University of Kansas
Liskin Swint-Kruse, Antonio Artigues, Aron Fenton, Joseph Fontes, Bruno Hagenbuch, John Karanicolas, Alexey Ladokhin, Audrey Lamb, Paul Smith
Protein engineering and personalized medicine would be significantly advanced if functional outcomes from amino acid mutations could be more accurately predicted. Although many computer algorithms have been developed for that purpose, they have met with limited success. Indeed, several common assumptions underlying these algorithms do not apply to a particular group of amino acid positions that are altered during evolution (nonconserved positions). Among the characteristics of these positions, amino acid substitutions can dial function “up” or “down,” in a manner analogous to changing the dials on a stereo. Since these “rheostat” positions appear to be wide-spread in many proteins, a team of investigators at the University of Kansas Medical Center, the University of Kansas – Lawrence, Kansas State University and Fox Chase Cancer Center seeks to develop a library of rheostat positions to guide and benchmark future computations. By comparing and contrasting key nonconserved positions in proteins of distinct structural classes (globular soluble, integral membrane, and intrinsically disordered), the investigators will develop general methods to identify rheostat positions within a protein. By generating biophysical data for representative proteins that have mutations at rheostat positions, the investigators will formulate new amino acid substitution rules for these locations. Results from these studies will provide the critical information needed to improve mutation prediction algorithms.
University of Washington
Judit Villen, Stanley Fields, William Noble
New DNA-sequencing methods have led to the discovery of millions of mutations that change the encoded protein sequences, but the impact of nearly all of these mutations on protein function is unknown. Current approaches that analyze the effects of mutations are inadequate, as they rely on computational predictions whose accuracy is questionable, or they interrogate only one protein at a time and thus would require hundreds of years to interpret existing data. To overcome this critical bottleneck, a team of investigators at the University of Washington seeks to develop Limited Mistranslation Mutagenesis (LMM). LMM is a technology that combines non-genetic mutagenesis with biochemical assays and mass spectrometry to assess the functional effects of amino acid substitutions on a proteome-wide basis in a timeframe of just days. LMM will generate the first mutational sensitivity maps that span the entire proteome, revealing deleterious amino acid substitutions that directly impact protein functions. These maps will provide an invaluable resource for biologists, serving as an essential companion guide to genome sequences. Development of this technology and resource will impact basic biology, by accelerating our knowledge of how proteins act; protein engineering, by guiding the design of novel proteins with enhanced properties; and genomics, by allowing the interpretation of millions of human mutations and establishing their relevance to disease..
Colorado State University
Timothy J. Stasevich, Brian Munsky
Fort Collins, CO
Translation of RNA to protein is a tightly regulated process that is fundamental to life. It is also one of the first processes that viruses hijack to replicate themselves in host cells. The molecular details of this hijacking remain poorly understood, as does the regulation of normal translation in vivo. Despite its undisputed importance in every biological and biomedical system, translation has never been measured in real-time within living cells. Two Colorado State University investigators will combine state-of-the-art single-molecule microscopy with RNA and protein tags and sophisticated computational modeling to quantify the expression and translation of a multitude of different RNA transcripts in real-time in living cells. The technological advances developed in this project will also make it possible to discover and quantify ribosomal frameshifting, a mechanism which allows distinctly different proteins to be translated from the same RNA strand and which is exploited by viruses for their replication. These results will help the biomedical community to better understand, control and predict the process of translation as it normally occurs in cells and when cells are infected by viruses.
FanWang, Kafui Dzirasa
Consciousness is a reversible brain state characterized by awareness and perception of one’s environment and self. Despite centuries of studies in neuroscience and medicine, the exact neural pathways and processes that reversibly switch the brain between conscious and unconscious states remain largely unknown. To begin to solve this long-standing puzzle, two Duke University investigators plan to examine the neural mechanisms whereby general anesthesia suppresses consciousness. Supported by preliminary results, the team hypothesizes that specific populations of anesthesia-activated neurons (AANs) are necessary and sufficient to initiate and maintain the unconscious brain state. The investigators plan to identify and manipulate AANs using a highly innovative technology recently developed in their laboratories, which allows them to specifically control the activity of AANs. These manipulations will be combined with large-scale multisite in vivo electrical recordings simultaneously from the output targets of AANs and from other cortical and sub-cortical regions. This will enable the analysis of the precise sequences of causal events that trigger the brain to transition between conscious and unconscious states. Finally, the investigators will use closed-loop optogenetic manipulations to instill putative conscious activity patterns into selected AAN targets to override the effects of anesthesia. Together, these studies are expected to unlock the neural gate to consciousness.
University of Virginia
Eyleen O'Rourke, Nathan Lewis, George Church
Multicellular organisms are ultimately complex genetic networks integrated in ascending levels of complexity (cells, tissues, organs). Human diseases are network diseases. Even family members exposed to the same agent or carrying the same mutation can develop mild to severe symptoms because the response of their individual gene networks is different, and those differences determine the severity of the disease. Thus, multilevel functional gene network reconstruction and modeling could transform our understanding of biology and our ability to discover disease mechanisms and rationalize the design and testing of new treatments. However, the technologies necessary to generate multilevel animal models have not yet been developed. Using the nematode C. elegans as their model system, a team of investigators from the University of Virginia, the University of California San Diego, and the Wyss Institute plans to develop the technologies to measure gene expression and fat metabolism that will make possible the first single-cell resolution metabolic model for an entire living animal. The tools and computational frameworks to be established will enable modeling of vertebrate animals, human tissues and organs, the microbiome, or any other complex communities of cells with critical multicellular structure. The detailed landscape of gene expression and function across a complex multicellular organism that will emerge from this work has never been accomplished before, and it is expected to open new doors in biology and biomedicine.
University of Washington
Cole Trapnell, Jay Shendure
A decade after the completion of the human genome project, the function of the vast majority of the 20,000 human genes remains largely unknown. Although forward genetic screens can coarsely implicate hundreds of genes in having some role in a phenotype, this almost always is the end of the story, rather than the beginning. There are at least two critical challenges to experimentally analyzing a gene. First, the phenotype measured upon perturbing a gene usually is based on cell growth or survival and therefore provides few or no clues about that gene’s molecular role. Second, perturbing a single gene often has no phenotypic effect because other genes buffer against the change. Two investigators from the University of Washington plan to develop a new paradigm, combinatorial forward genetics (CFG), which addresses both challenges at once. CFG aims to broadly capture the molecular consequences of perturbing thousands of genes in a multitude of combinations by evaluating the resulting signatures of cell states in a high-throughput manner. To analyze the data from the CFG experiments, the team will implement “deep neural algorithms” which can recognize features of increased complexity. While the risks are substantial, CFG has the potential to revolutionize the understanding of the human genome by markedly accelerating the discovery of sets of genes that functionally collaborate, e.g. in signaling pathways, as molecular machines, or for cellular reprogramming. Knowledge of such interactions will be invaluable in advancing the understanding of each gene’s role in human biology and disease.