Preprint Club
A cross-institutional Journal Club Initiative

Keywords

• CD8+ T cells

• Transcription factors

• Tumor immunology

Main Findings

The location and cues from the local microenvironment influence the differentiation state of T cells, resulting in a diverse array of functionality. Transcription factors (TF) control the differentiation of cells and understanding how and which TFs govern CD8+T cell differentiation can improve the outcome of CD8+ T cell-based therapies and help to uncover novel biology behind fundamental processes.

Chung et al (not peer reviewed) have generated a transcriptional atlas of CD8+T cell differentiation states by integrating transcriptomic and epigenomic data. This effort led to the identification of TFs that are associated with single-state-specific TFs and multi-state TFs that regulate more than one state. Their pipeline recapitulated some known TFs (e.g. TCF-1) but excitingly, revealed novel and previously unreported TFs (e.g. Jdp2, Zscan20, Zlp324 for terminally exhausted CD8+ T cells (TEXterm)).

As a proof-of-concept validation, the authors depleted the TFs predicted to be associated with TEXterm cells by introducing gRNA libraries to Cas9+ P14 donor T cells. These cells were adoptively transferred to mice infected with LCMV-Clone 13 and the authors assessed the differentiation of transferred cells to various CD8+T cell states. Depletion of novel TFs reduced the frequency of TEXterm cell frequency and expression of inhibitory receptors. Moreover, effector functions of CD8+T cells (e.g. frequency of IFNg+ TNFa+ P14 cells) and anti-viral responses have improved.

Similarly, the authors utilized B16-GP33 syngeneic tumor model to test if disrupting predicted TEXterm-specific TFs can enhance tumor control. Mice with established B16-GP33 tumors were adoptively transferred with Cas9+ P14 cells that were transduced with gRNAs against TFs predicted to be associated with TEXterm cells (e.g. Zscan20 and Zfp324). Depletion of TEXterm specific TFs improved tumor control, the transferred cells expressed effector markers more, and the transferred cells committed to progenitor state more than terminally exhausted state. Finally, anti-PD-1 treatment synergized with depletion of TEXterm-specific TFs and resulted in better tumor control than either treatment alone.

Overall, this study presents a framework to systematically identify novel TFs that drive various CD8+ T cell states and experimentally validates several of these TFs that are involved in TEXterm differentiation. With the growing sequencing datasets, similar frameworks can be for other cell types and/or disease states and improve our knowledge on transcriptional networks in immune cells.

Limitations

• CRISPR may result in unwanted gene edits, alternative methods to modulate the TF activity would be helpful.

• Human validation – either from TILs or chronic infections. Would modulating these TFs have a function in human CD8+T cell phenotypes?

• Human diseases develop in years, however the models used in this study are in weeks – what are the long-term effects of TF manipulation on T cell function? Do cells adapt to lack of these TFs?

Significance/Novelty

This study delineates the TF networks in different T cell states. With extensive validation of novel TFs, their observations can lead to more discoveries in T cells to enhance T-cell based therapies for cancer. Given the demonstration of the predictive power of this framework, this idea can be expanded to other cells/disease states.

Recommendations

• This study predicts various novel candidates and validates some – how are these TFs regulated, which signaling pathways result in the activation of these TFs?

• Combinatorial modulation? Can the anti-tumor activity improve if two or more TFs are depleted? If so, how can one select such combinations?

• Microenvironment that TRM develop may impact their TF landscape. How are these TFs impact TRM from different anatomical sites?

• Safety and efficacy of TF-modified T cells as therapeutics?

• There are other TF-activity prediction algorithms - comparing the results and looking for consensus TFs may result in higher confidence predictions.

Credit

Reviewed by Didem Ağaç Çobanoğlu as part of a cross-institutional journal club between the Icahn School of Medicine at Mount Sinai, the University of Oxford, the Karolinska Institute, the University of Toronto and MD Anderson Cancer Center.

The author declares no conflict of interests in relation to their involvement in the review.