Many human cancers are resistant to immunotherapy, for reasons that are poorly understood. We used a genome-scale CRISPR-Cas9 screen to identify mechanisms of tumor cell resistance to killing by cytotoxic T cells, the central effectors of antitumor immunity. Inactivation of >100 genes—including Pbrm1, Arid2, and Brd7, which encode components of the PBAF form of the SWI/SNF chromatin remodeling complex—sensitized mouse B16F10 melanoma cells to killing by T cells. Loss of PBAF function increased tumor cell sensitivity to interferon-γ, resulting in enhanced secretion of chemokines that recruit effector T cells. Treatment-resistant tumors became responsive to immunotherapy when Pbrm1 was inactivated. In many human cancers, expression of PBRM1 and ARID2 inversely correlated with expression of T cell cytotoxicity genes, and Pbrm1-deficient murine melanomas were more strongly infiltrated by cytotoxic T cells.

MICA and MICB are expressed by many human cancers as a result of cellular stress, and can tag cells for elimination by cytotoxic lymphocytes through natural killer group 2D (NKG2D) receptor activation. However, tumors evade this immune recognition pathway through proteolytic shedding of MICA and MICB proteins. We rationally designed antibodies targeting the MICA α3 domain, the site of proteolytic shedding, and found that these antibodies prevented loss of cell surface MICA and MICB by human cancer cells. These antibodies inhibited tumor growth in multiple fully immunocompetent mouse models and reduced human melanoma metastases in a humanized mouse model. Antitumor immunity was mediated mainly by natural killer (NK) cells through activation of NKG2D and CD16 Fc receptors. This approach prevents the loss of important immunostimulatory ligands by human cancers and reactivates antitumor immunity.

The rapid development of single-cell transcriptomic technologies has helped uncover the cellular heterogeneity within cell populations. However, bulk RNA-seq continues to be the main workhorse for quantifying gene expression levels due to technical simplicity and low cost. To most effectively extract information from bulk data given the new knowledge gained from single-cell methods, we have developed a novel algorithm to estimate the cell-type composition of bulk data from a single-cell RNA-seq-derived cell-type signature. Comparison with existing methods using various real RNA-seq data sets indicates that our new approach is more accurate and comprehensive than previous methods, especially for the estimation of rare cell types. More importantly, our method can detect cell-type composition changes in response to external perturbations, thereby providing a valuable, cost-effective method for dissecting the cell-type-specific effects of drug treatments or condition changes. As such, our method is applicable to a wide range of biological and clinical investigations.

Single-cell analysis is a powerful tool for dissecting the cellular composition within a tissue or organ. However, it remains difficult to detect rare and common cell types at the same time. Here, we present a new computational method, GiniClust2, to overcome this challenge. GiniClust2 combines the strengths of two complementary approaches, using the Gini index and Fano factor, respectively, through a cluster-aware, weighted ensemble clustering technique. GiniClust2 successfully identifies both common and rare cell types in diverse datasets, outperforming existing methods. GiniClust2 is scalable to large datasets.Electronic supplementary materialThe online version of this article (10.1186/s13059-018-1431-3) contains supplementary material, which is available to authorized users.