PT  - JOURNAL ARTICLE
AU  - Michal Slyper
AU  - Caroline B. M. Porter
AU  - Orr Ashenberg
AU  - Julia Waldman
AU  - Eugene Drokhlyansky
AU  - Isaac Wakiro
AU  - Christopher Smillie
AU  - Gabriela Smith-Rosario
AU  - Jingyi Wu
AU  - Danielle Dionne
AU  - Sébastien Vigneau
AU  - Judit Jané-Valbuena
AU  - Sara Napolitano
AU  - Mei-Ju Su
AU  - Anand G. Patel
AU  - Asa Karlstrom
AU  - Simon Gritsch
AU  - Masashi Nomura
AU  - Avinash Waghray
AU  - Satyen H. Gohil
AU  - Alexander M. Tsankov
AU  - Livnat Jerby-Arnon
AU  - Ofir Cohen
AU  - Johanna Klughammer
AU  - Yanay Rosen
AU  - Joshua Gould
AU  - Bo Li
AU  - Lan Nguyen
AU  - Catherine J. Wu
AU  - Benjamin Izar
AU  - Rizwan Haq
AU  - F. Stephen Hodi
AU  - Charles H. Yoon
AU  - Aaron N. Hata
AU  - Suzanne J. Baker
AU  - Mario L. Suvà
AU  - Raphael Bueno
AU  - Elizabeth H. Stover
AU  - Ursula A. Matulonis
AU  - Michael R. Clay
AU  - Michael A. Dyer
AU  - Natalie B. Collins
AU  - Nikhil Wagle
AU  - Asaf Rotem
AU  - Bruce E. Johnson
AU  - Orit Rozenblatt-Rosen
AU  - Aviv Regev
TI  - A single-cell and single-nucleus RNA-seq toolbox for fresh and frozen human tumors
AID  - 10.1101/761429
DP  - 2020 Jan 01
TA  - bioRxiv
PG  - 761429
4099  - http://biorxiv.org/content/early/2020/03/30/761429.short
4100  - http://biorxiv.org/content/early/2020/03/30/761429.full
AB  - Single cell genomics is essential to chart the complex tumor ecosystem. While single cell RNA-Seq (scRNA-Seq) profiles RNA from cells dissociated from fresh tumor tissues, single nucleus RNA-Seq (snRNA-Seq) is needed to profile frozen or hard-to-dissociate tumors. Each strategy requires modifications to fit the unique characteristics of different tissue and tumor types, posing a barrier to adoption. Here, we developed a systematic toolbox for profiling fresh and frozen clinical tumor samples using scRNA-Seq and snRNA-Seq, respectively. We tested eight tumor types of varying tissue and sample characteristics (resection, biopsy, ascites, and orthotopic patient-derived xenograft): lung cancer, metastatic breast cancer, ovarian cancer, melanoma, neuroblastoma, pediatric sarcoma, glioblastoma, pediatric high-grade glioma, and chronic lymphocytic leukemia. Analyzing 212,498 cells and nuclei from 39 clinical samples, we evaluated protocols by cell quality, recovery rate, and cellular composition. We optimized protocols for fresh tissue dissociation for different tumor types using a decision tree to account for the technical and biological variation between clinical samples. We established methods for nucleus isolation from OCT embedded and fresh-frozen tissues, with an optimization matrix varying mechanical force, buffer, and detergent. scRNA-Seq and snRNA-Seq from matched samples recovered the same cell types and intrinsic expression profiles, but at different proportions. Our work provides direct guidance across a broad range of tumors, including criteria for testing and selecting methods from the toolbox for other tumors, thus paving the way for charting tumor atlases.