - Using genomic database analysis to characterize pathogenicity of IDS missense variants in Mucopolysaccharidosis Type II

Publication Number: 387.7122

Shuhei Sako, The Jikei University School of Medicine, Tokorozawa-Shi, Saitama, Japan; Eri Imagawa, The Jikei University School of Medicine, Tokyo, Tokyo, Japan; Kimihiko Oishi, The Jikei University School of Medicine, Tokyo, Tokyo, Japan

Background: Mucopolysaccharidosis type Ⅱ (MPS Ⅱ; Hunter syndrome) is an X-linked recessive lysosomal storage disorder with complete penetrance caused by pathogenic variants in the IDS gene. Recently, MPS Ⅱ was added to the Recommended Uniform Screening Panel (RUSP); however, interpreting newborn screening (NBS) results remains challenging due to variable enzyme levels, pseudodeficiency alleles, and clinical variability, complicating severity prediction. Confirmatory genetic testing is needed to ensure accurate diagnosis and avoid unnecessary interventions. We hypothesized that a new population genetics-based in silico classification approach could provide a better diagnostic tool for MPS Ⅱ NBS.

Objective: To comprehensively classify the pathogenicity of IDS missense variants, often difficult to interpret and frequently classified as variants of uncertain significance (VUSs), by analyzing carrier rates in a genomic healthy control database.

Design/Methods: We assessed the allele counts (AC) of IDS missense variants in the Genome Aggregation Database v4.1.0 for each zygosity. Variants were categorized into three groups: 1) possibly pathogenic, with AC = 0 in hemizygous males and AC ≥ 3 in heterozygous females; 2) possibly benign, with AC ≥ 1 in hemizygous males; and 3) probably benign, with AC ≥ 1 in hemizygous males and AC ≥ 2 in homozygous females. Pathogenicity was evaluated by in silico analysis, with highly pathogenic candidates in group 1 variants defined by CADD ≥ 20, SIFT < 0.04, and PolyPhen-2 ≥ 0.9 scores.

Results: Of the 365 IDS variants in the database, 214 were classified into groups 2 and 3. Among these, over ten variants have been reported in MPS Ⅱ patients per HGMD, suggesting they are likely pseudodeficiency or mild phenotype alleles due to their presence in healthy males. Pathogenicity score distributions varied across the groups, with highly pathogenic predictions more common in group 1. Newly identified potentially pathogenic variants in group 1 included p.Leu359Val, p.Leu259Ile, p.Asp532Val, p.Val44Met. Notably, p.Leu259 and p.Asp532 may represent hot-spot variants in MPS II.

Conclusion(s): Our data suggest that analyzing genomic information, such as allele count and zygosity, from healthy control databases can be a new approach to distinguishing pathogenic from pseudodeficiency alleles in X-linked diseases with complete penetrance. Future in vitro analyses are necessary to confirm the pathogenicity of these variants.