bioRxiv Science⌕ Search

Biology subjects

Park, S.

Publications and source records attributed to Park, S..

At least 127 records · Page 7Linked to original sources

A model of inborn metabolism errors associated with adenine amyloid-like fiber formation reduces TDP-43 aggregation and toxicity in yeast

TDP-43 is linked to human diseases such as amyotrophic lateral sclerosis (ALS) and frontotemporal degeneration (FTD). Expression of TDP-43 in yeast is known to be toxic, cause cells to elongate, form liquid-like aggregates, and inhibit autophagy and TOROID formation. Here, we used the apt1{Delta} aah1{Delta} yeast model of disorders of inborn errors of metabolism, previously shown to lead to intracellular adenine accumulation and adenine amyloid-like fiber formation, to explore interactions with TDP-43. Results show that the double deletion shifts the TDP-43 aggregates from a liquid-like, toward a more amyloid-like, state. At the same time the deletions reduce TDP-43s effects on toxicity, cell morphology, autophagy, and TOROID formation without affecting the level of TDP-43. This suggests that the liquid-like and not amyloid-like TDP-43 aggregates are responsible for the deleterious effects in yeast. How the apt1{Delta} aah1{Delta} deletions alter TDP-43 aggregate formation is not clear. Possibly, it results from adenine/TDP-43 fiber interactions as seen for other heterologous fibers. The work offers new insights into the potential interactions between metabolite-based amyloids and pathological protein aggregates, with broad implications for understanding protein misfolding diseases.

genetics↗

Leucine zipper-based SAIM imaging identifies therapeutic agents to disrupt the cancer cell glycocalyx for enhanced immunotherapy

The abnormally thick glycocalyx of cancer cells can physically limit immune-cell engagement and reduce the efficacy of cellular immunotherapies, but strategies for identifying pharmacological agents that relieve this barrier remain limited. Here, we establish glycocalyx thickness as a pharmacologically actionable physical phenotype. Disruption of mucin-type O-glycosylation reduced glycocalyx thickness and increased NK- and CAR-NK-mediated killing, demonstrating that pharmacological glycocalyx remodeling can enhance innate and CAR-directed cytotoxicity. To screen this phenotype directly, we developed a scanning angle interference microscopy (SAIM) platform using complementary leucine zipper pairs as a stable extracellular plasma-membrane reference. A focused screen of glycosylation and metabolic inhibitors resolved compound- and cell-type-specific nanoscale changes in glycocalyx thickness and identified several thinning agents. Castanospermine, an inhibitor of early N-glycan processing, was among the strongest hits and reduced mucin-probe binding in an independent mucin-rich gastric cancer model. Castanospermine-treated gastric cancer cells retained MUC17 antibody recognition and showed increased physical avidity for MUC17 CAR-T cells, while castanospermine enhanced MUC17 CAR-T-mediated tumor control in real-time co-culture assays. These findings establish direct glycocalyx-thickness measurement as a physical-phenotype-guided screening strategy and identify castanospermine as a candidate modulator for combination with MUC17 CAR-T cells.

bioengineering↗

Investigating the consequences of chronic short sleep for metabolism and survival of oxidative stress

In previous work, we found that short sleep caused sensitivity to oxidative stress; here we set out to characterize the physiological state of a diverse group of chronically short-sleeping mutants during hyperoxia as an acute oxidative stress. Using RNA-sequencing analysis, we found that short-sleeping mutants had a normal transcriptional oxidative stress response relative to controls. In both short-sleeping mutants and controls, hyperoxia led to downregulation of glycolytic genes and upregulation of genes involved in fatty acid metabolism, reminiscent of metabolic shifts during sleep. We hypothesized that short-sleeping mutants may be sensitive to hyperoxia because of defects in metabolism. Consistent with this, short-sleeping mutants were sensitive to starvation. Using metabolomics, we identified a pattern of low levels of long chain fatty acids and lysophospholipids in short-sleeping mutants relative to controls during hyperoxia, suggesting a defect in lipid metabolism. Though short-sleeping mutants did not have common defects in many aspects of lipid metabolism (basal fat stores, usage kinetics during hyperoxia, respiration rates, and cuticular hydrocarbon profiles), they were all sensitive to dehydration, suggesting a general defect in cuticular hydrocarbons, which protect against dehydration. To test the bi-directionality of sleep and lipid metabolism, we tested flies with both diet-induced obesity and genetic obesity. Flies with diet-induced obesity had no sleep or oxidative stress phenotype; in contrast, the lipid metabolic mutant, brummer, slept significantly more than controls but was sensitive to oxidative stress. Previously, all short sleepers tested were sensitive and all long sleepers resistant to oxidative stress. brummer mutants, the first exceptions to this rule, lack a key enzyme required to mobilize fat stores, suggesting that a defect in accessing lipid stores can cause sensitivity to oxidative stress. Taken together, we found that short-sleeping mutants have many phenotypes in common: sensitivity to oxidative stress, starvation, dehydration, and defects in lipid metabolites. These results argue against a specific role for sleep as an antioxidant and suggest the possibility that lipid metabolic defects underlie the sensitivity to oxidative stress of short-sleeping mutants.

physiology↗

EasyPubPlot: a shiny web application for rapid omics data exploration and visualization

Computational toolkits for data exploration and visualization from widely used omics platforms often lack flexibility and customization. While many tools generate standardized output, advanced programming skills are necessary to create high-quality visualizations. Therefore, user-friendly tools that simplify this crucial yet time-consuming step are essential. We developed EasyPubPlot (Easy Publishable Plotting), a straightforward, easy-to-use, user experience-oriented, open-source, and shiny web application along with its associated R package to streamline data exploration and visualization for functional omics-empowered research. EasyPubPlot generates publishable scores plots, volcano plots, heatmaps, box plots, dot plots, and bubble plots with minimal necessary steps. The tool was designed to guide new users for accurate and efficient navigation. Step-by-step tutorials for each type of plot are also provided. Herein, we demonstrated EasyPubPlots competent functionality and versatility by showcasing metabolomics and transcriptomics data. Collectively, EasyPubPlot reduces the gap between data analysis and stunning visualization, thereby diminishing friction and focusing on science. The app can be downloaded and installed locally (https://github.com/Pharmaco-OmicsLab/EasyPubPlot) or used through a web application (https://pharmaco-omicslab.shinyapps.io/EasyPubPlot).

bioinformatics↗

In vivo CRISPR screens identify key modifiers of CAR T cell function in myeloma

Chimeric antigen receptor (CAR) T cells are highly effective in hematologic malignancies. However, loss of CAR T cells can contribute to relapse in a significant number of patients. These limitations could potentially be overcome by targeted gene editing to increase CAR T cell persistence. Here, we performed in vivo loss-of-function CRISPR screens in BCMA-targeting CAR T cells to investigate genes that influence CAR T cell persistence, function and efficacy in a human multiple myeloma model. We tracked the expansion and persistence of CRISPR-library edited T cells in vitro and then at early and late timepoints in vivo to track the performance of gene modified CAR T cells from manufacturing to survival in tumors. The screens revealed several context-specific regulators of CAR T cell expansion and persistence. Ablation of RASA2 and SOCS1 enhanced T cell expansion in vitro, while loss of PTPN2, ZC3H12A, and RC3H1 conferred early selective growth advantages to CAR T cells in vivo. Strikingly, we identified cyclin-dependent kinase inhibitor 1B (CDKN1B), a cell cycle regulator, as the most important factor limiting CAR T cell fitness at late timepoints in vivo. CDKN1B ablation increased BCMA CAR T cell proliferation and effector function in response to antigen, significantly enhancing tumor clearance and overall survival. Thus, our findings reveal differing effects of gene-perturbation on CAR T cells over time and in different selective environments, highlight CDKN1B as a promising target to generate highly effective CAR T cells for multiple myeloma, and underscore the importance of in vivo screening as a tool for identifying genes to enhance CAR T cell function and efficacy.

immunology↗

Human heart assembloids with autologous tissue-resident macrophages recreate physiological immuno-cardiac interactions

Interactions between the developing heart and the embryonic immune system are essential for proper cardiac development and maintaining homeostasis, with disruptions linked to various diseases. While human pluripotent stem cell (hPSC)-derived organoids are valuable models for studying human organ function, they often lack critical tissue-resident immune cells. Here, we introduce an advanced human heart assembloid model, termed hHMA (human heart-macrophage assembloid), which fully integrates autologous cardiac tissue- resident macrophages (MPs) with pre-existing human heart organoids (hHOs). Through multi-omic analyses, we confirmed that these MPs are phenotypically similar to embryonic cardiac tissue-resident MPs and remain viable in the assembloids over time. The inclusion of MPs significantly impacts hHMA development, influencing cardiac cellular composition, boosting cellular communication, remodeling the extracellular matrix, promoting ventricular morphogenesis, and enhancing sarcomeric maturation. Our findings indicate that MPs contribute to homeostasis via efferocytosis, integrate into the cardiomyocyte electrical system, and support catabolic metabolism. To demonstrate the versatility of this model, we developed a platform to study cardiac arrhythmias by chronic exposure to pro-inflammatory factors linked to arrhythmogenesis in clinical settings, successfully replicating key features of inflammasome-mediated atrial fibrillation. Overall, this work introduces a robust platform for examining the role of immune cells in cardiac development, disease mechanisms, and drug discovery, bridging the gap between in vitro models and human physiology. These findings offer insights into cardiogenesis and inflammation-driven heart disease, positioning the hHMA system as an invaluable tool for future cardiovascular research and therapeutic development.

bioengineering↗

Encoding opposing valences through frequency-dependent transmitter switching in single peptidergic neurons

Peptidergic neurons often co-express fast transmitters and neuropeptides in separate vesicles with distinct release properties. However, the release dynamics of each transmitter in various contexts have not been fully understood in behaving animals. Here, we demonstrate that calcitonin gene-related peptide (CGRP) neurons in the external lateral subdivision of the parabrachial nucleus (CGRPPBel) encode opposing valence via differential release, rather than corelease, of glutamate and neuropeptides, according to firing rate. Glutamate is released preferentially at lower firing rates with minimal release at higher firing rates, whereas neuropeptides are released at higher firing rates, resulting in frequency-dependent switching of transmitters. Aversive stimuli evoke high frequency responses with accompanying neuropeptide release to encode negative valence, whereas appetitive stimuli evoke low frequency responses with glutamate release to encode positive valence. Our study reveals a previously unknown capability of single CGRPPBel neurons to bidirectionally encode valence via frequency-dependent differential release of transmitters in vivo.

neuroscience↗

Aberrant expression of CD14 in breast tumor cells is associated with poor outcome

Emerging evidence suggests that cancer cells can mimic features of immune cells during oncogenic transformation to drive disease progression. We assessed the occurrence of immunological markers in breast cancer cells to determine their expression pattern. We initially analyzed 18 immune protein markers (CCR4, CCR6, CCR7, CD11, CD123, CD14, CD16, CD19, CD24, CD25, CD27, CD3, CD38, CD4, CD45, CD56, CD8 and CXCR3) expressed on the surface of 28 breast cancer cell lines using mass spectrometry. CD14 protein expression in tumor cells and its association with clinical outcomes was subsequently evaluated by tissue microarray (TMA) analysis of 346 breast tumors. Single-cell RNA sequencing data from breast cancer tumors and bulk transcriptomic data of breast cancer cell lines were interrogated for molecular signatures associated with CD14 tumor cell expression. Among the markers interrogated, CD14 protein was aberrantly expressed on the surface of 13 out of 15 triple-negative breast cancer (TNBC), one of six hormone receptor positive (HR+), one of five hormone receptors negative (HR-)/Her2+ cell lines. Likewise, RNA expression revealed higher levels of CD14 in TNBC cell lines compared to other subtypes. Tumor tissue microarray analysis revealed elevated levels of CD14 membrane expression predominantly in TNBC and was associated with higher tumor grade and increased incidence of disease recurrence compared to CD14-negative tumors. The CD14-positive subgroup exhibited Nuclear Factor Kappa Beta (NFkB) and TGF-{beta} centric networks at both the protein and Single-cell RNA levels. We have uncovered a novel subset of breast cancers characterized by aberrant surface expression of CD14 associated with aggressive disease. CD14 identifies a subset of breast cancers with poor outcome and is a potential therapeutic target.

cancer biology↗

Reengineering mRNA lipid nanoparticles for systemic delivery to pancreas

Lipid nanoparticles (LNPs) hold transformative potential for nucleic acid delivery, with applications ranging from clinical use, particularly in COVID-19 vaccines, to gene therapy and cancer immunotherapy. Traditional LNPs are composed of four components: ionizable lipids, cholesterol, helper lipids, and PEGylated lipids. However, a significant hurdle remains the need for more efficient and selective systemic delivery vehicles, as most targets are challenging to reach and primarily accumulate in the liver following intravenous administration, largely due to Apolipoprotein E (ApoE) mediated uptake in the blood. Recent studies have shown that introducing cationic or anionic lipids as a fifth component in the LNPs has resulted in lung- and spleen-specific mRNA expression. In this study, we report that incorporating endogenous ligands, such as vitamins, as a fifth component can enhance the extrahepatic localization of LNPs. Vitamins are highly biocompatible, possess excellent targeting potential, influence immune response, improve cellular uptake, and increase the stability of mRNA LNPs. We developed a library of 100 LNPs containing luciferase mRNA by partially replacing ionizable lipids with this fifth component and evaluated their efficacy both in vitro and in vivo. From comprehensive batch analysis screening, we identified two formulations with cholecalciferol (vitamin D3) as a fifth component that demonstrated selective systemic delivery to the pancreas (> 99%) with high efficacy. Among these, C-CholF3 emerged as the top formulation, exhibiting robust and sustained protein expression in the pancreas for up to 3 days in a dose-dependent manner with minimal toxicity that makes it suitable for repeated administration. Furthermore, C-CholF3 also demonstrated pancreas-specific gene editing in the Ai14 transgenic mouse model, showing high expression of TdTomato. This underscores its translational potential for protein replacement and CRISPR/Cas9-mediated gene editing in currently incurable pancreatic diseases.

bioengineering↗

Deep Gaussian Process with Uncertainty Estimation for Microsatellite Instability and Immunotherapy Response Prediction Based on Histology

Determining tumor microsatellite status has significant clinical value because tumors that are microsatellite instability-high (MSI-H) or mismatch repair deficient (dMMR) respond well to immune check-point inhibitors (ICIs) and oftentimes not to chemotherapeutics. We propose MSI-SEER, a deep Gaussian process-based Bayesian model that analyzes H&E whole-slide images in weakly-supervised-learning to predict microsatellite status in gastric and colorectal cancers. We performed extensive validation using multiple large datasets comprised of patients from diverse racial backgrounds. MSI-SEER achieved state-of-the-art performance with MSI prediction, which was by integrating uncertainty prediction. We achieved high accuracy for predicting ICI responsiveness by combining tumor MSI status with stroma-to-tumor ratio. Finally, MSI-SEERs tile-level predictions revealed novel insights into the role of spatial distribution of MSI-H regions in the tumor microenvironment and ICI response.

bioinformatics↗

APOBEC3A, not APOBEC3B, drives deaminase mutagenesis in human gastric epithelium

Cancer genomes frequently carry APOBEC (apolipoprotein B mRNA editing catalytic polypeptide-like)-associated DNA mutations, suggesting APOBEC enzymes as innate mutagens during cancer initiation and evolution. However, the pure mutagenic impacts of the specific enzymes among this family that are responsible for APOBEC-associated mutagenesis remain unclear in human normal cell lineage, particularly the comparative mutagenic activities of APOBEC3A and APOBEC3B. Here, we investigated the mutagenic contributions of these enzymes through whole-genome sequencing of human normal gastric organoid lines carrying doxycycline-inducible APOBEC3A or APOBEC3B cassettes. Our findings demonstrated that transcriptional APOBEC3A upregulation led to the acquisition of a massive number of genomic mutations in a few cell cycles. By contrast, APOBEC3B upregulation did not generate a substantial number of mutations in gastric epithelium. APOBEC3B-associated mutagenesis remained insignificant even after a combined inactivation of TP53. Based on the spectrum of acquired mutations after APOBEC3A upregulation, we further analyzed APOBEC3A-associated mutational signatures, encompassing indels mainly composed of 1bp deletions, characteristics of clustered mutations, and selective pressures operative on cells carrying the mutations. Our observations provide a clear foundation for understanding the mutational impact of APOBEC enzymes in human cells.

genomics↗

Mechanistic genotype-phenotype translation using hierarchical transformers

Genome-wide association studies have linked millions of genetic variants to biomedical phenotypes, but their utility has been limited by lack of mechanistic understanding and widespread epistatic interactions. Recently, Transformer models have emerged as powerful machine learning architectures with potential to address these and other challenges. Here we introduce the Genotype-to-Phenotype Transformer (G2PT), a framework for modeling hierarchical information flow among variants, genes, multigenic systems, and phenotypes. As proof-of-concept, we train G2PT to model the genetics of metabolic traits including insulin resistance (serum triglycerides-to-HDL ratio), LDL and type-2 diabetes. G2PT predicts these traits with accuracy exceeding state-of-the-art and, unlike other polygenic models, extends to distinct populations not used for training. Predictions of insulin resistance are based on >1,395 variants within 20 systems and include epistatic interactions among variants, e.g. between APOA4 and CETP in phospholipid transfer. This work positions hierarchical graph transformers as a next-generation approach to polygenic risk.

bioinformatics↗

Innovative Strategies for Enhancing Signal-to-Noise Ratio in Single-molecule Imaging: The Influence of Molecular Motion

Single-molecule fluorescence imaging has extensively revealed the dynamic and structural characteristics of biomolecules. However, its application is limited by the upper concentration of fluorophore-tagged biomolecules, which is in the sub-ten nanomolar range. We found that the signal-to-noise ratio (SNR) in single-molecule fluorescence imaging is strongly influenced by the size of fluorophore-labeled molecules in solution. Our computational simulations suggest that the faster diffusion of background fluorophores can enhance the SNR of target molecules. Moreover, we identified that the molecular motion through fluid flow can improve SNR. This study provides a novel perspective by emphasizing the importance of molecular motion in SNR and propose a rapid barrier-free method to increase the upper concentration limit in single-molecule imaging.

biophysics↗

The Composition of the Bacterial Community in Raw Milk from Holstein Dairy Cattle Correlated with the Occurrence of Klebsiella pneumoniae Clinical Mastitis Infections

Klebsiella pneumoniae is a common, opportunistic bacterial pathogen that can cause severe clinical mastitis in dairy cattle. Optimizing the bovine udder microbiome to resist mastitis pathogens is a growing area of research; however, previous work has not examined which members of the mammary microbiome may have antagonistic interactions with K. pneumoniae. In this study, we collected quarter-level milk samples from Holstein dairy cows in Quebec, Canada every two weeks for 14 months and analyzed differences in the milk microbiome between samples that were collected from healthy quarters, quarters that developed subclinical mastitis, and quarters that experienced K. pneumoniae clinical mastitis (KP-CM) (n = 512 milk samples). The occurrence of subclinical mastitis did not cause significant differences in the alpha-diversity of the milk microbiome, nor did subclinical mastitis alter the interactions between taxa in the microbiome. However, the occurrence of KP-CM caused reductions in Shannon diversity in raw milk relative to healthy milk and altered the interactions between taxa. Specifically, K. pneumoniae showed negative interactions with the genus Aerococcus. The negative interactions between Aerococcus spp. and K. pneumoniae in the context of the bovine milk microbiome should be analyzed further.

microbiology↗

Distinct Disruptions in CA1 and CA3 Place Cell Function in Alzheimer's Disease Mice

The hippocampus, a critical brain structure for spatial learning and memory, is susceptible to neurodegenerative disorders such as Alzheimers disease (AD). The APPswe/PSEN1dE9 (APP/PS1) transgenic mouse model is widely used to study the pathology of AD. Although previous research has established AD-associated impairments in hippocampal-dependent learning and memory, the neurophysiological mechanisms underlying these cognitive dysfunctions remain less understood. To address this gap, we investigated the activities of place cells in both CA1 and CA3 hippocampal subregions, which have distinct yet complementary computational roles. Behaviorally, APP/PS1 mice demonstrated impaired spatial recognition memory compared to wild-type (WT) mice in the object location test. Physiologically, place cells in APP/PS1 mice showed deterioration in spatial representation compared to WT. Specifically, CA1 place cells exhibited significant reductions in coherence and spatial information, while CA3 place cells displayed a significant reduction in place field size. Both CA1 and CA3 place cells in APP/PS1 mice also showed significant disruptions in their ability to stably encode the same environment. Furthermore, the burst firing properties of these cells were altered to forms correlated with reduced cognition. Additionally, the theta rhythm was significantly attenuated in CA1 place cells of APP/PS1 mice compared to WT. Our results suggest that distinct alteration in the physiological properties of CA1 and CA3 place cells, coupled with disrupted hippocampal theta rhythm in CA1, may collectively contribute to impaired hippocampal-dependent spatial learning and memory in AD.

neuroscience↗

Alzheimer's pathology disrupts flexible place cell coding and hippocampal-prefrontal neural dynamics during risky foraging decisions in mice

This study investigates how amyloid pathology influences hippocampal-prefrontal neural dynamics and decision-making in Alzheimers disease (AD) using 5XFAD mice, a well-established model system characterized by pronounced early amyloid pathology. Utilizing ecologically-relevant "approach food-avoid predator" foraging tasks, we observed that 5XFAD mice exhibited persistent risk-taking behaviors and reduced adaptability to changing threat conditions, indicative of impaired decision-making. Multi-regional neural recordings revealed rigid hippocampal CA1 place-cell fields, decreased sharp-wave ripple (SWR) frequencies, and disrupted medial prefrontal-hippocampal connectivity, all of which corresponded with deficits in behavioral flexibility during spatial risk scenarios. These findings highlight the critical role of SWR dynamics and corticolimbic circuit integrity in adaptive decision-making, with implications for understanding cognitive decline in AD in naturalistic contexts. By identifying specific neural disruptions underlying risky decision-making deficits, this work provides insights into the neural basis of cognitive dysfunction in AD and suggests potential targets for therapeutic intervention.

neuroscience↗

Inositol Hexaphosphate (InsP6) Activates the HDAC1/3 Epigenetic Axis toMaintain Intestinal Barrier Function

While HDAC inhibition shows promise in cancer treatment, pan-HDAC inhibitors cause gastrointestinal issues in 48% of patients. Understanding HDAC activation mechanisms is crucial to treat diverse diseases beyond cancer. Our study reveals the essential role of inositol polyphosphate multikinase (IPMK) and inositol hexakisphosphate (InsP6 or phytic acid), enriched in vegan diets, in activating the HDAC3 epigenetic axis and maintaining intestinal barrier integrity. IPMK binds to HDAC3, driving InsP6 synthesis, which selectively activates HDAC3 at 10nM concentration by recruiting the DAD domain of its corepressor protein. IPMK deletion diminishes HDAC3 activation, leading to histone hyperacetylation and MMP gene transcription, compromising intestinal barrier integrity. InsP6 treatment is sufficient to rescue these effects. In inflammatory bowel disease, diminished IPMK levels exacerbated intestinal permeability, while oral InsP6 treatment mitigated gut permeability by restoring the HDAC3 epigenetic axis, indicating the clinical implications of the IPMK-HDAC3 epigenetic axis and therapeutic potential of phytic acid.

cell biology↗

TAAR2-9 Knockout Mice Exhibit Reduced Wakefulness and Disrupted REM Sleep

Trace amine-associated receptor 1 (TAAR1) has gained attention for its roles in modulating neural systems, sleep/wake control, and as a therapeutic target for neuropsychiatric disorders. Although TAARs 2-9 were initially identified as non-canonical olfactory receptors, recent studies have identified extra-nasal receptor distribution of multiple TAARs. To evaluate whether TAARs 2-9 have a role in arousal state regulation, we investigated sleep/wake control in male TAAR2-9 knockout (KO) mice. After determination of baseline sleep/wake patterns, the homeostatic response to sleep deprivation and response to TAAR1 agonists were compared between KO and C57BL/6J mice. Although the EEG of TAAR2-9 KO mice had lower power in the delta and theta bands and higher power in the gamma range, sleep/wake states were readily identified. KO mice had more NREM sleep during the dark phase and more REM sleep during the light phase. Sleep/wake was fragmented in KO mice with shorter Wake and REM bouts during the dark phase and more REM bouts during the light phase. KO mice exhibited more REM sleep during a sleep latency test but the homeostatic response to sleep loss did not differ between the strains. A high dose of the TAAR1 agonist RO5256390 increased Wake and reduced NREM sleep in KO mice whereas RO5256390 and the partial TAAR1 agonist RO5263397 suppressed REM sleep. The number of tyrosine hydroxylase-immunoreactive neurons in the ventral tegmental area was significantly elevated in KO mice. These dopaminergic and sleep/wake alterations in TAAR2-9 KO mice highlight the need for further elucidation of the functions of TAAR2-9.

neuroscience↗