<p>Reductions in atmospheric deposition of sulfur (S) coupled with increases in yields of <i>Glycine Max</i> (L.) Merr. (soybean) has led to S deficiencies in Indiana. Poor nodulation due to limited S, and thus a decrease in nitrogen (N) supply, restricts the yield and quality of soybean grain (i.e., protein). Sulfur is a key component of methionine and cysteine, which are important amino acids in the nutrition of foodstuffs. The objective of the first study is to improve yield and composition of soybean through various applications of N and S. Ten N+S fertility treatments were factored by 2 planting dates (early vs. late) at West Lafayette, IN in 2018 and 2019. The same 10 N+S fertility treatments were factored by 2 varieties (Asgrow 24x7 and 34x6) at Wanatah, IN in 2018 and 2019. Soybean yield increases among the N+S fertility treatments of the May 11th planting (early) were 380 to 1006 kg ha^-1 over the untreated control, with no difference within the June 5th planting (late) in 2018. Cool and wet conditions that limited mineralization of N and S from the early planting are likely the source of yield improvements. Protein concentrations were maintained and even increased with N and S treatments that were coupled with yield improvements. The Wanatah location showed that protein levels were increased with the ATS and R4+ NS treatments, while the UAN Direct treatment had the lowest protein in both varieties, suggesting that having no source of S could limit protein development. Although variety did not affect yield, fertility improved yields with the V4R3 NS, Plant NS, R3 NS, R4+ NS, and V4 NS treatments. The yield improvements that developed with these treatments is interesting because each treatment contained a source of N equaling at least 44.8 kg N ha^-1.</p><p>Secondly, the optimal rate and timing of foliar S applications were determined at a S-deficient location (La Crosse, IN) in 2018 and 2019. Three target application timings; V4, R3, and V4 + R3, were crossed with 4 rates of foliar S at 1.12, 2.24, 4.49, 6.73 kg S ha^-1 with each application. Therefore, the sequential application (V4 + R3) received a total of 2.24, 4.49, 8.96, and 13.44 kg S ha^-1. The optimal rate with 2018 yields was 4.5 kg S ha^-1 at V4 or R3; whereas, the optimal rate was 7.9 kg S ha^-1 with the sequential V4 + R3 treatment in 2019. Leaf tissue concentrations of S were nearly deficient (0.25%) post-V4 and post-R3. Higher rates of S had greater S concentrations in the leaf; furthermore, most cases resulted in a linear increase of S concentration with the rate of S applied. Foliar applications of S also reduced N:S ratio. Protein levels in 2018 increased at an equal rate for both the V4 and the R3 timings. In 2019, at a 6 lb ac^-1 rate of S the protein levels were 39.5 and 39.8% for V4 and R3 timings, respectively. Foliar S applications at V4 vs. R3 timings had little variation in yield or protein levels, thereby resulting in flexibility for application timing for growers.</p>

Proteins that have been modified by attaching them to a surface or to a polyethylene glycol (PEG) molecule can see many uses in therapeutics and diagnostics -- these unique proteins are called protein devices. Current techniques can perform these functionalizations at a specific residue on the protein, but what remains is identifying what happens to protein structure when mutated, and where to perform the attachment. Both of these issues can be examined using molecular dynamic (MD) simulations. Currently, simulations of the unnatural amino acid (uAA) mutations necessary for protein device functionalization cannot be executed, and full-protein screens of all possible protein device models have never been attempted. Results from this dissertation first employs a new model for simulating PEGylated protein devices building off of previous studies that explore where to attach functional groups. Next, many current assumptions in the community regarding ideal attachment sites are examined. Some of these factors include primary chain location, amino acid type, solvent accessibility, and secondary structure. The focus then turns to novel tertiary structure factors that could influence how well attachment locations affect overall protein device stability. The usefulness of each factor is analyzed to show what factors provided the best predictive power for a site's performance in the screen. A general heuristic is given that could aid in future screens of other protein devices to reduce compute time and quickly identify sites for experimental examination. To explore uAA mutation effects on protein structure, parameters are developed for linear moiety R-groups present in these novel amino acids. The CHARMM and CGenFF force fields currently lack parameters for most linear-angle molecular moieties. This work proposes a method that (1) develops CHARMM parameters for four small molecules that contain terminal azido and alkynyl groups using ffTK, (2) addresses linearity issues, and (3) validates ffTK results via in silico MD simulation. Dihedral analysis examines the linear-angle-containing dihedrals and compares methods for the moiety parameterization. Next, the small molecule parameters are combined with CGenFF to generate parameters for unnatural amino acid MD simulation in a protein. Finally, validation confirms the parameters derived in this work to appropriately simulate unnatural amino acids and small molecules with azido and alkynyl groups.

Proteins that have been modified by attaching them to a surface or to a polyethylene glycol (PEG) molecule can see many uses in therapeutics and diagnostics -- these unique proteins are called protein devices. Current techniques can perform these functionalizations at a specific residue on the protein, but what remains is identifying what happens to protein structure when mutated, and where to perform the attachment. Both of these issues can be examined using molecular dynamic (MD) simulations. Currently, simulations of the unnatural amino acid (uAA) mutations necessary for protein device functionalization cannot be executed, and full-protein screens of all possible protein device models have never been attempted. Results from this dissertation first employs a new model for simulating PEGylated protein devices building off of previous studies that explore where to attach functional groups. Next, many current assumptions in the community regarding ideal attachment sites are examined. Some of these factors include primary chain location, amino acid type, solvent accessibility, and secondary structure. The focus then turns to novel tertiary structure factors that could influence how well attachment locations affect overall protein device stability. The usefulness of each factor is analyzed to show what factors provided the best predictive power for a site's performance in the screen. A general heuristic is given that could aid in future screens of other protein devices to reduce compute time and quickly identify sites for experimental examination. To explore uAA mutation effects on protein structure, parameters are developed for linear moiety R-groups present in these novel amino acids. The CHARMM and CGenFF force fields currently lack parameters for most linear-angle molecular moieties. This work proposes a method that (1) develops CHARMM parameters for four small molecules that contain terminal azido and alkynyl groups using ffTK, (2) addresses linearity issues, and (3) validates ffTK results via in silico MD simulation. Dihedral analysis examines the linear-angle-containing dihedrals and compares methods for the moiety parameterization. Next, the small molecule parameters are combined with CGenFF to generate parameters for unnatural amino acid MD simulation in a protein. Finally, validation confirms the parameters derived in this work to appropriately simulate unnatural amino acids and small molecules with azido and alkynyl groups.

System L (SL), a basolateral amino acid transporter, transports large neutral amino acids (LNAAs) in a Na+-independent manner. Previously, we identified two isoforms of transporters: L-type amino acid transporter 1 (LAT1) and 2 (LAT2) and revealed their distinct substrate selectivity and transport properties. In this study, to establish more stable human LAT1 (hLAT1) and LAT2 (hLAT2) in vitro assay systems, we established mouse cell lines stably expressing hLAT1 (S2-LAT1) and hLAT2 (S2-LAT2). Real-time quantitative RT-PCR analysis revealed that S2-LAT1 and S2-LAT2 cells express hLAT1 and hLAT2 mRNAs at 20 - 1000-fold higher levels than those of endogenous mouse Lat1 and Lat2. S2-LAT1 and S2-LAT2 mediated [14C]L-leucine transport properties were measured and corresponded to results observed via Xenopus oocytes. Using these cells, the data demonstrate that hLAT1 and hLAT2 exhibit different characters in the acceptance of α-methyl amino acids and amino acid-related compounds with bulky side chains such as thyroid hormones and melphalan. S2-LAT1 and S2-LAT2 cells are expected to facilitate hLAT1 and hLAT2 substrate recognition research and contribute to drug development by providing an efficient assay system to screen for chemical compounds that interact with hLAT1 and hLAT2.

Over the past several years, the fields of synthetic biology and machine learning have demonstrated marked advances in the scale of their capabilities and the success of their applications. The work presented in this thesis focuses on the translation of recent advances in machine learning toward new applications in synthetic biology. In particular it is argued that the needs of synthetic biology researchers and practitioners are well met by a class of generative machine learning models, and that the scale of synthetic biology capabilities allows for their successful application across multiple domains of interest. In Chapter 1, a novel algorithm utilizing Markov Random Fields is used to, for the first time, design functional synthetic overlapping pairs of genes with potential applications for improved biological robustness and biosafety. In Chapter 2, motivated by a desire to extend the scope of protein sequence modeling to a greater range and diversity of protein sequences, a variant of a variational autoencoder model is used to project hundreds of millions of protein sequences into a continuous latent space with potentially useful representation features. Finally, in Chapter 3, we move beyond the realm of protein sequences to define a probabilistic species-specific model of regulatory sequences and explore this model’s utility for the challenging task of gene expression prediction for non-model bacterial organisms. Machine learning models presented in this thesis represent novel applications of models traditionally applied to data in the domains of images, text or sound toward addressing challenging problems in biology. Particular attention is devoted to the challenging task of utilizing large amounts of unlabeled data present in metagenomic sequences and the genomes of poorly characterized bacteria in the hope of improving researchers’ abilities to manipulate complex biological phenomena.

Amino acid geochronology based on fossil molluscs provides a useful approach to determining the Quaternary history of Great Basin lakes. The Lahontan basin, Nevada, and the Chewaucan basin, Oregon, in the northwest corner of the Great Basin, both contained lakes during the Quaternary. The aim of this study is to improve the Quaternary geochronology in these two basins by measuring time-dependent changes in amino acids preserved in fossil molluscan shells. The abundance of D-alloisoleucine relative to Lisoleucine (All) characterizes the extent of racemization, which increases with age and Ul forms the basis of relative and correlated ages. An age-calibration curve for Vorticifex was developed using All ratios in shells from layers with radiocarbon-dated shells and with one thermoluminescence date in the Chewaucan basin. The All ratios from non-dated deposits were assigned ages from this calibration curve. The All ratios in 77 samples (-350 shells) of mainly Vorticifex were analyzed to improve the lake chronology in the Lahontan and Chewaucan basins. From the stratigraphic position, All ratios in the shells, and previously published radiometric ages, at least five and possibly six lake cycles were inferred in the Lahontan basin for the Quaternary period. Shells with highest All ratios ( -0.8) might correlate with the Rye Patch Alloformation, named for deep-lake sediments deposited in the Lahontan basin -630 ka. The next younger lake deposits are ascribed to the Eetza Alloformation. On the basis of the amino acid data, two and possibly three distinct lake expansions took place during the Eetza lacustrine episode, which lasted from -385 to 145 ka. Deposits of the Sehoo Alloformation ( -35 to 12 ka) can be differentiated from older deposits on the basis of All ratios in mollusc shells. Finally, a few shells with low All ratios near Pyramid Lake may indicate a minor lake expansion during the Holocene. Only two lake cycles were inferred from the amino acid data in the Chewaucan basin for the Quaternary period. Shells with the highest ratios correlated with the Eetza Alloformation and the shells with lowest ratios correlate with the Sehoo Alloformation. The amino acid data suggest that Lake Lahontan and Lake Bonneville experienced similar lake-level histories during the past -660 ka. The Sehoo Alloformation in the Lahontan basin broadly correlates with Bonneville Alloformation in the Bonneville basin based on All ratios and radiocarbon dates. The late and early aminozones within the Eetza Alloformation might correlate, respectively, with the Little Valley and Pokes Point Alloformations in the Bonneville basin.

abstract: Continuing and increasing reliance on fossil fuels to satisfy our population’s energy demands has encouraged the search for renewable carbon-free and carbon-neutral sources, such as hydrogen gas or CO2 reduction products. Inspired by nature, one of the objectives of this dissertation was to develop protein-based strategies that can be applied in the production of green fuels. The first project of this dissertation aimed at developing a controllable strategy to incorporate domains with different functions (e. g. catalytic sites, electron transfer modules, light absorbing subunits) into a single multicomponent system. This was accomplished through the rational design of 2,2’-bipyridine modified dimeric peptides that allowed their metal-directed oligomerization by forming tris(bipyridine) complexes, thus resulting in the formation of a hexameric assembly. Additionally, two different approaches to incorporate non-natural organometallic catalysts into protein matrix are discussed. First, cobalt protoporphyrin IX was incorporated into cytochrome b562 to produce a water-soluble proton and CO2 reduction catalyst that is active upon irradiation in the presence of a photosensitizer. The effect of the porphyrin axial ligands provided by the protein environment has been investigated by introducing mutations into the native scaffold, indicating that catalytic activity of proton reduction is dependent on axial coordination to the porphyrin. It is also shown that effects of the protein environment are not directly transferred when applied to other reactions, such as CO2 reduction. Inspired by the active site of [FeFe]-hydrogenases, the second approach is based on the stereoselective preparation of a novel amino acid bearing a 1,2-benzenedithiol side chain. This moiety can serve as an anchoring point for the introduction of metal complexes into protein matrices. By doing so, this strategy enables the study of protein interactions with non-natural cofactors and the effects that it may have on catalysis. The work developed herein lays a foundation for furthering the study of the use of proteins as suitable environments for tuning the activity of organometallic catalysts in aqueous conditions, and interfacing these systems with other supporting units into supramolecular assemblies. / Dissertation/Thesis / Doctoral Dissertation Chemistry 2019

Cleavage and polyadenylation are two necessary messenger RNA (mRNA) maturation steps for gene expression. The Cleavage and Polyadenylation Specificity Factor (CPSF) complex, which recognizes the AAUAAA polyadenylation signal and executes the cleavage reaction, is indispensable for these two processes. In this thesis, I describe my study of the regulation and functions of two non-canonical isoforms of the CPSF subunit WDR33. In addition, I provide detailed analyses on our current knowledge of CPSF subunits’ functions and their influences on a diverse collection of biological processes and conditions. In Chapter1, I provide a general introduction to cleavage and polyadenylation, WDR33, innate immune response via molecular pattern recognition, and the cGAS-STING pathway. Chapter 2 presents my original research on non-canonical WDR33 isoforms, termed WDR33v2 (V2) and WDR33v3 (V3). I determined that their mRNAs are produced by alternative polyadenylation. Both V2 and V3 proteins lack multiple WD repeats, but they can interact with and stabilize each other. This is a novel mode of protein-protein interaction, which I termed WD repeat complementation (WDRC). Unexpectedly, I found that even though V2 and V3 are isoforms of a polyadenylation factor, they are not themselves polyadenylation factors. Regulated by the NF-κB pathway, they are interestingly immune factors involved in the cGAS-STING pathway that induces immune responses against cytosolic double-stranded DNA. V2 decreases STING disulfide oligomerization and suppresses STING-mediated interferon β induction, but facilitates STING-mediated autophagy. Binding of V3 to V2 via WDRC prevents V2’s regulation of STING, suggesting that V3 is a V2 inhibitor. My findings thus further our understanding of STING-mediated immune responses. More broadly, these findings also demonstrate that isoforms produced by alternative mRNA processing can be functionally unrelated. In light of the versatility of the WDR33 gene, I performed a literature review in Chapter 3 on both the canonical and non-canonical functions of CPSF. I first summarize the general functions of CPSF subunits. Subsequently, I discuss their involvements in a variety of biological processes and conditions. This discussion reveals that different processes involve different CPSF subunits. Although CPSF is responsible for only two simple biochemical reactions, it has profound influences on cellular homeostasis. Together, my thesis studies reveal new insights into the molecular mechanism of the cGAS-STING pathway, underscore the importance of alternative mRNA processing, and provide the latest analyses of the functional significances of CPSF.

No description available.

No description available.