2021 Projects

Gibbons (Hylobatidae), the lesser apes, are the closest hominids to the last common ancestor of monkeys and apes (including humans), evolving and speciating in Asia around the same time period as human ancestors in Africa. Understanding gibbon evolutionary history is essential to understanding primate evolution, but previous studies have not reached a consensus on the evolutionary path of gibbons. We used the basicranium and endocranium–reflecting the bottom and overall shape of the brain, respectively–to investigate morphological variation across all four gibbon genera. We hypothesized that the relationships suggested by our data would corroborate the results of previous studies using genetic data. We quantified basicranium shape using 30 fixed landmarks on 105 adult gibbons. Overall endocranium shape was quantified with 926 fixed, semi, and sliding landmarks on 74 adult gibbons. For both data sets, we applied standard Geometric Morphometric methods and hierarchical cluster analysis to look at species average shapes in cladograms. Our cladograms generally grouped the genera Nomascus and Hylobates closer together and Symphalangus and Hoolock closer together, which correlates with the phylogeny concluded in some, but not the majority of, genetic studies. This discrepancy suggests that gibbon basicranium and endocranium variation is a result of natural selection rather than neutral evolution.

The 4th chromosome in Drosophila Melanogaster, or the Muller F element, exhibits expression levels similar to euchromatic genes, despite being packaged as heterochromatin. This indicates that F element genes have unique properties that allow expression in heterochromatic environments. Studies have shown that it has disproportionately expanded in at least 4 Drosophila species: D. ananassae, D. bipectinata, D. kikkawai, and D. takahashii. Our genomic analyses, in collaboration with Genomics Education Partnership (GEP), aim to assess the possible impact of this expansion on gene characteristics. In the first half of this project, we have annotated protein-coding spans in the select genomic regions (contigs) within D. ananassae and D. bipectinata F element using the tools available through GEP, such as the Genome Browser, as well as performed comparative genomics analyses through NCBI blast to fine-map the exon/intron boundaries of the genes. These coding region annotations have been submitted to GEP, which will provide a better understanding of the impacts of chromosomal expansion. While comparing the putative gene models to the D. melanogaster ortholog, novel gene characteristics have been found, such as extra introns/exons in the target species and differences in genomic neighborhoods. In the second half of this project, the focus was on annotating a single gene which participates in a highly conserved insulin signaling pathway, Tsc1, across multiple Drosophila species. Genomic neighborhoods and entire transcription spans of Tsc1 have been produced which will provide insight on how genes evolve in the context of their position within a signaling pathway. Tsc1 gene structure and genomic neighborhood have shown to be highly conserved in all annotated Drosophila species, indicating the evolutionary significance of this gene within the pathway.

As human activity and development continue to encroach upon natural areas, habitats are becoming more fragmented, causing edge effects that alter habitat structure and function. While a number of edge effects related to weather and habitat dynamics have been identified, including desiccation, increased tree mortality, and changes in community composition and species abundance, existing literature does not address any edge effects related to anthropogenic sound, which can have serious implications for animal health and communication. We investigated a sonic edge effect, along with other sources of noise pollution on the Vassar Ecological Preserve. In concert with forest vegetation surveys, soundscape recordings were taken at 47 points across 14 distinct forest communities on the Preserve. The recordings were processed for optimal audibility, analyzed for presence and sources of noise pollution, and assigned noise pollution index (NPI) ratings. The points and their NPI ratings were mapped to visualize spatial noise pollution patterns. Data indicate increased proximity to edge is correlated with higher NPI ratings throughout the Preserve. Primary causes of noise pollution include high-traffic suburban roadways, which surround the Preserve on all four boundaries, as well as airplanes and various sources relating to farming and utility work in and around the Preserve. These findings not only indicate that a sonic edge effect exists on the Vassar Ecological Preserve, but also provide information on prevalence, location, and sources of noise pollution that may inform future research and stewardship.

The microbiome of wood frogs (Rana sylvatica) has long been utilized for pathogen defense research. Most research has focused on how the gut microbiome is affected by infection by ranaviruses, which have caused declines in numerous species of frogs and salamanders . However, limited research has been conducted on the microbiome at the tadpole stage and in relation to the skin microbial community. Our research investigated how the skin microbiome differs across the body of individual tadpoles. Tadpoles were collected from a single pond at two time points 1 week apart. Tissue samples were taken from two body locations -- the tail and the ventral surface -- to characterize the bacterial community diversity and structure. This research will help to illuminate the differences in the skin microbiome throughout the body and give us insight into how and why these differences may occur. We hope that this work will also be helpful for future studies by revealing the significance of sampling from different body locations to characterize the skin microbiome.

Herbaria, or historical collections of pressed and dried plant specimens, serve as physical records of plant phenology, distribution, morphology, and genetic information. As the climate crisis continues, herbaria have become increasingly relevant to the study of plant responses to climate change and as documentation of biodiversity. Digitization of herbaria increases their accessibility, enabling the creation of larger data sets that can span centuries and hundreds of individual herbaria. In our study, we specifically considered herbaria that belong to the Hudson Valley Environmental Monitoring & Management Alliance (EMMA). Our project consisted of travelling to four EMMA herbaria, where we verified and updated nomenclature through annotation labels, affixed unique barcodes, and captured uniform, research quality images of each specimen. We followed standards established by the New England Vascular Plants Network, making the resulting database consistent with international standards. The resulting database will provide researchers with one central resource to view herbarium specimens from many locations across the Hudson Valley simultaneously, enabling further climate and biodiversity studies.

In collaboration with the Genomics Education Partnership (GEP), I am analyzing how the Drosophila genome has changed over evolutionary time by 1) assessing the impact of expansion of the Drosophila Muller F element in reference to the non-expanding Muller D element, and 2) comparing a specific gene (dawdle) across multiple Drosophila genomes. The smallest chromosome in the D. melanogaster species is the fourth chromosome, or Muller F element, and it displays heterochromatic properties usually associated with chromosomal regions that contain fewer genes and tighter packaging. Despite being packaged in this way, some yet unknown features of the genes on the F element allow them to still be highly expressed. While the F element is similar in size in many Drosophila species, it has undergone extreme expansion in several species such as D. ananassae and D. bipectinata. For the first part of the project, I annotated coding regions on two overlapping stretches of DNA (contigs) within the D. ananassae Muller D element – contig28 and contig29. On contig29, annotation of the bru3 gene ortholog showed that only 13/18 isoforms were present, indicating five isoforms were lost. There was no large change in gene size given that this contig was on the D element, which has not expanded in the same way the F element has. For the second part of the project, I chose a common gene to compare across ten Drosophila species with varying evolutionary relationships. The dawdle gene (daw) is a TGF-beta/Activin-type ligand involved in insulin-regulated larval development. My analysis across all ten species involved comparison of the synteny (local genomic neighborhood) and its orientation, and annotation of coding regions and transcription start sites. Despite the varying synteny, I found that the relationships between the resulting protein sequence for daw reflected the hypothesized evolutionary relationships for the species at large.

A long-lasting hypothesis in behavioral ecology is that the detachable scales of moths serve as a defense mechanism against the glue droplets of orb-weaving spider webs that trap most other aerial prey. This defense mechanism is overcome by two distinct moth-specialist families of spiders. One species, C. akirai, found in Japanese rice fields, have been found to have glue droplets that exhibit low viscosity coupled with high adhesive strength and toughness. The low viscosity allows the glue to quickly permeate the scales, seeping to the moth’s base cuticle and gluing the scales to the underlying integument. In contrast, the glues produced by most orb-weaving spiders only stick to the surface of the scale, allowing moths to shed the attached scales and escape. In our study, we identified vertical orb-webs built on a metal gate at the entrance of Vassar Farm. At night, we used an ultraviolet light to attract and capture nearby moths which we subsequently released under the webs, Araneus, with hopes that the moths would come in contact with it. Moths of various species were released on each identified web. Webs were illuminated with infrared light and attempts were filmed using a Basler Ace 2 -Pro NIR high-speed video camera at 62 fps. Attempts were additionally made to collect the web. In the lab, videos were digitized using LoggerPro; we assessed the success rate of moth capture and the retention time of moths on the web. The resulting data shows that all 15 moths successfully escaped from the web, retention time of moth on web spanned from less than 0.04 seconds to 3.76 seconds. These data suggest an evolutionary significance selection pressure by spiders on moths and biomechanical significance of glue droplets that are able to overcome the scale shedding defense mechanism. This project is funded by the National Science Foundation, proposal number #2031962.

Invasive species pose a great threat to native ecosystems. By outcompeting native plants invasive plants have started to overtake native ecosystems and are slowly becoming the dominant species present, seriously harming the local biodiversity. Vassar Farm and Ecological Preserve is the home to many unique Forest Communities with different species that are at risk due to these invasions. To observe the impact of invasives on native forest generation we surveyed a total of 51 20m by 20m Forest Plots. The plots were divided into smaller subplots and the plants present in each stratum: the herbaceous layer, the shrub layer, the vine layer, and the canopy layer were identified and were ranked 1 to 4 on an Invasive Index based on the presence and percentage of invasive plants in each layer. The data was analyzed to determine which community types were more susceptible to invasion at different layers to understand which communities are at higher risk. We found that Southern Successional Hardwood Malus and Successional Northern Hardwood/Southern Successional Southern Hardwood were the most invaded overall, indicating that older successional communities were most invaded. However different communities were more susceptible to invasives at different layers, with Red Maple Hardwood Swamps being as invaded on the vine layer and Beech Maple Mesic Variant being just as invaded on the Shrub layer. These communities were less invaded on the canopy layer, however the shrub and vine layers being heavily invaded suggests that these communities are at a high risk of being invasive dominated. Conversely some communities like Beech Maple Mesic Forest were found to be mostly stable and so will require less attention. These data will inform our decisions on what communities to prioritize for the Conservation Action Plan that will see the preserve become a carbon sink.

Comparative gene annotations provide important data sets that can be further analyzed to address many scientific questions, especially those involving the effects of evolutionary time on chromosomes and genes. This study consisted of gene annotations and genomic analysis for two separate projects under the Genomics Education Partnership – the F element and the Drosophila Pathways. The Muller F element or the fourth chromosome is the smallest chromosome in Drosophila melanogaster. Although it is heterochromatic, the F element has a high gene density and exhibits high gene expression. Annotations of the coding region exons (CDSs) and gene models of the genes in contigs 2, 7, and 11 of the F element of Drosophila bipectinata were produced. It was found that both of the genes annotated in contig 7, Ekar and Gat, have an additional CDS compared to their D. melanogaster orthologs. Since the F element of D. bipectinata shows substantial expansion, the data from all these annotations is expected to facilitate comparative analyses that will help us to gain a more comprehensive understanding of the impact of evolution on chromosome and gene size. The Pathways Project involved examining and annotating a single gene, Sdr, across several Drosophila species. The secreted decoy of InR (Sdr) is part of the insulin receptor signaling pathway. The coding regions of Sdr were found to be well-conserved across D. ananassae, D. erecta, D. grimshawi, D. suzukii, D. yakuba, D. persimilis, D. mojavensis, D. virilis, and D. willistoni, and synteny was almost always maintained, especially in species that are more closely related to D. melanogaster. It is anticipated that these comparative gene annotations, in addition to network analysis methods, will be able to provide insight into the evolution and function of this important biological pathway.

Orb-weaving spiders in the Araneidae family are common predators of flying insects. Their webs contain capture threads lined with aggregate glue droplets composed of a glycoprotein core, the adhesive mechanism, and a surrounding aqueous solution of salts that helps the glue spread. Though these allow spiders to catch most prey, moths remain elusive. Moths escape capture using a protective layer of removable scales which stick to the glue and then peel off the body. From studying the glue droplets of a moth-specialist species Cyrtarachne akirai, we hypothesize Cyrtarachne is able to capture moths due to the rapid spreading and self-curing nature of their glue. To understand the biochemical relationship between spreading and curing, we analyzed the spatial distribution of glue chemical constituents in glue droplets. Using Raman Spectroscopy, we analyzed multiple silk strands from webs of Cyrtarachne akirai. The threads were analyzed at 5 different locations: drop core, drop edge, central thread, tendrils, and intermediate space. Samples were then washed to separate the water-soluble components from insoluble proteins, and both were reanalyzed. Comparing the peaks across these 3 conditions, we identified glycoprotein and salt specific peaks and measured their distribution across the initial glue droplets. We hypothesize that the glue's behavior is due to the separation of salts and proteins in its droplets as it spreads, which would be shown as separate gradients along the droplet. These findings were compared to previous spectra of Larinioides cornutus, a common orb weaving spider whose glue does not exhibit a separation between glycoproteins and salts, nor this self-curing behavior. As a result of our experiment, we were able to identify the distinction between the protein and salt peaks present in Cyrtarachne glue. This research is supported by the National Science Foundation #2031962.

Gene expression is a fundamental biological process that cells use to transcribe DNA into proteins, macromolecules that perform the many duties of life. Translation is the final major event of gene expression during which a messenger RNA (mRNA) template is read to synthesize the protein that it encodes. Translation initiation is the rate-limiting, most regulated step of translation and sets the reading frame for protein synthesis. It is characterized by the formation of the pre-initiation complex (PIC), which binds to one end of the mRNA to begin scanning for the start codon. Initiation is mediated by at least 12 eukaryotic initiation factors (eIFs), of which eIF3 is the largest and most complex. In human cells, the eIF3d subunit binds to specific classes of mRNAs, many of which code for proteins that regulate cell proliferation. This could possibly explain why eIF3d is overexpressed in many cancers, including prostate cancer in humans. The Aitken lab has previously used tools to study the mechanistic features of the eIF3 complex in budding yeast, which lacks the eIF3d subunit. Currently, we are studying eIF3d in fission yeast instead of budding yeast because fission yeast’s eIF3 complex exhibits a closer similarity to that of humans. Fission yeast’s eIF3 complex contains many of the additional subunits present in the human complex, including eIF3d. Specifically, we are investigating the effects of an eIF3d deletion using polysome profiling to capture a snapshot of global translation in vivo. We further employ ribosome profiling to identify and investigate mRNAs that are sensitive to this deletion and compare their biological and structural relationships. With this, we hope to compare eIF3d’s role in fission yeast with that of humans, potentially elucidating eIF3d’s role in eukaryotic translational regulation.

A forest is a highly dynamic community composed of mature trees, newly established saplings, and young seedlings that form the foundation of future forest populations. Differences in species composition among these groups may illustrate how a forest community is adapting to changing conditions. When younger forest strata compositionally differ from mature strata, mature tree species are not regenerated, and the forest community may shift to an alternative community type or even collapse entirely. The Vassar Farm and Ecological Preserve (VFEP) is home to over 15 different forest community types, each with its own unique set of tree species that have and will continue to respond to changes in climate, land use, and management differently. In order to understand how these forest communities are regenerating, we conducted a survey of 51 20x20 meter plots across 15 communities. We collected tree species richness and composition data in each plot across 5 different maturity levels, or strata. We then calculated the Jaccard similarity index between strata in order to illuminate differences between the species composition of the mature and young strata. We found that some community types are exhibiting more stable regeneration trajectories than others. Notably, the Beech-Maple Mesic Forest communities, which are co-dominated by sugar maple (Acer saccharum) and American beech (Fagus grandifolia), exhibit similar species in the sapling and seedling strata as the mature strata, whereas the Beech-Maple Mesic Variant Forest communities, which are co-dominated by sugar maple and oak (Quercus spp.) exhibit different species between mature and sapling strata. These findings have significant implications on the biodiversity and ecosystem services that these forest communities support and may help inform management strategies to reach the goals of Vassar College’s Conservation and Climate Action Plans.

Plants encounter a host of environmental stresses from which they must protect themselves to ensure survival under suboptimal conditions. One such molecular mechanism involves post-translational modification by Small Ubiquitin-related Modifier (SUMO), a small protein that modifies the activity, protein localization, and stability of other proteins. In plants, SUMOylation is induced within minutes of exposure to heat, cold, drought, and oxidative stress, but how this modification confers stress resilience remains unclear. Thus, we developed the tools to genetically interrogate the importance of the SUMO system in the moss Physcomitrium patens and identify the proteins that are modified by SUMO. Moss was the model system used, since it is highly amenable to genetic and cell biological analyses. We generated CRISPR/Cas9 constructs to knockout SUMO1b and used Sanger sequencing to identify SUMO1b loss-of-function plants. Immunoblot analysis showed the absence of SUMO1b protein in mutant plants, thereby confirming that we have generated lines to dissect the importance of this gene in plants. Furthermore, we screened and identified plants in which the SUMO1a gene was replaced with 10His-tagged SUMO1a at the native locus by homology-directed repair. Nickel affinity chromatography demonstrated that the 10His-SUMO1a can be purified, thereby enabling the subsequent identification of SUMOylated proteins by proteomic methods. Collectively, these provide new tools that should facilitate our understanding of the importance of SUMOylation in plant stress protection.

In recent years there has been a shift towards carbon neutrality and a focus on mitigating human carbon emissions in order to lower carbon emissions. Trees sequester carbon by holding the carbon in tree biomass, and the ability of forests to act as carbon sinks is an integral component of efforts to remove and store carbon dioxide from the atmosphere. The climate is already changing and as it continues to be altered by human impacts, there is a reciprocal need for humans to sequester their impact. In this study, we measured the diameter at breast height (DBH) for all tree species greater than 5 meters in height located in 51 different forest plots on the Vassar Farm and Ecological Preserve. Using the DBH measurements and the Jenkin’s model for calculating the total biomass of a tree, the weight (kg) of the carbon dioxide sequestered of each tree in the sampled forest plots was able to be estimated. The data shows the tree species and vegetation cover types that store or sequester the most carbon dioxide on the farm, with the genus Quercus leading for tree carbon storage. These data provide important insights into what trees should be planted on Vassar’s Preserve to reduce and compensate for the college’s carbon emissions, such as oak trees. The data has important implications for the future and can be further analyzed to incorporate more complex tree properties such as climate resiliency, growth rate, longevity of tree species, and the importance that specific tree species play in ecosystem functions, all of which will inform efforts to restore and manage the Preserve for carbon sequestration.

Due to climate change, plants are experiencing greater fluctuations in their environment that will lead to major impacts on agricultural output. Therefore, it is essential that we understand the molecular underpinnings by which plants mitigate damage inflicted by abiotic stressors such as heat, cold, and drought. Small ubiquitin-related modifier (SUMO) is a post-translational modification that enables organisms to respond to and manage environmental stress by rapidly altering protein function, stability, and localization. How SUMOylation provides protection against environmental stress is not well understood. Here, we are developing genetic tools in the moss Physcomitrium patens to better understand SUMOylation in plants. Moss is a useful model organism because it is compatible with both genetic and cell biological studies, and as a basal land plant it provides insights into gene evolution. We generated CRISPR/Cas9 constructs to knockout various genes involved in the attachment or regulation of SUMOylation including: SUMO and SUMO variants, ligases that adhere SUMO to target proteins, and a conserved SUMO-targeted Ubiquitin Ligase that mediates degradation of SUMOylated proteins. I screened plants by Sanger Sequencing to identify lesions predicted to generate loss-of-function mutants. Immunoblot analysis of protein extracts from SUMO1a mutants confirmed the absence of the SUMO1a protein. Collectively, this work provides new mutants with which to uncover the SUMO system in plants to gain a better appreciation for how SUMOylation shields plants from stress.

The establishment of symbiotic microbial communities on hosts is influenced by a variety of factors, one of which is the environment. In this study, we sought to determine how soil and water alter the assembly of the microbiome of wood frog tadpoles. Further, we assessed if the natural environment can be mimicked relatively well in the laboratory. We collected newly laid wood frog egg masses, soil, and water from a pond on the Vassar College campus. Once embryos hatched from their eggs, we randomly sorted them into four treatment groups designed to tease apart the contributions of pond water and soil microbes. Tadpoles were reared in individual containers until just prior to metamorphosis and were then dissected to characterize the bacterial communities from their gut and tail. This research furthers our understanding of the influence microbes in the environment have on the communities assembled on tadpoles. Lower microbial diversity and abundance have been linked to higher susceptibility to pathogen infection, thus understanding how different environments impact microbial assembly is essential. Insight into the microbial communities assembled with water and soil alone versus in a natural environment not only signifies how considerable these two microbe sources are, but also provides a basis for future laboratory studies involving tadpole rearing.

Gene expression is a process that is necessary for cellular growth, bodily response, and metabolic activity. Translation, an event that interprets messenger RNA into proteins, is highly regulated to ensure levels of gene expression are ideal. The misregulation of translation is implicated in a variety of human diseases, including cancer. Translation initiation is the most regulated, rate-limiting phase of translation. It begins with the formation of the pre-initiation complex (PIC), which is mediated by 12 eukaryotic initiation factors (eIFs). Among these eIFs, eIF3 is the most expansive, large enough to physically surround the PIC. Partially due to its complexity, there is much to be understood about the exact role of eIF3 despite its involvement in all phases of translation initiation. One challenge to studying eIF3 is that it is essential in Saccharomyces cerevisiae, our model organism. Significant mutations to eIF3 tend to be lethal, preventing us from studying the effects of these mutations directly. With our collaborators, we are purifying individual initiation factors and subunits, grown via plasmids in E. coli, to assemble a recombinant system in vitro. Through biochemical assays, we can monitor key events of translation initiation, such as PIC formation and mRNA recruitment, in the presence of modifications to the system. We hope to show that the activity of the recombinant system is indistinguishable from that of the system purified endogenously. This will allow us to use the system to dissect the initiation complex, mutating, testing, and illuminating the roles of individual subunits of eIF3. Ultimately, these experiments may improve our understanding of translation and its regulation, potentially shedding light on how its misregulation contributes to disease development.

Tau protein is a microtubule associated protein that is important for regulating microtubule stability and spacing in neurons. The protein has six naturally-occurring isoforms that are formed through alternative splicing, and that are regulated throughout development and in some diseases. Additionally, abnormal aggregation of the protein correlates with the progression of several neurodegenerative disorders including Alzheimer's disease and chronic traumatic encephalopathy. An important factor in tau aggregation is phosphorylation which leads to several effects such as impairing tau’s ability to bind on microtubules and stabilize them. One approach that has been successfully applied to study tau phosphorylation in vitro is pseudophophorylation, which mimics phosphorylation by having seven point mutations that change serines and threonines to aspartic acids and glutamic acid in the amino acid sequence. Previous work in our lab has shown that pseudophosphorylated tau mimics phosphorylated tau in both physical and biochemical ways. It was also found that in different solution conditions pseudophosphorylated tau behaves differently than the wild type tau and adopts a more extended conformation than the wild type protein. In the current study we are examining other isoforms of pseudophosphorylated tau to look at how different solution conditions affect tau as well as performing different assays to see how tau aggregates. Two pseudophosphorylated isoforms were produced that deleted N-terminal inserts to mimic the phosphorylated shorter isoforms. This included 1N which was created through the deletion of the N2 region and the 0N isoform which was created through the deletion of both the N1 and N2 regions. The N1 and N2 regions were first identified in the DNA sequence then these regions were deleted using PCR. The resulting PCR product was transformed into E. coli which were used to express the mutant proteins. The proteins were purified using heat denaturation, ion exchange and gel filtration chromatography. Future studies of these constructs will include AFM analysis of these constructs, to examine their conformation, intermolecular interactions and responses to changes in solution conditions.

The rapid emergence of antibiotic resistant bacteria is a pressing concern, threatening public health and food security worldwide. Antibiotic resistance primarily originates through conjugative plasmid transfer (CPT), where plasmids are replicated and transferred between bacteria. Conjugative plasmids encode the necessary components for their transfer. Although the overall mechanism of CPT is known, most plasmid-encoded proteins are unstudied. Here, we present the initial characterization of orf90, a protein located on plasmid pSK41. In addition to being a known dsDNA binding protein, Orf90 is suspected to regulate entry-exclusion protein Orf77, whose function is to prevent bacteria from receiving a plasmid twice. In this investigation, we successfully expressed and purified two orf90 constructs, His-orf90 and His-MBP-orf90, using affinity and size exclusion chromatography. His-orf90 was purified twice to improve purity. However, the second purification yielded an incomplete His tag cleavage. Crystallography trials did not yield macromolecular crystals, likely due to the uncleaved His tag. His-MBP-orf90 was successfully purified and will hopefully yield crystals. An EMSA (Electrophoretic Mobility Shift Assay), which detects nucleic acid-protein interactions, was conducted, revealing that His-Orf90 binds to the 205 bp pSK41 oriT DNA. This confirms that the DNA sequence contains the specific orf90 binding site. Future studies include crystallization trials of His-MBP-orf90 and re-purifying His-orf90 to improve purity. This will allow His-orf90’s use in a DNAseI footprinting assay to pinpoint the specific orf90 binding site and increase the chances of obtaining crystals. These studies will improve our CPT understanding, which may yield new methods to thwart antibiotic resistance.

Tau is a multifunctional protein found in the brain, most known for its role in the formation and stabilization of microtubules. Its six naturally occurring isoforms differ in the presence or absence of repeat regions in the N-terminus projection domain and microtubule-binding region. Irregular tau polymerization correlates highly with the extent and localization of neurodegeneration in frontotemporal dementia and Parkinsonism linked to chromosome 17 (FTDP-17). Mutations in the gene encoding tau have been shown to alter affinity for microtubules as well as kinetics, extent, and morphology of tau aggregation, and the extent of these differences depend on the tau isoform. This project focuses on the characterization of R5L, a mutant originally found in a progressive supranuclear palsy patient. R5L prepared in full-length human tau was examined through in vitro protein aggregation, microtubule binding and formation, as well as atomic force microscopy (AFM), which records relative extension of the projection domain. Compared with the wild type, R5L shows lowered total aggregation at all concentrations, though microtubule binding and morphology remains similar. These data indicate that the pathogenic missense mutation alters tau function in the full-length isoform. Further research will focus on characterization of the extension of the projection domain in full-length R5L, as well as analysis of the mutant in other disease-relevant isoforms.

Two-dimensional (2D) transition metal dichalcogenides (TMDs) are attractive semiconductors for use in electronic, optoelectronic, and spintronic devices. In this project, we examined the combined effects of alloying and strain on the electronic properties of group-VIB TMD monolayers using density-functional theory. Our goal was to identify an alloy that has a highly sensitive response to strain and an equilibrium band gap in the target energetic range for pump-probe experiments. We modeled one alloy in the metal sublattice (Mo_xW_(1-x)S₂) and one alloy in the chalcogen sublattice (MoS_2xSe_2(1-x)), with the alloy concentrations ranging from 0-100% at intervals of 12.5%. Each configuration was subjected to tensile and compressive strains ranging from 0-10%, in intervals of 2%. Density functional theory was then used to analyze the equilibrium structure, derive the band gap, and examine the electronic structure at the varying levels of concentration and strain, providing a better understanding of the combined effects of these modifications on the TMDs’ structural and optoelectronic observables.

Chirality is an important characteristic of molecules, especially those used in pharmaceutical applications such as amines. This study aims to find safe and more effective ways to stereoselectively reduce the two diphenylphosphinic protected imines of 1-aminoindan oxime and 3,4-dihydronaphthalen-1(2H)-one oxime using various metal borohydride complexes. The two oximes were allowed to react with diphenylphosphinic chloride and triethylamine in order to produce deactiviated imines with a diphenylphosphinic substituent where the imine C=N bonds have been shown to be inert towards reduction by simple borohydride sources such as NaBH₄. Combinations of borohydride based reducing agents and solvent mixtures were tested in the reduction of these deactivated imines. Gas chromatography/mass spectrometry (GC-MS) was used to screen for the production of amine. Using the activated borohydride complex Zr(BH₄)₄ as a reducing agent in tetrahydrofuran (THF) successfully yielded amine products. (R) or (S) isomers of 1,1′-bi-2-naphthol (BINOL) were introduced during the reaction in order to yield enantiomerically enriched products. NMR shift studies point toward enriched amine products, but determining enantiomeric excess will require further study.

Trinuclear gold(I) complexes (CTCs) have in recent years been studied for their solvolumincescence and photoluminescence thermochromism properties with applications to organic light emitting diodes. Despite this recent interest in CTCs, a full exploration into their structure and reactivity has not been conducted. Gold(I) centers are known to undergo oxidative addition, resulting in a gold(III) center. Additionally, the stepwise oxidative addition of dihalogens to gold(I) centers has been observed in the past. In this study, we aim to synthesize samples of various CTCs and perform the oxidative addition of alkyl halides to these gold centers. These trinuclear gold(I) complexes were synthesized by dissolving dimethylsulphidechlorogold(I) or chloro(triphenylphosphine)gold(I) in solutions of tetrahydrofuran and n-butyllithium with various imidazoles. The oxidative addition was attempted by adding methyl iodide to a solution of dichloromethane containing CTCs under an inert atmosphere. Synthesized compounds were characterized through nuclear magnetic resonance and X-ray structural characterization.

Increasing prevalence of antibiotic resistant bacterial infections contributes to financial and global public health issues. Antibiotic resistance spreads predominantly through conjugative plasmid transfer (CPT), the process in which bacteria transfer segments of their genetic material on plasmids to other bacteria. CPT is mediated by a protein complex known as the relaxosome which contains a DNA binding integrative host factor, a DNA processing relaxase, and related accessory proteins. The relaxosome of pCU1, a plasmid isolated from Salmonella Typhimurium, contains an accessory protein known as TraK. The structure, function, and activity of TraK are relatively unknown. Here, we present studies towards the structural characterization of TraK. TraK was successfully overexpressed in E. coli cells and purified using nickel affinity and size exclusion chromatography. The purified TraK bound to a short 35bp pCU1 DNA sequence which was used in crystallization trials and crystals of ~ 200​​μm were grown. Further work will include x-ray data collection and analysis of the TraK-DNA crystals, hopefully leading to the elucidation of the structure. Resolving the structure of TraK bound to DNA will provide further insight into CPT and potentially contribute to new strategies combating the spread of antibiotic resistance.

Tau is a protein in the brain which regulates microtubule assembly and is implicated in many neurodegenerative conditions such as Alzheimer disease. Atomic force microscopy (AFM) is a valuable tool for investigating this protein, as it can provide information about the conformations and intermolecular interactions of tau. AFM utilizes a tip on a microscopic cantilever to probe surfaces at the molecular level. When tau is adsorbed to the tip and surface, the AFM observes the forces between the proteins as the tip approaches the surface. This results in force vs distance curves which can be analyzed to reveal information about the interactions between tau proteins. In order to directly and quantitatively compare force curves taken with different AFM tips, an important consideration is the radius of curvature of the apex of the AFM tip. Slight variations between the sizes of cantilever tips arise due to manufacturing or wear from use. This project has focused on characterizing AFM tips by imaging a rough, polycrystalline titanium surface with many high aspect ratio peaks. We have focused on the optimization of an AFM tip characterization algorithm that can determine the radius of the tips based on features in images of the titanium surface. We have assessed image processing parameters that affect the resultant measurement of the tip radius and analyzed the precision of the tip characterization function. Future work will focus on characterizing how the tip radius affects the fitting parameters of tau force curves.

Scientific research faces an ongoing replication crisis, in which many published results are not being independently verified—and, when they are, often do not replicate successfully. The Center for Open Science has organized the SCORE project to address this weakness in the social and behavioral sciences.

The aim of SCORE is to examine the extent to which it is possible to predict the likelihood that a given study will replicate, reproduce, and/or generalize successfully. If the reliability of a result can be predicted without requiring an independent replication, it would substantially improve our ability to move past the replication crisis. To generate data for this meta-analysis, SCORE has created a database of individual studies for review by independent research teams. This summer, we took on two of these projects: a replication and a generalization.

A replication involves running an experiment again, while matching the original methodology as closely as possible. The goal is simply to determine whether the same pattern of results in a published experiment can be recreated. To run our replication, we re-implemented the experiment code, collected a new sample of participants, and performed our own analysis. We created a participant experience similar to that of the original experiment by using the same graphics and instructions for each task.

The goal of a generalizability study is to determine the extent to which a published claim can be replicated when changes are made to the original experimental design. A successful generalization demonstrates the relevance of the original claim to conditions beyond those of the initial experiment. We attempted to generalize a number estimation task by testing it in a different scenario and on a much larger scale in order to examine whether changing these factors causes significant deviation from the expected experimental results.

Alexander Eisert ’22, Nhan Nguyen ’22, Cherrie Chang ’24 and Professor Josh de Leeuw

Conventional cognitive science experiments can be long and tedious; people typically need to be incentivized to participate in them, even those who are excited to aid in the advancement of science. In addition, when participants, even enthusiastic ones, reach the tail end of such an experiment they start to lose focus. This loss of focus likely affects their minds in myriad ways, complicating the researchers’ task: understanding minds.

At the Brain Game Lab, we seek to strike a balance between constructing experimental games that are simultaneously enjoyable and scientifically valuable. We have been constructing a website containing a suite of such games, and anyone wanting to participate can easily log on and compete with their friends by simply entering the same room code.

Further, by developing interesting social games that users can play from the comfort of their own home, we hope to attract large and heterogeneous groups of participants who remain invested in the experiment throughout the data-gathering process. With a large enough sample size, we will be able to split participants into many different conditions, varying many different factors and testing to see if effects previously reported in the cognitive science literature generalize across a wide range of situations.

This summer, our primary focus was to improve the design, user experience, and functionalities of the Brain Game Lab website and the included games. We especially wanted to make the user experience more engaging and intuitive. For the currently supported games such as Rainbow Run, we revamped the setups to make them more conducive to experimental manipulation by making the games more customizable for the experimenter. We hope that, with these improvements, the website will be ready for data collection in the near future. For now, you can try out the beta versions of the currently supported games.

We use simulated robots to test hypotheses about basic principles governing the evolution of intelligent agents. Specifically, we hypothesize that mechanisms of gene duplication and subsequent differentiation increase the modularity of a robotic agent’s sensor/neural network/motor system over generational time, resulting in populations that are more evolvable — that is, better able to survive and even prosper when fitness landscapes shift. This summer, we began the process of building a simulation environment, realistic robotic models, and the complex network of software necessary to study the evolution of populations of these robots across hundreds of generations. This summer’s work focused on the three core components of the software system, which will be optimized to run on Vassar’s multicore supercluster, Hopper. First, we designed and coded a genotype to phenotype (G-to-P) mapping algorithm that uses an unusual process-oriented (rather than part descriptive) mechanism to turn any randomly generated robot genotype into its corresponding phenotypic embodiment of sensors, motors, and neural network. This G-to-P mapping scheme is designed to allow simulation of gene duplication events as well as the mutation events typical in evolutionary robotics simulations. Next, we used the Gazebo simulation environment, which implements the Bullet physics engine, to construct a representation of the robot specified by the genome as well as the environment in which it must operate. Finally, we developed a fitness function that requires the solution to a behavior problem logically equivalent to the exclusive OR, a problem known to require neural networks of more than two layers to solve. Next steps will include the construction of the software infrastructure necessary to create a population of genotypes that will then be evaluated in the simulated environment and allowed to reproduce based on their success in that environment. A comparison of the evolvability of two populations, one evolved with and the other without the gene duplication mechanism turned on, will allow a clear test of our hypothesis.

Learned Categorical Perception (LCP) describes a phenomenon wherein learning to categorize objects affects how they are perceived. After learning, stimuli that share a category label are perceived as more similar and stimuli belonging to different categories are perceived as more dissimilar. While previous research suggests that LCP effects occur in a wide range of cases, it is unclear how robust LCP effects really are. We conducted a partial replication of previous research using AI-generated faces produced from the StyleGAN2 neural network architecture. Subjects in the experimental group learned to categorize stimuli into one of two novel categories followed by a task in which they were asked to discriminate between pairs of stimuli. The control group only performed the discrimination task. Using Bayesian sequential sampling, we contrasted the discrimination performance of experimental and control subjects on exemplars that either shared category membership or did not. We observed strong evidence for LCP when subjects did not receive feedback on their performance during the discrimination task; however, LCP did not occur when feedback was provided. The sensitivity of LCP to this methodological variation suggests that concerns about its robustness are justified. We will continue to investigate this issue by exploring whether incorporating random variation in the stimuli, which is not typical of LCP research but is found in natural categories, impacts the presence of LCP.

The primary goal of this project is to create a public repository including implementations of several algorithms for manipulating Simple Temporal Networks (Dechter, Meiri, & Pearl, 1991). A Simple Temporal Network (STN) is a data structure for representing and reasoning about time, pairing real-valued variables called time-points with sets of binary difference constraints which regulate distances between pairs of time-points. Each STN has a graphical representation where the nodes correspond to the time-points and the edges correspond to the constraints. One of the most important problems concerning STNs is the Simple Temporal Problem, or the question of determining whether an STN is consistent. A network is consistent if it has a solution, i.e. a set of values for all time-points that satisfy all constraints. We implemented a variety of consistency-checking algorithms from the literature on STNs and evaluated their performance on several benchmarks. Each benchmark contained a selection of randomly generated STNs with increasing numbers of time-points and varying densities. From repeated testing we concluded that the algorithm that performed best was a “stop early” version of the Bellman-Ford algorithm. We also implemented and tested a competitor to Bellman-Ford, which performed better than Bellman-Ford on dense networks with over 400 time-points. This is significant because across our testing we found that variations of Bellman-Ford (including speed-ups due to Yen) tend to be the better performing algorithms, whereas our Bellman-Ford competitor is based on a different algorithm, due to Ramalingam et al (1999).

With the increase in popularity of serverless computing, purpose-built Linux kernels with fast and efficient boot processes are becoming the focus of cloud providers for running micro virtual machines (microVMs). In the pursuit of minimal-overhead, AWS Firecracker, a lightweight virtual machine monitor (VMM), has eliminated steps in the Linux kernel bootstrapping process, with the unfortunate side-effect of removing kernel address space layout randomization (KASLR), a widely adopted first line of defense against code reuse and data-only attacks. This project moves the implementation of KASLR from the Linux kernel to the Firecracker VMM with only a minimal increase in boot time over a non-KASLR kernel. We also implemented and characterized the performance of in-monitor function-granular KASLR (FG-KASLR), a feature that has yet to be adopted upstream into the Linux kernel but would also be bypassed by Firecracker. While the impact of FG-KASLR on boot time is non-negligible, FG-KASLR greatly increases entropy over KASLR and decoupling FG-KASLR from the kernel could allow kernels to be pre-randomized and loaded to boot when they are needed, eliminating the overhead of randomization from the VMM startup time while maintaining the levels of defense provided by both KASLR and FG-KASLR. These implementations demonstrate the feasibility of moving processes traditionally done in the kernel to the VMM and future work consists of exploring the potential for more computation that is traditionally done in the kernel to be moved into the VMM for performance benefits.

Ada King ’23, Elise Poniatowski ’23 and Professor Laura Haynes

Paleoclimatology uses ancient climate and biogeochemical systems as case studies for future projections as these global processes cannot be adequately reconstructed in a lab. The Eocene hot house (56-47 million years ago) serves as a partial analog for projected anthropogenic warming. We analyzed deep-sea sediments from IODP Site U1553, located off the coast of New Zealand, to investigate the impact of global warming on sensitive high latitude ocean systems. We used the shells of single celled calcifying organisms called foraminifera to explore ecosystem dynamics and carbon cycling in ancient oceans by examining the isotopic signatures of the shells and their population dynamics. After extracting foraminifera from Site U1553 we weighed the coarse particle fraction (>63µm) to determine what percentage of the core is foraminifera. The lack of correlation between foraminifera percent weight through time and documented carbon excursions in our samples suggests that assemblage results are not influenced by dissolution. Switches in abundance in the genera Morozovella and Acarinina correlate to carbon excursions, supporting the theory that ecosystems in the southern ocean changed in response to warming. Carbon isotope gradients from surface to deep dwelling species will give further insight to carbon sequestration in the water column of this location during warming. Reconstructing the water column’s carbon isotope chemistry and population dynamics during the Eocene allows us to extrapolate from past ecosystem function and biological carbon cycling to anticipate the functioning of the ocean in a warmer world.

Understanding the locomotion of Caenorhabditis elegans (C. elegans) allows scientists to explore and explain the neuronal dynamics of humans. These nematodes have tissues that are similar to humans' form, function, genetics and also share essential core mechanisms. C. elegans are considered model organisms because they are easily cultured in a lab, have a short generation time, and can easily access their gene functions during an experiment. In our study, C. elegans were grown under monoxenic conditions in a laboratory using an E. coli strain. A transfer was made every four days of three to five nematodes to a new agar to ensure continued growth and to ensure viable nematodes are available at any given time. They were kept in a fridge at 20 degrees Celsius to maintain continuity and slow down their life cycle. While there are different ways to observe at locomotion, we are focusing specifically on microscopic video tracking. To track the C. elegans, we recorded videos of the adult worms. The methodology for recording the worms included creating contrast, selecting the correct microscope, and correct numerical aperture. Poor video resolution increases errors when tracking the locomotion of the nematodes. Our team determined that the microscope needed a higher numerical aperture to resolve the C. elegans at a high resolution. Our research is continuous and ongoing as we understand the neuronal dynamics of C. elegans; however, we were able to produce trackable videos using Matlab, which will further our collective understanding of the locomotion.

Transition metal dichalcogenides (TMDs) are an emerging class of 2-D materials with applications in electronics and nanoscale devices, but many properties remain unknown. We used picosecond laser ultrasonics to investigate the thermal and acoustic properties of three semiconductor TMDs: MoSe₂, WSe₂, and Bi₂Te₂Se. First, we mechanically exfoliated crystalline flakes onto silicon substrates, and imaged variously sized flakes with a microscope. We then examined properties such as sound velocity and attenuation using pump-probe laser spectroscopy. This technique uses ultrafast pump pulses to induce strain waves within the crystal lattice, while time-delayed probe pulses reflect off the sample’s surface. The changes in reflectivity over time provide insight into the structural and elastic properties of the material. We measured the thickness of the flakes, performed frequency comb analysis, calculated attenuation, and confirmed our results with Matlab simulations. We also attempted to measure thermal conductivity by depositing a layer of aluminum over the sample and repeating our experiment. Ongoing work will require fitting this thermal data to a nanoscale heat flow model.

We analyze the locomotion of the nematode C. elegans using optical diffraction. C. elegans is a microscopic worm the diameter of a human hair. These nematodes are commonly used as a simplified model for more complex organisms including humans, due to their biological similarities, and aid in studying bodily systems such as neuronal pathways. For locomotion analysis, a laser is directed at the live worm, producing a dynamic diffraction pattern. Analysis of these patterns is useful because variations in light intensity at just one point in the diffraction pattern can reveal information about the entire time series, and therefore the C. elegans’ overall movement. We modified and updated our optical setup by using a spatial filter and an assortment of lenses to prepare the beam so that a highly resolved diffraction pattern can be produced, using a human hair as a simulation of the worm. The spatial filter cleans the beam by shaping it into a consistent plane wave. It contains an objective lens, which diffracts the non-uniform beam into a concentric pattern of rings around a central circle of light. Next, a pinhole allows only the desired central circle to pass through, improving the quality of the beam as it exits the spatial filter. As a result, the intensity distribution is uniform before it interacts with the diffracting object, allowing for the straightforward comparison to models of diffraction patterns.

Topological insulators are a type of material that allows the existence of unidirectional currents at the quantum scale. These currents, called edge states, are unaffected by material imperfections, which makes topological insulators an active research topic with potential applications in quantum computing. Protected edge states - analogous to the unidirectional current of electronic topological insulators - have recently been demonstrated not only with electrons, but also with photons, sound waves, and mechanical waves. Using a well-known model of topological insulators known as the Su-Schrieffer-Heeger (SSH) model, we constructed a one-dimensional mechanical model of topological insulators that propagates longitudinal waves in a metallic spring. Our device demonstrated a phase transition between the insulator, conductor, and topological insulator phases of the SSH model, with clearly defined edge states in the topological insulator phases. Our findings are supported by a mathematical model that we derived from the mechanics of harmonic oscillators. The simplicity of our device facilitates its construction and observation and allows undergraduate students to become familiar with state-of-the-art condensed matter research.

Topological insulators are a type of material that allows the existence of unidirectional currents at the quantum scale. These currents, called edge states, are unaffected by material imperfections, which makes topological insulators an active research topic with potential applications in quantum computing. Protected edge states - analogous to the unidirectional current of electronic topological insulators - have recently been demonstrated not only with electrons, but also with photons, sound waves, and mechanical waves. Using a well-known model of topological insulators known as the Su-Schrieffer-Heeger (SSH) model, we constructed a one-dimensional mechanical model of topological insulators that propagates transverse waves in an elastic string. Our device demonstrated a phase transition between the insulator, conductor, and topological insulator phases of the SSH model, with clearly defined edge states in the topological insulator phases. Our findings are supported both by numerical simulations using the finite element method and a mathematical model that we derived from the mechanics of harmonic oscillators. The simplicity of our device facilitates its construction and observation and allows undergraduate students to become familiar with state of the art condensed matter research.

Studying the locomotory patterns of the nematode C. elegans helps researchers better understand how the neuronal dynamics of more complex animals like humans function through the simplified model organism. We observe worm locomotion experimentally using video tracking and laser diffraction. The latter of these techniques is particularly advantageous because diffraction can resolve subtle changes in motion to the level of the wavelength of the light—greater resolution and precision than that of an optical microscope. Additionally, one point in the laser diffraction pattern is a superposition of all points in the worm and therefore can give information about the entire shape and dynamics of the sample. We created an optical setup to illuminate a model hair and used a CCD camera to record dynamic diffraction patterns. Experimental data is analyzed using Matlab code that creates a time series of pixel intensity at each frame for two specifically chosen pixels in the videos. We found that this method of creating and analyzing diffraction patterns accurately represents the sample’s movement as indicated by chaotic markers such as Largest Lyapanov Exponents. This experimental time series is compared with that of computer simulations of the worm run on Vassar’s Grace Hopper computer cluster to gauge and improve the accuracy of our current understanding.

Men with adolescent delinquency records experience 1.5 to 7 times the population mortality risk into their 70’s (Nieuwbeerta & Piquero, 2008; Robins & O’Neal, 1958; Sailas et al., 2006; Teplin et al., 2005; Trumbetta et al., 2010). This wide inter-sample variation in mortality risk by age 70 may be due to variation across subtypes of adolescent delinquents, specifically, across the temporal typologies of adolescent-limited (AL) versus lifetime-persistent (LP) ABS (Moffitt, 2017; Moffitt et al., 2002; Piquero et al., 2007). Most ABS is AL and relatively “normative,” yet the less frequent subtype of LP ABS emerges in childhood and predicts increased risk for ongoing criminal behavior, psychiatric comorbidity, and general maladjustment (Kjelsberg & Dahl 1998; Krasnova et al., 2019; Loab & Vaillant 2000; Moffitt & Caspi 2005; Piquero et al., 2007). We tested our hypothesis that, relative to non-delinquent or adolescent-onset delinquent individuals, individuals with childhood-onset delinquency, typically a marker of more severe ASB, will experience earlier mortality, on average, using a 70-year follow-up of ~5,500 men from Hathaway and Monachesi’s statewide adolescent Minnesota Multiphasic Personality Inventory (MMPI) sample. We also examined whether childhood-onset, remitted ABS increases early mortality risk at the same level as non-remitted childhood-onset ABS, or if remission from ABS by adolescence confers some protection. We also examined covariates of mortality and delinquency, such as socioeconomic status, cognitive ability, and personality for their potential coaction and interaction with delinquency onset status in predicting mortality risk. Potential personality risks will include the MMPI-A-RF Restructured Clinical (RC) scales for Antisocial Behavior (RC4), Aberrant Experiences (RC8), and Hypomanic Activation (RC9) and potential protectors will include Conscientiousness, Agreeableness, and Extraversion.

Previous research has shown that intranasal oxytocin (OT) administration increases sociability in mice. Interestingly, we found that OT decreased sociability in the hypersocial fmr1 knockout mouse, a model of Fragile X Syndrome and Autism. We hypothesized that altered OT receptor (OTR) function in the ventral tegmental area (VTA) of these mice may account for observed behavioral differences. To test this idea, we administered various doses of OT directly into the VTA. Specifically, bilateral cannulae were surgically implanted into the VTA and one week later, mice were tested using a social interaction assay. To administer OT, mice were briefly anesthetized and then given 1 µL of either saline,100ng, or 500ng OT infused over 1 minute. Mice were placed into the test arena immediately following a post-injection 1 minute diffusion. The sociability assay consisted of a 5 minute exploration followed by a recorded 5-minute “no contact” and then a 5-minute “contact” condition. Social behavior was video recorded by a ceiling mounted camera and assessed using Smart v3.0, a behavioral tracking program (Panlab, Harvard Apparatus). Each subject received all three doses counterbalanced across three weeks of experimental testing, with 3-7 days between trials to allow for drug wash out.

We observed that 500ng OT increased sociability by 1.9 fold relative to saline, while the 100ng dose had no effect (1.01 fold change). We are currently running additional experiments to determine whether this OT effect is indeed mediated by OTR activation as OT also has weak activity at vasopressin receptors. To do so, 1 µL of 4µg/µL of Atosiban (Tocris), an OTR antagonist, is infused into the VTA 3 minutes prior to the intra-VTA injection of 500ng OT. Data acquisition is still ongoing, but if the observed increase in sociability following OT administration is indeed mediated by increased activation of OT receptors in the VTA, we expect that atosiban pretreatment will block this effect

Astrocytes are a type of glial cell within the central nervous system that have a pivotal role in maintaining the blood brain barrier, helping metabolic processes, and supporting synaptic transmission. All of these functions are critical components of memory. The goal of this study is to investigate how activating astrocytes in the hippocampus, an area associated with spatial working memory, will affect their behavior during a delayed spontaneous alternation task. Our hypothesis is that by activating astrocytes, it will increase their supportive functions, hence improve cognition. To activate astrocytes, we used a chemogenetic technique, Designer Receptors Exclusively Activated by Designer Drugs (DREADDs), to target the hippocampus. We performed a stereotaxic surgery of bilateral AAV injection where either a hM3D(Gq)-mCherry with an astrocyte specific GFAP promoter or a GFAP-eGFP control virus was injected into the dorsal hippocampus area of female and male rats. After 2 weeks of recovery period and allowing viral expression, the rats underwent delayed spontaneous alternation testing where rats remember extra-maze cues to effectively navigate a four-arm maze. Half an hour before testing, they received an intraperitoneal injection of either the corresponding hM3D(Gq) receptor agonist, Compound 21 (C21), or the vehicle in a counterbalanced order. Extra-maze cues were changed to provide novel stimuli for each testing session. Previous work done in the prelimbic cortex showed a sex difference such that astrocyte activation improved memory in males and impaired in females. Our preliminary results suggest that there is a possibility that activating astrocytes in the hippocampus with C21 decreased memory performance. These findings suggest while astrocytes clearly play a role in spatial memory, their specific functions need further investigation.

Astrocytes are complex glial cells known for their supportive roles in the CNS and position to influence neuronal communication and activity, further suggesting a role in cognitive processes such as spatial working memory (SWM). While previous research has supported the view that astrocytes play a supportive role in the hippocampus, a brain region associated with SWM (Newman et al., 2011), their role in the prefrontal cortex (PFC) is not well characterized, thus elucidating the dynamics of astrocytic activity within this region is crucial. When astrocytes are active, increases in concentrations of calcium are seen in the cytosol. Additionally, astrocytes communicate with each other through calcium waves, thereby making calcium signaling an important measure for astrocytic activity. We aim to achieve a broader understanding of the role of astrocytes in the PFC by imaging astrocytic calcium activity in adult Long-Evans rats during a delayed spontaneous alternation task (dSAT) using a miniature microscope and an astrocyte-specific virus that leads to astrocytes expressing a calcium sensor that fluoresces when calcium is present. After finding inconsistent imaging outcomes, we hypothesize that many factors can prevent acquiring clear astrocyte imaging, such as lack of virus concentration and expression, unaligned lens placement and baseplating, and excessive background noise. Based on troubleshooting and histology, lack of virus expression may be the main issue. Thus, further modification of procedures is needed in order to obtain consistent imaging data during behavioral testing and proceed with data analysis.

Recent advancements in machine learning and microscopy have enabled neuroscientists to begin decoding neural activity responsible for various behaviors. Our lab is particularly interested in mechanisms of social behavior, the dysregulation of which is a core symptom of many neurodevelopmental disorders. Our work focuses on deciphering neural activation patterns in the ventral tegmental area (VTA) that correlate with social behaviors in adult mice. To achieve this, we simultaneously record neural activity and social behavior using fluorescent mini-endoscopic microscopy and a ceiling mounted camera, respectively. Both sets of videos are processed using open source programs, MiniAn (Dong et al., 2021) and DeepLabCut (Mathis et al., 2018). MiniAn identifies active cells based on changes in fluorescence and outputs the inferred activity of each cell over trial time. Thus, MiniAn provides a quantifiable measure of neural activity in the brain region of interest. To understand social behavior in mice, we analyze behavioral videos using an open-source neural network called DeepLabCut, which allows us to implement machine learning to automatically and reproducibly mark and label 8 body parts of the mouse. This creates a digital body skeleton which we can use to detect and classify behavior based on orientation and velocity of tracked body points. Using these two programs, we’ve quantified neural activity and social behavior from 3 mice under conditions of social and non-social exploration. Next, we will input these datasets into unsupervised machine learning program(s) that will reduce the multi-dimensionality of our data and potentially identify underlying neuronal activity patterns linked with various social behaviors.

Neuronal circuits that control reward and avoidance behaviors converge in the dorsal medial prefrontal cortex (prelimbic cortex (PrL)). Previous analyses in the lab have shown a significant increase in neuronal activity in the PrL following fear conditioning at a remote timepoint in adolescents treated with alcohol. Circuits that control reward and avoidance are dysregulated in alcohol use disorder (AUD) and Post-traumatic stress disorder (PTSD), respectively. Given the comorbid diagnosis of these conditions, the PrL has emerged as an important area of study. Dendrite morphology, spine morphology, and spine density are important contributors to neuron function. Layer 2/3 pyramidal neurons were visualized using a modified Golgi-Cox stain. Basilar and apical dendritic arbors were constructed in 3D using a brightfield microscope under a 63X/0.75 air objective. In order to reconstruct these cells, points of the dendritic tree were encoded into a data file using Neurolucida software and analyzed in Neuroexplorer. The location and classification of dendritic spines (mushroom, thin, stubby) that appear on the dendritic tree was measured under a 100X/1.3 oil objective. Analysis of these structures includes measuring nodes, dendritic length, and the Sholl analysis. Based on previous research in the lab, we expect an increase in dendrite complexity in the PrL for subjects treated with alcohol versus control. These results will help us understand how alcohol impacts dendrite morphology in the PrL.

Astrocytes are star-shaped cells within the brain known for their roles in supportive functions. These include transporting resources in and out of the brain and controlling the extracellular concentrations of ions and chemical signals such as neurotransmitters. Recent data suggest that astrocytes may significantly affect cognition through their involvement in glutamate recycling with the enzyme glutamine synthetase (GS). GS has been implicated in various neurodegenerative diseases, such as schizophrenia, which changes with GS in the prefrontal cortex. In this study, male and female rats were trained in a sustained attention task designed to tax attention over time. This task presents trials at variable intervals where they must detect whether a light signal has been presented and press one of two available levers in response. Based on their performance in the sustained attention task, the rats were sorted into three groups: attending, which demonstrated the best-sustained attention, unattending, which demonstrated chance performance, and side bias, where rats favored one lever response. The prefrontal cortex, hippocampus, and striatum were then assessed using GS immunohistochemistry (IHC). The GS IHC showed higher levels of glutamine synthetase for attending rats than either unattending or side biased rats. This demonstrates that higher levels of glutamine synthetase are correlated with greater sustained attention.

Social behaviors are mediated by numerous brain regions. Our lab focuses on the ventral tegmental area (VTA), which is made up primarily of dopaminergic (DAergic) and GABAergic neurons. Oxytocinergic neurons projecting to this region modulate social behavior by activating oxytocin receptors (OTRs) on these neurons. DAergic OTR activation promotes, while GABAergic OTR activation inhibits sociability by respectively increasing or reducing DA neuron activity. We have previously shown that prosocial effects of intranasal oxytocin are attenuated in male mice lacking fmr1 gene expression (a mouse model of Fragile X Syndrome and Autism) as well as those reared in a fmr1 deficient environment (maternal haploinsufficiency-dependent programming). Dysregulated OTR expression in the VTA is a possible mechanism for this reduced behavioral response. Our aim is to determine if behavioral differences are associated with altered OTR expression in DAergic and GABAergic cells within the VTA. Brains of 21 male mice from three genotypes, WT_offspring(WT_maternal), WT(H), KO(H) were frozen, sliced, and processed using established fluorescent in situ hybridization (fISH) protocol. Tissues were probed for OTR mRNA and either that of tyrosine hydroxylase (TH) for DAergic or vesicular GABA transporter (VGAT) for GABAergic cells respectively, and then imaged using a Leica confocal microscope with a 20x objective at 1.5 zoom with a 2.5 μm z-step. To quantify OTR in each cell type, acquired images were processed using ImageJ plugin BaSiC (Peng et al. 2017) to eliminate vignetting artifacts, and then a MATLAB program for automated colocalization in multiplex fISH tissues, Dotdotdot (Maynard et al. 2020). Image analysis is ongoing. We intend to assess percent colocalization and average OTR pixels across cell type and genotype, allowing us to determine what pattern of VTA OTR expression, if any, is associated with fmr1-related behavioral variance.

Voluntary exercise can have positive effects on mental health. In mice, voluntary wheel running increases stress resilience. Our lab found that protective effects of exercise may be intergenerational. Adult male offspring of dams with postpartum access to a running wheel showed greater resilience to acute, but not chronic stress relative to offspring of dams housed without a running wheel. These data suggest that maternal exercise may program resiliency to acute stress. Acute stress leads to reduced dendritic arborization, dendritic length (Trinchero et al, 2019), and spine density in the apical dendrites of hippocampal pyramidal cells of the CA3 region (Zhuang et al, 2019). Exercise leads to an increase in these parameters (Serra et al, 2019). Since runner dam offspring exhibited resilience to acute stress, we examined the structure of offspring CA3 pyramidal cells to determine whether maternal exercise is associated with morphological changes which may account for this observation. Hippocampal coronal sections taken from 6 mice of each group (Runner/Sedentary dam, chronically stressed/unstressed offspring) were stained using the Golgi method. Using a brightfield light microscope and Neurolucida (MBF Bioscience), we reconstructed 4 CA3 pyramidal cells/subject (24 neurons per group). We’ll conduct Sholl analysis for dendritic arborization and length and assess spine density by characterizing spines over certain dendritic length. We hypothesize that runner dam offspring will show increased dendritic arborization, dendritic length, and spine density relative to sedentary dam offspring. We expect that exposure to chronic stress will result in reduced dendritic arborization and spine density as chronically stressed mice failed to show acute stress resilience regardless of maternal exercise. This would suggest that maternal running may program acute stress resilience by promoting dendritic maturation, but that this effect is compromised by chronic stress exposure.

Post-traumatic stress disorder (PTSD) is characterized by overgeneralization of traumatic memories. Generalization is the transfer of learned responses to stimuli that are similar to, but not the same as, the original stimuli. Little is known about how sparse groups of neurons in the brain, known as “neuronal ensembles,” mediate fear memory generalization. To study these neuronal ensembles, ArcCreER^T2 x EYFP mice were bred and genotyped using a Polymerase Chain Reaction (PCR) assay developed in the lab. ArcCreER^T2 X EYFP positive mice were placed in dark housing for 24 hours. Following dark housing, mice were injected with 4-OHT to permanently genetically label cells during fear conditioning with Green Fluorescent Protein (GFP). Mice were fear conditioned using a 5-kHz tone (Conditioned Stimulus) and foot shock (Unconditioned Stimulus). Six (recent) or 30 (remote) days following fear conditioning, mice were re-exposed to the “target” 5-kHz tone or a novel 3-kHz tone in a disguised context. Brains were then extracted for immunohistochemistry. Two antibodies were used to visualize active cells during the original learning period (EYFP, 488 nm), and fear memory retrieval (Arc/arg3.1, 594 nm). Following immunohistochemistry, brain sections were imaged using a confocal microscope. Co-labelling of the two fluorophores indicated neurons active at both time points, allowing us to quantify neuronal ensembles underlying generalization. Preliminary results show lower freezing rates in the 3-kHz versus the 5-kHz recent group during fear memory recall, indicating discrimination. Experiments are ongoing to characterize fear memory generalization at a remote time point following learning.

Existing literature suggests that intergenerational storytelling is associated with better emotional regulation and improved problem solving abilities among children (Fivush, Bohanek, & Zaman, 2011). Acknowledging that there has been limited research exploring the content and impact of self-generated biographical knowledge, the present study looks into the intergenerational transmission of biographical knowledge in relation to mental wellbeing. We hypothesized that individuals who are able to recall more stories from their parents and grandparents would have better mental health outcomes overall. Additionally, previous work by Fivush (1989) suggested that stories passed along from fathers and mothers may differ in themes. We therefore expected to see differences reflected in the themes of generated stories and their relation to mental health variables depending on the gender of the selected parent. A total of 108 individuals aged between 18 to 25 who currently reside in the U.S. were recruited. Individuals were asked to recall 10 important events of their parents and 3 of their grandparents. Results show that individuals who recalled more categorical events (i.e. with less specific details) scored significantly higher for distress, depression, PTSD, stress, and anxiety. More negative, trauma and marriage related stories are generated for mothers whereas more career related stories are generated for fathers. A significantly positive correlation was also discovered between a higher proportion of career stories and worse mental health outcomes for male but not female participants. The results support the importance of intergenerational storytelling and suggests the benefits of detail-oriented exchange. Future research could incorporate multi-disciplinary efforts to better understand why storytelling impacts people of different genders differently.