Cells use a variety of mechanisms to regulate gene expression. One way cells control gene expression is by forming structures such as DNA loops to position genes in 3D space near regulators. My lab is developing tools to synthetically control 3D genome structure and assess the relationship between positioning and expression. Our strategy is to engineer DNA loops by fusing DNA binding domains to targeting domains that dimerize and bring the two sites together. To ensure the binding domains bind to each other to form a loop, we are developing an allosteric sensor of DNA binding where the dimerization motifs are only active when bound to DNA. Without this switch, the looping interaction would be outcompeted by free binding domains that are not attached to DNA. We used LOCKR, a bioactive protein switch comprised of a protein cage that switches to an ON-state in the presence of a key protein. These proteins are tethered to DNA using dCas9 as a programmable binding domain. I have tested different system parameters, including the length of the linker between dCas9 and the key protein, for increased activation of green fluorescent protein, a reporter gene that is expressed when the switch is activated. My data show that the switch is effective with a variety of linker lengths. I am also exploring other parameters for optimization, such as changing the relative orientation of the dCas9 complexes. Exploring these parameters is important because the switch might be sensitive to small changes in structure or orientation, and we want to identify the optimal arrangement of dCas9 and switch proteins for effective switch function. After this system is optimized, we will be able to direct our efforts toward looping DNA, which will allow us to address broad questions about the relationship between gene position and expression.

Scaffold proteins, which assemble enzymes and their substrates into multiprotein complexes, are critical for many cellular functions. By bringing enzymes and their substrates into close proximity, scaffold proteins are thought to enhance the rates of enzymatic reactions. For instance, Axin is a scaffold protein in the Wnt pathway that binds the transcription factor, ß-catenin and the kinase, GSK3ß. This tethering method is predicted to enhance the rate of ß-catenin phosphorylation. An outstanding question is whether a scaffold protein that binds to both GSK3ß and ß-catenin is sufficient to observe these effects or whether Axin has other structural properties that are not currently understood. Although there is no detailed structural information about Axin, it is predicted to be disordered. To address this gap, we have engineered a synthetic scaffold protein that can bind both GSK3ß and ß-catenin and contains a linker that is also presumed to be disordered. Previous work in the lab has reconstituted the reaction between Axin, ß-catenin and GSK3ß in vitro and found that Axin enhances the rate of ß-catenin phosphorylation. Using enzyme kinetics, I have compared the rate of ß-catenin phosphorylation of the engineered scaffold to that of Axin. Identifying the quantitative kinetic comparisons between the engineered scaffold protein with the natural scaffold protein could tell us more about what aspects of a tethering system make it effective at enhancing the rates of reactions. We are expecting to see similar tethering effects between the two scaffold proteins, as their linkers are both disordered and equivalent in length. 

Harnessing the power of enzymes to carry out synthetically relevant reactions is rapidly emerging as a powerful tool for sustainable chemistry. Iron-dependent enzymes have been of particular interest to the field because of their ability to facilitate complex reactions with broad substrate scope. Recently, we found that Fe(II) 2-oxoglutarate dependent hydroxylases (Fe(II)/2OGs) exhibit non-native activity towards either epoxidation or allylic hydroxylation. Oxyfunctionalizations are important in chemical synthesis as they allow for diversification to a range of more complex molecular scaffolds from simple olefinic precursors. Because of this, reactions that produce hydroxides or epoxides are of high interest, especially in an asymmetric fashion. However, current methods are not broadly applicable, as they are limited in substrate scope, selectivity, and tolerance towards other functional groups. Enzymes could rival traditional methods, offering greater selectivity and substrate scope while using inexpensive and Earth-abundant reagents. Here, we explore the ability for Fe(II)/2OGs to catalyze non-native asymmetric epoxidations or allylic hydroxylations. First, we will determine whether the reaction taking place is an epoxidation or allylic hydroxylation. Second, we will identify any additional Fe(II)/2OGs capable of oxyfunctionalization beyond our initial hit. Third, we will use directed enzyme evolution to improve oxyfunctionalization activity and enantioselectivity of a selected enzyme candidate. Fourth, we hope to expand substrate scope of an evolved Fe(II)/2OG to develop a general “epoxidase” or hydroxylase capable of reacting with a variety of functionalized olefinic small molecules in an asymmetric fashion. Generally, our lab seeks to use Fe(II)/2OGs mutagenesis libraries and a developed liquid chromatography-mass spectrometry screening platform to exploit the existing wide catalytic diversity of Fe/2OGs into a useful array of synthetically relevant transformations.

The hippocampus (HPC) and lateral habenula (LHb) work together to guide flexible responding as one’s environment changes. The HPC plays a critical role in learning and remembering events. While the LHb has been shown to encode information about rewards and aversive events, it also enables flexible responding during HPC-dependent tasks. Therefore, we hypothesized that the LHb and HPC communicate during choices, as this should be when the two structures combine reward information and spatial memory to make a decision. It is generally thought that brain structures are communicating when their neural oscillatory activity is coupled. Thus, we measured the activity of large populations of cells in the LHb and HPC while rats performed a spatial memory task. We expected to see higher coherence of oscillatory activity during choices than during other phases of the task, as this would suggest the HPC and LHb are interacting to process reward information and spatial memory in a flexible manner. However, our preliminary results showed that the population activity recorded in the LHb is likely inherited from the HPC by volume conduction, which means that the LHb oscillation is not locally generated. A future direction is to improve our recording method and investigate single-cell characteristics in the LHb with respect to hippocampal oscillations. Overall, this work is important because it guides future steps for studying the communication between the LHb and HPC, which could provide insights on pathological conditions where a person is unable to flexibly respond in a changing context (such as in depression, addiction, or memory disorders). Therefore, future findings from our research may have important clinical applications, with the potential to inform and improve upon existing interventions for disorders of memory and mood.