Novel Translational Methodologies

  1. What's It All About?
  2. Who is Responsible?
  3. Tell Me More.
  4. Resources.

The diversity of our GHUCCTS portfolio is built upon the strengths and resources of the collaborative team who come together to address ideas and challenges that only as collaborative efforts are realistic to tackle in an efficacious manner. We believe this has been conceived from the genuine spirit of the goals of the CTSA program from its inception, and represents team science at its best. The current NTM portfolio includes 8 projects, each with a thrust of seeding high impact, moderate risk interdisciplinary, collaborative projects with the objective of developing novel technologies that will advance current standards in medicine with new methodologies.


Project #1. Structure-Based Drug Discovery in the Petascale and Exascale Supercomputing Age.

Jerome Baudry, Jeremy Smith, Sally Ellison; Oak Ridge National Laboratory
Milton Brown, Siva Dakshanamurthy; Georgetown University

The return of investment for this project has been outstanding. The joint GU-ORNL effort has: submitted 9 grants; generated over $3 million in funding; built 15 collaborations including with other CTSAs; resulted in 4 publications (50-53) and 23 presentations at national and international conferences; been recognized by BBC and NPR for their work; developed 2 software tools for High Performance Computing; and created an HPC new undergraduate level course specifically on conducting HPC docking (one of their URM STEM graduate students was awarded a prize by the American Chemical Society). With limited seed funding from GHUCCTS, this team has made a tremendous impact in the world of predictive pharmacology utilizing HPC; a small business venture was born from this seed project and is well on its way to impacting the science of predicative pharmacology.

Project #2. Hybrid Nanostructured/Microstructured Platforms for the Enrichment of Circulating Tumor Cells.

Tim McKnight; Oak Ridge National Laboratory
Minnetta Liu; Georgetown University

This seed project has evolved in an exciting fashion, from using a microfluidics technology platform developed to capture and characterize circulating tumor cells into a rapid and potentially field deployable diagnostic tool for detection of RNA viral infection (e.g., Ebola, Dengue, Marburg, Hepatitis C, and others). This diagnostic assay is anticipated to cost less than $1 per test and is nearly reagentless. As a functional assay of viral replication, the approach will be more resilient to rapidly mutating viruses, which may be missed by conventional genotypic and antigen-based approaches. For Dengue and HepC, and possibly for Ebola, this diagnostic targets monocytes, isolated from a pinprick of blood. Active interest in the approach has been expressed by the CDC, NIH, Gates Foundation, DOD, and DOE, and a licensing agreement for the technology is currently being negotiated by UT-Battelle. The PI has developed over 13 collaborations, including two with other CTSAs (Boston University and University of Texas Medical Branch), and this project has generated over 400K in funding. UT-Battelle has applied for a U.S. non-provisional patent for this technology (#20140272943).

Project #3. Systematic Characterization of the Endogenous Cohesion Complex in Untransformed Human Cells.

Chongle Pan, Ph.D.; Oak Ridge National Laboratory
Todd Waldman M.D., Ph.D.; Lombardi Cancer Center, Georgetown School of Medicine

The first common cause of aneuploidy in human cancer was discovered in 2011 by Dr. Todd Waldman – inactivating mutations of the STAG2 gene, which encodes a subunit of the poorly-characterized cohesin complex (Science 333:1039, 2011). The multi-protein cohesin complex encircles sister chromatids, keeping them aligned until the cohesin complex is degraded by separase during the metaphase/anaphase transition, leading to the separation of sister chromatids into the two daughter cells. The cohesin complex is known to be composed of (at least) nine discrete proteins – five proteins that compose the cohesin ring (SMC1, SMC3, Rad21, and STAG1/2), and four proteins that are accessory factors thought to regulate the proteins in the ring (WAPAL, Pds5A/B, sororin). The protein composition and the biochemical features of the cohesin complex have been defined mostly in yeast; few experiments have been performed in human cells. In this project, Dr. Pan and Dr. Waldman combine state-of-the-art genetic approaches with high-resolution mass spectrometry to identify the completeendogenous protein composition of the cohesin complex, as well as identify post-translational modifications present in each member of the cohesin complex. We have successfully performed Endogenous Epitope Tagging (EET) and subsequent proteomics on three members of the cohesin complex – STAG1, STAG2, and Pds5A. In each case, affinity purification of the EET-modified protein led to co-purification of all known members of the cohesin complex. In addition, purification of STAG1 (but not STAG2) led to co-purification of transcription factors including key components of the Mediator complex. Purification of Pds5A led to co-purification of a number of novel putative interactors, including an extremely robust co-purification of two lipid biosynthesis proteins – long chain fatty acid CoA ligase 5 (ACSL5) and fatty aldehyde dehydrogenase (ALDH3A2). Of note, somatic mutations of ACSL5 of unknown functional significance have been reported in colon cancer.

Project #4: Elucidation of the Structure of Disordered C-terminal Domain of B-catenin by Integrating Neutron Scattering, Atomistic and Statistical Inference

Arvind Ramanathan; Oak Ridge National Laboratory
Salim Shaw; Georgetown University

Intrinsically disordered proteins (IDPs) represent a novel class of proteins that challenge the traditional protein structure-function paradigm; they lack stable tertiary structure under physiological conditions but undergo synergistic folding with substrate-specific conformations when bound. IDP dysfunction is implicated in several diseases including cancers, neurodegenerative and cardiovascular disorders, and diabetes. IDP binding and their aggregation mechanisms can help design self-assembly of supra-molecular structures. Understanding the biophysical mechanisms of how IDP conformational diversity influences its function is one of the forefront problems in structural biology with impact in health, environmental and energy biosciences.

The disordered C-terminal domain (CTD) of β-catenin (residues 664-781 following the ARM repeat domain) represents one such IDP, which functions as transcriptional co-activator. As a key regulator of the wnt signaling pathway, β-catenin is activated in many human cancers, including hepatocellular carcinoma and colorectal cancer. β-catenin is a 781-residue protein, with a N-terminal regulatory domain, followed by a central region of 12 armadillo repeats (ARM) and a structurally flexible CTD. The CTD of β-catenin is known to be a transcription co-activator and its transcriptional activity is largely responsible for β-catenin’s oncogenic effect. Although the structure and function of several of β-catenin ARM interactions have been characterized, similar structural information is not available for the CTD. However, no structural information for this IDP exists and it is currently unknown (1) whether it interacts with the ARM domain and (2) whether it interacts with other binding partners of β-catenin. Absence of the CTD results in β-catenin being transcriptionally inactive and truncation of residues from 723-781 results in a reduction of β-catenin activity by nearly 80%. Given that β-catenin’s transcriptional activity is largely responsible for its oncogenic effects, understanding the structural details of whether and how the C-terminal domain folds is important to design protein mutations and/or small molecule inhibitors that can target various types of cancers, including colorectal and hepatocellular cancer. Drs. Stephen Byers and Salim Shah at Georgetown University Medical Center and Lombardi Cancer Center have been interested in understanding the structure and interaction of the CTD of β-catenin.

In this project, we will investigate the structural flexibility of the intrinsically disordered CTD of β-catenin protein. Since the CTD is highly flexible and is intrinsically disordered, its full conformational range is not observable by any one structure determination technique. Therefore, we envision developing integrated experimental and computational techniques to elicit high-resolution structural details of intrinsically disordered protein ensembles. Specifically, we will (1) develop methods to prepare selectively deuterated C-terminal β-catenin and utilize SANS contrast variation techniques to refine high- resolution conformational ensembles; (2) build parallel ensemble simulation strategies on heterogeneous computer architectures to generate millisecond timescale atomistic simulations; and (3) design Bayesian inference techniques for statistical characterization of IDP conformational ensembles. The proposed work will improve our understanding of the structure and function of the β-catenin CTD, especially in its ability to activate transcription and guide the design of novel small-molecules to inhibit β-catenin function in oncogenesis. The integrated framework proposed here would have broad applications for SANS analysis of flexible and order-disorder phenomena in polymer and materials science research.

Project #5. Developing Computer-aided Support Algorithms to Mitigate the Disruptiveness of Interruptions in Radiology

Georgia Tourassi; Oak Ridge National Laboratory
RM Ratwani; MedStar Health Research Institute

Ten years ago the Agency for Healthcare Research and Quality (AHRQ) published a report on healthcare working conditions that highlighted the patient safety risks associated with task interruptions. Interruptions are particularly detrimental in radiology; radiologists must examine numerous images and make complex interpretations that require focused attention resulting in significant cognitive demands. In radiology, interruptions not only increase the time required to examine an image, which reduces efficiency, interruptions may also increase the likelihood of perceptual or diagnostic errors resulting in serious threats to patient safety.

Interruptions during the search process have been shown to eradicate search memory, so it follows that the interruption of radiologists while they are interpreting images and searching for abnormalities may be particularly risky. To understand the cognitive processes underlying visual search researchers have examined radiologist’s eye movements. Eye tracking technology provides detailed information on where radiologists are looking and provides insight into the cognition of the radiologist. While eye tracking methods have been used to understand how radiologists search images, little is known about how interruptions influence the search process.

The overarching goal of this proposal is to leverage eye tracking to develop a process model of how radiologists handle interruptions during the search process and to use this model to develop algorithms that can facilitate resumption and reduce the likelihood of perceptual errors. In a joint effort between MedStar Health’s National Center for Human Factors in Healthcare (NCHFH) and the Biomedical Science and Engineering Center (BSEC) at the Oak Ridge National Laboratory, experts in cognitive psychology, medical imaging, and computer science have formed a collaboration to achieve these goals. While little is known about how radiologists resume an interrupted search and the types of errors that might result from interruptions, research scientists at the NCHFH have extensive experience studying interruptions and examining both the memory and perception processes during the cycle of search, interruptions, and resumption. In addition to developing models of these processes the NCHFH team has also worked to develop mechanisms to facilitate resumption which have reduced time costs and lowered error rates. The BSEC team has extensive experience working with radiologists to develop cognitive decision support mechanisms and have developed several different computer-based support algorithms to facilitate detection. We will leverage the previous work of the NCHFH and BSEC to create a synergistic collaboration. Novel methods will be developed to mitigate the disruptiveness of interruptions and
thus increase patient safety in radiology.

Project #6. Integrated Data, Analytics, Modeling and Simulation (IDAMS): Elucidating the Pathway to Eliminate Hepatitis C Viral Infection

Mallikarjun Shankar, Ozgur Ozmen, Arvind Ramanathan; Oak Ridge National Laboratory
Dawn Fishbein, Coleman Smith, Nawar Shara, Ian Brooks; MedStar Health Research Institute

The goal of this project is to model and simulate the cascade of care involving the treatment of patients with hepatitis-C to determine factors influencing positively and negatively the end treatment goals. The project utilizes the rich clinical experience of GHUCCTS in combination with ORNL’s big data analytics and counterfactual simulation of IDAMS-HC to predict outcomes with computer simulated trials.

Project #7. Real-Time Adaptive Motion Correction for Proton Therapy

James S. Goddard, Shaun S. Gleason, Stephen D. Miller, Dustin R. Osborne, Niek Schreuder, Mark Artz; Oak Ridge National Laboratory
Anatoly Dritschilo, Sean P. Collins; Georgetown University

This project capitalizes on the unique resources of GHUCCTS institutions. Georgetown University/Lombardi Cancer Center built a new proton-beam facility that became operational in 2017, and Provision Health in Knoxville TN had just completed their facility and began treating patients in 2015. This allowed the assembly of an experienced multi-site and multidisciplinary team that can accomplish the primary goal of improving proton beam targeting by utilizing the super-computer resources of ORNL. The collaboration utilizes existing clinical data to develop innovative algorithms to better target and deliver proton therapy. This project is in the final stage of data analysis, and has developed several abstracts and posters.

Project #8. Genomic and Functional Characterization of the Microbiota of Obese African Americans

Chongle Pan; Oak Ridge National Laboratory
Hassan Brim; Howard University

In this study, the gut microbiota of an obese group (BMI: 35.0–39.9) with those of a lean group (BMI: 18.5–24.9) will be compared. Both groups will be made up of 10 African American subjects who are 18 to 22 years old, female, and generally healthy. Stool samples from the recruited subjects are collected on day 0 before any intervention.  This will inform on the gut microbiota of obese and lean individuals before the experiment. To control for the effect of diet on the gut microbiota, the same meal plan is provided for the subjects on both groups over a period of 10 days. After the first 7 days, we expect the microbiota of all subjects will be fully shifted to being sustained by the provided balanced diet and the energy intake and overall metabolism of both obese and lean subjects will converge on those based on the same diet. Fecal samples are collected on days 8, 9 and 10.  The integrated metagenomics, metatranscriptomics and metaproteomics analyses of these fecal samples will help to identify the community composition and functional structure of the microbiota in obese and lean subjects. Through comparisons of the obese and lean subjects, these analyses will identify taxonomic groups and metabolic signatures that are associated with or indicative of obesity. The results are expected to yield a holistic genes-to-systems view of the taxonomic and functional underpinnings of the microbiota associated with obesity. Specifically, the expected results are: 1) identification of genes that are candidates for biomarker or significant risk factor for obesity; 2) identification of taxonomic groups and associated metabolic pathways that show differences in obese subjects; and 3) identification of gene content in obese versus lean subjects. Currently this project has completed sample collection in 40% (8/20) of the planned enrollment. The extracted DNA samples are presently being sequenced for metagenomics to assess similarities and differences in bacterial genes’ functions. RNA extracts are presently undergoing Next Generation Sequencing for transcriptomic analysis to assess the dynamics of gene expression in the two groups of subjects (Obese vs. Lean) under different diets.