Understanding cell-specific cellular networks: a presentation by Dr Aviv Regev at Genentech Research and Early Development
Single-cell genomics is a revolutionary technology able to unveil cell types, differentiation status, gene programs, physical location, and cell interactions in space. Dr Aviv Regev, Executive Vice President at Genentech Research and Early Development delivered a comprehensive summary of her group’s research during the conference.
She postulated that single-cell genomics can help to unveil new relationships between the genotype and phenotype, and ultimately enhance our understanding of diseases. To illustrate her point, she first outlined her use of single-cell genomics to elucidate the cellular basis of a complex genetic disease, ulcerative colitis (UC), a form of irritable bowel syndrome (IBS).
Genome-wide association studies (GWASes) previously identified multiple loci linked to IBS. However, the cellular faction of the variants or the mechanism of action of the variants remains unknown.
To address this problem, Dr Smillie from Dr Regev’s lab created a single-cell atlas of the healthy and ulcerative colitis (UC) colonic mucosa (Smillie et al., 2018, bioarXiv).
Through sequencing of 251,133 cells from 17 patients with UC and 10 healthy biopsies, scPhere deep-learning algorithm separated the cells into epithelial, stroma, and immune cells.
The study demonstrated a change of cellular profile, an enrichment in M-like stromal cells, and expansion of inflammatory fibroblasts in UC compared to healthy tissue.
Dr Regev’s group next looked into the enrichment of specific GWAS genes in certain cellular populations (Smillie et al., 2019, Cell). They found that specific GWAS genes are enriched in M-like cells in UC.
"There is a need for more comprehensive cell atlases, because of the techniques used to recover cells used for downstream sequencing"
A connectome analysis between cells identified a change in cellular networks in UC, driven by GWAS genes and cell types enriched in such genes acting as hubs of decompartmentalization. Dr Regev highlighted that most UC-risk genes are only expressed in specific mucosal cells, which may be lacking in their research (Drokhlyansky et al., Cell 2020).
There is a need for more comprehensive cell atlases, because of the techniques used to recover cells used for downstream sequencing. Dr Regev’s lab developed such techniques, named RAISIN-Seq (nuclei + ribosomes) and MIRCL-Seq (label-free rare cell enrichment).
Dr Drokhlyansky combined the two methods to construct an enhanced single-cell atlas, unraveling additional genes constricted to new cell types related to enteric neuropathies and extra-intestinal disorders with gastrointestinal dysmotility, such as Parkinson’s disease.
Dr Regev suggested that this technique could be used to study conditions that are CNS-related, but affect neurons outside of the CNS, posing barriers to their current research. She highlighted the need to have comprehensive catalogs of diseases available.
In the second part of her talk, Dr Regev focused on the potential use of single-cell sequencing for understanding cellular programs and the ways genes relate to each other.
To illustrate her point, Dr Regev introduced gene C1org106 commonly expressed in enterocytes (Smillie et al., 2019, Cell).
By studying C1org106 co-expression signatures, Dr Regev’s group found that C1org106 expression was found to vary across tight junction enterocytes, suggesting the gene’s potential new function.
Related to intercellular co-expression variability, Dr Regev’s group studied co-expressing of C1org106 within the cells, finding that 10 meta-modules span over 50% of GWAS-implicated IBD risk genes.
Dr Regev next asked whether causal genes for the risk regions could be identified. Smillie et al. (2019) found that single-cell expression and co-expression help nominate causal genes in associated regions, identifying concrete disease-related cellular programs.
Dr Regev’s lab created a scLinker technique to be able to identify cell programs from GWAS studies combined with scRNA-Seq sequencing (Jagadeesh et al., 2021, BioarXiv).
Using the technique, Dr Regev found that pathway-related programs can be found in UC, and also Alzheimer’s disease and multiple sclerosis. Using the analysis, IBD heritability cell programs could be found across monocytes and fibroblasts.
Building upon cell programs, Dr Regev explained that the focus of her lab was next on ‘multi-cellular programs’ (Jerby et al., 2020, BioarXiv) as cells need to coordinate cell-type specific response across tissues.
Jerby and Regev et al. (2020, bioarXiv) developed a two-step approach for sequencing called Dialogue, that allows inference of multicellular programs. Dialogue can be applied onto dissociated tissues, including the single-cell atlas of healthy and UC colonic mucosa.
Jerby and Regev identified IBD-associated multicellular program across T cells, epithelial cells and macrophages using the scRNA-Seq and Dialogue.
In the last part of her talk, Dr Regev focused on using genetic perturbations to decipher cellular function. Dr Regev’s lab used Perturb-Seq technique of engineered perturbations, by combining single-cell RNA-Seq with CRISPR-Cas9 engineering (Dixit and Parnas et al., Cell, 2016).
Using computational analyses, Dr Regev was able to identify transcription factor modules controlling five cellular programs in dendritic cells treated with lipopolysaccharides.
Another use of Perturb-Seq was to study gain-of-function coding variants in cancer. Dr Regev’s lab investigated p53 and KRAS variants, including common missense mutations.
To conclude her talk, Dr Regev summarized that Perturb-Seq is a family of methods that couples pooled perturbation with a high content readout that can be applied to one or multiple perturbations per cell and can be applied in a variety of contexts.
This includes in vivo or in organoids, and in combination with other techniques such as genome-wide and interactions screens.
