1 Hye-Won Song, 1 Gisele Baracho, 2 Nidhanjali Bansal, 3 Ian Taylor, 2 Eleen shum, 1 Stephanie Widman, 2 Stefanie Mortimer 1 BD Biosciences, San Diego, CA; 2 BD Biosciences, San Jose, CA; 3 FlowJo LLC, Ashland, OR Simultaneous Analysis of mRNA and Proteins in Mouse Immune Cells Using the BD TM Mouse Single- Cell Multiplexing Kit and the BD TM AbSeq Assay on the BD Rhapsody TM System Abstract • Obesity can weaken the body’s immune system and reduce its ability to fight off infections. • To understand the role immune cells play in obesity-related disorders, thousands of immune cells isolated from lymphoid organs and adipose tissue from a diet-induced obesity mouse model were examined using the BD Rhapsody TM , a high-throughput single cell analysis system. • DNA-barcoded universal antibodies from the BD TM Mouse Single-Cell Multiplexing Kit (SMK) allowed us to combine a total of 8 samples into a single pool, significantly reducing experimental scale and cost while eliminating any potential batch effect. • The targeted immune-focused mRNA (~400 genes) and BD TM AbSeq panel (30 DNA-barcoded antibodies) provided robust clustering of immune cell types and showed genes such as Pdcd1 and Tigit that are critical to functional immune responses. • Using this multiomics approach, we propose a model to explain certain obesity-related disease phenotypes. BD TM AbSeq BD TM Single-Cell Multiplexing Kit 1A 1B Technology BD Rhapsody TM mRNA, SMK and AbSeq captured on bead 1C cDNA synthesis on bead High multiplex-targeted amplification 1D Figure 1. A) BD TM SMK can be used to multiplex up to 12 different samples into one cartridge and bioinformatically demultiplex samples after sequencing. B) BD TM AbSeq uses DNA-barcoded antibodies to simultaneously analyze mRNA and protein with high- throughput sequencing. C) mRNA and oligos from BD TM SMK and BD TM AbSeq are captured on BD Rhapsody™ Cell Capture Beads. D) Captured mRNA and oligos are now associated with a cell label and unique molecular index (UMI) and are amplified by targeted multiplexed PCR. Experimental Design 1. Broad BD TM AbSeq panel identifies immune cell populations across mouse tissues CD45 + cells sorted BD SMK, BD TM AbSeq and CD45- FITC BD Rhapsody TM Tissues isolated: Fat Thymus Spleen Bone Marrow Amplification: • Sample Tag • BD TM AbSeq • BD Rhapsody™ Immune Response Panel Mm Sequencing SeqGeq TM analysis Figure 2. At 6 weeks of age 2 mice were placed on a control diet (Control, 10% DIO) and 1 mouse was placed on a high fat diet (HFD, 60% DIO) for 17 weeks before the experiment. Epididymal fat, thymus, spleen, and bone marrow cells were dissociated and stained with a 30-plex BD AbSeq panel, SMK, and CD45-FITC. CD45 + immune cells from fat tissue were then sorted. All 8 different samples were pooled prior to loading onto two BD Rhapsody™ Cartridges. The two replicates showed consistent data and so were combined for analysis. CD19 TCR beta All Tissues (HFD and Control) CD11b CD49b 3C Figure 3. A) Sample Tag metrics table shows the 8 different samples that were successfully demultiplxed from both replicates. B) SeqGeq™-generated t-SNE plot of all cells, combined using mRNA and protein profile. The similarity and difference between tissues and obesity condition are shown. Use of Sample Tags helps to minimize the technical noise from batch effect for easier sample-to-sample comparisons. C) t-SNE analysis with heat map showing expression levels of BD TM AbSeq lineage markers shows cell type specific clustering. D) Major immune cell types were identified using the BD TM AbSeq lineage markers shown in 3C. t-SNE, t-distributed stochastic neighbor embedding. 2. Unbiased clustering reveals disease-associated cell phenotypes 4B Figure 4. A) PhenoGraph identified 12 cell clusters that showed tissue- specific enrichment. This unbiased clustering uncovered 4 clusters (PG-1, PG-5, PG-7, PG-12) from fat and thymus showing differences in t-SNE localization between control and obese mice. B) Cell composition in 12 clusters of control and obese mice fat tissue consistently shows differences in 4 identified clusters. C) t-SNE plot of fat tissue with PhenoGraph clustering. Using marker-expression analysis, cell subtypes for each cluster were identified; (ATM: Adipose tissue macrophages). PG- 7 B cells were increased in obese mouse fat and monocle and marker analysis showed that these B cells resemble B-2 regulatory cells. Results Sample name Sample tag ID Rep1 Rep2 Total Control Fat ST1 972 928 1900 Control Bone marrow ST2 1501 1527 3028 Control Spleen ST3 1566 1337 2903 Control Thymus ST4 1074 1006 2080 HFD Fat ST5 1196 1029 2225 HFD Bone marrow ST6 1471 1308 2779 HFD Spleen ST7 1389 1229 2618 HFD Thymus ST8 1129 1039 2168 Multiplet 1048 818 1866 Undetermined 72 42 114 Total 11418 10263 21681 3A 3D • BD SMK TM allows high confidence of relationship between samples without batch effects. • BD TM AbSeq allows effective cell clustering, identification of obesity related phenotypes, and investigation of the underlying mechanism behind these phenotypes. • SeqGeq TM provides various tools to allow deeper analysis of single cell multiomics data. 3. Adipose tissue macrophages (ATM) and CD11c + myeloid cells in obese mice exhibit inflammatory and adipogenic gene signatures Figure 5. A) Heat map showing differentially expressed genes (DEG) in HFD ATM. Genes involved in inflammation such as Cd9, Cxcl1, and Il1b were upregulated in HFD ATM. B) Trem2 + ATM increased while Cd163 + anti- inflammatory ATM decreased in the HFD mouse. C) Heat map showing DEG in HFD CD11c + myeloid cells. Genes inducing adipogenesis (i.e. Lgals1) and recruiting inflammatory cells (i.e. Clec10a, Clec4e) were upregulated in HFD CD11c + myeloid cells. BD, the BD Logo, BD AbSeq, BD Rhapsody, SeqGeq and BD Single Cell Multiplexing Kit are trademarks of Becton, Dickinson and Company or its affiliates. All other trademarks are the property of their respective owners. © 2019 BD. All rights reserved 23-21496-00 4. Adipose tissue infiltrating cytotoxic cells from obese mouse showed signs of exhaustion Figure 6. A) SeqGeq TM -generated Monocle trajectory from CD8 + cells from all tissues. The most differentiated CD8 + cells were located at the end of the trajectory and corresponded to cells in the fat. B) Monocle trajectory of fat CD8 + cells. (left) Monocle defined 5 states. Direction of differentiation states were marked by an arrow. The most differentiated cells marked by an oval. (right) HFD CD8 + cells accumulated at the end of the trajectory. C) Exhaustion markers, PD-1, Pdcd1 (PD-1 mRNA), and Tigit were co- expressed in states 3 through 5 while CD49a and CD69 were co-expressed in state 2 cells. This suggests that HFD cytotoxic cells were transitioning from an activated to an exhausted state. Working Model BD Rhapsody Cell- Capture Bead NK Cells B Cells T Cells Myeloid Thymus – HFD Thymus – Control Spleen – HFD Spleen – Control Fat – HFD Fat – Control Bone marrow – HFD Bone marrow - Control 4A Fat t-SNE Control HFD Fat Control HFD 4C ATM Control ATM HFD Lgals3 Trem2 Cd9 Cxcl1 Il1b Mmp12 Cd72 Spp1 Il1m Qpct I-A_I-E (BD AbSeq) CD64 (BD AbSeq) CD1d (BD AbSeq) CD49b (BD AbSeq) Fcna Cd163 F13a1 Tfrc Cd38 Mmp9 Control HFD Trem2 Cd163 (promotes diet-induced obesity) (control inflammation) Up in HFD Down in HFD 5A 5B CD11c + Control CD11c + HFD CD44 (BD AbSeq) Lgals1 Mmp12 Cd63 Trem2 Clec10a Cd9 Clec4e Clec4d Chil3 CD5 (BD AbSeq) CD49b (BD AbSeq) Il6 Up in HFD Down in HFD 5C Monocle of fat CD8 + cells Fat Bone Marrow Thymus Spleen Control HFD Monocle of all tissue CD8 + cells 6A 6B 6C 1 2 3 4 5 CD103 (AbSeq) PD-1 (AbSeq) Pdcd1 Lag3 Tigit CD49a (AbSeq) CD69 (AbSeq) CD11c Myeloid cells F4/80 Adipocytes Cytotoxic cells B cells Adipocytes CD11c Myeloid cells F4/80 CD163 Trem2 Clec10a Healthy fat Obese fat Cell exhaustion CD8 CD49a PD-1 CD11b CD11b Anti-inflammatory signals CD1d CD1d CD11b CD11b Inflammatory signals Galectin-1 (Lgals1) Cytotoxic cells CD8 CD49a PD-1 CD8 PD-1 PD-1 Tigit Tigit Lag3 Adipogenesis Cell activation Cell migration Conclusions Introduction 3B Class 1 Laser Product. For Research Use Only. Not for use in diagnostic or therapeutic procedures.