The kappa opioid receptor (KOR) system is a promising target for substance use disorder, yet its role in long-term addiction regulation remains unclear. This project investigates how selective activation of the KOR/c-Jun N-terminal kinase (JNK) pathway activates the enzyme peroxiredoxin VI (PRDX6), triggering the release of reactive oxygen species (ROS) and resulting in long-lasting KOR inactivation distinct from its canonical Gαi pathway. I investigate whether JWT-101, a repurposed ligand, acts as a long-term KOR antagonist by inducing JNK-mediated ROS production, potentially offering new therapeutic avenues. KOR-Cre mice were injected in the prefrontal cortex with oROS-Gr, a fluorescent tag that senses ROS concentrations, for selective expression in KOR-positive neurons. Using high-resolution two-photon microscopy, I monitored ROS levels in live brain slices after 2 weeks from these mice. Bath application of JWT-101 led to increased fluorescence, indicating elevated ROS production and thus, JNK activation. To confirm JNK path specificity, I applied MJ33, an inhibitor of PRDX6. Fluorescence was reduced following MJ33 treatment, indicating that JWT-101 acts in a KOR/JNK manner. These findings suggest that JWT-101 induces KOR inactivation through ROS-mediated signaling. This research provides insights into KOR/JNK signaling in substance use disorders, with implications for developing targeted therapies for recovery and relapse prevention.

Frontline communities across and beyond the United States are continuously affected by the US nuclear legacy: from mines to test sites to radioactive waste disposal and beyond. Activists advocating for reparations and community safety often have to engage in their own research to provide the evidence the government requires as proof of radioactive harm. Most of this evidence exists only within physical documents that are behind multiple barriers, sometimes only recently declassified, and scattered across US nuclear sites. My research goal focuses on the creation of a living digital archive to help facilitate the digitization and sharing of important documentation among frontline communities. I use GitHub’s Collection Builder platform to create the foundation for an archive that can grow over time: including the first collection, the metadata schema, and information regarding the archive and how to get involved. My work includes selective digitization of documents of the now defunct UW Radiation Ecology Department, which was primarily funded by the US Atomic Energy Commission (AEC) and involved in a multitude of nuclear operations. These materials are housed in the UW Libraries Special Collections. Based on preliminary findings, I expect to find important evidence in this material of neglect to the harm of long-term exposure to low levels of radiation, and the specific testing done and results produced by the UW Radiation Ecology Department. My research also prototypes an archive that community researchers can continue to build and involves community feedback that helps shape the design. It is a step in the process of achieving necessary safety measures that protect people and ecosystems from continued radioactive harm.

Pregnant women infected with influenza A virus (IAV) are at higher risk of morbidity, mortality, and poor fetal outcomes. However, the difference in the pathogenesis of IAV between pregnant women and non-pregnant women remains inadequately understood, primarily due to the lack of animal studies that use a translational model of infection. I hypothesized that higher IAV viral load and Type I interferon concentrations would be observed in the lungs and bronchoalveolar lavage of pregnant pig-tail macaques compared to non-pregnant macaques, and that correlating these metrics would yield different results across groups. We inoculated pregnant (n=11) and non-pregnant female (n=18) pig-tail macaques (Macaca nemestrina) with IAV H1N1 (A/California/07/2009) and euthanized them at 5 days post-inoculation, when we expected to observe peak lung pathology. We tested pulmonary function at baseline and study endpoint and conducted clinical assessments daily. I extracted RNA and performed quantitative polymerase chain reactions on the samples to calculate viral load. I also performed enzyme-linked immunosorbent assays to quantify concentrations of Type I interferons (IFN-α, IFN-β). Lastly, I analyzed pulmonary physiology data and clinical assessment scores as a reliable measure of disease severity. A bi-modal distribution of viral load was observed in the lungs of pregnant animals (high>9e5 copies/mg; low<2e4 copies/mg), which was not observed in non-pregnant animals.  When correlating viral load at 5 days post-inoculation with Type I IFN in the lung of the pregnant animals, I found a significant positive correlation between IFN-β and viral load in both the lungs (ρ=0.8, p=0.03) and BAL (ρ=0.9, p=0.02). These results suggest that despite a strong IFN-β response in the lung, a high viral load persisted in the pregnant animals. Next steps could explore whether the kinetics of the pulmonary innate immune response is delayed in pregnancy, which impairs viral clearance. 

Chloroplasts are central to plant immunity and act as a hub for immune signalling and defence-related hormone synthesis. The essential chloroplast-localized protein FtsHi1 is a component of the FtsHi import motor and is vital to translocating proteins across the chloroplastic membrane. Viral-induced gene silencing (VIGS) of FtsHi1 in Nicotiana benthamiana results in a bleached phenotype, indicative of decreased translocation of essential chloroplastic proteins and decreased chlorophyll synthesis. Previous work identified herbivore-induced kinase 1 (HIK1) as a potential interactor of FtsHi1. HIK1 is a receptor-like cytoplasmic kinase (RLCK) implicated in the immune response to caterpillars. FtsHi1 contains a predicted RLCK phosphorylation site, indicating possible phosphorylation of FtsHi1 by HIK1, which could promote defence signalling over photosynthesis. This project aims to test the impact of phosphorylation on FtsHi1 function and its role in protein translocation across the chloroplastic membrane. This work utilises engineered FtsHi1 variants, which mutate the serine phosphosite to either mimic FtsHi1 phosphorylation (Ser→Asp), prevent it (Ser→Ala), or recapitulate the wild-type protein sequence. To test the phosphosite mutation effects, FtsHi1 VIGS knockdown of N. benthamiana leaves will be transiently modified using Agrobacterium tumefaciens bacteria to express WT or phosphorylation-modified FtsHi1 constructs. The resulting colour phenotype for each construct will then be compared to the bleached phenotype of TRV2:FtsHi1 plants and the phenotype of wild-type plants. I hypothesise that the FtsHi1 phospho-null mutant will result in a rescue phenotype similar to wild-type FtsHi1, whereas FtsHi1(phospho-mimic) will retain the bleached phenotype.Examining FtsHi1 phosphorylation enhances our understanding of its potential interaction with HIK1 in herbivory-induced plant defences. Future studies will explore FtsHi1's role in defence mechanisms, with implications for engineering herbivory-resistant crops. 

Kappa opioid receptor (KOR) ligands have been explored for anti-anxiolytic, anti-depressive, pain, and substance use disorder therapeutics. These therapeutic effects are partly due to biased signaling through the cJun N-terminal Kinase (JNK) pathway, which involves complex molecular interactions and downstream effects that inactivate the receptor by producing reactive oxygen species (ROS). JWT-101, a clinically approved compound, has been shown to produce therapeutic effects for these conditions. We hypothesize that its mechanisms of action are through KOR antagonism. I previously assessed KOR agonist-induced analgesia by measuring the latency of tail withdrawal from 52.5°C water after treatment with U50,488, a KOR agonist. Pretreatment with 15mg/kg JWT-101 24 hours before U50,488 injection effectively blocked KOR-induced analgesia in wild-type male mice. This effect was reversed by the short-acting, KOR-selective antagonist Aticaprant (5 mg/kg), suggesting that JWT-101’s action is mediated through KOR. Further investigation using in-vivo fiber photometry with the novel peroxide sensor AAV oROS-Gr revealed that JWT-101 significantly increases ROS production in KOR-expressing cells. Injection of 15 mg/kg of JWT-101 increases oROS fluorescence compared to control post-injection. Pretreatment with Aticaprant 15 minutes prior to JWT-101, blocks oROS fluorescence, suggesting that JWT-101’s activity is mediated by KOR. Pretreatment with MJ33 (a PRDX6 inhibitor upstream of JNK activity) 50 minutes before treatment with JWT-101 blocked oROS fluorescence, suggesting that this ROS production is through the JNK/PRDX6 pathway of KOR activation. This study provides insights into the mechanism of action of JWT-101 and examines the underlying molecular mechanisms of KOR-associated effects.

Plants utilize molecular patterns in order to detect threats to the plant. Through the recognition of molecular patterns by their associated receptor, plants are able to initiate an appropriate immune response, measurable by the increased production of reactive oxygen species (ROS). In the model organism Arabidopsis thaliana, the pathogen associated molecular pattern flg22 is detected by the pattern recognition receptor Flagellin Sensitive 2 (FLS2) to initiate intracellular signaling. The immune signaling kinase Botrytis-Induced Kinase 1 (BIK1) is then phosphorylated by FLS2 to transduce the immune signal, initiating ROS production. However, A. thaliana lacks a group of immune signaling kinases related to BIK1 called Herbivory-Induced Kinase Like Kinases (HLKs), which are present in multiple species including tobacco (Nicotiana benthamiana) and common bean (Phaseolus vulgaris). The goal of this project is to determine the role of HLKs in immune signaling. To investigate the role of HLKs in immune signaling, A. thaliana were transformed with transgenes to express HLKs or overexpress BIK1. Stably transformed A. thaliana lines were then bred to produce progeny that are homozygous for the transgenes. These lines are treated with flg22 in order to initiate an immune response. ROS is used to measure the immune response of each transgenic line, where HLK expressing individuals are compared with BIK1 overexpressing individuals as a control group. I anticipate that HLKs will increase the ROS production when compared with the controls, signifying an increased immune response, since HLKs are related to the native BIK1. An understanding of the role of HLKs in FLS2 immune signaling in the model organism A. thaliana can be applied to crop species that employ HLK mediated immune signaling.

Chronic pain affects about 20% of the adult population in the US, with more than 25% of these being pain that severely limit a person’s daily activities. In recent years, scientists in the field have been classifying pain as both a sensory response and emotional experience influenced by physiological and social factors. Newer research on pain behaviors and social behaviors have indicated that there is a positive association between the presence of cage mate in pain and the sensitivity to pain for a mouse. Although the behavioral responses are observed, the neural circuits mechanisms have yet to be examined. I will inject wild type mice with GCaMP in the medial prefrontal cortex (mPFC) and RCaMP in the basolateral amygdala (BLA). GCaMP and RCaMP are both genetically encoded Calcium indicators and are sensitive proxies for measuring excitatory transmission between brain regions. I will then implant fibers in both brain regions of all mice for fiber photometry recordings. After sensor expression time, I will check Calcium signals using a stressful stimulus known to stimulate excitatory pathways in mice then surgically induce pain in half of the mice. Mice will be split into chronic pain and pain-free groups, with their cage mate being either in pain or pain-free. I will perform a triad of behavioral pain testing simultaneously with fiber photometry recording, including tests for mechanical and thermal pain. I predict that for pain-free mice housed with a cage mate in pain, their pain threshold will decrease, as measured by all behavioral experiments. This should be accompanied by a stronger increase in BLA to mPFC Calcium signal when the mice are receiving painful stimuli. 

The ongoing opioid epidemic has made the need for alternative pain management strategies more urgent than ever. Nearly 1 in 5 Americans suffer from chronic pain, which has traditionally been treated with opioids and non-steroidal anti-inflammatory drugs (NSAIDs). However, both classes of drugs come with significant drawbacks. NSAIDs are often ineffective for managing chronic pain and can cause kidney and liver damage with prolonged use. Meanwhile, opioids lose their effectiveness over time, contributing to misuse, substance use disorders, and an increased risk of overdose. With few alternatives available that don't carry these risks, researchers are exploring new pain management options. One promising avenue is the use of cannabinoids, which are known for their anti-inflammatory and analgesic properties. In this study, I employ machine learning to create an unbiased kinematic and behavioral profile of mice experiencing chronic neuropathic pain using a custom-built linear track. Chronic pain and limb impairment are induced through partial sciatic nerve ligation, and a deep learning system analyzes videos of the mice to assess their movement patterns before and after treatment. I then compare these profiles to those of mice treated with NSAIDs, opioids, and cannabinoids, evaluating the effects of each treatment on behavioral measures like body position, which serves as a proxy for pain state and stress. We expect the mice treated with analgesics to show increased rearing and grooming behaviors. This research not only compares the analgesic effectiveness of cannabinoids to traditional pain-relief drugs but also helps reduce the stigma surrounding cannabinoid-based treatments.

The impact of microbial communities plays a large role in the lives of all organisms. Numerous thriving communities of microorganisms are present in places we often overlook. Our research investigates the identification of microbial communities present in the University of Washington Bothell's land, specifically the soil, leaves, and mushrooms commonly found around campus. We aim to support future microbiology lab students by identifying additional species that can be used as unknowns and address common issues students face when sequencing and identifying their known microbes. Our team collected 3 samples from 3 different environments, and each microbial strain was isolated, sequenced, and analyzed using Polymerase Chain Reaction (PCR) to amplify the 16s rRNA gene. The resulting genetic sequences obtained were then matched with NCBI BLAST to determine species identity. Additionally, microscopy and gram-staining were used to classify the bacterial isolates based on their structural characteristics. All collected data will be provided to future students to contribute to their understanding of microbial growth and increase the success rate of replicating and identifying their microbes. We aim to encourage further exploration of microbial life in commonly encountered environments and emphasize the significance of microbes in shaping the ecosystems around us. Understanding these microbial interactions can contribute to a broader understanding in fields such as microbial ecology, medicine, and public health.

Heartistry is a digital art gallery project dedicated to showcasing the journey of healing and empowerment for victim-survivors of relationship abuse. Heartistry, as a concept and in practice, embodies the powerful intersection of heart and artistry. As a victim-survivor myself, art became an outlet for me to express my feelings, reclaim power, and honor the complexity of resilience. My goal with this project is to establish a general understanding of how recovery is non-linear and deeply personal while also a universal human experience. I produced nine art pieces, utilizing three theoretical frameworks to explore different healing directions. Each framework -- Resilience Theory, Narrative Therapy, and Trauma-Informed Care -- guides three art pieces in direction, meaning, and intentionality. Project Heartistry consists of three sections: the art and description connecting it to the framework, resources for victim-survivors and their supporters as well as resources to learn more about each framework, and an about page with the intention of the project. My hope is that other victim-survivors may feel empowered to share their art with me and the site can adapt into a gallery of many artist's work, but for now, it is just my nine pieces. It was incredibly important for me to approach the art, my language, and site design sensitively, intentionally, and inclusively. The biggest takeaways from Heartistry is that healing is not linear, victim-survivors are not alone, and art is therapeutic. 

The rapid evolution of antimicrobial resistance (AMR) in bacteria poses a critical global health challenge, predicted to cause 10 million deaths annually by 2050 if left unaddressed. AMR genes frequently reside on plasmids– small, circular DNA separate from bacterial chromosomes. These plasmids spread between bacteria through horizontal gene transfer (HGT), where genetic material moves directly from one cell to another, rapidly disseminating resistance genes across populations and species. In contrast, vertical gene transfer (VGT) occurs during bacterial reproduction, passing genes from parent to daughter cells. The machinery plasmids use for HGT imposes a fitness cost on the host, slowing its growth and reproduction (VGT). This means plasmids typically face a trade-off: investing resources in HGT limits the host’s ability to reproduce efficiently through VGT. My research uncovered a “trade-off-breaking mutation” that simultaneously enhances both HGT and VGT, accelerating the spread of AMR genes. Such mutations have significant public health implications, potentially leading to highly virulent, drug-resistant bacterial strains. I am creating a genotype-to-phenotype map to link specific plasmid mutations to their effects on HGT and VGT rates, aiming to understand the dynamics of resistance spread. This work involves verifying mutations in our mutant plasmid library using targeted sequencing techniques and applying the Luria-Delbrück method, a specialized approach developed by my mentor, Dr. Olivia Kosterlitz, to measure gene transfer rates. By analyzing these mutations, I seek to uncover how some plasmids avoid the typical trade-offs, enabling them to reproduce quickly while spreading resistance efficiently. Understanding the relationship between HGT and VGT is critical for predicting how antibiotic resistance evolves and for developing strategies to slow its spread. This research reveals the importance of trade-off-breaking mutations in resistance management, providing new insights into how we might combat one of our time's greatest public health challenges.

Bacteria can shuttle pieces of DNA between unrelated cells via a process called horizontal gene transfer (HGT). Genes that undergo HGT (i.e. mobile genes) evolve in different host bacteria with different genomic backgrounds, which can influence the types of mutations the mobile gene acquires. Studying the effect of HGT on mobile gene evolution is important as many clinically relevant antibiotic resistance genes are mobile. In a prior study, we used a simple model to simulate mobile gene evolution as they engage in HGT. Under the simple model, the mobile gene evolves in only one species at a time. With this model, we found that fitness landscape similarity between two host species engaging in HGT is highly indicative of the effect HGT has on mobile gene fitness outcomes (i.e. whether performing HGT has a positive, negative, or neutral effect on fitness). We expanded the simple model into a more ecologically realistic consumer-resource model (CRM), in which the mobile gene continuously transfers between species. We observed similar outcomes between the two models; however, in the CRM there was an increase in cases in which performing HGT had a positive fitness effect. We hypothesize that the CRM highlights features like the continuous existence of host species, resulting in constant gene flow between the two species. To further probe how gene flow influences the effect HGT has on mobile gene evolution, I tested how varying the HGT rate with the CRM (effectively allowing us to control the amount of gene flow) affects mobile gene fitness outcomes. I used the same host landscape pairs used in our pilot study while varying the HGT rate along a biologically relevant range. I expect to find a positive correlation between HGT rate and the magnitude of positive fitness effects conferred by a mobile gene that has undergone HGT.

In our biochemistry teaching labs, students conduct 10-week projects using recombinant protein expression and purification protocols, adapted from Fred Hutch, distributed and tracked via GENI-ACT.org, to identify immunoproteins of research or biomedical interest. We hypothesize they can produce antigen fragments for antibody studies and siderocalin proteins, which bind bacterial siderophores, yielding different amounts and results. In Winter 2023, students modeled antibody fragments with I-TASSER, expressed top constructs with His-tags, and purified them using Ni-NTA resin. In Winter and Fall 2024, siderocalins were expressed as GST-tagged constructs in BL21 and DH5alpha cells using longer expression. The human siderocalin in DH5alpha formed an orange solution, consistent with known siderocalin-enterobactin-Fe complexes. Unexpectedly, other species’ siderocalins appeared yellow, pink, or blue, suggesting functional diversity. Students produced enough immunoproteins for viability tests and are now expressing homologs of the blue siderocalin. They participated in all stages, developing spectroscopy and protein crystallization skills for research careers.

Mutations, which arise spontaneously, are the foundation of genetic variation and play a key role in evolution. Understanding mutation dynamics has relevance for public health, as antibiotic resistance in bacteria often results from genetic mutations that allow them to thrive in the presence of drugs that would typically inhibit their growth. Our research builds on the Luria-Delbrück method, originally designed to estimate mutation rates phenotypically, by using Next-Generation Sequencing (NGS) to measure base-level mutation rates in Escherichia coli that confer resistance to rifampicin. Rifampicin targets the β-subunit of RNA polymerase, and resistance arises from single nucleotide mutations in the rpoB gene. My team and I conducted experiments by inoculating E. coli populations, exposing them to rifampicin at specific times, and sequencing resistant mutants to calculate mutation rates for each base change. Interestingly, our data revealed that identical base changes at different genomic positions can have significantly different mutation rates. However, our mutation rate estimation does assume that every mutant cell has the same probability of establishing a lineage in the presence of rifampicin. If a certain mutant has a lower probability of lineage survival, its mutation rate will be underestimated. Thus, to determine whether the mutation rate variability we found is due to actual differences and not survival differences, I developed an assay to measure the probability that a mutant fails to establish a lineage. To date, I have isolated nine distinct rifampicin-resistant mutants and tested the extinction rates of two, finding no observable extinction, supporting the accuracy of our mutation rate estimates for these mutants. This research refines mutation rate calculations and enhances our understanding of bacterial adaptation, with implications for developing strategies to predict and mitigate antibiotic resistance. Additionally, it contributes to evolutionary biology by revealing the complexities of mutation and survival in microbial populations

Using temperature reconstruction to understand how the Earth’s climate responded to external forcing from factors such as CO2 in the past can inform predictions about future climate change due to global warming. This project aims to examine a recent paleoclimate data assimilation study of the past 24,000 years from the Last Glacial Maximum (LGM) to the present day. Paleoclimate data assimilation combines both proxy data and climate model simulations to address the discrepancies in climate reconstructions produced by each. For the LGM to present, discrepancies between model simulations and proxy data include the timing and characteristics of climate events like deglaciation. While data assimilation helps to resolve some of these discrepancies, it also makes assumptions about the uncertainty of the proxy data used. Processes that introduce proxy uncertainty such as bioturbation–sediment mixing by marine organisms–and calibration errors are often not characterized as time scale-dependent which could potentially introduce bias and affect the accuracy of these data assimilation studies. We examine the proxy uncertainty within this data assimilation to identify timescale-dependent errors and measure their impact on the accuracy of the temperature reconstruction. We do this by producing a set of pseudoproxies, which are synthetic datasets of different sediment proxies such as δ¹⁸O, to create hypothetical systems of past climate. By isolating and controlling different uncertainty characteristics, we are able to measure their overall impact on the climate reconstructions.