Single-cell RNA sequencing (scRNA-seq) technologies are poised to reshape the current cell-type classification system. However, a transcriptome-based single-cell atlas has not been achieved for complex mammalian systems. Here, we developed Microwell-seq, a high-throughput and low-cost scRNA-seq platform using simple, inexpensive devices. Using Microwell-seq, we analyzed more than 400,000 single cells covering all of the major mouse organs and constructed a basic scheme for a mouse cell atlas (MCA). We reveal a single-cell hierarchy for many tissues that have not been well characterized previously. We built a web-based "single-cell MCA analysis" pipeline that accurately defines cell types based on single-cell digital expression. Our study demonstrates the wide applicability of the Microwell-seq technology and MCA resource.

It has come to our attention that in preparing the final version of this paper, we inadvertently misspelled the first name of an author Ziming Zhou as ''Zimin Zhou''. In addition, we have made two errors in describing the reagents in the STAR Methods. First, under the subheading of ''Synthesis of barcoded beads'' in the Method Details section, the supplier of the magnetic beads coated with carboxyl groups should be Suzhou Knowledge & Benefit Sphere Tech. Co., Ltd. (diameter 20-25 mm, http://www.kbspheretech. com/), instead of Zhiyi. Second, under the subheading of ''Cell collection and lysis'' in the Method Details section, the concentration of Tris-HCL for the cold lysis buffer should be 0.1 M, instead of 1 M. These errors have been corrected online, and we apologize for any confusions we may have caused.

BackgroundHuman pluripotent stem cells (hPSCs) provide powerful models for studying cellular differentiations and unlimited sources of cells for regenerative medicine. However, a comprehensive single-cell level differentiation roadmap for hPSCs has not been achieved.ResultsWe use high throughput single-cell RNA-sequencing (scRNA-seq), based on optimized microfluidic circuits, to profile early differentiation lineages in the human embryoid body system. We present a cellular-state landscape for hPSC early differentiation that covers multiple cellular lineages, including neural, muscle, endothelial, stromal, liver, and epithelial cells. Through pseudotime analysis, we construct the developmental trajectories of these progenitor cells and reveal the gene expression dynamics in the process of cell differentiation. We further reprogram primed H9 cells into naïve-like H9 cells to study the cellular-state transition process. We find that genes related to hemogenic endothelium development are enriched in naïve-like H9. Functionally, naïve-like H9 show higher potency for differentiation into hematopoietic lineages than primed cells.ConclusionsOur single-cell analysis reveals the cellular-state landscape of hPSC early differentiation, offering new insights that can be harnessed for optimization of differentiation protocols.Electronic supplementary materialThe online version of this article (10.1186/s13059-018-1426-0) contains supplementary material, which is available to authorized users.

Unlike hematological malignancies, solid tumors have proved to be less susceptible to chimeric antigen receptor (CAR)-T cell therapy, which is partially caused by reduced accumulation of therapeutic T cells in tumor site. Since efficient trafficking is the precondition and pivotal step for infused CAR-T cells to exhibit their anti-tumor function, strategies are highly needed to improve the trafficking ability of CAR-T cells for solid tumor treatment. Here, based on natural lymphocyte chemotaxis theory and characteristics of solid tumor microenvironments, we explored the possibility of enhancing CAR-T cell trafficking by using chemokine receptors. Our study found that compared with other chemokines, several CXCR2 ligands showed relatively high expression level in human hepatocellular carcinoma tumor tissues and cell lines. However, both human peripheral T cells and hepatocellular carcinoma tumor infiltrating T cells lacked expression of CXCR2. CXCR2-expressing CAR-T cells exhibited identical cytotoxicity but displayed significantly increased migration ability in vitro.In a xenograft tumor model, we found that expressing CXCR2 in CAR-T cells could significantly accelerate in vivo trafficking and tumor-specific accumulation, and improve anti-tumor effect of these cells.

Chimeric antigen receptor T (CAR-T) cell therapies have demonstrated high response rate and durable disease control for the treatment of B cell malignancies. However, in the case of solid tumors, CAR-T cells have shown limited efficacy, which is partially attributed to intrinsic defects in CAR signaling. Here, we construct a double-chain chimeric receptor, termed as synthetic T cell receptor (TCR) and antigen receptor (STAR), which incorporates antigen-recognition domain of antibody and constant regions of TCR that engage endogenous CD3 signaling machinery. Under antigen-free conditions, STAR does not trigger tonic signaling, which has been reported to cause exhaustion of traditional CAR-T cells. Upon antigen stimulation, STAR mediates strong and sensitive TCR-like signaling, and STAR-T cells exhibit less susceptibility to dysfunction and better proliferation than traditional 28zCAR-T cells. In addition, STAR-T cells show higher antigen sensitivity than CAR-T cells, which holds potential to reduce the risk of antigen loss–induced tumor relapse in clinical use. In multiple solid tumor models, STAR-T cells prominently outperformed BBzCAR-T cells and generated better or equipotent antitumor effects to 28zCAR-T cells without causing notable toxicity. With these favorable features endowed by native TCR-like signaling, STAR-T cells may provide clinical benefit in treating refractory solid tumors.