Genetic Inactivation of CD33 in Hematopoietic Stem Cells to Enable CAR T Cell Immunotherapy for Acute Myeloid Leukemia.

The absence of cancer-restricted surface markers is a major impediment to antigen-specific immunotherapy using chimeric antigen receptor (CAR) T cells. For example, targeting the canonical myeloid marker CD33 in acute myeloid leukemia (AML) results in toxicity from destruction of normal myeloid cells. We hypothesized that a leukemia-specific antigen could be created by deleting CD33 from normal hematopoietic stem and progenitor cells (HSPCs), thereby generating a hematopoietic system resistant to CD33-targeted therapy and enabling specific targeting of AML with CAR T cells. We generated CD33-deficient human HSPCs and demonstrated normal engraftment and differentiation in immunodeficient mice. Autologous CD33 KO HSPC transplantation in rhesus macaques demonstrated long-term multilineage engraftment of gene-edited cells with normal myeloid function. CD33-deficient cells were impervious to CD33-targeting CAR T cells, allowing for efficient elimination of leukemia without myelotoxicity. These studies illuminate a novel approach to antigen-specific immunotherapy by genetically engineering the host to avoid on-target, off-tumor toxicity.

Base editing of haematopoietic stem cells rescues sickle cell disease in mice.

Sickle cell disease (SCD) is caused by a mutation in the ß-globin gene HBB1. We used a custom adenine base editor (ABE8e-NRCH)2,3 to convert the SCD allele (HBBS) into Makassar ß-globin (HBBG), a non-pathogenic variant4,5. Ex vivo delivery of mRNA encoding the base editor with a targeting guide RNA into haematopoietic stem and progenitor cells (HSPCs) from patients with SCD resulted in 80% conversion of HBBS to HBBG. Sixteen weeks after transplantation of edited human HSPCs into immunodeficient mice, the frequency of HBBG was 68% and hypoxia-induced sickling of bone marrow reticulocytes had decreased fivefold, indicating durable gene editing. To assess the physiological effects of HBBS base editing, we delivered ABE8e-NRCH and guide RNA into HSPCs from a humanized SCD mouse6 and then transplanted these cells into irradiated mice. After sixteen weeks, Makassar ß-globin represented 79% of ß-globin protein in blood, and hypoxia-induced sickling was reduced threefold. Mice that received base-edited HSPCs showed near-normal haematological parameters and reduced splenic pathology compared to mice that received unedited cells. Secondary transplantation of edited bone marrow confirmed that the gene editing was durable in long-term haematopoietic stem cells and showed that HBBS-to-HBBG editing of 20% or more is sufficient for phenotypic rescue. Base editing of human HSPCs avoided the p53 activation and larger deletions that have been observed following Cas9 nuclease treatment. These findings point towards a one-time autologous treatment for SCD that eliminates pathogenic HBBS, generates benign HBBG, and minimizes the undesired consequences of double-strand DNA breaks.

CRISPR-targeted MAGT1 insertion restores XMEN patient hematopoietic stem cells and lymphocytes.

XMEN disease, defined as "X-linked MAGT1 deficiency with increased susceptibility to Epstein-Barr virus infection and N-linked glycosylation defect," is a recently described primary immunodeficiency marked by defective T cells and natural killer (NK) cells. Unfortunately, a potentially curative hematopoietic stem cell transplantation is associated with high mortality rates. We sought to develop an ex vivo targeted gene therapy approach for patients with XMEN using a CRISPR/Cas9 adeno-associated vector (AAV) to insert a therapeutic MAGT1 gene at the constitutive locus under the regulation of the endogenous promoter. Clinical translation of CRISPR/Cas9 AAV-targeted gene editing (GE) is hampered by low engraftable gene-edited hematopoietic stem and progenitor cells (HSPCs). Here, we optimized GE conditions by transient enhancement of homology-directed repair while suppressing AAV-associated DNA damage response to achieve highly efficient (>60%) genetic correction in engrafting XMEN HSPCs in transplanted mice. Restored MAGT1 glycosylation function in human NK and CD8+ T cells restored NK group 2 member D (NKG2D) expression and function in XMEN lymphocytes for potential treatment of infections, and it corrected HSPCs for long-term gene therapy, thus offering 2 efficient therapeutic options for XMEN poised for clinical translation.

Enhanced homology-directed repair for highly efficient gene editing in hematopoietic stem/progenitor cells.

Lentivector gene therapy for X-linked chronic granulomatous disease (X-CGD) has proven to be a viable approach, but random vector integration and subnormal protein production from exogenous promoters in transduced cells remain concerning for long-term safety and efficacy. A previous genome editing-based approach using Streptococcus pyogenes Cas9 mRNA and an oligodeoxynucleotide donor to repair genetic mutations showed the capability to restore physiological protein expression but lacked sufficient efficiency in quiescent CD34+ hematopoietic cells for clinical translation. Here, we report that transient inhibition of p53-binding protein 1 (53BP1) significantly increased (2.3-fold) long-term homology-directed repair to achieve highly efficient (80% gp91phox+ cells compared with healthy donor control subjects) long-term correction of X-CGD CD34+ cells.

Prediction and validation of hematopoietic stem and progenitor cell off-target editing in transplanted rhesus macaques.

The programmable nuclease technology CRISPR-Cas9 has revolutionized gene editing in the last decade. Due to the risk of off-target editing, accurate and sensitive methods for off-target characterization are crucial prior to applying CRISPR-Cas9 therapeutically. Here, we utilized a rhesus macaque model to compare the predictive values of CIRCLE-seq, an in vitro off-target prediction method, with in silico prediction (ISP) based solely on genomic sequence comparisons. We use AmpliSeq HD error-corrected sequencing to validate off-target sites predicted by CIRCLE-seq and ISP for a CD33 guide RNA (gRNA) with thousands of off-target sites predicted by ISP and CIRCLE-seq. We found poor correlation between the sites predicted by the two methods. When almost 500 sites predicted by each method were analyzed by error-corrected sequencing of hematopoietic cells following transplantation, 19 off-target sites revealed insertion or deletion mutations. Of these sites, 8 were predicted by both methods, 8 by CIRCLE-seq only, and 3 by ISP only. The levels of cells with these off-target edits exhibited no expansion or abnormal behavior in vivo in animals followed for up to 2 years. In addition, we utilized an unbiased method termed CAST-seq to search for translocations between the on-target site and off-target sites present in animals following transplantation, detecting one specific translocation that persisted in blood cells for at least 1 year following transplantation. In conclusion, neither CIRCLE-seq or ISP predicted all sites, and a combination of careful gRNA design, followed by screening for predicted off-target sites in target cells by multiple methods, may be required for optimizing safety of clinical development.

CHANGE-seq-BE enables simultaneously sensitive and unbiased in vitro profiling of base editor genome-wide activity.

Base editors ( BE ) enable programmable conversion of nucleotides in genomic DNA without double-stranded breaks and have substantial promise to become new transformative genome editing medicines. Sensitive and unbiased detection of base editor off-target effects is important for identifying safety risks unique to base editors and translation to human therapeutics, as well as accurate use in life sciences research. However, current methods for understanding the global activities of base editors have limitations in terms of sensitivity or bias. Here we present CHANGE-seq-BE, a novel method to directly assess the off-target profile of base editors that is simultaneously sensitive and unbiased. CHANGE-seq-BE is based on the principle of selective sequencing of adenine base editor modified genomic DNA in vitro , and provides an accessible, rapid, and comprehensive method for identifying genome-wide off-target mutations of base editors.

Optimization of base editors for the functional correction of SMN2 as a treatment for spinal muscular atrophy.

Spinal muscular atrophy (SMA) is caused by mutations in SMN1. SMN2 is a paralogous gene with a C•G-to-T•A transition in exon 7, which causes this exon to be skipped in most SMN2 transcripts, and results in low levels of the protein survival motor neuron (SMN). Here we show, in fibroblasts derived from patients with SMA and in a mouse model of SMA that, irrespective of the mutations in SMN1, adenosine base editors can be optimized to target the SMN2 exon-7 mutation or nearby regulatory elements to restore the normal expression of SMN. After optimizing and testing more than 100 guide RNAs and base editors, and leveraging Cas9 variants with high editing fidelity that are tolerant of different protospacer-adjacent motifs, we achieved the reversion of the exon-7 mutation via an A•T-to-G•C edit in up to 99% of fibroblasts, with concomitant increases in the levels of the SMN2 exon-7 transcript and of SMN. Targeting the SMN2 exon-7 mutation via base editing or other CRISPR-based methods may provide long-lasting outcomes to patients with SMA.

Susceptibility of Brazilian isolates of Trypanosoma evansi to suramin sodium: test in experimentally infected mice.

This study aimed to evaluate the susceptibility of Brazilian isolates of Trypanosoma evansi to suramin sodium. For this purpose, three isolates of T. evansi (LPV-2005, LPV-2009 and LPV-2010) and seventy mice were used, with the animals divided in 10 groups (A, B, C, D, E, F, G, H, I and J) with seven animals each group. Mice of groups A, B, and C were infected with LPV-2005; Groups D, E and F with LPV-2009 and the groups G, H and I with LPV-2010. The group J was composed by healthy mice or uninfected. The parasitemia was monitored daily through blood smear, and the treatment of all groups was performed three days post-infection (PI), when all mice showed increased parasitemia. Groups A, D and G represented the positives controls, while groups B, E and H received a single dose of suramin sodium at 10 mgkg(-1) intramuscularly. Groups C, F and I were treated with three doses of suramin sodium at 10 mgkg(-1), respecting an interval of 24 h between each dose. Negative blood smears from all animals were obtained 24 h after treatment (AT), status maintained until the end of the experiment (50 days PI). The specific PCR for T. evansi was carried out from blood, showing negative results AT. Therefore, this study showed that a single dose of suramin sodium at 10 mgkg(-1) has the same efficacy of three doses, as recommended by the therapeutic literature. Furthermore, we observed that Brazilian isolates did not show resistance to the drug.

Trichostatin A for Efficient CRISPR-Cas9 Gene Editing of Human Pluripotent Stem Cells.

Genome-edited human-induced pluripotent stem cells (iPSCs) have broad applications in disease modeling, drug discovery, and regenerative medicine. Despite the development of clustered regularly interspaced short palindromic repeats (CRISPR)-Cas9 system, the gene editing process is inefficient and can take several weeks to months to generate edited iPSC clones. We developed a strategy to improve the efficiency of the iPSC gene editing process via application of a small-molecule, trichostatin A (TSA), a Class I and II histone deacetylase inhibitor. We observed that TSA decreased global chromatin condensation and further resulted in increased gene-editing efficiency of iPSCs by twofold to fourfold while concurrently ensuring no increased off-target effects. The edited iPSCs could be clonally expanded while maintaining genomic integrity and pluripotency. The rapid generation of therapeutically relevant gene-edited iPSCs could be enabled by these findings.

Base editing as a genetic treatment for spinal muscular atrophy.

Spinal muscular atrophy (SMA) is a devastating neuromuscular disease caused by mutations in the SMN1 gene. Despite the development of various therapies, outcomes can remain suboptimal in SMA infants and the duration of such therapies are uncertain. SMN2 is a paralogous gene that mainly differs from SMN1 by a C•G-to-T•A transition in exon 7, resulting in the skipping of exon 7 in most SMN2 transcripts and production of only low levels of survival motor neuron (SMN) protein. Genome editing technologies targeted to the SMN2 exon 7 mutation could offer a therapeutic strategy to restore SMN protein expression to normal levels irrespective of the patient SMN1 mutation. Here, we optimized a base editing approach to precisely edit SMN2, reverting the exon 7 mutation via an A•T-to-G•C base edit. We tested a range of different adenosine base editors (ABEs) and Cas9 enzymes, resulting in up to 99% intended editing in SMA patient-derived fibroblasts with concomitant increases in SMN2 exon 7 transcript expression and SMN protein levels. We generated and characterized ABEs fused to high-fidelity Cas9 variants which reduced potential off-target editing. Delivery of these optimized ABEs via dual adeno-associated virus (AAV) vectors resulted in precise SMN2 editing in vivo in an SMA mouse model. This base editing approach to correct SMN2 should provide a long-lasting genetic treatment for SMA with advantages compared to current nucleic acid, small molecule, or exogenous gene replacement therapies. More broadly, our work highlights the potential of PAMless SpRY base editors to install edits efficiently and safely.

Potent and uniform fetal hemoglobin induction via base editing.

Inducing fetal hemoglobin (HbF) in red blood cells can alleviate ß-thalassemia and sickle cell disease. We compared five strategies in CD34+ hematopoietic stem and progenitor cells, using either Cas9 nuclease or adenine base editors. The most potent modification was adenine base editor generation of γ-globin -175A>G. Homozygous -175A>G edited erythroid colonies expressed 81 ± 7% HbF versus 17 ± 11% in unedited controls, whereas HbF levels were lower and more variable for two Cas9 strategies targeting a BCL11A binding motif in the γ-globin promoter or a BCL11A erythroid enhancer. The -175A>G base edit also induced HbF more potently than a Cas9 approach in red blood cells generated after transplantation of CD34+ hematopoietic stem and progenitor cells into mice. Our data suggest a strategy for potent, uniform induction of HbF and provide insights into γ-globin gene regulation. More generally, we demonstrate that diverse indels generated by Cas9 can cause unexpected phenotypic variation that can be circumvented by base editing.

Biochemistry detection of acetylcholinesterase activity in Trypanosoma evansi and possible functional correlations.

Several chemical and immunohistochemical techniques can be used for the detection of acetylcholinesterase (AChE) activity. In this experiment we aimed to detect AChE activity in Trypanosoma evansi. For this, the parasites were isolated from the blood of experimentally infected rats using a DEA-cellulose column. Enzymatic activity was determined in trypomastigote forms at 0, 0.2, 0.4, 0.8 and 1.2 mg/mL of protein concentrations by a standard biochemical protocol. At all concentrations tested, the study showed that T. evansi expresses the enzyme AChE and its activity was proportional to the concentration of protein, ranging between 0.64 and 2.70 µmol of AcSCh/h. Therefore, we concluded that it is possible to biochemically detect AChE in T. evansi, an enzyme that may be associated with vital functions of the parasite and also can be related to chemotherapy treatments, as further discussed in this article.

In vivo engineered B cells secrete high titers of broadly neutralizing anti-HIV antibodies in mice.

Transplantation of B cells engineered ex vivo to secrete broadly neutralizing antibodies (bNAbs) has shown efficacy in disease models. However, clinical translation of this approach would require specialized medical centers, technically demanding protocols and major histocompatibility complex compatibility of donor cells and recipients. Here we report in vivo B cell engineering using two adeno-associated viral vectors, with one coding for Staphylococcus aureus Cas9 (saCas9) and the other for 3BNC117, an anti-HIV bNAb. After intravenously injecting the vectors into mice, we observe successful editing of B cells leading to memory retention and bNAb secretion at neutralizing titers of up to 6.8 µg ml-1. We observed minimal clustered regularly interspaced palindromic repeats (CRISPR)-Cas9 off-target cleavage as detected by unbiased CHANGE-sequencing analysis, whereas on-target cleavage in undesired tissues is reduced by expressing saCas9 from a B cell-specific promoter. In vivo B cell engineering to express therapeutic antibodies is a safe, potent and scalable method, which may be applicable not only to infectious diseases but also in the treatment of noncommunicable conditions, such as cancer and autoimmune disease.

Production and characterization of virus-free, CRISPR-CAR T cells capable of inducing solid tumor regression.

BACKGROUND: Chimeric antigen receptor (CAR) T cells have demonstrated high clinical response rates against hematological malignancies (e.g., CD19+ cancers) but have shown limited activity in patients with solid tumors. Recent work showed that precise insertion of a CAR at a defined locus improves treatment outcomes in the context of a CD19 CAR; however, it is unclear if such a strategy could also affect outcomes in solid tumors. Furthermore, CAR manufacturing generally relies on viral vectors for gene delivery, which comprise a complex and resource-intensive part of the manufacturing supply chain. METHODS: Anti-GD2 CAR T cells were generated using CRISPR/Cas9 within 9 days using recombinant Cas9 protein and nucleic acids, without any viral vectors. The CAR was specifically targeted to the T cell receptor alpha constant gene (TRAC). T cell products were characterized at the level of the genome, transcriptome, proteome, and secretome using CHANGE-seq, targeted next-generation sequencing, scRNA-seq, spectral cytometry, and ELISA assays, respectively. Functionality was evaluated in vivo in an NSG™ xenograft neuroblastoma model. RESULTS: In comparison to retroviral CAR T cells, virus-free CRISPR CAR (VFC-CAR) T cells exhibit TRAC-targeted genomic integration of the CAR transgene, elevation of transcriptional and protein characteristics associated with a memory-like phenotype, and low tonic signaling prior to infusion arising in part from the knockout of the T cell receptor. On exposure to the GD2 target antigen, anti-GD2 VFC-CAR T cells exhibit specific cytotoxicity against GD2+ cells in vitro and induce solid tumor regression in vivo. VFC-CAR T cells demonstrate robust homing and persistence and decreased exhaustion relative to retroviral CAR T cells against a human neuroblastoma xenograft model. CONCLUSIONS: This study leverages virus-free genome editing technology to generate CAR T cells featuring a TRAC-targeted CAR, which could inform manufacturing of CAR T cells to treat cancers, including solid tumors.

CRISPR-Cas9-AAV versus lentivector transduction for genome modification of X-linked severe combined immunodeficiency hematopoietic stem cells.

Introduction: Ex vivo gene therapy for treatment of Inborn errors of Immunity (IEIs) have demonstrated significant clinical benefit in multiple Phase I/II clinical trials. Current approaches rely on engineered retroviral vectors to randomly integrate copy(s) of gene-of-interest in autologous hematopoietic stem/progenitor cells (HSPCs) genome permanently to provide gene function in transduced HSPCs and their progenies. To circumvent concerns related to potential genotoxicities due to the random vector integrations in HSPCs, targeted correction with CRISPR-Cas9-based genome editing offers improved precision for functional correction of multiple IEIs. Methods: We compare the two approaches for integration of IL2RG transgene for functional correction of HSPCs from patients with X-linked Severe Combined Immunodeficiency (SCID-X1 or XSCID); delivery via current clinical lentivector (LV)-IL2RG versus targeted insertion (TI) of IL2RG via homology-directed repair (HDR) when using an adeno-associated virus (AAV)-IL2RG donor following double-strand DNA break at the endogenous IL2RG locus. Results and discussion: In vitro differentiation of LV- or TI-treated XSCID HSPCs similarly overcome differentiation block into Pre-T-I and Pre-T-II lymphocytes but we observed significantly superior development of NK cells when corrected by TI (40.7% versus 4.1%, p = 0.0099). Transplants into immunodeficient mice demonstrated robust engraftment (8.1% and 23.3% in bone marrow) for LV- and TI-IL2RG HSPCs with efficient T cell development following TI-IL2RG in all four patients' HSPCs. Extensive specificity analysis of CRISPR-Cas9 editing with rhAmpSeq covering 82 predicted off-target sites found no evidence of indels in edited cells before (in vitro) or following transplant, in stark contrast to LV's non-targeted vector integration sites. Together, the improved efficiency and safety of IL2RG correction via CRISPR-Cas9-based TI approach provides a strong rationale for a clinical trial for treatment of XSCID patients.

Biochemical detection of adenosine deaminase in Trypanosoma evansi.

Biochemical and molecular research on parasites has increased considerably in trypanosomes in the recent years. Many of them have the purpose of identify areas, proteins and structures of the parasite which are vulnerable and could be used in therapy against the protozoan. Based on this hypothesis this study aimed to detect biochemically the enzyme adenosine deaminase (ADA) in Trypanosoma evansi, and to adapt an assay to the measurement of its activity in trypomastigotes. Firstly, the parasites were separated from the blood of mice experimentally infected with a DEAE-cellulose column. The ADA activity in trypomastigotes was evaluated at concentrations of 0.1, 0.2, 0.5, 0.6 and 0.8mg of protein by spectrophotometry. ADA activity was observed in the parasites at all concentrations tested and its activity was proportional to the concentration of protein, ranging between 0.64 and 2.24U/L in the lowest and highest concentration of protein, respectively. Therefore, it is possible to detect biochemically ADA in T. evansi, an enzyme that may be associated with vital functions of the parasite, similar to what occurs in mammals. This knowledge may be useful in the association of the chemotherapic treatment with specific inhibitors of the enzyme, in future studies.

Susceptibility of mice to Trypanosoma evansi treated with human plasma containing different concentrations of apolipoprotein L-1.

The aim of this study was to test the susceptibility of mice to Trypanosoma evansi treated with human plasma containing different concentrations of apolipoprotein L-1 (APOL1). For this experiment, a strain of T. evansi and human plasma (plasmas 1, 2, and 3) from 3 adult males clinically healthy were used. In vivo test used 50 mice divided in 5 groups (A to E) with 10 animals in each group. Animals of groups B to E were infected, and then treated with 0.2 ml of human plasma in the following outline: negative control (A), positive control (B), treatment with plasma 1 (C), treatment with plasma 2 (D), and treatment with plasma 3 (E). Mice treated with human plasma showed an increase in longevity of 40.9 ± 0.3 (C), 20 ± 9.0 (D) and 35.6 ± 9.3 (E) days compared to the control group (B) which was 4.3 ± 0.5 days. The number of surviving mice and free of the parasite (blood smear and PCR negative) at the end of the experiment was 90%, 0%, and 60% for groups C, D, and E, respectively. The quantification of APOL1 was performed due to the large difference in the treatments that differed in the source plasma. In plasmas 1, 2, and 3 was detected the concentration of 194, 99, and 115 mg/dl of APOL1, respectively. However, we believe that this difference in the treatment efficiency is related to the level of APOL1 in plasmas.

Defining genome-wide CRISPR-Cas genome-editing nuclease activity with GUIDE-seq.

Genome-wide unbiased identification of double-stranded breaks enabled by sequencing (GUIDE-seq) is a sensitive, unbiased, genome-wide method for defining the activity of genome-editing nucleases in living cells. GUIDE-seq is based on the principle of efficient integration of an end-protected double-stranded oligodeoxynucleotide tag into sites of nuclease-induced DNA double-stranded breaks, followed by amplification of tag-containing genomic DNA molecules and high-throughput sequencing. Here we describe a detailed GUIDE-seq protocol including cell transfection, library preparation, sequencing and bioinformatic analysis. The entire protocol including cell culture can be completed in 9 d. Once tag-integrated genomic DNA is isolated, library preparation, sequencing and analysis can be performed in 3 d. The result is a genome-wide catalog of off-target sites ranked by nuclease activity as measured by GUIDE-seq read counts. GUIDE-seq is one of the most sensitive cell-based methods for defining genome-wide off-target activity and has been broadly adopted for research and therapeutic use.

Prime editing in mice reveals the essentiality of a single base in driving tissue-specific gene expression.

BACKGROUND: Most single nucleotide variants (SNVs) occur in noncoding sequence where millions of transcription factor binding sites (TFBS) reside. Here, a comparative analysis of CRISPR-mediated homology-directed repair (HDR) versus the recently reported prime editing 2 (PE2) system was carried out in mice over a TFBS called a CArG box in the Tspan2 promoter. RESULTS: Quantitative RT-PCR showed loss of Tspan2 mRNA in aorta and bladder, but not heart or brain, of mice homozygous for an HDR-mediated three base pair substitution in the Tspan2 CArG box. Using the same protospacer, mice homozygous for a PE2-mediated single-base substitution in the Tspan2 CArG box displayed similar cell-specific loss of Tspan2 mRNA; expression of an overlapping long noncoding RNA was also nearly abolished in aorta and bladder. Immuno-RNA fluorescence in situ hybridization validated loss of Tspan2 in vascular smooth muscle cells of HDR and PE2 CArG box mutant mice. Targeted sequencing demonstrated variable frequencies of on-target editing in all PE2 and HDR founders. However, whereas no on-target indels were detected in any of the PE2 founders, all HDR founders showed varying levels of on-target indels. Off-target analysis by targeted sequencing revealed mutations in many HDR founders, but none in PE2 founders. CONCLUSIONS: PE2 directs high-fidelity editing of a single base in a TFBS leading to cell-specific loss in expression of an mRNA/long noncoding RNA gene pair. The PE2 platform expands the genome editing toolbox for modeling and correcting relevant noncoding SNVs in the mouse.

Deleting DNMT3A in CAR T cells prevents exhaustion and enhances antitumor activity.

Chimeric antigen receptor (CAR) T cell therapy is revolutionizing cancer immunotherapy for patients with B cell malignancies and is now being developed for solid tumors and chronic viral infections. Although clinical trials have demonstrated the curative potential of CAR T cell therapy, a substantial and well-established limitation is the heightened contraction and transient persistence of CAR T cells during prolonged antigen exposure. The underlying mechanism(s) for this dysfunctional state, often termed CAR T cell exhaustion, remains poorly defined. Here, we report that exhaustion of human CAR T cells occurs through an epigenetic repression of the T cell's multipotent developmental potential. Deletion of the de novo DNA methyltransferase 3 alpha (DNMT3A) in T cells expressing first- or second-generation CARs universally preserved the cells' ability to proliferate and mount an antitumor response during prolonged tumor exposure. The increased functionality of the exhaustion-resistant DNMT3A knockout CAR T cells was coupled to an up-regulation of interleukin-10, and genome-wide DNA methylation profiling defined an atlas of genes targeted for epigenetic silencing. This atlas provides a molecular definition of CAR T cell exhaustion, which includes many transcriptional regulators that limit the "stemness" of immune cells, including CD28, CCR7, TCF7, and LEF1. Last, we demonstrate that this epigenetically regulated multipotency program is firmly coupled to the clinical outcome of prior CAR T cell therapies. These data document the critical role epigenetic mechanisms play in limiting the fate potential of human T cells and provide a road map for leveraging this information for improving CAR T cell efficacy.