Driving Biomedical Projects

DBP1) ¹³C NMR for improved metabolite identification, Arthur S. Edison, University of Georgia

Metabolomics studies in biomedical and biological research are growing rapidly, fueled by advances in analytical technology that provide exceptional sensitivity and resolution. NMR is an emerging technique, but is often over-shadowed by LC-MS. The major advantage of LC-MS is it captures a very large number of “features”, here defined as reproducible peaks in an LC-MS or NMR spectrum. The primary shortcomings of LC-MS are difficulty in quantification, relatively high technical variability, and ambiguities in feature identification. In contrast, NMR is less sensitive and yields fewer features but can provide absolute quantification, has low technical variability, and can provide complete covalent structures and relative stereochemistry for unknown molecules. NMR is also non-destructive to the sample. Thus, NMR and LC-MS are highly complementary technologies, but it remains challenging to integrate them.

The NIH Metabolomics Common Fund established five compound identification cores (CIDCs) to develop new approaches for identifying metabolites. Our CIDC incorporates analytical measurements by both LC-MS and NMR with genetic pathway analysis and quantum mechanical calculations. We use the model organism Caenorhabditis elegans, with thousands of available mutants and hundreds of fully sequenced natural isolates freely available, to compare a known mutant in a metabolic pathway with a reference strain. Every sample is analyzed by LC-MS and NMR, and we select unknown features that have the biggest changes between the strains. These measurements create an extremely rich dataset that can be used for the duration of the study and by others to facilitate compound ID. Fractionation allows us to simplify the association of a specific NMR feature with a corresponding LC-MS feature and enables compound ID to be performed like traditional natural products chemistry. This enables discovery of related unknown nodes. Importantly, we can link the NMR and LC-MS data from our fraction library with the GNPS network. For ¹³C, we will add 2D ¹³C J-resolved spectroscopy and “gold standard” carbon-carbon correlation experiments that will directly map the covalent structure of all the metabolites in our library.

DBP2) Imaging Hepatic Gluconeogenesis with Hyperpolarized Dihydroxyacetone, Matt Merritt, University of Florida

There is an urgent need for development of metabolic imaging methods that are sensitive to nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH), and to the transi- tion linking the two pathophysiological states. NAFLD and NASH are now a worldwide epidemic, with some estimates of NAFLD prevalence as high as 24% of the world population. NASH is expected to surpass hepatitis as the number one cause of liver transplant in the United States within the next 5 years. Over the next 10 years, this disease is projected to be a 1 trillion dollar burden to the healthcare system.

While imaging of fibrosis is somewhat diagnostic of NASH progression, there is no metabolic imaging technique that is sensitive to the inflammation endemic to the transition of NAFLD to NASH. Current step-wise paradigms for identifying NASH lack the sensitivity to correctly classify early development. When NASH is diagnosed, clinical management of the disease changes dramatically, becoming much more expensive. Development of a metabolic imaging method for diagnosis and staging of NASH would significantly enhance healthcare practice, with prospects for improving patient care and decreasing costs. We are developing magnetic resonance based metabolic imaging using hyperpolarized ¹³C and thermally polarized ²H-labeled substrates. The HTS probes for liquid state NMR supports our demands for high resolution ¹H and ¹³C NMR as methods for confirming the results measured in vivo.

DBP3) RNA Recognition by Intrinsically Unstructured Proteins, Robert Silvers, Florida State University

Post-transcriptional gene regulation at the mRNA level is essential for cellular function. A fundamental aspect of post-transcriptional gene regulation is the formation of ribonucleoprotein (RNP) complexes involving RNA-binding domains (RBDs).Despite a variety of studies in several related systems, there is still a fundamental lack of knowledge regarding the molecular basis of RNA recognition. A set of proteins, the La-related protein (LARP) superfamily, has recently emerged as important players in gene regulation at mRNA level. Their importance is underlined by the fact that most known eukaryotic organisms have retained LARPs in their genomes.

This project focuses on characterizing the structural mechanism of RNA recognition by IURs found in LARPs. These biomolecules are challenging systems due to low chemical shift dispersion in both RNA and IURs. Therefore, this project is highly dependent on the development of novel approaches in NMR spectroscopy, particularly ¹⁵N and ¹³C detected experiments. NMR spectra of RNA as well as IURs/IUPs suffer from poor chemical shift dispersion leading to significant signal overlap, particularly with increasing molecular size. While the dynamic character of these biomolecules make NMR spectroscopy uniquely suited for investigations at atomic resolution, the poor chemical shift dispersion hampers their spectral analysis and investigation tremendously. The development of ¹³C-optimized high sensitivity solution (HTS) probes at high magnetic fields (>800 MHz) is of considerable use to enable the investigation of these highly dynamic protein and RNA systems suffering from low chemical shift dispersion. Furthermore, investigating the role of proline cis/trans isomerization and protein phosphorylation will be enabled by the development of ¹³C-optimized HTS probes.

DBP4) Structural and proton dynamics of pyridoxal-5′-phosphate dependent enzymes, Leonard J. Mueller, UC Riverside

Knowledge of active site protonation states is central to a full mechanistic understanding of enzyme catalysis. Accurate knowledge of proton locations is critical, for example, to establishing the mechanism of acid-base catalysis and as input for molecular dynamics simulations, molecular docking routines, and structure-based drug design. Over the last five years, we have developed the tools of NMR crystallography and applied them to multiple intermediates in the TS catalytic cycle. Key to the success of our approach has been the measurement of site-specific chemical shifts for kinetically-competent intermediates within this 143 kDa protein complex. We enrich specific ¹³C, ¹⁵N, and ¹⁷O NMR spins into the active site with mixed labeling schemes and double resonance NMR experiments to unambiguously assigning chemical shifts.

DNP experiments at ~100 K increase sensitivity and indefinitely stabilize the more transient intermediates at the NHMFL, where we see sensitivity increases of ~60 fold with DNP. These enhancements permit the efficient measurement of chemical shifts and chemical shift tensors. Working at cryogenic temperatures also permits a test for tautomeric exchange as the significantly lower temperature can dramatically alter the relative populations of tautomers. Somewhat unique to our work is the use of ¹⁷O quadrupole central transition (QCT) NMR spectroscopy, including on the SCH magnet. ¹⁷O is not yet considered a high-resolution probe for biomolecular NMR, but takes advantage of the narrowing of the central transition as the size of the protein increases; we have found that ¹⁷O QCT NMR is remarkably robust and an important reporter on the electronic structure of the substrate-cofactor complex. Based on our experience, for ¹⁷O NMR, there is no substitute for the SCH to separate out the large contribution of the dynamic frequency shift to the observed resonance position and recover the isotropic shift. This requires spectra at 3-4 magnetic fields and has the added benefit that chemical shift anisotropy (CSA) and quadrupole tensors can also be extracted. The larger the span of the field strengths, the more accurately the NMR parameters can be determined. By adding the SCH, the error bars on the CSA are improved by a factor of 5 and those on the isotropic CS by a factor of 10!

DBP5) Molecular organization and drug responses of fungal cell walls, Tuo Wang, Louisiana State University

Life-threatening fungal infections occur to 2-3 million people worldwide every year. The insufficient efficacy of commercially available drugs, the substantial rise of azole-resistant strains, and the extensive application of immunosuppressive agents call for the development of novel antifungal compounds with novel modes of action, for example, those targeting the cell wall polysaccharides absent in humans. It remains difficult to characterize the complex structure of cell walls with high resolution, despite the well-studied carbohydrate composition. Recently, my lab has explored the use of solid-state NMR (ssNMR) for investigating polysaccharide structure and cell wall organization. These studies provided the first molecular-level model of fungal cell walls and revealed an overlooked but prominent role of α-glucans in the cell wall structuration. This project focuses on characterizing the structural remodeling mechanism of the fungal cell wall and its relevance to fungal virulence and drug resistance.

The sensitivity enhancement provided by MAS-DNP and the recently optimized AsympolPOK- family biradicals will allow us to efficiently detect the lowly populated molecules, for example, the remaining molecules that could not be completely removed by the inhibition of antifungal drugs. DNP also allows us to visualize the packing interface between chitin and glucans, and between glycans and pigment to obtain an in-depth understanding of the 3D organization of fungal cell walls. The ³¹P-detection of the new MAS- DNP probe will be used to characterize wall-associated phospholipids. Coupling fast MAS (40 kHz) with DNP will allow us to efficiently detect those biomolecules with large chemical shift anisotropy, for example, the aromatic-rich pigments in fungal cells. Second, the ultrahigh-field SCH magnet (35.2 T) provides sufficient resolution for deciphering the structural polymorphism of cellular carbohydrates as induced by the conformational distribution and linkage diversity. Representative 2D ¹³C-¹³C/¹⁵N correlation spectra collected on the 1.5 GHz SCH magnet have shown sharp lines with 0.3-0.5 ppm ¹³C linewidths, allowing us to resolve the heavily overlapped signals and obtain atomic resolution on the complex carbohydrates. The technologies are applicable to the under-investigated carbohydrates in other biosystems, such as bacterial, mammalian, and human cells.

DBP6) Functional amyloid formation in Streptococcus mutans, L. Jeannine Brady, University of Florida

Amyloid represents a fibrous cross β-sheet quaternary structure of ordered peptide or protein aggregates that demonstrate common biophysical properties, and many microorganisms produce functional amyloids that influence biofilm development. We showed amyloid is present in human dental plaque and produced by both lab strains and clinical isolates of Streptococcus mutans. S. mutans is a major etiologic agent of human dental caries, the most common infectious disease in the world.

This project focuses on characterizing the structures of S. mutans amyloidogenic proteins important to biofilm development to understand structural transitions underlying amyloid aggregation in order to develop therapeutic agents able to modulate biofilm formation. The significantly lower concentration of these proteins in their native environment puts such measurements beyond the reach of conventional ssNMR. Sensitivity gains of DNP provide a unique opportunity to undertake these studies. Initial data collected on isotopically enriched C123 in fibril form and attached to cell walls exhibited an SNR gain of ~20 with DNP using AMUPol as the PA making 2D and 3D NMR experiments feasible. However, the spectral resolution for these samples was poor (line widths of 4-6 ppm). A significant effort of this DBP will focus on improving sample preparation to minimize DNP NMR resonance linewidths for C123 in fibril form and when attached to cell walls. The VT range of the proposed DNP-MAS probes combined with fast- MAS will enable us to explore temperature and MAS regimes that may provide an optimal compromise between sensitivity and spectral resolution. For characterizing intact biofilms containing C123 and AgA, we will test new methods of introducing stable biradicals for DNP.