Calbindin D9k is a calcium binding protein that contains a helix-loop-helix motif, called EF-Hand. Members of the EF-Hand family are involved in calcium uptake and regulation of functions within the neuroendocrine and muscular systems. As calcium is involved in many cellular processes, such as muscle contraction and nucleotide metabolism, the function of these calcium-binding proteins is of great biological significance. Nuclear magnetic resonance spectroscopy (NMR) can provide a measure of protein dynamics, such as order parameters, that describe the residue-level mobility of a protein, including that of the backbone amide groups. These order parameters are often translated into conformational entropies, enabling an assessment of the entropic changes associated with unfolding or binding events. NMR studies that probe the stability and cooperative binding of calbindin D9k provide evidence that a temperature dependence of order parameters can identify regions of the protein that exhibit linear or nonlinear stability as a function of temperature. I will use molecular dynamics simulations to characterize the dynamics of Calbindin D9k and provide atomic level detail of the protein’s temperature dependent behavior over a temperature range of 3 to 63 Celsius. I will be comparing order parameters calculated from the simulations with experimentally derived order parameters for both the calcium loaded and unloaded states to determine the effects the calcium ions have on protein stability and conformation. Residues instrumental in calcium binding and structural stability are expected to exhibit smaller reductions in order parameters while non-structural regions should exhibit larger and more erratic reductions through increasing temperature. This study will further our knowledge of the dynamics of calcium binding proteins and offer an estimation of the free energy contribution that conformational entropy provides ligand binding. It may also be possible to extend these results to other calcium binding proteins known to act in disease and calcium deficiencies.

Proteins are functionally dependent on their 3-D structure. Many proteins share similar structural folds arising from either convergent evolution or a common ancestor. Understanding a protein’s folding pathway is a fundamental problem of biochemistry. The Daggett Lab created the Consensus Domain Dictionary to categorize common 3-D protein folds, termed fold families, based on various arrangements of secondary protein structure. Evidence suggests there may be common mechanisms in the folding and unfolding of members of a fold family. Here, we look for such characteristic mechanisms within 14 members of the Ubiquitin-like family, the 8th most commonly found protein fold. These proteins have a common core structure with sequence identities ranging from approximately 25% to 85%. Using Molecular Dynamics simulations to model the proteins denaturing at 498K, we compare transition states in the unfolding pathway, various atom-atom contact data, and the alpha-helical content amongst protein fold family members to analytically determine the existence of family-specific folding mechanisms. We anticipate that the early pathway will revolve around helix-turn-helix rotations or splays, sequential alpha-helical melting, and spatial separation of the four main beta sheets.

It has been hypothesized that there is a correlation between the flexibility of amino acids in a protein and their conservation. Inflexible residues tend to be structurally important and thus less tolerant of change. By this theory, mutations at these positions are expected to typically be deleterious. The Dynameomics project is an ongoing effort by the Daggett Research Group to characterize protein folding space through molecular dynamics simulations of representatives of all known protein structures or folds. The representative domains are organized such that the lowest rank fold (rank 1) represents the highest populated fold – with each subsequent rank representing fewer structures. We have used a novel software module to calculate residue flexibility directly from the relevant molecular dynamics simulation files. The resulting data were used to test the hypothesized correlation between conservation and flexibility. Analysis of the entire dataset showed that hydrophobic residues and Cysteine were relatively conserved on average. Hydrophobic residues are generally found in the less flexible core of the protein while Cysteine residues form highly inflexible disulfide bonds. Thus, residues expected to be inflexible were conserved on average – providing support for the tested hypothesis, with the caveat that the standard deviations were large. To avoid these large variations, analysis of individual proteins was needed. Calculated correlation coefficients allowed for the quantification of how well each simulated structure agreed with the hypothesis. Work is now being done to characterize the difference between the various ranks of Dynameomics in terms of this conservation-flexibility relationship. We anticipate that a better understanding of evolution at the amino acid level of proteins will help us to better predict the impact of mutation. Combining this information with rotamer information can allow for the creation of tools useful in the field of protein design and engineering.

Proteins that share similar 3D structures, or folds, can be grouped together into fold families, and evidence suggests that there may be common folding mechanisms that exist within these fold families. Finding these common folding mechanisms will shed light on the way in which folding is determined among proteins, which is a major question in biochemistry. Here I investigate the dynamics and unfolding pathways of 18 members selected from the sixth most populated fold family, the SH3-like barrel domain. The SH3-like barrel fold is an all-beta fold, constructed of beta strands only. Using molecular dynamic simulations at high temperatures, the unfolding pathways, including the characterization of transition states, can be studied for each of these proteins. A broad analysis of native and non-native contacts and secondary structure content will allow common folding patterns to be determined across the fold family. It is anticipated that a hierarchical loss of contacts between the beta-strand of the fold, with those between the turns being maintained longer will be observed. Here I report an initial analysis of the unfolding of these 18 members of the SH3-like barrel fold family and present the observed commonalities, as well as variances across the pathways.

 Acute myeloid leukemia (AML) is a blood cancer characterized by the failure of differentiation of white blood cells to their progeny (myeloid differentiation). MicroRNAs (miRNAs) are small noncoding RNAs that regulate gene expression by targeting messenger RNAs (mRNAs) for degradation or translational repression. In cancer, miRNA expression is altered and these changes contribute to impaired differentiation. In previous work we examined over 600 miRNAs to establish which miRNAs could promote differentiation or limit growth of AML cells. This work identified that miR-150 expression induces myeloid differentiation and that miR-150 expression is decreased in many AML patients. Because transcription factors are important regulators of myeloid differentiation, we examined the transcriptional regulation of miR-150 in hematopoietic cells to determine how miR-150 expression is decreased. We identified the minimal promoter region of miR-150 using luciferase reporter assays in THP-1 (AML) and K562 (chronic myeloid leukemia) cell lines. We identified DNA binding sites for the Krüppel-like factor (KLF) family of transcription factors as necessary for miR-150 promoter activity using site-directed DNA mutagenesis of the luciferase reporters. The current study aims to identify which KLF isoforms regulate miR-150 expression. We cloned KLFs 2, 3, 4, 5, 6, 7, 9, and 10 into lentiviral expression vectors. KLF2 and KLF4 overexpression increases miR-150 expression as measured by luciferase assays and quantitative PCR in THP-1 and K562 cells. Furthermore, we confirmed published results that KLF2 and KLF4 promote myeloid differentiation of THP-1 cells using flow cytometry and gene expression studies. Previous studies have determined that KLF2 and KLF4 expression are decreased in AML cells. This loss contributes to pathogenesis; however how miRNAs contribute remains unknown. Consequently, determining how miR-150, KLF2, and KLF4 promote differentiation will provide insight as to how loss of KLFs contributes to leukemia and how we may use miRNAs to overcome block in differentiation therapeutically.

Reward presentation reinforces behaviors and increases the likelihood of a particular behavior occurring in the future. This increase in behavior is one of the bases underlying reward-based learning. Reward-based learning involves several brain areas, of particular interest to this study is the periaqueductal grey (PAG). The PAG is anatomically linked to other brain structures that are involved in reward processing, such as the ventral tegmental area (VTA). When presented with rewards and novel stimuli, neurons in the VTA exhibit burst firing (brief periods of high activity), and the intensity of this firing is modulated by reward size and expectation. It is possible that incoming glutamatergic signals (excitatory signals) from the PAG contribute to this burst firing as it is the third heaviest subcortical supplier of glutamate to the VTA; however this has not been previously investigated. Previous studies have recorded burst firing in the VTA during a working memory task on a radial 8-arm maze. These past studies also showed velocity correlated firing by VTA neuron of rats as they approached or retreated from rewards. The current study had rats perform the same working memory task, but recorded neural activity from the PAG to examine whether the PAG showed similar burst firing patterns as the VTA. Recordings of single unit (one cell) responses were collected from the PAG as rats performed the task under different conditions, including omission of expected rewards or expected reward magnitude reversal. The rat’s velocity as it was approaching and retreating from the reward was also recorded. It was found that a subset of PAG neurons did show reward-related firing, similar to that of the VTA, and we anticipate also finding velocity correlates. Thus, this study will better characterize the function of the PAG and the neural processes through which reward-based learning occur.