TL;DR
This paper presents a high-throughput yeast-based assay to functionally test nearly all possible single amino acid substitutions in the human OTC gene. The assay accurately predicts variant pathogenicity and clinical presentation, significantly improving the interpretation of genetic variants of uncertain significance and paving the way for more precise genetic diagnostics.
What Problem Does This Paper Address?
Ornithine transcarbamylase (OTC) deficiency is a serious, yet treatable, metabolic disorder. While DNA sequencing can identify potential disease-causing variants, a significant challenge lies in interpreting the clinical significance of these variants, especially those of uncertain significance (VUS). Existing diagnostic methods have limitations, and a large number of identified variants lack clear clinical interpretation. This hinders timely diagnosis and treatment, particularly for severe forms of OTC deficiency that can manifest in newborns. The paper aims to bridge this gap by developing a robust functional assay to assess the impact of a vast number of OTC variants, thereby aiding in the classification of VUS and improving diagnostic accuracy.
Methods
The researchers developed a novel high-throughput functional assay using Saccharomyces cerevisiae (yeast) as a model system. They engineered yeast strains that lack their native arginine biosynthesis gene (ARG3), which is functionally analogous to human OTC. The human OTC gene, codon-optimized for yeast expression and with its mitochondrial targeting sequence removed, was then introduced into these yeast strains. This "yOTC" construct allowed yeast to grow in the absence of arginine, with growth rate serving as a proxy for OTC enzyme activity.
To test a large number of variants, they created a library of yOTC constructs, each containing a single amino acid substitution accessible by single nucleotide polymorphisms (SNPs). This library included 1,570 variants, representing 84% of all possible SNP-accessible missense mutations in OTC. The growth of yeast strains carrying these variants was measured quantitatively. This approach allowed them to assess the functional impact of each amino acid substitution independently of its effect on splicing, a common confounder in other methods.
Datasets
The primary dataset generated in this study is a functional assay dataset for 1,570 SNP-accessible missense variants in the human OTC gene. This dataset was used to:
- Validate the assay: Compared against known pathogenic and benign variants from ClinVar.
- Stratify variants by clinical presentation: Correlated with neonatal, female, and late-onset disease presentations.
- Reclassify Variants of Uncertain Significance (VUS): Used in conjunction with ACMG/AMP guidelines to re-evaluate variants.
The study also references existing databases such as ClinVar for variant classifications and gnomAD for population-level variant frequencies.
Key Insights and Ideas
- Yeast as a powerful model for human protein function: The study demonstrates that a yeast system can effectively recapitulate the functional consequences of human OTC variants, providing a scalable and cost-effective platform for large-scale functional analysis.
- OTC is highly sensitive to amino acid substitutions: A striking finding is that a large proportion (27%) of tested amino acid substitutions resulted in very low or no detectable enzyme activity, suggesting that the OTC protein's function is highly dependent on specific amino acid sequences.
- Functional data dramatically improves VUS interpretation: The inclusion of their functional assay data as PS3/BS3 evidence, according to ACMG/AMP guidelines, led to the reclassification of a significant number of VUS variants to likely-pathogenic, highlighting the critical role of functional assays in clinical genetics.
- Distinct functional profiles correlate with disease presentation: Variants causing severe loss of function (near-zero growth) were strongly associated with neonatal and female presentations of OTC deficiency, while variants with moderate impairment were linked to late-onset disease. This suggests that the degree of functional deficit directly influences the clinical phenotype.
- Separating splicing from coding effects: The use of an intronless construct was crucial for distinguishing the impact of amino acid changes from potential effects on mRNA splicing, providing a clearer understanding of the underlying molecular mechanisms of pathogenicity.
Key Results
- 1,570 variants assayed: The study functionally characterized 1,570 SNP-accessible missense variants in the human OTC gene, representing 84% of all possible such variants.
- High proportion of deleterious variants: 27% of tested amino acid substitutions exhibited less than 5% of wild-type OTC activity, indicating a complete or near-complete loss of function.
- Strong correlation with ClinVar: The assay results showed a strong agreement with existing ClinVar classifications, accurately distinguishing between pathogenic and benign variants.
- Clinical stratification validated: Variants associated with neonatal and female presentations of OTC deficiency showed significantly impaired function in the assay, while late-onset variants showed less severe impairment.
- Reclassification of VUS: By applying their functional data, the researchers reclassified 21 out of 39 VUS variants to likely-pathogenic, demonstrating the clinical utility of their assay.
Limitations and Future Work
- Limitations of the yeast model: While powerful, the yeast system is a simplified model. The study acknowledges that human OTC function is not fully complemented by the yeast ortholog (77% growth), meaning that variants enhancing activity beyond wild-type yOTC could potentially be detected, but this is less likely to be a major issue for pathogenic variants.
- Ambiguous growth range: The 71-90% growth range in the assay remains ambiguous for functional interpretation, requiring additional clinical information.
- Small sample size for benign variants: The number of known benign variants available for comparison was relatively small, which could limit the robustness of conclusions drawn from this subset.
- Future Work: The authors suggest that their dataset is of high clinical utility and illustrates the power of functional assays to inform interpretation of existing and novel genetic variation. This work lays the groundwork for applying similar high-throughput functional assays to other genes implicated in genetic diseases, particularly for interpreting VUS. The ability to distinguish between splicing defects and coding sequence effects opens avenues for more targeted mechanistic studies.
Summary updated