2017-LB: Comparison of Nuclear Preparation for Single-Nucleus RNA-seq of Adipose Tissue

Introduction & Objective: To appreciate the pathophysiological changes in the transcriptome and cell type landscape during obesity, single-nucleus (sn)RNA-seq overcomes the difficulty of maintaining intact transcripts. Sedimentation of dissociated live cells is used to concentrate adipocytes (Ads) and stromal cells. The current study compared the single cell level transcriptome in libraries with and without the fractionation. Methods: C57BL/6J mice were fed an HFD for 12 weeks starting at 8 weeks of age. Dissected epididymal fat from the same donor was divided into three conditions: 1) nuclei from unfractionated frozen fat (UF), 2) nuclei from the Ad fraction (AF) after enzymatic digestion and centrifugation, and 3) cells from the stromal fraction corresponding to the AF. As a dietary control, nuclei were prepared from the UF of an age-matched normal chow-fed mouse. Single cell/nucleus transcriptome libraries were sequenced using short-read NGS. Results: A total of 32,949 cells were recovered by snRNA-seq. Fractionation lost recovery of certain cell types, including epithelial, mesothelial, and B cells. Fractionation barely enriched Ads in the AF at 12.4%, comparable to the Ads (11.2%) in the UF. Macrophages (Mφs), rather than Ads, accounted for 54.6% of the AF. The Mφs were clustered into six subtypes by Seurat independent of dietary conditions or fractionation. However, certain cytokines were more highly expressed in Mφs from fractions than those from UF. In support of this, functional annotation-driven unsupervised clustering, ASURAT, mapped three Mφ subtypes in the libraries from fractions that differed from those from the UF, where there were four other subtypes. Conclusion: The current study suggests that the conventional fractionation could bias the transcriptome and a nuclear library from unfractionated frozen fat would represent a naïve comprehensive transcriptome that better reflects the pathophysiology of obesity, which should be important for understanding cell function. Disclosure Y. Onogi: None. T. Wada: None. S. Fujisaka: None. T. Sasaoka: None. K. Iida: None. Y. Okada: None. M. Kadowaki: None. S. Saito: None. K. Tobe: None. Funding JST Moonshot R&D (JPMJMS2021)