CRISPR-Cas9 mediated LAG-3 disruption in CAR-T cells

Abstract

T cells engineered with chimeric antigen receptor (CAR) have been successfully applied to treat advanced refractory B cell malignancy. However, many challenges remain in extending its application toward the treatment of solid tumors. The immunosuppressive nature of tumor microenvironment is considered one of the key factors limiting CAR-T efficacy. One negative regulator of T cell activity is lymphocyte activation gene-3 (LAG-3). We successfully generated LAG-3 knockout T and CAR-T cells with high efficiency using CRISPR-Cas9 mediated gene editing and found that the viability and immune phenotype were not dramatically changed during in vitro culture. LAG-3 knockout CAR-T cells displayed robust antigen-specific antitumor activity in cell culture and in murine xenograft model, which is comparable to standard CAR-T cells. Our study demonstrates an efficient approach to silence immune checkpoint in CAR-T cells via gene editing.

Keywords

CAR-T / CRISPR-Cas9 / LAG-3

Introduction

Chimeric antigen receptor T (CAR-T) cell therapy has gained tremendous popularity as a promising treatment for cancer in recent years. As an artificial T cell receptor, CAR is composed of a single-chain variable fragment (scFv) derived from a monoclonal antibody recognizing an antigen, a transmembrane domain, and an intracellular domain derived from CD3Z and/or co-stimulatory molecules [1]. CAR-T cell therapy has been exceptionally successful in treating B cell acute lymphoblastic leukemia (B-ALL) in early phase clinical studies, with a complete response rate of 70%–90% [2–4]. However, the clinical response, when targeting other CD19-positive hematological malignancies such as chronic lymphoblastic leukemia (CLL), is less effective, with a response rate of approximately 50%. The reason for the low response rate was possibly because of the immune suppression through T cell checkpoint inhibitory receptors [5]. Similarly, the tumor immunosuppressive microenvironment (TME) poses serious challenge to CAR-T’s efficacy when treating solid tumors.

The inhibitory pathways of the immune system that modulate immune responses are referred to as immune checkpoints, which are utilized by tumor to gain immune resistance, particularly by inhibiting tumor-specific T cells [6]. The cell surface protein LAG-3 is an important inhibitory receptor with structural homologies to CD4. The protein binds MHC class II molecules with a cell with higher affinity than CD4 [7,8]. LAG-3 also interacts with LSECT in a surface lectin of the DC-SIGN family, which is expressed on dendritic cells and also on tumor tissue [9]. LAG-3 is expressed on activated CD4 and CD8 T cells, regulatory T cells (Tregs), natural killer (NK) cells, B cells, and plasmacytoid dendritic cells [9–15].

Accumulating evidence indicates that LAG-3 exhibits a negative impact on effector function of T cells in vivo and in vitro [9,16–18] and enhances the suppressive function of Treg cells [13,19]. As a T cell exhaustion marker, LAG-3 is upregulated in cancer and chronic infections [20,21]. LAG-3 blockade in vitro augments T cell proliferation and cytokine productions, leading to an increase in memory cells due to a delayed cell cycle arrest [22,23]. Dual blockade of PD-1 and LAG-3 reverses T cell exhaustion and improves therapeutic activity in preclinical models of chronic infection and cancers in a synergistic manner [20,21,24–28]. Clinical trials, which explore the use of anti-LAG-3 antibodies either alone or in combination with anti-PD-1 in both solid and hematologic cancers, are in progress (ClinicalTrials.gov, numbers: NCT02061761, NCT01968109).

Permanently silencing immune checkpoint receptors, such as PD-1 and LAG-3 in CAR-T cells through gene editingin vitro before adoptive cell transfer, represents an exciting approach to improve the efficacy of CAR-T cells, while avoiding the potential toxicities associated with the long-term administration of anti-LAG-3 antibodies. While PD-1 has been eliminated from T cells using gene editing and showed elevated T cell function [29–32], human T and CAR-T cells with LAG-3 knockout have not been described.

Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and the CRISPR-associated protein (Cas) (CRISPR-Cas9) system is a powerful system for gene editing. We have established methods to disrupt single and multiple genes in CAR-T cells with high efficiency [30,31]. In this study, we eliminated LAG-3 expression on human T and CAR-T cells using CRISPR-Cas9 and characterized the phenotype and functions of these modified cells.

Materials and methods

Primary human umbilical cord blood (UCB)-derived T cells

Fresh UCB units were obtained from the Beijing Cord Blood Bank (Beijing, China) with informed consent from healthy volunteer donors. After Histopaque-1077 (Sigma-Aldrich) gradient separation, mononuclear cells were collected, and T cells were isolated using the EasySep human T cell enrichment kit (Stemcell Technologies). T cells were stimulated with anti-CD3/CD28 Dynabeads (Thermo Fisher Scientific) at a bead to T cell ratio of 1:1 and cultured in X-vivo15 medium (Lonza) supplemented with 5% (v/v) heat-inactivated fetal bovine serum, 2 mmol/L L-glutamine, and 1 mmol/L sodium pyruvate in the presence of 300 IU/ml recombinant human IL-2 (all from Thermo Fisher Scientific).

Cell lines

K562 (erythroleukemia cell line) and Raji (Burkitt’s lymphoma cell line) were purchased from American Type Culture Collection (ATCC). Raji-ffluc for bioluminescent imaging and K562-CD19 cells were generated as previously described [31]. All cell lines were grown under standard conditions in RPMI1640 medium (Thermo Fisher Scientific). Lentiviral producer cell lines 293T (ATCC-CRL3216) were maintained in DMEM (Thermo Fisher Scientific). All media were supplemented with 10% (v/v) fetal bovine serum, 100 U/ml penicillin and streptomycin, 2 mmol/L L-glutamine, and 1 mmol/L sodium pyruvate. All cell lines were grown at 37 °C in a 5% CO₂ atmosphere.

sgRNA design andin vitro T7 transcription of sgRNA

We obtained the first exon sequence of LAG-3 from NCBI and used the CRISPR Design Tool (http://crispr.mit.edu) to design sgRNAs. Oligonucleotides containing T7 promoter and 20 bp targeting sequences were synthesized as forward primer (Supplementary Table S1). T7-sgRNA PCR product was amplified using pX330 plasmid (Addgene plasmid #4223) as template, column-purified, and used as the template for in vitro transcription using MEGAshortscript T7 kit (Thermo Fisher Scientific). RNAs were purified using MEGAclear columns (Thermo Fisher Scientific) and eluted in RNase-free water.

Generation of LAG-3 knockout T cells

T cells were stimulated for 3 days, and then 1 × 10⁶ cells were electroporated with 3 mg Cas9 protein (Thermo Fisher Scientific) and 3 mg sgRNAs targeting the LAG-3 exon1 by 4D-Nucleofector System N (Lonza) using the P3 Primary Cell 4D-Nucleofector X Kit (V4XP-3024, Lonza) according to manufacturer instructions. Program EO-115 was used. After electroporation, cells were resuspended in 2 ml pre-warmed T cell medium and transferred into 12-well cell plate and incubated at 37 °C in 5% CO₂. The transfection efficiency was evaluated 3 days after electroporation. Cell culture medium was half replaced by fresh complete medium every 2 days to 3 days. A similar procedure was used to generate LAG-3 knockout CAR-T cells. Freshly purified primary T cells were activated for 1 day and then transduced with lentiviral vectors, harboring a second-generation CD19 CAR. The structure of the CAR was previously described [31]. Two days after transduction, CAR-T cells were electroporated with Cas9 protein and sgRNAs using the previously described procedure.

Flow cytometry

All samples were analyzed with CytoFLEX (Beckman Coulter Inc.) on the following days after transfection. The mouse anti-human antibodies to the following antigens were used: CD45RO-PE (UCHL1, Biolegend), CD45RA-PerCP/Cy5.5 (HI100, Biolegend), CD62L-Brilliant Violet421 (DREG-56, Biolegend), CD223 (LAG-3)-PE (11C3C65, Biolegend), CD8-APC (HIT8a, Biolegend), and CD4-PE (A161A1, Biolegend).

Sequencing of the mutations

Indels were quantified by TIDE (tracking of indels by decomposition) analysis and clonal sequence analysis. Cells were harvested, and the genomic DNA was extracted with 100 mg/ml proteinase K in lysis buffer (10 mmol/L Tris-HCl, 2 mmol/L EDTA, 2.5% Tween-20, and 2.5% Triton-X 100). The primers used for the amplification of target locus and sequencing are listed in Table S1. The PCR products were sequenced for TIDE analysis using web tool (available at http://tide.nki.nl). The purified PCR segments were cloned into pEASY vector using pEASY Blunt Cloning Kit (Transgen Biotech) to detect mutant alleles. A total of 40–70 colonies per sample of the transformed PCR ligation products are sequenced using universal primer M13F. All PCR sequencing methods used followed the instructions provided by the manufacturer or standard molecular cloning protocols.

Cytokine enzyme-linked immunosorbent assay (ELISA)

Cytokine release assays were performed by coculture of effector (CAR-T, LAG-3-KO-CAR-T, T) with target tumor cells (Raji, K19, K562) at a 1:1 ratio (10⁴ cells each) per well, in duplicate in 96-well V-bottom plates, in a final volume of 200 ml complete RPMI1640 medium. After 24 h, supernatants were assayed to produce IL-2 and IFN-g using ELISA kit (Biolegend).

Flow cytometry-based cytotoxicity assay

The cytotoxicity of the CAR-T cell was assessed by the flow cytometry-based cytotoxicity assay [31]. The lytic activities of effector cells were tested by Violet/Annexin V and 7-AAD labeling assay. Target tumor cells were labeled with 1 mmol/L Celltrace Violet and then incubated with effector cells by different effectors to target ratio for 4 h. FITC-Annexin V and 7-AAD (Biolegend) were added to determine the ratio of dead target cells. Samples were analyzed by flow cytometry. Target cells were selected by gating the Violet-positive cell population and further analyzed for different subpopulations. The percentages of cytotoxic activity were calculated using the following equation: % specific lysis= {[% (Violet⁺Annexin V⁺ + Violet⁺Annexin V⁻€7-AAD⁺)−% spontaneous (Violet⁺ Annexin V⁺ + Violet⁺Annexin V⁻€7-AAD⁺)]/[100%−% spontaneous (Violet⁺Annexin V⁺ + Violet⁺Annexin V⁻€7-AAD⁺)]}×100%.

Murine xenograft studies

To establish the Raji-ffluc tumor model, 6−12 weeks old NOD-Prkdcscid Il2rgnull (NPG) mice (VITALSTAR, Beijing, China) were injected with 2 × 10⁵ Raji-ffluc cells via intraperitoneal injection on day 0. Three days after injection, tumor engraftment was evaluated by serial biophotonic imaging using the NightOWL LB983in vivo Imaging System (Berthold Technologies). Mice with comparable tumor loads were segregated into different treatment groups. T cells were administered at a dose of 1 × 10⁷ cells/mouse via intraperitoneal injection. The tumor loads were evaluated 7 days after treatment.

Results

Screening for the most efficient sgRNAs, targeting LAG-3 in T cells

To achieve efficient gene disruption, we designed five sgRNAs, targeting the first exon of LAG-3(Table S1). The gene editing efficiency using each sgRNA was quantified by TIDE analysis [33], and the most efficient sgRNA (sgRNA5) was selected for further experiments (Fig. 1A and 1B). Gene editing using sgRNA5 in another donor-derived T cells achieved similar high LAG-3 knockout efficiency (Fig. 1B). We amplified and sub-cloned the sgRNA5 target region and identified mutant alleles. A total of 30 out of 49 sequenced alleles are mutants, confirming the high gene disruption efficiency. As shown in Fig. 1C, all mutations recovered occurred precisely at the sgRNA targeting region. These results indicate that LAG-3 can be efficiently eradicated in primary T cells.

Effect of electroporation and LAG-3 knockout on T cell proliferation and phenotype

Next, the effect of LAG-3 knockout on T cell proliferation and surface phenotype was determined. Upon anti-CD3 and anti-CD28 antibody stimulations, T cells with LAG-3 knockout maintained normal proliferation (Fig. 2A). The immune phenotype of the gene-edited T cell was evaluated by the expression of CD4 and CD8, as well as the characteristics of naïve (CD45RO^-/CD62L⁺, TN), central memory (CD45RO⁺ /CD62L⁺, TCM), and effector memory (CD45RO⁺/CD62L^-, TEM) T cell subsets. Compared with control T cells, a higher fraction of CD4 and TCM and a lower faction of CD8 cells were observed in LAG-3 knockout T cells. However, this effect is probably caused by electroporation given that this feature was also observed in T cells receiving electroporation without Cas9 RNP (Fig. 2B). Overall, LAG-3 knockout T cells displayed characteristics similar to non-edited T cells.

Off-target analysis

The potential off-target mutation is a major concern of CRISPR-Cas9 gene editing. The top five potential off-target sites for sgRNA5 were predicted using Benchling software [34], and the five loci in the T cells with LAG-3 knockout were genotyped (Table S1). We did not detect mutation at any of these loci using TIDE analysis (Table S2).

Preparation and characterization of LAG-3 knockout CD19 CAR-T cells

CD19 is presented in most B cell malignancies, and anti-CD19 CAR-T cells were best characterized and successfully applied to treating CD19⁺ B cell leukemia and lymphomas [2–4]. We successfully generated LAG-3 knockout anti-CD19 CAR-T cells, and LAG-3 knockout efficiency is approximately 45%–70% in CAR-T cells produced from three different donors (Fig. 3A and 3B). Loss of LAG-3 function was confirmed by detecting LAG-3 protein expression using flow cytometry (Fig. 3C). The CAR-T cells electroporated with Cas9 RNP effectively proliferated. However, we did not observe much robust cell proliferation in LAG-3 knockout CAR-T cells compared to control cells (Fig. 3D).

The surface expression of LAG-3 is low after expansion; thus, performing enrichment for LAG-3 knockout CAR-T cells is difficult. Therefore, the prepared LAG-3 knockout CAR-T cells are a mixture of cells with and without LAG-3 mutations. Samples from donors 2 and 3 with LAG-3 knockout efficiency of more than 70% were used for the following functional analysis (Fig. 3A).

In vitro characterization of LAG-3 knockout CD19 CAR-T cells

To characterize CAR-T cells with LAG-3 disruption phenotypically, we harvested CAR-T cells on day 14 post transfection. Similar to the primary T cells, the LAG-3 knockout CAR-T cells did not show major change in the expression of CD4 and CD8 as well as the memory T cell phenotypic features compared with control CAR-T cells electroporated without Cas9 RNP (Fig. 4A). Consistent with the results in T cells, these data suggest that although electroporation has some effect, CRISPR-Cas9-mediated LGA-3 disruption does not interfere with the immunophenotype of CAR-T cells.

The cytotoxic function of LAG-3 knockout CAR-T cells was first evaluated using in vitro culture. Similar to standard CAR-T cells, LAG-3 knockout CAR-T cells released IL-2 and IFN-g only when they were cocultured with CD19-expressing Raji or K562-CD19 cells (Fig. 4B). This cytokine release was dependent on CAR given that non-transduced T cells failed to release cytokines in the presence of CD19 positive cells. Furthermore, the amount of cytokines released by LAG-3 knockout CAR-T cells is similar to standard CAR-T cells (Fig. 4B). In cytotoxicity assay, LAG-3 knockout CAR-T cells showed CAR-dependent lysis of CD19 tumor targets (Fig. 4C). These results indicate that CAR-T cells with LAG-3 disruption maintain antitumor activity equivalent to standard CAR-T cells.

LAG-3 knockout CD19 CAR-T cells eradicate tumor in murine xenograft model

To further evaluate the antitumor function of the LAG-3 knockout CAR-T cells, we employed Raji-ffluc lymphoma xenograft models. Except for one mouse (likely due to technical error) in CAR-T treated group, tumor size was substantially smaller in all mice treated with standard CAR-T cells and LAG-3 knockout CAR-T cells, whereas progressive tumor growth was observed in the control group of mice treated with un-modified T cells or PBS (Fig. 5A and 5B). The survival of mice that received LAG-3 knockout and standard CAR-T cells is also longer (Fig. 5C). The tumor burden of mice treated with standard CAR-T and LAG-3 knockout CAR-T was similar, indicating that CAR-T cells with and without LAG-3disruption have similar antitumor efficacy in this Raji-ffluc lymphoma murine xenograft model.

To evaluate the engraftment of CAR-T cells in vivo, we sacrificed one of the LAG-3 knockout CAR-T cells treated mice 71 days after treatment, and no tumor was detectable. The peripheral blood, peritoneal wash fluid, and spleen cells were collected, and human T cells and CAR-T cells were detected using flow cytometry. As shown in Fig. S1, extremely limited human T and CAR-T cells were found in mouse peripheral blood, whereas some human T and CAR-T cells exist in the peritoneal cavity and spleen, indicating that these cells can achieve long-term engraftment in vivo.

Discussion

With a complete response rate of 70%–90% [2–4], CAR-T therapy offers an effective treatment for B cell acute lymphoblastic leukemia. Building on this success, efforts are ongoing to extend it toward the treatment of solid tumors; however, the clinical response is extremely limited [1]. Recent studies suggest that CAR-T cells and endogenous T cells are likely to suffer the loss of cytotoxic functionality in immunosuppressive tumor microenvironment [6]. Strategies that improve the efficacy of CAR-T cells, by preventing the negative regulation of effector functions mediated by the immunosuppressive tumor microenvironment, are critical for solid tumor immunotherapy. The well-known features and functions of LAG-3 allow it to be an attractive target for immune modulation.

To this end, we described, for the first time, the disrupting LAG-3 expression in human primary T cells and CAR-T cells using CRISPR-Cas9 system. We demonstrate that the delivery of CRISPR-Cas9 system via electroporation provides an efficient platform to knockout LAG-3 in T and CAR-T cells. Different from in vivo study using mouse model [22], we did not find any significant difference in the cell expansion rate between LAG-3 knockout cells and control cells. In addition, the LAG-3 knockout CAR-T cells were not adversely affected by the electroporation process and were rich in central memory subtype. LAG-3 knockout CAR-T cells maintained their antigen-specific cytokine release and antitumor potency in vitro and in vivo.

Previous studies demonstrated that the blockade of LAG-3 by monoclonal antibody enhanced cytokine release and cytoxicity in vivo and in vitro [9,16–18]. The present study did not find a positive effect on CAR-T effector function following LAG-3 disruption. One possibility is that our murine tumor model does not recapitulate primary tumor immunosuppressive environment. Therefore, LAG-3-disrupted CAR-T cells do not show advantage of resisting T cell exhaustion. In addition, LAG-3 and PD-1 work in a synergetic manner; thus, only blocking LAG-3 function might not be sufficient to show superior efficacy [25]. Therefore, future efforts are required to explore PD-1 and LAG-3 double knockout CAR-T cells, which are evaluated in a murine model with better TME setup, such as PDX model. We only achieved ~70% LAG-3 knockout efficiency; thus, the remaining 30% WT cells might also affect the functional evaluations.

In conclusion, we established a CRISPR-Cas9-based method for highly efficient disruption of LAG-3 in human primary T cells and CAR-T cells. While the potential advantages of LAG-3 knockout CAR-T cells require further study, using gene editing to silence immune check points promises to improve the efficacy of CAR-T cells treating solid tumors.

References

[1] Kakarla S, Gottschalk S. CAR T cells for solid tumors: armed and ready to go? Cancer J 2014; 20(2): 151–155

[2] Davila ML, Riviere I, Wang X, Bartido S, Park J, Curran K, Chung SS, Stefanski J, Borquez-Ojeda O, Olszewska M, Qu J, Wasielewska T, He Q, Fink M, Shinglot H, Youssif M, Satter M, Wang Y, Hosey J, Quintanilla H, Halton E, Bernal Y, Bouhassira DC, Arcila ME, Gonen M, Roboz GJ, Maslak P, Douer D, Frattini MG, Giralt S, Sadelain M, Brentjens R. Efficacy and toxicity management of 19-28z CAR T cell therapy in B cell acute lymphoblastic leukemia. Sci Transl Med 2014; 6(224): 224ra25

[3] Maude SL, Frey N, Shaw PA, Aplenc R, Barrett DM, Bunin NJ, Chew A, Gonzalez VE, Zheng Z, Lacey SF, Mahnke YD, Melenhorst JJ, Rheingold SR, Shen A, Teachey DT, Levine BL, June CH, Porter DL, Grupp SA. Chimeric antigen receptor T cells for sustained remissions in leukemia. N Engl J Med 2014; 371(16): 1507–1517

[4] Lee DW, Kochenderfer JN, Stetler-Stevenson M, Cui YK, Delbrook C, Feldman SA, Fry TJ, Orentas R, Sabatino M, Shah NN, Steinberg SM, Stroncek D, Tschernia N, Yuan C, Zhang H, Zhang L, Rosenberg SA, Wayne AS, Mackall CL. T cells expressing CD19 chimeric antigen receptors for acute lymphoblastic leukaemia in children and young adults: a phase 1 dose-escalation trial. Lancet 2015; 385(9967): 517–528

[5] McClanahan F, Riches JC, Miller S, Day WP, Kotsiou E, Neuberg D, Croce CM, Capasso M, Gribben JG. Mechanisms of PD-L1/PD-1-mediated CD8 T-cell dysfunction in the context of aging-related immune defects in the Eµ-TCL1 CLL mouse model. Blood 2015; 126(2): 212–221

[6] Khalil DN, Smith EL, Brentjens RJ, Wolchok JD. The future of cancer treatment: immunomodulation, CARs and combination immunotherapy. Nat Rev Clin Oncol 2016; 13(5): 273–290

[7] Triebel F, Jitsukawa S, Baixeras E, Roman-Roman S, Genevee C, Viegas-Pequignot E, Hercend T. LAG-3, a novel lymphocyte activation gene closely related to CD4. J Exp Med 1990; 171(5): 1393–1405

[8] Huard B, Prigent P, Tournier M, Bruniquel D, Triebel F. CD4/major histocompatibility complex class II interaction analyzed with CD4- and lymphocyte activation gene-3 (LAG-3)-Ig fusion proteins. Eur J Immunol 1995; 25(9): 2718–2721

[9] Xu F, Liu J, Liu D, Liu B, Wang M, Hu Z, Du X, Tang L, He F. LSECtin expressed on melanoma cells promotes tumor progression by inhibiting antitumor T-cell responses. Cancer Res 2014; 74(13): 3418–3428

[10] Baixeras E, Huard B, Miossec C, Jitsukawa S, Martin M, Hercend T, Auffray C, Triebel F, Piatier-Tonneau D. Characterization of the lymphocyte activation gene 3-encoded protein. A new ligand for human leukocyte antigen class II antigens. J Exp Med 1992; 176(2): 327–337

[11] Huard B, Gaulard P, Faure F, Hercend T, Triebel F. Cellular expression and tissue distribution of the human LAG-3-encoded protein, an MHC class II ligand. Immunogenetics 1994; 39(3): 213–217

[12] Workman CJ, Rice DS, Dugger KJ, Kurschner C, Vignali DA. Phenotypic analysis of the murine CD4-related glycoprotein, CD223 (LAG-3). Eur J Immunol 2002; 32(8): 2255–2263

[13] Huang CT, Workman CJ, Flies D, Pan X, Marson AL, Zhou G, Hipkiss EL, Ravi S, Kowalski J, Levitsky HI, Powell JD, Pardoll DM, Drake CG, Vignali DA. Role of LAG-3 in regulatory T cells. Immunity 2004; 21(4): 503–513

[14] Kisielow M, Kisielow J, Capoferri-Sollami G, Karjalainen K. Expression of lymphocyte activation gene 3 (LAG-3) on B cells is induced by T cells. Eur J Immunol 2005; 35(7): 2081–2088

[15] Workman CJ, Wang Y, El Kasmi KC, Pardoll DM, Murray PJ, Drake CG, Vignali DA. LAG-3 regulates plasmacytoid dendritic cell homeostasis. J Immunol 2009; 182(4): 1885–1891

[16] Hannier S, Triebel F. The MHC class II ligand lymphocyte activation gene-3 is co-distributed with CD8 and CD3-TCR molecules after their engagement by mAb or peptide-MHC class I complexes. Int Immunol 1999; 11(11): 1745–1752

[17] Workman CJ, Dugger KJ, Vignali DA. Cutting edge: molecular analysis of the negative regulatory function of lymphocyte activation gene-3. J Immunol 2002; 169(10): 5392–5395

[18] Sierro S, Romero P, Speiser DE. The CD4-like molecule LAG-3, biology and therapeutic applications. Expert Opin Ther Targets 2011; 15(1): 91–101

[19] Okamura T, Fujio K, Shibuya M, Sumitomo S, Shoda H, Sakaguchi S, Yamamoto K. CD4⁺CD25^-LAG3⁺ regulatory T cells controlled by the transcription factor Egr-2. Proc Natl Acad Sci USA 2009; 106(33): 13974–13979

[20] Wherry EJ, Ha SJ, Kaech SM, Haining WN, Sarkar S, Kalia V, Subramaniam S, Blattman JN, Barber DL, Ahmed R. Molecular signature of CD8⁺ T cell exhaustion during chronic viral infection. Immunity 2007; 27(4): 670–684

[21] Matsuzaki J, Gnjatic S, Mhawech-Fauceglia P, Beck A, Miller A, Tsuji T, Eppolito C, Qian F, Lele S, Shrikant P, Old LJ, Odunsi K. Tumor-infiltrating NY-ESO-1-specific CD8⁺ T cells are negatively regulated by LAG-3 and PD-1 in human ovarian cancer. Proc Natl Acad Sci USA 2010; 107(17): 7875–7880

[22] Workman CJ, Cauley LS, Kim IJ, Blackman MA, Woodland DL, Vignali DA. Lymphocyte activation gene-3 (CD223) regulates the size of the expanding T cell population following antigen activation in vivo. J Immunol 2004; 172(9): 5450–5455

[23] Maçon-Lemaître L, Triebel F. The negative regulatory function of the lymphocyte-activation gene-3 co-receptor (CD223) on human T cells. Immunology 2005; 115(2): 170–178

[24] Blackburn SD, Shin H, Haining WN, Zou T, Workman CJ, Polley A, Betts MR, Freeman GJ, Vignali DA, Wherry EJ. Coregulation of CD8⁺ T cell exhaustion by multiple inhibitory receptors during chronic viral infection. Nat Immunol 2009; 10(1): 29–37

[25] Richter K, Agnellini P, Oxenius A. On the role of the inhibitory receptor LAG-3 in acute and chronic LCMV infection. Int Immunol 2010; 22(1): 13–23

[26] Butler NS, Moebius J, Pewe LL, Traore B, Doumbo OK, Tygrett LT, Waldschmidt TJ, Crompton PD, Harty JT. Therapeutic blockade of PD-L1 and LAG-3 rapidly clears established blood-stage Plasmodium infection. Nat Immunol 2011; 13(2): 188–195

[27] Woo SR, Turnis ME, Goldberg MV, Bankoti J, Selby M, Nirschl CJ, Bettini ML, Gravano DM, Vogel P, Liu CL, Tangsombatvisit S, Grosso JF, Netto G, Smeltzer MP, Chaux A, Utz PJ, Workman CJ, Pardoll DM, Korman AJ, Drake CG, Vignali DA. Immune inhibitory molecules LAG-3 and PD-1 synergistically regulate T-cell function to promote tumoral immune escape. Cancer Res 2012; 72(4): 917–927

[28] Goding SR, Wilson KA, Xie Y, Harris KM, Baxi A, Akpinarli A, Fulton A, Tamada K, Strome SE, Antony PA. Restoring immune function of tumor-specific CD4⁺ T cells during recurrence of melanoma. J Immunol 2013; 190(9): 4899–4909

[29] Menger L, Sledzinska A, Bergerhoff K, Vargas FA, Smith J, Poirot L, Pule M, Hererro J, Peggs KS, Quezada SA. TALEN-mediated inactivation of PD-1 in tumor-reactive lymphocytes promotes intratumoral T-cell persistence and rejection of established tumors. Cancer Res 2016; 76(8): 2087–2093

[30] Ren J, Liu X, Fang C, Jiang S, June CH, Zhao Y. Multiplex genome editing to generate universal CAR T cells resistant to PD1 inhibition. Clin Cancer Res 2017 ; 23(9): 2255–2266 

[31] Liu X, Zhang Y, Cheng C, Cheng AW, Zhang X, Li N, Xia C, Wei X, Liu X, Wang H. CRISPR-Cas9-mediated multiplex gene editing in CAR-T cells. Cell Res 2017; 27(1): 154–157

[32] Su S, Hu B, Shao J, Shen B, Du J, Du Y, Zhou J, Yu L, Zhang L, Chen F, Sha H, Cheng L, Meng F, Zou Z, Huang X, Liu B. Corrigendum: CRISPR-Cas9 mediated efficient PD-1 disruption on human primary T cells from cancer patients. Sci Rep 2017; 7: 40272

[33] Brinkman EK, Chen T, Amendola M, van Steensel B. Easy quantitative assessment of genome editing by sequence trace decomposition. Nucleic Acids Res 2014; 42(22): e168

[34] Hsu PD, Scott DA, Weinstein JA, Ran FA, Konermann S, Agarwala V, Li Y, Fine EJ, Wu X, Shalem O, Cradick TJ, Marraffini LA, Bao G, Zhang F. DNA targeting specificity of RNA-guided Cas9 nucleases. Nat Biotechnol 2013; 31(9): 827–832

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg