CLASH of the Titans: How CAR-T Cells Can Triumph Over Tumors.

Abstract

The CRISPR JournalAhead of Print First CutFree AccessCLASH of the Titans: How CAR-T Cells Can Triumph Over TumorsRianne Opstelten and Julian J. Freen-van HeerenRianne Opsteltenhttps://orcid.org/0000-0003-0596-9876Immunomonitoring Services, Pharma and Biotech Services, Sanquin Health Solutions, Amsterdam, The Netherlands.Search for more papers by this author and Julian J. Freen-van Heeren*Address correspondence to: Julian J. Freen-van Heeren, Immunomonitoring Services, Pharma and Biotech Services, Sanquin Health Solutions, Amsterdam, The Netherlands. E-mail Address: j.freenvanheeren@sanquin.nlhttps://orcid.org/0000-0003-0506-2756Immunomonitoring Services, Pharma and Biotech Services, Sanquin Health Solutions, Amsterdam, The Netherlands.Search for more papers by this authorPublished Online:21 Mar 2023https://doi.org/10.1089/crispr.2023.0003AboutSectionsPDF/EPUB Permissions & CitationsDownload CitationsTrack CitationsAdd to favorites Back To Publication ShareShare onFacebookTwitterLinked InRedditEmail CLASH, a new CRISPR-based screening tool, enables the identification of target genes that can enhance T cell therapies through multiplex gene editing in an antigen-specific setting.Cytotoxic T cells are instrumental in fighting cancer. They recognize tumor antigens through their T cell receptor (TCR) and subsequently kill the cancer cells that express this antigen. T cells can only target tumor cells for a limited amount of time, however. Prolonged exposure to their cognate antigen, as well as contributing factors from the tumor and its microenvironment, will eventually inactivate T cells, for example, through inhibitory receptor signaling, leading to quiescence.Immunotherapy has revolutionized cancer treatment. In one form of immunotherapy, patient T cells or allogeneic T cells are isolated, expanded, and (re)infused. These cells can be further potentiated by using genetic editing to introduce a chimeric antigen receptor (CAR). These CARs have been selected for optimal tumor antigen recognition and providing superior activatory signaling. Although these CAR-T cells have shown remarkable efficiency in hematological cancers,1 for solid cancers, efficacy of CAR-T therapy is still lacking. These underwhelming effects have been linked to suboptimal T cell proliferation and failure of CAR-T persistence due to inhibitory signaling.T cell effector function can efficiently be enhanced through the use of CRISPR-mediated genome editing,2 for example, by knocking out genes of inhibitory pathways. Usually, CRISPR-based screenings are limited to a preselected handful of genes. However, in a recent report in Nature Biotechnology, Dai et al. from Yale University take it to the next level.3 They have developed a new platform—CRISPR-based library-scale adeno-associated virus (AAV) perturbation with simultaneous HDR knock-in (CLASH)—to perform unbiased large-scale CRISPR screening experiments with the aim to boost CAR-T cell efficiency (Fig. 1A). The authors applied CLASH to screen for the effect of knocking out a library of 8047 crRNAs targeted to 901 genes (“Descartes library”), which are involved in T cell exhaustion, epigenetic regulation, T cell co-stimulation, memory T cell differentiation, TCR signaling, adaptive immune responses, immune responses to tumor cells, or T cell proliferation.FIG. 1. Models employed by Dai et al. (A) Schematic representation of the CLASH construct and its introduction into human T cells. A crRNA targeting the TRAC locus (crRNA TRAC), one or more crRNAs targeting genes of interest (crRNA Gene X) and a CD22 CAR, flanked by HDR sequences specific for the TRAC locus are packaged into AAV6 vectors. T cells are electroporated to introduce Cpf1 mRNA, and transduced with AAV6 to introduce the crRNAs and TRAC. CRISPR-mediated cleavage disrupts the TRAC locus, resulting in downregulation of the TCR. Integration of the CD22 CAR in the TRAC locus facilitates expression of the CD22 CAR molecule. The cRNA for gene X is then also integrated and transcribed and subsequently disrupts gene X at its respective locus. (B) Culture system employed for target identification through selective culture. CAR-T cells were cocultured with antigen-bearing NALM-6 target cells. Upon tumor eradication, new tumor cells were added for a total of seven (for CD4+ T cells) or eight (for CD8+ T cells) rounds of culture. Cells from each culture round were preserved, and knockout selection was compared with a control library by next-generation sequencing. Further analyses included phenotypical assessment (e.g., expression of CTLA-4 and LAG-3), functional assessment (e.g., proliferative capacity or IFN-γ production), and validation in a xenograft model (for PRDM1 knockout only). AAV, adeno-associated virus; CAR, chimeric antigen receptor; CLASH, CRISPR-based library-scale AAV perturbation with simultaneous HDR knock-in; IFN-γ, interferon-γ.To do so, the authors introduced Cas12a/Cpf1 mRNA through electroporation and used AAVs as a vector for delivering a construct to interfere with multiple genetic targets, whereas simultaneously redirecting T cell antigen specificity. Each AAV encoded for a crRNA targeting the TCR alpha locus (TRAC). This construct also carried a CAR targeting CD22, a well-defined tumor antigen in leukemia, which was flanked with homology-directed repair sequences specific for the TRAC locus, resulting in construct integration into the TRAC locus. This disrupts the generic TCR and replaces it with the CAR.Furthermore, each AAV simultaneously carried a crRNA cassette that could be used to test virtually any number of variable crRNAs for genetic screening purposes. The resulting CAR-T cells have several features of interest: (1) they only recognize one epitope (in this case CD22); (2) they will only exhibit knockout of a library-based target if the CAR was integrated successfully; and (3) they are equipped with a CAR whose expression is regulated similarly as the endogenous TCR would have been.As a first test of the CLASH platform, Dai et al. designed a crRNA cassette with a pair of crRNAs targeting two surface molecules expressed by T cells: CD5 and CD226. TCR knockout was fairly efficient at >60% in two independent donors, whereas CAR expression was achieved in, respectively, 37.4% and 51% of cells. Furthermore, both CD5 and CD226 were knocked out at significant rates. Although knockout and knock-in rates could be improved, authors proceeded to test a larger library.When continuously exposed to antigen, T cells gradually lose the capacity to produce effector molecules. By performing multiple rounds of coculture with NALM-6 leukemia cells, authors selected for knockouts that were superior effector cells in a setting where they are continuously exposed to antigen (Fig. 1B). Upon comparison of the nontargeting control crRNA vector group with the Descartes crRNA-containing constructs, a marked increase in tumor killing capacity and interferon-γ/tumor necrosis factor-α production, both hallmarks of effective antitumor responses, were observed in Descartes-transduced CAR-T cells in both CD4+ and CD8+subsets when comparing with the nontargeting control.Next, the authors used next-generation sequencing to determine which knockout clones were enriched during the prolonged coculture. Over time, they found enrichment for a decreasing number of knockouts, indicating a selective benefit for particular knockout phenotypes. Surprisingly, although genes associated with checkpoint inhibition (e.g., PDCD1, BTLA, and LAIR1) were potential hits early during culture, these hits were depleted by the end of the experiment, indicating that these genes are not necessarily the most promising targets for enhancing CAR-T persistence. Top hits instead included TET2, PRDM1, PELI1, and JADE1. TET2 has been previously shown to inhibit CAR-T persistence.4 Similarly, PRDM1 (encoding for BLIMP-1), a key transcription factor associated with terminal T cell differentiation, has also been shown to be an important target for enhancing CAR-T effector function.5 Furthermore, PELI1 has previously been identified as an inhibitor of antitumor function in CD8+ T cells.6Further analysis revealed that most of the enriched knockouts resulted in increased cellular proliferation and lower expression of the inhibitory receptors CTLA-4 and LAG-3. The most pronounced effect was observed upon PRDM1 knockout. The authors next validated PRDM1 as a potential target and showed that, as previously reported,5PRDM1 perturbation results in enhanced antitumor function of CAR-T cells and superior treatment efficacy in mouse leukemia and colon cancer models when compared with PRDM1 wild-type CAR-T cell therapy.Interestingly, while PRDM1 was identified as a target in a screen with NALM-6 tumor cells, a hematological cancer, PRDM1 knockout also enhanced antitumor efficacy of CAR-T cells in a solid cancer model for colon cancer. As the role of PRDM1/BLIMP-1 in T cell effector function and differentiation has been well characterized, this is not surprising. However, the environment in which these tumor types reside are significantly different, and thus T cells are exposed to different signals. For instance, to allow access to colon tissue, T cells need to express gut homing receptors such as CCR9 and α4β7.7 Potentially, other targets more specific to the intended tumor environment would have been identified if a different target cell or culture system had been employed during the primary screen.Of note, CAR-T cell therapy is associated with significant side effects. One way to mitigate this is the incorporation of a “kill switch”—a molecular mechanism that can be engaged to specifically remove CAR-T cells after eradication of the tumor.8 This should especially be considered upon enhanced T cell effector function, for example, after PRDM1 knockout.In summary, although this study did not identify truly novel targets for enhancing antitumor immunity of CAR-T cells, the CLASH platform enables screening of an impressive tally of target genes in an antigen-specific setting. The produced data set, also containing data on, for example, co-stimulatory molecules, can be used for in silico applications, for instance, to identify novel signaling moieties that could be used when designing new CARs.In addition, hits from this study could be investigated as potential T cell effector function enhancement strategies in other T cell therapies, such as transgenic TCR-bearing T cell therapy9 or chimeric-switch receptor T cell therapy.10 Of note, CLASH could be of interest not only in cytotoxic T cells but also in other cells in which CARs have been established, such as natural killer cells11 and regulatory T cells.12 Finally, CLASH could also be employed to investigate other (biological) processes, such as factors involved in germinal center responses,13 potentially enhancing CD4+ T cell help to antibody producing B cells.