Complementary computational and experimental evaluation of missense variants in the ROMK potassium channel

Abstract

The renal outer medullary potassium (ROMK) channel is essential for potassium transport in the kidney, and its dysfunction is associated with a salt-wasting disorder known as Bartter syndrome. Despite its physiological significance, we lack a mechanistic understanding of the molecular defects in ROMK underlying most Bartter syndrome-associated mutations. To this end, we employed a ROMK-dependent yeast growth assay and tested single amino acid variants selected by a series of computational tools representative of different approaches to predict each variants' pathogenicity. In one approach, we used in silico saturation mutagenesis, i.e. the scanning of all possible single amino acid substitutions at all sequence positions to estimate their impact on function, and then employed a new machine learning classifier known as Rhapsody. We also used two additional tools, EVmutation and Polyphen-2, which permitted us to make consensus predictions on the pathogenicity of single amino acid variants in ROMK. Experimental tests performed for selected mutants in different classes validated the vast majority of our predictions and provided insights into variants implicated in ROMK dysfunction. On a broader scope, our analysis suggests that consolidation of data from complementary computational approaches provides an improved and facile method to predict the severity of an amino acid substitution and may help accelerate the identification of disease-causing mutations in any protein. Author summaryAs the number of sequenced human genomes rises, a major challenge is to identify which single amino acid variations in a protein affect function and predispose individuals to disease. While predictive algorithms are available for this purpose, a comparative analysis of recently developed algorithms has not been adequately performed, nor is it clear whether combining algorithms would improve predictive power. To this end, we compared the efficacy of three publicly available algorithms and applied the results to Bartter syndrome, PLOS COMPUTATIONAL BIOLOGY PLOS Computational Biology | https://doi.Citation: Ponzoni L, Nguyen NH, Bahar I, Brodsky JL (2020) Complementary computational and experimental evaluation of missense variants in the ROMK potassium channel. PLoS Comput Biol 16 (4): e1007749. https://doi.org/10.a human disease for which numerous poorly-characterized single amino acid variants have been identified and for which there is no cure. In silico saturation mutagenesis, i.e., the computational prediction of pathogenesis for every possible amino acid substitution, allowed us to experimentally test predictions by measuring the activity of an ion channel linked to Bartter syndrome. Based on data from blinded experiments, we discovered that Rhapsody and EVmutation successfully predicted deleterious mutations. Moreover, Rhapsody-which takes into account evolutionary as well as structural and dynamic considerations-predicted that >90% of known Bartter syndrome mutations are deleterious. Overall, our data will aid investigators who wish...

Genome mining yields putative disease-associated ROMK variants with distinct defects

“…The accuracy of this method was previously corroborated in silico on a dataset of ~20,000 labelled human variants, and when compared with alternative approaches using multiple accuracy metrics, Rhapsody’s performance at that time consistently ranked among the highest [ 30 , 40 ]. We further experimentally verified the predictive power of Rhapsody using a yeast growth assay that reports on ROMK plasma membrane residence and channel function (see Introduction ) [ 31 ]. Using the experimental data from the yeast screen, we also previously computed receiver operating characteristic (ROC) curves and found that Rhapsody had the highest accuracy compared to Polyphen-2 [ 41 ] and EVmutation [ 42 ] (Computed AUROC was 0.86 for Rhapsody, as opposed to 0.81 and 0.77 for Polyphen-2 and EVmutation).…”

“…Using the experimental data from the yeast screen, we also previously computed receiver operating characteristic (ROC) curves and found that Rhapsody had the highest accuracy compared to Polyphen-2 [ 41 ] and EVmutation [ 42 ] (Computed AUROC was 0.86 for Rhapsody, as opposed to 0.81 and 0.77 for Polyphen-2 and EVmutation). Importantly, Rhapsody predicted the severity of known disease-linked ROMK mutations with >90% accuracy [ 31 ].…”

Machine learning in computational modelling of membrane protein sequences and structures: From methodologies to applications

“…For example, twelve tools were systematically benchmarked for evaluating the pathogenic impact of the variants of a voltage-gated sodium channel, hNav1.5 [226] , and determined more than 70 potential pathogenic variants in both TM and extracellular regions. In another study, several tools (e.g., Rhapsody and EVmutation) were employed to determine deleterious mutations in the renal outer medullary potassium channel [227] . Crucially, training these tools with high prediction capacity depends heavily upon the availability of high-confidence experimental data.…”

How Functional Genomics Can Keep Pace With VUS Identification

“…However, only rarely are KCNJ2 pathogenic variants associated with trafficking defects ( 80 ) and alternative DMS approaches for measuring Kir2.1 might be better suited. One possible DMS approach is to adapt yeast-based functional assays that demonstrate normal and LoF variants based on growth patterns, which have been reported for ATS-associated Kir2.1 ( 73 ) and Barter Syndrome-associated Kir1.1 (a Kir2.1 paralog that is not arrhythmia related) ( 81 ). Regardless, development of a potassium channel DMS is needed for this important arrhythmogenic and underrepresented DMS data group.…”