Longitudinal high-dimensional analysis identifies immune features associating with response to anti-PD-1 immunotherapy.

Response to immunotherapy widely varies among cancer patients and identification of parameters associating with favourable outcome is of great interest. Here we show longitudinal monitoring of peripheral blood samples of non-small cell lung cancer (NSCLC) patients undergoing anti-PD1 therapy by high-dimensional cytometry by time of flight (CyTOF) and Meso Scale Discovery (MSD) multi-cytokines measurements. We find that higher proportions of circulating CD8+ and of CD8+CD101hiTIM3+ (CCT T) subsets significantly correlate with poor clinical response to immune therapy. Consistently, CD8+ T cells and CCT T cell frequencies remain low in most responders during the entire multi-cycle treatment regimen; and higher killer cell lectin-like receptor subfamily G, member 1 (KLRG1) expression in CCT T cells at baseline associates with prolonged progression free survival. Upon in vitro stimulation, CCT T cells of responders produce significantly higher levels of cytokines, including IL-1ß, IL-2, IL-8, IL-22 and MCP-1, than of non-responders. Overall, our results provide insights into the longitudinal immunological landscape underpinning favourable response to immune checkpoint blockade therapy in lung cancer patients.

Integrated sequencing of exome and mRNA of large-sized single cells.

Current approaches of single cell DNA-RNA integrated sequencing are difficult to call SNPs, because a large amount of DNA and RNA is lost during DNA-RNA separation. Here, we performed simultaneous single-cell exome and transcriptome sequencing on individual mouse oocytes. Using microinjection, we kept the nuclei intact to avoid DNA loss, while retaining the cytoplasm inside the cell membrane, to maximize the amount of DNA and RNA captured from the single cell. We then conducted exome-sequencing on the isolated nuclei and mRNA-sequencing on the enucleated cytoplasm. For single oocytes, exome-seq can cover up to 92% of exome region with an average sequencing depth of 10+, while mRNA-sequencing reveals more than 10,000 expressed genes in enucleated cytoplasm, with similar performance for intact oocytes. This approach provides unprecedented opportunities to study DNA-RNA regulation, such as RNA editing at single nucleotide level in oocytes. In future, this method can also be applied to other large cells, including neurons, large dendritic cells and large tumour cells for integrated exome and transcriptome sequencing.

ChIP-Array 2: integrating multiple omics data to construct gene regulatory networks.

Transcription factors (TFs) play an important role in gene regulation. The interconnections among TFs, chromatin interactions, epigenetic marks and cis-regulatory elements form a complex gene transcription apparatus. Our previous work, ChIP-Array, combined TF binding and transcriptome data to construct gene regulatory networks (GRNs). Here we present an enhanced version, ChIP-Array 2, to integrate additional types of omics data including long-range chromatin interaction, open chromatin region and histone modification data to dissect more comprehensive GRNs involving diverse regulatory components. Moreover, we substantially extended our motif database for human, mouse, rat, fruit fly, worm, yeast and Arabidopsis, and curated large amount of omics data for users to select as input or backend support. With ChIP-Array 2, we compiled a library containing regulatory networks of 18 TFs/chromatin modifiers in mouse embryonic stem cell (mESC). The web server and the mESC library are publicly free and accessible athttp://jjwanglab.org/chip-array.

CTCF controls HOXA cluster silencing and mediates PRC2-repressive higher-order chromatin structure in NT2/D1 cells.

HOX cluster genes are activated sequentially in their positional order along the chromosome during vertebrate development. This phenomenon, known as temporal colinearity, depends on transcriptional silencing of 5' HOX genes. Chromatin looping was recently identified as a conserved feature of silent HOX clusters, with CCCTC-binding factor (CTCF) binding sites located at the loop bases. However, the potential contribution of CTCF to HOX cluster silencing and the underlying mechanism have not been established. Here, we demonstrate that the HOXA locus is organized by CTCF into chromatin loops and that CTCF depletion causes significantly enhanced activation of HOXA3 to -A7, -A9 to -A11, and -A13 in response to retinoic acid, with the highest effect observed for HOXA9. Our subsequent analyses revealed that CTCF facilitates the stabilization of Polycomb repressive complex 2 (PRC2) and trimethylated lysine 27 of histone H3 (H3K27me3) at the human HOXA locus. Our results reveal that CTCF functions as a controller of HOXA cluster silencing and mediates PRC2-repressive higher-order chromatin structure.

dbPSHP: a database of recent positive selection across human populations.

The dbPSHP database (http://jjwanglab.org/dbpshp) aims to help researchers to efficiently identify, validate and visualize putative positively selected loci in human evolution and further discover the mechanism governing these natural selections. Recent evolution of human populations at the genomic level reflects the adaptations to the living environments, including climate change and availability and stability of nutrients. Many genetic regions under positive selection have been identified, which assist us to understand how natural selection has shaped population differences. Here, we manually collect recent positive selections in different human populations, consisting of 15,472 loci from 132 publications. We further compiled a database that used 15 statistical terms of different evolutionary attributes for single nucleotide variant sites from the HapMap 3 and 1000 Genomes Project to identify putative regions under positive selection. These attributes include variant allele/genotype properties, variant heterozygosity, within population diversity, long-range haplotypes, pairwise population differentiation and evolutionary conservation. We also provide interactive pages for visualization and annotation of different selective signals. The database is freely available to the public and will be frequently updated.

GWAS3D: Detecting human regulatory variants by integrative analysis of genome-wide associations, chromosome interactions and histone modifications.

Interpreting the genetic variants located in the regulatory regions, such as enhancers and promoters, is an indispensable step to understand molecular mechanism of complex traits. Recent studies show that genetic variants detected by genome-wide association study (GWAS) are significantly enriched in the regulatory regions. Therefore, detecting, annotating and prioritizing of genetic variants affecting gene regulation are critical to our understanding of genotype-phenotype relationships. Here, we developed a web server GWAS3D to systematically analyze the genetic variants that could affect regulatory elements, by integrating annotations from cell type-specific chromatin states, epigenetic modifications, sequence motifs and cross-species conservation. The regulatory elements are inferred from the genome-wide chromosome interaction data, chromatin marks in 16 different cell types and 73 regulatory factors motifs from the Encyclopedia of DNA Element project. Furthermore, we used these function elements, as well as risk haplotype, binding affinity, conservation and P-values reported from the original GWAS to reprioritize the genetic variants. Using studies from low-density lipoprotein cholesterol, we demonstrated that our reprioritizing approach was effective and cell type specific. In conclusion, GWAS3D provides a comprehensive annotation and visualization tool to help users interpreting their results. The web server is freely available at http://jjwanglab.org/gwas3d.

EpiRegNet: constructing epigenetic regulatory network from high throughput gene expression data for humans.

The advances of high throughput profiling methods, such as microarray gene profiling and RNA-seq, have enabled researchers to identify thousands of differentially expressed genes under a certain perturbation. Much work has been done to understand the genetic factors that contribute to the expression changes by searching the over-represented regulatory motifs in the promoter regions of these genes. However, the changes could also be caused by epigenetic regulation, especially histone modifications, and no web server has been constructed to study the epigenetic factors responsible for gene expression changes. Here, we pre-sent a web tool for this purpose. Provided with different categories of genes (e.g., up-regulated, down-regulated or unchanged genes), the server will find epigenetic factors responsible for the difference among the categories and construct an epigenetic regulatory network. Furthermore, it will perform co-localization analyses between these epigenetic factors and transcription factors, which were collected from large scale experimental ChIP-seq or computational predicted data. In addition, for users who want to analyze dynamic change of a histone modification mark under different cell conditions, the server will find direct and indirect target genes of this mark by integrative analysis of experimental data and computational prediction, and present a regulatory network around this mark. Both networks can be visualized by a user friendly interface and the data are downloadable in batch. The server currently supports 12 cell types in human, including ESC and CD4+ T cells, and will expand as more public data are available. It also allows user to create a self-defined cell type, upload and analyze multiple ChIP-seq data. It is freely available to academic users at http://jjwanglab.org/EpiRegNet.