Entering a New Field! Zhang Feng's Team Has Published Its First Cell Paper in 2023

Transcription factors (TFs) regulate gene programs, thereby controlling different cellular processes and cell states.

On January 5th, 2023, the team led by Feng Zhang from the Broad Institute/Harvard Medical School/Massachusetts Institute of Technology published a research paper titled "A transcription factor atlas of directed differentiation" online in Cell. In order to comprehensively understand TF and its controlled processes, the study created a barcode library (>3500) containing all annotated human TF splice isoforms and applied it to construct a TF map, mapping the expression profiles of human embryonic stem cells (hESCs) overexpressing each TF at single cell resolution.

This study mapped the TF-induced expression profile to reference cell types and validated candidate TFs for generating different cell types, covering all three germ layers and the trophoblast layer. Targeted screening with a library subset enabled researchers to create customized cell disease models and integrate mRNA expression and chromatin accessibility data to identify downstream regulatory factors. Finally, the study described the impact of overexpressed combination of TFs by developing and validating a strategy for predicting TF combinations, which generated target expression profiles matching reference cell types to accelerate cell engineering work.

Understanding the gene regulatory networks (GRNs) that control cellular states is a fundamental goal of molecular biology. Transcription factors (TFs) bind to specific sequences in the genome to alter gene expression and specify cellular states. Overexpression of a single TF can lead to profound changes in cell fate. For example, a single TF has been shown to guide the differentiation of pluripotent stem cells towards many different types of cells, including muscle and neurons. Overexpression of a combination of TFs, such as the four "Yamanaka factors" (Oct4, Sox2, Klf4 and c-Myc), can produce even greater changes in the GRN, such as reprogramming fibroblasts into stem cells. These findings highlight the power of TFs to drive changes in cellular states, and underscore the utility of TF overexpression in understanding the gene expression programs that control cell fate.

Establishing a directional differentiation TF atlas (image source from Cell) 

The human genome contains over 1,800 TF loci, encoding more than 3,500 isoforms, creating vast prospects for possible regulatory outcomes. Previous studies have explored various aspects of this landscape, such as mapping quantitative trait loci that link TFs to phenotypes and studying perturbations of overexpression or inhibition of TFs in model systems.

Article mode diagram (image sourced from Cell)

Perturbation studies typically require a trade-off between the breadth of perturbations applied and the depth of phenotypic readouts obtained, either by performing large-scale screening with simple readouts or small-scale focused screening with detailed readouts. For instance, in the context of TF overexpression screening, a recent study screened a large library of 1,732 TF isoforms for pluripotency marker expression using a focused readout, whereas a smaller-scale analysis of 61 TFs assessed their overall impact through single-cell analysis.

However, to fully decipher the regulatory circuit, a systematic approach is needed that combines the breadth of screening with the depth of readouts, particularly for the transcriptional changes induced by each TF. Here, the study systematically mapped the expression changes driven by all human TF isoforms at single-cell resolution and used these data to identify TFs and their combinations that guide the differentiation of human embryonic stem cells (hESCs).

This study generated a barcode ORF library containing 3548 TF splice isoforms and developed a screening platform to construct a TF map of one million cells, depicting the transcriptome changes caused by TF overexpression in hESCs. The comprehensive TF map generated from this study can systematically identify TFs that drive cell state changes and classify orphan TFs through gene programs and broad observations. Additionally, the TF map from this study can be used to predict and validate TF combinations for target reference cell types. Therefore, the TF library and atlas generated from this study provide valuable resources for systematically elucidating TF gene programs and comprehensively understanding the GRNs that control cell states.

Reprinted from NetEase News