Date Published: February 9, 2017
Publisher: Public Library of Science
Author(s): Michael T. Zimmermann, Raul Urrutia, Gavin R. Oliver, Patrick R. Blackburn, Margot A. Cousin, Nicole J. Bozeck, Eric W. Klee, Freddie Salsbury.
Variants in the TGFBR2 kinase domain cause several human diseases and can increase propensity for cancer. The widespread application of next generation sequencing within the setting of Individualized Medicine (IM) is increasing the rate at which TGFBR2 kinase domain variants are being identified. However, their clinical relevance is often uncertain. Consequently, we sought to evaluate the use of molecular modeling and molecular dynamics (MD) simulations for assessing the potential impact of variants within this domain. We documented the structural differences revealed by these models across 57 variants using independent MD simulations for each. Our simulations revealed various mechanisms by which variants may lead to functional alteration; some are revealed energetically, while others structurally or dynamically. We found that the ATP binding site and activation loop dynamics may be affected by variants at positions throughout the structure. This prediction cannot be made from the linear sequence alone. We present our structure-based analyses alongside those obtained using several commonly used genomics-based predictive algorithms. We believe the further mechanistic information revealed by molecular modeling will be useful in guiding the examination of clinically observed variants throughout the exome, as well as those likely to be discovered in the near future by clinical tests leveraging next-generation sequencing through IM efforts.
The transforming growth factor-β (TGFβ) superfamily of signaling proteins is comprised of a diversity of TGFβ receptors, TGFβ ligands, activins, inhibins, and bone morphogenic proteins which collectively regulate a broad spectrum of biologic functions including wound healing, cellular differentiation, and deposition of extracellular matrix proteins [1–3]. Given their role in mediating embryonic development and maintaining the homeostasis of most tissues, the proper function of these signaling proteins is vital for all multicellular organisms. Genetic variants within these molecules or the downstream proteins that mediate and integrate their signals have been shown implicit with human disease including developmental disorders, vascular diseases, and cancer [2, 4–6]. Technological advances in DNA sequencing have fostered a new era of Individualized Medicine (IM), which among other effects is increasing the rate at which new variants in these pathways are being discovered and associated with disease phenotypes . While the total number of known TGFβ family variants has increased, those characterized by experimental information enabling conclusions as to pathogenicity or the lack thereof are substantially fewer. While well designed functional studies provide a high level of confidence in classifying a variant as pathogenic , they are typically costly and time consuming, thus limiting wide-spread use to systematically characterize variants of unknown significance (VUSs). Subsequently, a need exists for higher-throughput computational and experimental methods to evaluate the functional impact of variants at the molecular, biochemical, cellular, and organismal levels.
We aim to gain insights into the effects of amino acid variants on the TGFBR2 kinase domain and to provide mechanistic interpretations. Using a molecular model of the protein structure to predict changes in stability and dynamic behavior upon mutation, we present the case for greater application of these methods. Hypothesizing that variants leading to more severe structural effects will be evidenced by alterations in folding energy, local flexibility, regulatory loop positioning, or loss of important structural contacts including ligand binding site conformation, relatively short simulations were used. We believe that the widespread adoption of these methods to the prioritization and interpretation of clinically observed variants within the context of IM initiatives is likely to have a significant positive impact on the biomedical community.
The interpretation of novel variants in the TGFBR2 kinase domain is important for furthering our understanding of several human diseases. This task has increased in scope due to the widespread application of clinical next generation sequencing, which is uncovering disease-associated variants in many proteins at a faster rate than ever before. Consequently, in this work, we evaluated the utility of short MD simulations for assessing the potential impact of variants, revealing various mechanisms by which they may lead to functional alteration. Our results also underscore that the function most likely affected by each variant may be allosteric in nature. Differentiating which variants may lead to dysfunction and the mechanism underlying these alterations is not possible from current sequence-based analysis. Therefore, we believe that the mechanistic information revealed by molecular modeling will be critical for the examination of variants discovered by clinical sequencing tests, particularly for individual patient cases as resulting from ongoing IM efforts. Hence, we are optimistic that the methodology and information gathered in this study will have clinical utility.