Brain extracellular matrix (ECM) structure mediates many aspects of neuronal function. When ECM structure becomes dysregulated in neurological disease, one resulting impact is impaired neuronal function. Therefore, probing changes in ECM structure could provide insights into disease mechanisms and expose potential therapeutic pathways. Previous work in our group determined that degrading neural ECM structures, including perineuronal nets (PNNs), leads to a significant increase in the diffusive ability of nanoparticles navigating the brain extracellular space. However, this diffusion-based analysis provides little insight into changes in PNN-specific morphology or structure; it only predicts whether or not they are present and the degree to which they may be altered from normal. With this project, we aim to quantify changes in PNN structure with high spatial resolution. PNNs are stained using a fluorescently labeled lectin (Wisteria floribunda agglutinin) and images are acquired via confocal microscopy. Using Python, a coding language, we developed an automated image processing workflow to characterize morphological and structural features associated with PNNs, including total number of branches, average branch length, average mesh size of the net, and the areal density of fluorescence. This approach was applied to brains that span a range of chronological ages, from 14 days old to adult. PNNs are known to increase in counts early on in life, so this age-based study served as a proof of concept for our methodology. This same approach can be applied to study the effect of various neurological diseases on PNN structure. Collectively, this work aims to enhance our understanding of neurological disease mechanisms and open new avenues of therapeutic intervention.

The blood-brain barrier (BBB) prevents the accumulation of most drugs in the brain due to brain capillary endothelium tight junctions and efflux mechanisms, encouraging the use of a delivery platform to enhance drug delivery. Nanoparticles can encapsulate therapeutic cargo to improve diffusion through physiological obstacles and ultimately assist in improving treatment of neurological diseases. However, the formulation process must first be improved. The goal of my project is to 1) optimize nanoparticle formulation parameters to maximize therapeutic enzyme activity, and 2) characterize the extent of nanoparticle degradation due to sonication. Both formulation methods were composed of poly(lactic-co-glycolic) acid (PLGA, 45kDa) copolymerized with poly(ethylene glycol) (PEG, 5kDa), cholic acid (CHA) or polyvinyl alcohol (PVA) surfactant, and the enzyme catalase. We specifically assessed the effect of varied sonication time (1s, 15s, 30s, 60s) during the emulsion method. The 30s sonication double emulsion and the nanoprecipitation methods yielded the greatest enzymatic activity for both CHA and PVA surfactants (2.10% and 1.79% CHA, and 3.72% and 3.58% PVA, respectively). However, in the presence of degradative proteins in the form of pronase, double emulsion-formulated nanoparticles with CHA exhibited better retention of enzymatic activity than nanoprecipitation at 2h (75.66% and 9.22% activity, respectively). We monitored the heat flow as a function of temperature to identify the melting temperature phase transition in a sample of PEG exposed to varying lengths of sonication to assess PEG degradation with differential scanning calorimetry, where a lower melting temperature corresponds to a decreased PEG molecular weight. 5kDa PEG exhibited a melting temperature of 60.792C, while 5kDa PEG exposed to 2x150s of sonication exhibited a melting temperature of 59.450C correlating to 4.8kDa. The 30s sonication CHA double emulsion formulation yielded the highest enzymatic activity with protection from external proteases, yielding a promising polymeric nanoparticle platform for neurological applications.

In trying to understand biology’s dynamic heterogeneity, scientists have sought to recapitulate the spatial complexity and temporal presentation in which proteins are naturally presented to cells. Currently, the most promising strategies in this regard exploit sequential ligation/cleavage reactions, each controlled in time and space using light so as to reversibly immobilize proteins within synthetic biomaterials. Though our lab has utilized these approaches to spatially control complex biological fates with micron-scale resolutions, previous methods suffer from complex syntheses, as well as requirements for specialized equipment and skillsets rarely available in bio-based laboratories. Improving upon these fundamental limitations, our group has developed a scalable system wherein proteins can be bound/released from hydrogels using light, without the need for such expertise/equipment, by being fully genetically encodable. In this approach, biology performs all the modifications necessary to photopattern protein binding to gels, as well as install the reactive species requisite for the protein’s photo-mediated release. We have accomplished this using a photoactivatable protein-peptide ligation reaction developed by our lab, wherein UV irradiation “activates” the protein to ligate specifically with the peptide tag. Additionally, we exploit co-translational chemoenzymatic modification strategies to install a functional handle for tethering the protein into polymeric hydrogels during protein expression. To the peptide tag, we append a photocleavable protein that cleaves when irradiated by visible light, fused to a protein of interest (POI) to be tethered to the hydrogel. Expressing these proteins in E. coli yields the first-ever fully genetically encodable system which can reversibly pattern proteins into hydrogels, by first shining UV light to tether POIs into biomaterials, then subsequently shining visible light to photocleave the protein and trigger POI release. Highlighting the system’s versatility, we demonstrate that the approach is compatible with fluorescent proteins and bioactive growth factors to direct 4D cell fate.

The most effective semiconductors used as absorber layers for solar cells have concerns regarding earth abundance, toxicity, cost-volatility of the materials, or high capital expenditure (CapEx) for new manufacturing facilities. Solution processing is a low cost, low temperature development method leading to lower CapEx. The exploration of "new" photovoltaic materials seeks to develop an earth abundant, non-toxic semiconductor via solution processing with efficiencies comparable to market-leading materials like silicon or CdTe. Bismuth rudorffites (chemical formula A_aBi_bX_a+3b) are a category of new materials proven to be solution processable, to have high absorption, and to be capable of cell efficiencies over 5%. One of the limitations of bismuth rudorffites thus far is current flow; the electrons that are capable of providing electrical power are not being extracted from the absorber material effectively before they return to their stable low-energy state. A way in which this limitation can be explored is via photoconductivity, the difference between material conductivity under illumination versus in the dark. My project seeks to characterize the photoconductivity of silver bismuth iodide (Ag_aBi_bI_a+3b) as a function of the ratio a/b, identifying the composition(s) that best facilitate electron transport and the crystal phases to which they correspond. Results indicate that ion migration within the crystal lattice occurs when a/b is large, and that high a/b ratios introduce AgI impurities in the film that dramatically increase the photoconductivity, among other important phenomena. This presentation gives procedures, results, and analyses from this photoconductivity exploration, working toward a more advanced understanding of bismuth rudorffite material properties.

Since discovery in 2009, hybrid organic-inorganic perovskite photovoltaics (HOIP’s) have proven to be highly-efficient, inexpensive photovoltaics. However, their ability to compete in commercial markets is hindered by their poor stability. The overall goal of the project is to develop a machine-learning model that will use data collected on perovskite degradation to predict the stability of the perovskite based on composition and environmental stressors. By testing spray-coated gradients that transition from one perovskite composition to another, data can be collected on the degradation of a range of materials simultaneously. Spray-coating deposits a thicker layer of “ink” onto the substrate than spin-coating, and the resultant film is more prone to de-wetting and the formation of microscopic pinholes. The aim of this project is to improve the morphology of spray-coated films so that it matches that of conventional spin-coated films. To do so, solvent evaporation during deposition was modeled off of data collected on the physical properties of the perovskite inks. Increasing the viscosity of the ink and decreasing the volume of ink deposited onto the substrate was found to produce MAPbI₃ films with significantly less de-wetting and no microscopic pinholes. However, when cations and halides aside from MA⁺, Pb²⁺, and I^- are incorporated into the chemical structure, the microscopic pinholes are still present. The second and current stage of this investigation is examining the solvent-perovskite complexes that form within the inks and determining their impact on the film-formation of perovskites with FA⁺, Cs⁺, and Br^-. This study has demonstrated the viability of spray-coating for the deposition of MAPbI₃. Although further research is required to produce microscopically homogeneous films of different compositions, this has demonstrated that spray-coated films can emulate spin-coated ones; opening the door for the large scale data collection required for machine learning.

The human body does not fully metabolize a pharmaceutical dose. Consequently, these stable compounds are excreted through human waste, and many pass through wastewater treatment plants untreated, polluting aquatic ecosystems and drinking water supplies. Point-source electrochemical oxidation of fresh human urine can be a low cost, versatile option for eliminating these biologically active compounds before their discharge into the environment. To design an effective device to do this, it is important to understand the mechanisms of pharmaceutical degradation in a complex system containing solution-phase and interfacial chemistry as well as kinetic and mass transport limited degradation rates. We used the limiting current technique to characterize the mass transport conditions in a cylindrical electrochemical cell stirred by a magnetic stir bar. By conducting a steady state potential scan on electrolyte containing a kinetically facile redox couple, [Fe(CN)₆]^3-/[Fe(CN)₆]^4-, we measured the mass transport limiting current. The limiting current reflects the reaction rate of an electroactive compound with the electrode surface, and this rate is limited by the diffusion of the compound through a boundary layer with a thickness defined by the system’s convective conditions. We derived a Sherwood number correlation as a function of Reynolds number and Schmidt number by varying stir rate, stir bar dimensions and shape and measured the corresponding limiting current. Knowledge of the mass transport coefficient demonstrated that pharmaceutical degradation can be mass transport limited on boron-doped diamond (BDD) anodes in full synthetic urine matrixes. For the same stir conditions, observed rate constants on iridium(IV) oxide (IrO₂) anodes fell below the mass transport limiting rate indicating a kinetic limit for pharmaceutical degradation. The limiting current technique is an easy method to characterize and predict mass transport conditions for any geometry and helps differentiate between interfacial and solution-phase degradation pathways for environmental pollutants.

Water-swollen polymeric networks (i.e., hydrogels) provide a structural platform for the manipulation of chemical and mechanical signals that mimics the complex heterogeneous environment experienced by cells in vivo. Photoresponsive chemistries have been of particular interest to this end, as they allow for precise spatiotemporal control of physiochemical properties and, thus, cell behavior. Here, we present a novel protein-based network that will allow for the photo-mediated stiffening of genetically-encoded hydrogels. In this system, we exploit a biochemical technology recently pioneered by our lab in which two pairs of proteins undergo irreversible, covalent heterodimerization after photoactivation. Through the incorporation of an inert, unstructured polypeptide backbone, we have exploited the aforementioned reaction to induce gelation in response to light through the formation of four-arm protein crosslinks. Unlike previous synthetic polymer-based hydrogel systems, this system is entirely genetically encoded, which provides significant advantages in terms of cost, time, and production simplicity. As we intend to demonstrate through photorheometry, this reaction proceeds in a dose-dependent manner, providing step-wise control of both where and when gel stiffening occurs. Such 4D control of a gel’s mechanical properties can be used to influence cell migration, growth, and differentiation, and, thus, could have applications in tissue engineering. Furthermore, we anticipate our system could be utilized to model the stiffening of the extracellular matrix, which is commonly associated with pathologies such as cancer, fibrosis, and cardiovascular disease.

Extracellular Vesicles (EVs) are group of cell-derived structures including exosomes, microvesicles, and apoptotic bodies, which have been found to play a key role in intercellular communication, through the biological cargo that these EVs can carry. Their ability to deliver proteins and nucleic acids from donor cells to their target cells has led to growing interest in the potential of EVs being used as biomarkers for disease. But, a comprehensive understanding of EVs behavior is lacking, especially in neuroscience, which may hinder the development for further application of the EVs. Thus, we are interested in investigating the effect of brain-derived EVs (bEVs) on brain cells, especially microglia, the brain’s primary resident immune cells. To do this, we first extracted the bEVs from the rat brain through ultracentrifugation and purified them through size exclusion chromatography (SEC). We then applied the bEVs to cultured mouse BV-2 microglial cells and incubated for 24 hours before performing quantitative reverse transcription PCR (RT-qPCR) on the treated BV-2 cells to explore any bEV induced inflammation response. Preliminary results have shown that bEVs play a role in inducing both pro and anti-inflammatory responses in microglial cells, both to varying degrees in the cytokine markers expressed after incubation for 24 hrs. Furthermore, to better understand the interaction between bEVs and microglial cells, we labeled bEVs with fluorescent nano-sized semiconductor quantum dots (QDs). Through fluorescent confocal microscopy and time-lapse imaging, we were able to explore the time-dependent interaction of bEVs and BV-2 cells at high resolution. Our study can provide insights into bEV behavior, which can be used to better understand their potential use as biomarkers for specific brain disease models.