Research Projects

Background: Approximately 11% of all genetic diseases are caused by premature stop codons resulting from nonsense mutations (1). These mutations introduce aberrant termination signals that prematurely halt translation, leading to the production of truncated, non-functional proteins or complete loss of protein expression. Such disruptions can result in severe clinical manifestations, as seen in disorders like β-thalassemia. Currently, there are no curative treatments for these conditions (2). Problem statement: Certain molecules, such as Ataluren and antisense oligonucleotides, can promote translational readthrough of premature stop codons, thereby restoring full-length protein expression and offering therapeutic potential. However, the efficacy of these agents varies widely because of which none of such agents are approved for clinical use, yet. Some premature stop codons are inherently "leaky," allowing low levels of natural readthrough and resulting in milder clinical phenotypes. In contrast, others exhibit strong termination signals, leading to more severe manifestations. Unfortunately, the readthrough potential (or termination efficiency) of a given stop codon is currently unpredictable. It is strongly influenced by the nucleotide context surrounding the stop codon, particularly the upstream and downstream sequences. While it is well established that sequence context plays a critical role in determining termination efficiency, the precise rules governing this process remain elusive. Moreover, the substantial clinical heterogeneity observed among affected individuals further complicates prognosis and early-stage assessment, as the underlying determinants of variability are still not fully understood. Hypothesis: Based on the results of research from our lab and other groups, we hypothesize that the sequence context surrounding premature stop codons encode specific rules that govern translation termination efficiency, thereby influencing clinical severity and efficacy of readthrough-based treatment. To uncover these rules, we propose an integrative approach combining machine learning with experimental validation. Objectives: 1. To identify the nucleotide sequence rules that determine the efficiency of translation termination at a stop codon. Approach: Computational analysis of ribosome profiling data using machine learning 2. To experimentally validate the nucleotide sequence rules identified by computational methods Approach: Using standard readthrough assays based on reporters in mutant cells. Anticipated clinical impact: Uncovering the underlying rules that determine the termination efficiency of disease-causing premature stop codons could have significant clinical implications. First, it would enable the identification of patients who are likely to benefit from readthrough-inducing therapies, allowing for more personalized and effective treatment strategies. Second, it could allow clinicians to predict disease prognosis based solely on the location and context of the nonsense mutation, even at an early stage. Together, these advances have the potential to transform the clinical management of genetic disorders caused by nonsense mutations, shifting the paradigm toward more precision-based care.