Introduction & Objective: Lower body fat distribution decreases risk for T2D, but the mechanisms involved are poorly understood. The objective of this study was to compare AP populations of paired samples of upper body abdominal (A) vs lower body thigh (T) adipose tissues that may contribute to this association. Methods: Single nuclei RNA sequencing (snRNASeq) identified AP subclusters (Seurat); bioinformatic and within subject statistical comparisons used pseudobulk analyses and paired ttest. Cell-Cell communication analysis (CellChat) addressed depot-dependent mechanisms tissue homeostasis. Results: Healthy, premenopausal women in this study were n=7, BMI (28.7+/- 6.3 kg/m2), age 30 +/-5.5 y, waist to hip ratio 0.79 +/- 0.02 (mean ± SD). APs (6489 (A), 7194 (T)) in 9 subclusters varied in their degree of adipogenic commitment. An early AP population expressed higher levels of CD55 in A than T, and low MGP and AREG/F3. Two other ‘early’ AP clusters expressed the highest levels of anti-adipogenic factors, AREG/F3 AP were more abundant in T (9% vs 6%, p<0.007; 5.4 vs 3.7%, p=0.025). A small population (5%) of committed ‘preadipocytes’ was in both depots. Additional clusters had high expression of secreted adipokines and extracellular matrix factors and secretory pathways (Collagens, Lumicans). We found evidence for higher cell-cell communication in T (FGF-FGFR, IGF-IGFR1, PDGF-PDGFR). Conclusion: Our results suggest that higher expression of anti-adipogenic factors in TSAT maintain a larger pool of APs that can respond to pro-adipogenic signals, as needed, to better preserve tissue and systemic metabolic homeostasis in response to a stress (e.g. obesity). Disclosure Y. Sun: None. G. Smith: None. J. Albu: None. M. Walsh: None. S.K. Fried: None. Funding National Institutes of Health (R01DK121547)

The US prevalence of severe obesity [(SevO), body mass index [BMI] ≥40 kg/m 2 )] is increasing at an alarming rate; women and Hispanic Latino populations have experienced among the greatest increases. While genome-wide association studies (GWAS) have identified >1000 loci associated with body mass index, the function of much of this variation is unknown. Gene expression measures can ] link genetic variation and disease highlighting pathways for targeted therapeutic intervention, but to-date, few studies have examined the role of gene expression to identify molecular signatures associated with SevO. To this end, we leveraged extant whole blood (WB) RNA sequencing (RNAseq) data in 75 SevO cases and 116 controls (with BMI = 18-25) collected from randomly selected Mexican Americans in the Cameron County Hispanic Cohort (CCHC) to identify patterns associated with SevO. We used established protocols and alignment, yielding 18,565 genes after quality control. We applied DESeq2 to assess DE associated with SevO, using a negative binomial regression model with a gene-specific dispersion parameter, adjusted for sex, age, T2D, hypertension, hypercholesterolemia, and 10 probabilistic estimation of expression residual (PEER) factors. After FDR correction, 124 genes were significantly DE, including top genes C1RL , IL4R, and RGS16 . We identified several replications of the 124 genes in a transcriptomic follow-up study of 52 SevO cases and 59 normal weight controls- 20 genes displayed directionally consistent and FDR significant evidence of replication. We additionally identified several replications of the 124 genes in subcutaneous adipose tissue (SAT) from 19 NYC community volunteers, including for RGS16 , C1RL , and IL4R. 2- sample MR assessed causal associations between SevO and gene dysregulation, using SevO GWAS from DIAMANTE and CCHC eQTLs. Upregulated IL4R demonstrate pleiotropy, among other genes (e.g., REM2, ENGASE, and SCAP ). Collectively, these data demonstrate how transcriptomic studies may elevate understanding of SevO and inform efforts to reduce health disparities associated with SevO in HL populations.

IMPACT: Human exhaled breath is rich in metabolomic content that represents pulmonary function and gas exchange with blood, which can provide insights into an individual’s state of health. OBJECTIVES/GOALS: Human exhaled breath is rich in metabolomic content that represents pulmonary function and gas exchange with blood. It contains a mixture of compounds that offer insight into an individual’s state of health. Here, we present two novel non-invasive breath sampling devices for use in basic medical practice. METHODS/STUDY POPULATION: The two breath samplers have a disposablemouthpiece, a set of inhale and exhale one-way flap valves to allow condensation of exhaled breath only, and a saliva filter. The housing is constructed out of Teflon®, a chemically inert material to reduce chemical absorbance. The first device condenses exhaled breath into a frozen condensate using dry ice pellets and the other is a miniaturized design that liquifies exhaled breath on a condenser surface with micropatterned features on a cooling plate. Both designs have individual strategic and analytical advantages: frozen exhaled breath condensate (EBC) has high retention of analytes and sample volume; EBC collected in liquid phase offers facilitated sample collection and device portability. RESULTS/ANTICIPATED RESULTS: We investigated if breath aerosol size distribution affects the types or abundances ofmetabolites.Wemodified the geometry of the first device to redirect aerosol trajectories based on size. The trapping of larger aerosols increases with filter length, thus altering the aerosol size distribution although no significant changes in the metabolite profiles were found. With the miniaturized device, metabolite abundances were measured in a small cohort of healthy control and mild asthmatic subjects. Differences among subjects were found, as well as main differences between control and asthmatic groups. All analyses of EBC were performed with liquid chromatography mass spectrometry. Inflammatory suppression found in asthmatic subjects can be explained by prescribed daily use of inhaled corticosteroids. DISCUSSION/SIGNIFICANCE OF FINDINGS: Breath collection devices can be used in intensive care units, outpatient clinics, workplaces, and at home. EBC analysis has been used to monitor asthma and chronic obstructive pulmonary disease. It can be applied to infectious respiratory diseases (e.g. influenza, COVID-19) and for monitoring environmental and occupational chemical exposures.

ABSTRACT IMPACT: Characterizing the cellular composition of human adipose tissue may contribute to the prevention and/or treatment of obesity-associated metabolic diseases. OBJECTIVES/GOALS: Our aims in this study were to use single-cell techniques 1. to characterize cell types within the stromavascular fraction of human adipose tissue, 2. to identify subsets of cells within each type (sub-clustering), 3. to identify gene sets and pathways that may provide information on the function and significance of each cell cluster. METHODS/STUDY POPULATION: Abdominal subcutaneous adipose tissue samples from n=6 healthy volunteers (1M, 5F, age 28-38 y, BMI 24.5-63.0 kg/m2) were collected by aspiration or during surgery. In 3 subjects, all females, paired femoral samples were also collected. After collagenase digestion approximately n=10,000 cells/sample were used for single-cell RNA sequencing using the 10X Genomics platform. After QC and downstream analysis, data were analyzed in Seurat v.3.1.5. We identified first different cell types and then subclusters in an unbiased fashion. Gene Set Enrichment Analysis (GSEA) was used for pathway analysis. RESULTS/ANTICIPATED RESULTS: Progenitor markers are related to extracellular matrix, eg DCN (logFC progenitors vs other cells types=3.17, expressed in 99.7% (pct.1) of progenitors, 26.1% (pct.2) of others). Endothelial and pericytes shared markers like RBP7 (logFC=1.67, pct.1 0.88, pct.2 0.39); pericytes also showed unique markers, eg RGS5 (logFC=2.29, pct.1 0.89, pct.2 0.17). Progenitors are further divided into 11 sub-clusters, one of which showed enrichment of CD36 (high proliferation potential), FABP4 (differentiation), and of the novel marker PALMD (logFC=7.13, pct.1 0.94, pct.2 0.48). All p<10E-5. GSEA analysis suggests that inflammatory pathways are downregulated in both adipose progenitors and endothelial/pericyte cells in the femoral compared to the abdominal depot. DISCUSSION/SIGNIFICANCE OF FINDINGS: Single cell RNA sequencing provides unique insights into the molecular profile of cell types and the identification of novel subsets of cells within the human adipose tissue. Such cellular heterogeneity may explain differences in adipose function between individuals and eventually in the risk of obesity-associated metabolic diseases.

Glucocorticoids (GC) regulate the distribution and function of human adipose tissue, and their pathophysiological effects are attributed to changes in their concentration. We hypothesized that the diurnal cortisol rhythm contributes to its effects. Microarray analysis and qPCR were used to identify genes with expression changes as circulating cortisol declines between 9:00 and 15:00, in human abdominal subcutaneous adipose tissue of healthy volunteers (n=5). Placebo or a single, physiological dose of cortisol were administered out of phase (11:30) to identify genes with expression changes dependent on the cortisol rhythm. A novel in vitro model was established to study the long‐term effects of pulsatile versus constant GC in human adipocytes. We identified 280 genes significantly upregulated and 292 downregulated from 9:00 to 15:00. Cortisol changed the expression pattern of 191 (33.4%) genes. We verified changes in the clock gene PER1 , the GC receptor co‐chaperone FKBP5 , and KLF15 , a regulator of circadian nitrogen homeostasis. In vitro, pulsatile GC established rhythmic clock gene expression in human adipocytes, led to enlarged lipid droplets and lower basal lipolysis and de novo fatty acid synthesis compared to constant GC. These findings suggest that higher turnover of triglycerides may occur in patients with disrupted glucocorticoid rhythm and contribute to increased cardiometabolic risk. Research support: NIH DK080448, P30 DK046200 and UL1‐TR000157.