Evaluation of enzyme activity predictions for variants of unknown significance in Arylsulfatase A.

Continued advances in variant effect prediction are necessary to demonstrate the ability of machine learning methods to accurately determine the clinical impact of variants of unknown significance (VUS). Towards this goal, the ARSA Critical Assessment of Genome Interpretation (CAGI) challenge was designed to characterize progress by utilizing 219 experimentally assayed missense VUS in the Arylsulfatase A (ARSA) gene to assess the performance of community-submitted predictions of variant functional effects. The challenge involved 15 teams, and evaluated additional predictions from established and recently released models. Notably, a model developed by participants of a genetics and coding bootcamp, trained with standard machine-learning tools in Python, demonstrated superior performance among submissions. Furthermore, the study observed that state-of-the-art deep learning methods provided small but statistically significant improvement in predictive performance compared to less elaborate techniques. These findings underscore the utility of variant effect prediction, and the potential for models trained with modest resources to accurately classify VUS in genetic and clinical research.

A New Set of in Silico Tools to Support the Interpretation of ATM Missense Variants Using Graphical Analysis.

Establishing the pathogenic nature of variants in ATM, a gene associated with breast cancer and other hereditary cancers, is crucial for providing patients with adequate care. Unfortunately, achieving good variant classification is still difficult. To address this challenge, we extended the range of in silico tools with a series of graphical tools devised for the analysis of computational evidence by health care professionals. We propose a family of fast and easy-to-use graphical representations in which the impact of a variant is considered relative to other pathogenic and benign variants. To illustrate their value, the representations are applied to three problems in variant interpretation. The assessment of computational pathogenicity predictions showed that the graphics provide an intuitive view of prediction reliability, complementing and extending conventional numerical reliability indexes. When applied to variant of unknown significance populations, the representations shed light on the nature of these variants and can be used to prioritize variants of unknown significance for further studies. In a third application, the graphics were used to compare the two versions of the ATM-adapted American College of Medical Genetics and Genomics and Association for Molecular Pathology guidelines, obtaining valuable information on their relative virtues and weaknesses. Finally, a server [ATMision (ATM missense in silico interpretation online)] was generated for users to apply these representations in their variant interpretation problems, to check the ATM-adapted guidelines' criteria for computational evidence on their variant(s) and access different sources of information.

Choosing Variant Interpretation Tools for Clinical Applications: Context Matters.

Pathogenicity predictors are computational tools that classify genetic variants as benign or pathogenic; this is currently a major challenge in genomic medicine. With more than fifty such predictors available, selecting the most suitable tool for clinical applications like genetic screening, molecular diagnostics, and companion diagnostics has become increasingly challenging. To address this issue, we have developed a cost-based framework that naturally considers the various components of the problem. This framework encodes clinical scenarios using a minimal set of parameters and treats pathogenicity predictors as rejection classifiers, a common practice in clinical applications where low-confidence predictions are routinely rejected. We illustrate our approach in four examples where we compare different numbers of pathogenicity predictors for missense variants. Our results show that no single predictor is optimal for all clinical scenarios and that considering rejection yields a different perspective on classifiers.

Gain and loss of function variants in EZH1 disrupt neurogenesis and cause dominant and recessive neurodevelopmental disorders.

Genetic variants in chromatin regulators are frequently found in neurodevelopmental disorders, but their effect in disease etiology is rarely determined. Here, we uncover and functionally define pathogenic variants in the chromatin modifier EZH1 as the cause of dominant and recessive neurodevelopmental disorders in 19 individuals. EZH1 encodes one of the two alternative histone H3 lysine 27 methyltransferases of the PRC2 complex. Unlike the other PRC2 subunits, which are involved in cancers and developmental syndromes, the implication of EZH1 in human development and disease is largely unknown. Using cellular and biochemical studies, we demonstrate that recessive variants impair EZH1 expression causing loss of function effects, while dominant variants are missense mutations that affect evolutionarily conserved aminoacids, likely impacting EZH1 structure or function. Accordingly, we found increased methyltransferase activity leading to gain of function of two EZH1 missense variants. Furthermore, we show that EZH1 is necessary and sufficient for differentiation of neural progenitor cells in the developing chick embryo neural tube. Finally, using human pluripotent stem cell-derived neural cultures and forebrain organoids, we demonstrate that EZH1 variants perturb cortical neuron differentiation. Overall, our work reveals a critical role of EZH1 in neurogenesis regulation and provides molecular diagnosis for previously undefined neurodevelopmental disorders.

Mechanism of KMT5B haploinsufficiency in neurodevelopment in humans and mice.

Pathogenic variants in KMT5B, a lysine methyltransferase, are associated with global developmental delay, macrocephaly, autism, and congenital anomalies (OMIM# 617788). Given the relatively recent discovery of this disorder, it has not been fully characterized. Deep phenotyping of the largest (n = 43) patient cohort to date identified that hypotonia and congenital heart defects are prominent features that were previously not associated with this syndrome. Both missense variants and putative loss-of-function variants resulted in slow growth in patient-derived cell lines. KMT5B homozygous knockout mice were smaller in size than their wild-type littermates but did not have significantly smaller brains, suggesting relative macrocephaly, also noted as a prominent clinical feature. RNA sequencing of patient lymphoblasts and Kmt5b haploinsufficient mouse brains identified differentially expressed pathways associated with nervous system development and function including axon guidance signaling. Overall, we identified additional pathogenic variants and clinical features in KMT5B-related neurodevelopmental disorder and provide insights into the molecular mechanisms of the disorder using multiple model systems.

PirePred: An Accurate Online Consensus Tool to Interpret Newborn Screening-Related Genetic Variants in Structural Context.

PirePred is a genetic interpretation tool used for a variety of medical conditions investigated in newborn screening programs. The PirePred server retrieves, analyzes, and displays in real time genetic and structural data on 58 genes/proteins associated with medical conditions frequently investigated in the newborn. PirePred analyzes the predictions generated by 15 pathogenicity predictors and applies an optimized majority vote algorithm to classify any possible nonsynonymous single-nucleotide variant as pathogenic, benign, or of uncertain significance. PirePred predictions for variants of clear clinical significance are better than those of any of the individual predictors considered (based on accuracy, sensitivity, and negative predictive value) or are among the best ones (for positive predictive value and Matthews correlation coefficient). PirePred predictions also outperform the comparable in silico predictions offered as supporting evidence, according to American College of Medical Genetics and Genomics guidelines, by VarSome and Franklin. Also, PirePred has very high prediction coverage. To facilitate the molecular interpretation of the missense, nonsense, and frameshift variants in ClinVar, the changing amino acid residue is displayed in its structural context, which is analyzed to provide functional clues. PirePred is an accurate, robust, and easy-to-use tool for clinicians involved in neonatal screening programs and for researchers of related diseases. The server is freely accessible and provides a user-friendly gateway into the structural/functional consequences of genetic variants at the protein level.

Loss of microRNA-135b Enhances Bone Metastasis in Prostate Cancer and Predicts Aggressiveness in Human Prostate Samples.

About 70% of advanced-stage prostate cancer (PCa) patients will experience bone metastasis, which severely affects patients' quality of life and progresses to lethal PCa in most cases. Hence, understanding the molecular heterogeneity of PCa cell populations and the signaling pathways associated with bone tropism is crucial. For this purpose, we generated an animal model with high penetrance to metastasize to bone using an intracardiac percutaneous injection of PC3 cells to identify PCa metastasis-promoting factors. Using genomic high-throughput analysis we identified a miRNA signature involved in bone metastasis that also presents potential as a biomarker of PCa progression in human samples. In particular, the downregulation of miR-135b favored the incidence of bone metastases by significantly increasing PCa cells' migratory capacity. Moreover, the PLAG1, JAKMIP2, PDGFA, and VTI1b target genes were identified as potential mediators of miR-135b's role in the dissemination to bone. In this study, we provide a genomic signature involved in PCa bone growth, contributing to a better understanding of the mechanisms responsible for this process. In the future, our results could ultimately translate into promising new therapeutic targets for the treatment of lethal PCa.

New genes involved in Angelman syndrome-like: Expanding the genetic spectrum.

Angelman syndrome (AS) is a neurogenetic disorder characterized by severe developmental delay with absence of speech, happy disposition, frequent laughter, hyperactivity, stereotypies, ataxia and seizures with specific EEG abnormalities. There is a 10-15% of patients with an AS phenotype whose genetic cause remains unknown (Angelman-like syndrome, AS-like). Whole-exome sequencing (WES) was performed on a cohort of 14 patients with clinical features of AS and no molecular diagnosis. As a result, we identified 10 de novo and 1 X-linked pathogenic/likely pathogenic variants in 10 neurodevelopmental genes (SYNGAP1, VAMP2, TBL1XR1, ASXL3, SATB2, SMARCE1, SPTAN1, KCNQ3, SLC6A1 and LAS1L) and one deleterious de novo variant in a candidate gene (HSF2). Our results highlight the wide genetic heterogeneity in AS-like patients and expands the differential diagnosis.

Case Report: Characterizing the Role of the STXBP2-R190C Monoallelic Mutation Found in a Patient With Hemophagocytic Syndrome and Langerhans Cell Histiocytosis.

Hemophagocytic lymphohistiocytosis (HLH) is a life-threatening hyperinflammatory disorder. HLH can be considered as a threshold disease depending on the trigger and the residual NK-cell cytotoxicity. In this study, we analyzed the molecular and functional impact of a novel monoallelic mutation found in a patient with two episodes of HLH. A 9-month-old child was diagnosed at 2 months of age with cutaneous Langerhans cell histiocytosis (LCH). After successful treatment, the patient developed an HLH episode. At 16 month of age, the patient went through an HSCT losing the engraftment 5 months later concomitant with an HLH relapse. The genetic study revealed a monoallelic mutation in the STXBP2 gene (.pArg190Cys). We transfected COS7 cells to analyze the STXBP2-R190C expression and to test the interaction with STX11. We used the RBL-2H3 cell line expressing STXBP2-WT-EGFP or R190C-EGFP for degranulation assays. Mutation STXBP2-R190C did not affect protein expression or interaction with syntaxin-11. However, we have demonstrated that STXBP2-R190C mutation diminishes degranulation in the RBL-2H3 cell line compared with the RBL-2H3 cell line transfected with STXBP2-WT or nontransfected. These results suggest that STXBP2-R190C mutation acts as a modifier of the degranulation process producing a decrease in degranulation. Therefore, under homeostatic conditions, the presence of one copy of STXBP2-R190 could generate sufficient degranulation capacity. However, it is likely that early in life when adaptive immune system functions are not sufficiently developed, an infection may not be resolved with this genetic background, leading to a hyperinflammation syndrome and eventually develop HLH. This analysis highlights the need for functional testing of new mutations to validate their role in genetic susceptibility and to establish the best possible treatment for these patients.

Towards a New, Endophenotype-Based Strategy for Pathogenicity Prediction in BRCA1 and BRCA2: In Silico Modeling of the Outcome of HDR/SGE Assays for Missense Variants.

The present limitations in the pathogenicity prediction of BRCA1 and BRCA2 (BRCA1/2) missense variants constitute an important problem with negative consequences for the diagnosis of hereditary breast and ovarian cancer. However, it has been proposed that the use of endophenotype predictions, i.e., computational estimates of the outcomes of functional assays, can be a good option to address this bottleneck. The application of this idea to the BRCA1/2 variants in the CAGI 5-ENIGMA international challenge has shown promising results. Here, we developed this approach, exploring the predictive performances of the regression models applied to the BRCA1/2 variants for which the values of the homology-directed DNA repair and saturation genome editing assays are available. Our results first showed that we can generate endophenotype estimates using a few molecular-level properties. Second, we show that the accuracy of these estimates is enough to obtain pathogenicity predictions comparable to those of many standard tools. Third, endophenotype-based predictions are complementary to, but do not outperform, those of a Random Forest model trained using variant pathogenicity annotations instead of endophenotype values. In summary, our results confirmed the usefulness of the endophenotype approach for the pathogenicity prediction of the BRCA1/2 missense variants, suggesting different options for future improvements.

The histone demethylase PHF8 regulates astrocyte differentiation and function.

Epigenetic factors have been shown to play a crucial role in X-linked intellectual disability (XLID). Here, we investigate the contribution of the XLID-associated histone demethylase PHF8 to astrocyte differentiation and function. Using genome-wide analyses and biochemical assays in mouse astrocytic cultures, we reveal a regulatory crosstalk between PHF8 and the Notch signaling pathway that balances the expression of the master astrocytic gene Nfia. Moreover, PHF8 regulates key synaptic genes in astrocytes by maintaining low levels of H4K20me3. Accordingly, astrocytic-PHF8 depletion has a striking effect on neuronal synapse formation and maturation in vitro. These data reveal that PHF8 is crucial in astrocyte development to maintain chromatin homeostasis and limit heterochromatin formation at synaptogenic genes. Our studies provide insights into the involvement of epigenetics in intellectual disability.

A Collaborative Effort to Define Classification Criteria for ATM Variants in Hereditary Cancer Patients.

BACKGROUND: Gene panel testing by massive parallel sequencing has increased the diagnostic yield but also the number of variants of uncertain significance. Clinical interpretation of genomic data requires expertise for each gene and disease. Heterozygous ATM pathogenic variants increase the risk of cancer, particularly breast cancer. For this reason, ATM is included in most hereditary cancer panels. It is a large gene, showing a high number of variants, most of them of uncertain significance. Hence, we initiated a collaborative effort to improve and standardize variant classification for the ATM gene. METHODS: Six independent laboratories collected information from 766 ATM variant carriers harboring 283 different variants. Data were submitted in a consensus template form, variant nomenclature and clinical information were curated, and monthly team conferences were established to review and adapt American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) criteria to ATM, which were used to classify 50 representative variants. RESULTS: Amid 283 different variants, 99 appeared more than once, 35 had differences in classification among laboratories. Refinement of ACMG/AMP criteria to ATM involved specification for twenty-one criteria and adjustment of strength for fourteen others. Afterwards, 50 variants carried by 254 index cases were classified with the established framework resulting in a consensus classification for all of them and a reduction in the number of variants of uncertain significance from 58% to 42%. CONCLUSIONS: Our results highlight the relevance of data sharing and data curation by multidisciplinary experts to achieve improved variant classification that will eventually improve clinical management.

FHLdb: A Comprehensive Database on the Molecular Basis of Familial Hemophagocytic Lymphohistiocytosis.

Background: Primary immunodeficiencies (PIDs) are a heterogeneous group of disorders. The lack of comprehensive disease-specific mutation databases may hinder or delay classification of the genetic variants found in samples from these patients. This is especially true for familial hemophagocytic lymphohistiocytosis (FHL), a life-threatening PID classically considered an autosomal recessive condition, but with increasingly demonstrated genetic heterogeneity. Objective: The aim of this study was to build an open-access repository to collect detailed information on the known genetic variants reported in FHL. Methods: We manually reviewed more than 120 articles to identify all reported variants related to FHL. We retrieved relevant information about the allelic status, the number of patients with the same variant, and whether functional assays were done. We stored all the data retrieved in a PostgreSQL database and then built a website on top of it, using the Django framework. Results: The database designed (FHLdb) (https://www.biotoclin.org/FHLdb) contains comprehensive information on reported variants in the 4 genes related to FHL (PRF1, UNC13D, STXBP2, STX11). It comprises 240 missense, 69 frameshift, 51 nonsense, 51 splicing, 10 in-frame indel, 7 deep intronic, and 5 large rearrangement variants together with their allelic status, carrier(s) information, and functional evidence. All genetic variants have been classified as pathogenic, likely pathogenic, uncertain significance, likely benign or benign, according to the American College of Medical Genetics guidelines. Additionally, it integrates information from other relevant databases: clinical evidence from ClinVar and UniProt, population allele frequency from ExAC and gnomAD, and pathogenicity predictions from well-recognized tools (e.g., PolyPhen-2, SIFT). Finally, a diagram depicts the location of the variant relative to the gene exon and protein domain structures. Conclusion: FHLdb includes a broad range of data on the reported genetic variants in familial HLH genes. It is a free-access and easy-to-use resource that will facilitate the interpretation of molecular results of FHL patients, and it illustrates the potential value of disease-specific databases for other PIDs.

Molecular analysis of the novel L243R mutation in STXBP2 reveals impairment of degranulation activity.

The presence of mutations in PRF1, UNC13D, STX11 and STXBP2 genes in homozygosis or compound heterozygosis results in immune deregulation. Most such cases lead to clinical manifestations of haemophagocytic lymphohistiocytosis (HLH). In the present study, we analyzed degranulation and cytotoxicity in a pediatric patient with a late presentation of HLH associated with Epstein-Barr virus infection. Remarkably, the results of the degranulation assay showed reduction of CD107a median fluorescence intensity (MFI) and absent cytotoxicity. Genetic analysis identified compound heterozygous mutations in STXBP2 gene: a previously reported splicing defect in exon 15 (c.1247-1G>C, p.V417LfsX126) and a novel missense mutation in exon 9 (c.728T>G, p.L243R). Transfection experiments of STXBP2-L243R or STXBP2-WT constructs showed an undetectable protein expression of the STXBP2-L243R mutation. The residue L243 is highly preserved evolutionarily; moreover, computational analysis of its structure revealed its participation in the rich network of interactions that stabilizes domains 2 and 3 of the protein. Altogether, we demonstrated by molecular and in silico analysis that the new L243R mutation in STXBP2 plays a pathogenic role that, together with the p.Val417Leufsc mutation, shows the synergistic negative effect of these two mutations on STXBP2 function, leading to a decrease of degranulatory activity in vivo.

PHF2 histone demethylase prevents DNA damage and genome instability by controlling cell cycle progression of neural progenitors.

Histone H3 lysine 9 methylation (H3K9me) is essential for cellular homeostasis; however, its contribution to development is not well established. Here, we demonstrate that the H3K9me2 demethylase PHF2 is essential for neural progenitor proliferation in vitro and for early neurogenesis in the chicken spinal cord. Using genome-wide analyses and biochemical assays we show that PHF2 controls the expression of critical cell cycle progression genes, particularly those related to DNA replication, by keeping low levels of H3K9me3 at promoters. Accordingly, PHF2 depletion induces R-loop accumulation that leads to extensive DNA damage and cell cycle arrest. These data reveal a role of PHF2 as a guarantor of genome stability that allows proper expansion of neural progenitors during development.

Assessment of blind predictions of the clinical significance of BRCA1 and BRCA2 variants.

Testing for variation in BRCA1 and BRCA2 (commonly referred to as BRCA1/2), has emerged as a standard clinical practice and is helping countless women better understand and manage their heritable risk of breast and ovarian cancer. Yet the increased rate of BRCA1/2 testing has led to an increasing number of Variants of Uncertain Significance (VUS), and the rate of VUS discovery currently outpaces the rate of clinical variant interpretation. Computational prediction is a key component of the variant interpretation pipeline. In the CAGI5 ENIGMA Challenge, six prediction teams submitted predictions on 326 newly-interpreted variants from the ENIGMA Consortium. By evaluating these predictions against the new interpretations, we have gained a number of insights on the state of the art of variant prediction and specific steps to further advance this state of the art.

Compensated pathogenic variants in coagulation factors VIII and IX present complex mapping between molecular impact and hemophilia severity.

Compensated pathogenic deviations (CPDs) are sequence variants that are pathogenic in humans but neutral in other species. In recent years, our molecular understanding of CPDs has advanced substantially. For example, it is known that their impact on human proteins is generally milder than that of average pathogenic mutations and that their impact is suppressed in non-human carriers by compensatory mutations. However, prior studies have ignored the evolutionarily relevant relationship between molecular impact and organismal phenotype. Here, we explore this topic using CPDs from FVIII and FIX and data concerning carriers' hemophilia severity. We find that, regardless of their molecular impact, these mutations can be associated with either mild or severe disease phenotypes. Only a weak relationship is found between protein stability changes and severity. We also characterize the population variability of hemostasis proteins, which constitute the genetic background of FVIII and FIX, using data from the 1000 Genome project. We observe that genetic background can vary substantially between individuals in terms of both the amount and nature of genetic variants. Finally, we discuss how these results highlight the need to include new terms in present models of protein evolution to explain the origin of CPDs.

BRCA1- and BRCA2-specific in silico tools for variant interpretation in the CAGI 5 ENIGMA challenge.

BRCA1 and BRCA2 (BRCA1/2) germline variants disrupting the DNA protective role of these genes increase the risk of hereditary breast and ovarian cancers. Correct identification of these variants then becomes clinically relevant, because it may increase the survival rates of the carriers. Unfortunately, we are still unable to systematically predict the impact of BRCA1/2 variants. In this article, we present a family of in silico predictors that address this problem, using a gene-specific approach. For each protein, we have developed two tools, aimed at predicting the impact of a variant at two different levels: Functional and clinical. Testing their performance in different datasets shows that specific information compensates the small number of predictive features and the reduced training sets employed to develop our models. When applied to the variants of the BRCA1/2 (ENIGMA) challenge in the fifth Critical Assessment of Genome Interpretation (CAGI 5) we find that these methods, particularly those predicting the functional impact of variants, have a good performance, identifying the large compositional bias towards neutral variants in the CAGI sample. This performance is further improved when incorporating to our prediction protocol estimates of the impact on splicing of the target variant.

Structural and Computational Characterization of Disease-Related Mutations Involved in Protein-Protein Interfaces.

One of the known potential effects of disease-causing amino acid substitutions in proteins is to modulate protein-protein interactions (PPIs). To interpret such variants at the molecular level and to obtain useful information for prediction purposes, it is important to determine whether they are located at protein-protein interfaces, which are composed of two main regions, core and rim, with different evolutionary conservation and physicochemical properties. Here we have performed a structural, energetics and computational analysis of interactions between proteins hosting mutations related to diseases detected in newborn screening. Interface residues were classified as core or rim, showing that the core residues contribute the most to the binding free energy of the PPI. Disease-causing variants are more likely to occur at the interface core region rather than at the interface rim (p < 0.0001). In contrast, neutral variants are more often found at the interface rim or at the non-interacting surface rather than at the interface core region. We also found that arginine, tryptophan, and tyrosine are over-represented among mutated residues leading to disease. These results can enhance our understanding of disease at molecular level and thus contribute towards personalized medicine by helping clinicians to provide adequate diagnosis and treatments.

Increased dNTP pools rescue mtDNA depletion in human POLG-deficient fibroblasts.

Polymerase γ catalytic subunit (POLG) gene encodes the enzyme responsible for mitochondrial DNA (mtDNA) synthesis. Mutations affecting POLG are the most prevalent cause of mitochondrial disease because of defective mtDNA replication and lead to a wide spectrum of clinical phenotypes characterized by mtDNA deletions or depletion. Enhancing mitochondrial deoxyribonucleoside triphosphate (dNTP) synthesis effectively rescues mtDNA depletion in different models of defective mtDNA maintenance due to dNTP insufficiency. In this study, we studied mtDNA copy number recovery rates following ethidium bromide-forced depletion in quiescent fibroblasts from patients harboring mutations in different domains of POLG. Whereas control cells spontaneously recovered initial mtDNA levels, POLG-deficient cells experienced a more severe depletion and could not repopulate mtDNA. However, activation of deoxyribonucleoside (dN) salvage by supplementation with dNs plus erythro-9-(2-hydroxy-3-nonyl) adenine (inhibitor of deoxyadenosine degradation) led to increased mitochondrial dNTP pools and promoted mtDNA repopulation in all tested POLG-mutant cells independently of their specific genetic defect. The treatment did not compromise POLG fidelity because no increase in multiple deletions or point mutations was detected. Our study suggests that physiologic dNTP concentration limits the mtDNA replication rate. We thus propose that increasing mitochondrial dNTP availability could be of therapeutic interest for POLG deficiency and other conditions in which mtDNA maintenance is challenged.-Blázquez-Bermejo, C., Carreño-Gago, L., Molina-Granada, D., Aguirre, J., Ramón, J., Torres-Torronteras, J., Cabrera-Pérez, R., Martín, M. Á., Domínguez-González, C., de la Cruz, X., Lombès, A., García-Arumí, E., Martí, R., Cámara, Y. Increased dNTP pools rescue mtDNA depletion in human POLG-deficient fibroblasts.