CRISPR-Cas9 mediated efficient PD-1 disruption on human primary T cells from cancer patients

Strategies that enhance the function of T cells are critical for immunotherapy. One negative regulator of T-cell activity is ligand PD-L1, which is expressed on dentritic cells (DCs) or some tumor cells, and functions through binding of programmed death-1 (PD-1) receptor on activated T cells. Here we described for the first time a non-viral mediated approach to reprogram primary human T cells by disruption of PD-1. We showed that the gene knockout of PD-1 by electroporation of plasmids encoding sgRNA and Cas9 was technically feasible. The disruption of inhibitory checkpoint gene PD-1 resulted in significant reduction of PD-1 expression but didn’t affect the viability of primary human T cells during the prolonged in vitro culture. Cellular immune response of the gene modified T cells was characterized by up-regulated IFN-γ production and enhanced cytotoxicity. These results suggest that we have demonstrated an approach for efficient checkpoint inhibitor disruption in T cells, providing a new strategy for targeting checkpoint inhibitors, which could potentialy be useful to improve the efficacy of T-cell based adoptive therapies.


Results
The design and validation of sgRNA targeting hPD-1. Previous works have shown that simultaneous use of dual sgRNAs to target an individual gene significantly improved the Cas9-mediated genome targeting efficiency and reduce the site-dependent off-target effect 22 . So we selected 2 pairs of targeted sites on exon 2 (Fig. 1a). All sgRNA expression vectors were co-transfected into HeLa cells separately with Cas9 expression vector to test their efficiency. Blasticidin (5 μ g/ml) and puromycin (1 μ g/ml) were added 24 h after transfection for positive selection. Genomic DNA was isolated from cells harvested 72 h after transfection and screened for the presence of site-specific gene modification by PCR amplification of regions surrounding the target sites as well as T7EN1 cleavage assay. The cleavage bands were all clearly observed (Fig. 1b) and various indels were detected by the sequencing (Supplemental Fig. 1). Mutation sizes ranged from − 86 to + 51 at the efficiency of 61.9% for sg1, 52.6% for sg2, 40% for sg3, 52.6% for sg4, 47.6% for sg (1 + 2) and 38.9% for sg (3 + 4), respectively. These data demonstrated that the selected sgRNAs worked effectively with Cas9 on human genomes. We chose sg (1 + 2) in our following experiments.
Optimizing Cas9/sgRNAs delivery of primary T cells. We next aimed at optimizing conditions for the co-transfection of Cas9/sgRNA expression plasmids on primary T cells via electroporation. For this purpose, we used reporter plasmids pST1374-Cas9-GFP encoding green fluorescent protein (GFP) tagged Cas9 to evaluate the transfection efficacy through the co-electroporation with pST1374-Cas9-GFP and pGL3-U6-hPD-1-sgRNA (1 + 2). Here we referred to several related research based on primary T cells transfection by the Nucleofecter platform and expanded our experimental program beyond the Lonza protocol. To this end, a mixture of 5 μ g of pST1374-Cas9-GFP and 10 μ g of pGL3-U6-hPD-1-sgRNA plasmids was used, and different programs (Y-001, T-007, T-023, X-001, U-014, V-024) were applied. GFP expression was determined 24 h later by fluorescence microscope or flow cytometry. We found that higher transfection efficacy was obtained with program U-014 than with V-024 and T-007 ( Fig. 2a,b, p = 0.0231 and p = 0.0316). The T7EN1 cleavage assay confirmed the PD-1 mutation of T cells (Fig. 2c). Notably, during the following days, significant improved cell viability was observed with program T-007 than with U-014 and V-024 shown by Trypan blue exclusion assay (Fig. 2d, p = 0.0353 and p = 0.0058). Therefore, T-007 was used as the optimal program for the following experiments. The transfection efficacy of different molar ratio of pST1374-Cas9-GFP and pGL3-U6-hPD-1-sgRNA evaluated by GFP expression. (f) Detection of sgRNA1:Cas9-mediated cleavage of hPD-1 by T7EN1 cleavage assay using program T-007 with a molar ratio of 1:4 (pST1374-Cas9-GFP:pGL3-U6-hPD-1-sgRNA). Data shown are mean ± SD of 3 independent experiments. p < 0.05.
In addition, we found significant increase of GFP expression when plasmids were delivered at a molar ratio of 1:4 (pST1374-Cas9-GFP:pGL3-U6-hPD-1-sgRNA) (Fig. 2e). To determine the mutation of PD-1, T7EN1 cleavage assay was applied and the gene knockout efficacy was proved (Fig. 2f). These results indicated that human PBMC-derived T cells were efficiently modified by sgRNA:Cas 9 system through the electroporation of plasmids.

Cas9-mediated efficient PD-1 KO in primary T cells of patients.
To test whether the Cas9-mediated gene knockout was as efficient in patients as in healthy donors, primary T cells from two late stage cancer patients and healthy donors were transfected. Freshly isolated PBMC were activated for 3 d. A total of 10 μ g of pST1374-Cas9-GFP and 20 μ g of pGL3-U6-hPD-1-sgRNA plasmids were used for each reaction. The GFP expression was evaluated by fluorescence microscope 24 h after electroporation. We found that PBMC-derived T cells from patients (Donor 1 and Donor 2) (Fig. 3a) achieved less transfection efficacy comparing with T cells from healthy donor (Donor 3) (Fig. 3a). Over 85% viable cells were observed in all three donors evaluated by Trypan blue exclusion assay (Table 1). To determine the subsets of T cells, surface expression of CD3, CD4 and  CD8 were calculated on day 7 of culture. The results showed that over 85% of cells were CD3 + T cells and over 80% of CD3 + T cells were CD8 + cells in all three donors (Table 1). PCR results showed there harbored large fragment deletion on samples from healthy donor (H2), also indicating the higher efficacy of gene disruption than the other two donors (G2 and Z2). The subsequent T7EN1 assay and sequencing further confirmed that the mutation of PD-1 was successfully achieved in all three donors (Fig. 3b). Further characterization of the cleavage by sequencing showed, different indels were detected in all the donors with various mutation sizes, the efficacy varied among donors (10.71% in Donor 1, 14.81% in Donor 2, 66.67% in Donor 3) (Fig. 3c). These results demonstrated that the efficient disruption of gene PD-1 is achieved by sgRNA:Cas9 system although the efficacy of the gene modifications varied among different individuals.
The proliferation of gene edited primary T cells and the sustained knockout of PD-1 during the prolonged culture conditions. As adoptive T cell therapies require relatively long culture of T cells in vitro, we are curious how long the down regulation of PD-1 lasted by our gene disruption method. To this end, we first assessed the surface expression of PD-1 48 h post transfection. Here we depicted a representative out of three experiments yielding similar results. The percentage of PD-1 + T cells was 2.96% on control T cells and was 1.37% on the sgRNA:Cas9 modified T cells (Fig. 4a), though the baseline expression of PD-1 was relatively low. Then, we assessed the capacity of sgRNA:Cas9-treated T cells to proliferate in vitro upon stimulation with IL-2 by counting the total cell numbers. Over a period of 21 d, we found the proliferation of primary T cells of all these donors was not significantly affected by the disruption of PD-1 during the prolonged culture period, ranging from 2 × 10 7 cells on day 7 to 20 × 10 7 on day 21, and the fold increase was 1.68 ± 0.03, 7.57 ± 0.09 and 10.64 ± 0.27 on day 7 (p = 0.8816), day 14 (p = 0.8557) and day 21 (p = 0.7705), respectively (Fig. 4b,c). T cell clones were observed around day 7 and grew largely during the following days, indicating good proliferation and activation of the T cells (Fig. 4d). Meanwhile, T cells from a healthy donor were transfected and stimulated by autologous DCs loaded with immunogenic peptides for another two weeks. The surface expression of PD-1 of CD3 + T cells increased from 2.44% at the baseline to 4.48% of the PD-1 KO group while it increased from 5.92% to 15.9% in control group (Fig. 4e). To further mimic the in vivo environment when T cells encounter with whole tumor antigen, we co-cultured the gene modified T cells with irradiated PD-L1-high tumor cells. We observed that only 3.4% of sgRNA:Cas9 modified T cells express PD-1 while 17.5% of control T cells express PD-1 by the stimulation of whole tumor antigen (Fig. 4f). Taken together, these data represented that the proliferation capability was not affected by the disruption of PD-1 on primary T cells and that the efficient disruption of PD-1 sustained during the prolonged culture with the stimulation of tumor antigens.

The characterization of the cultured T cells by sgRNA:Cas9 mediated knock out of PD-1.
To characterize the cultured T cells with PD-1 disruption, we harvested T cells on day 21 post transfection. The T cell subsets were determined by the expression of CD4, CD8 and characterized with memory or activation markers, including CD28, CD27, CD69 and HLA-DR. The results showed, the PD-1 knock out T cells did not exhibit any significant change on the expression of CD4 (7.43 ± 2.78% on control T cells vs 13.36 ± 6% on sgRNAhPD-1 T cells, p = 0.1957) or CD8 (89.10 ± 4.5% on control T cells vs 81.87 ± 7.82% on sgRNAhPD-1 T cells, p = 0.2367) during the extended in vitro culture after gene editing (Fig. 5a). Although the decrease of CD4 + CD25 + cells on PD-1 KO T cells (4.48% vs. 2.42%) was observed in one of the donor (Fig. 5b), there was no statistical significance, p = 0.4604. It is noteworthy that, we did not find any significant change of the memory markers ( Fig. 5c), including central memory CD45RO + CD62L + T cell (26.87 ± 3.48% vs 31.53 ± 3.1%, p = 0.1583), effector memory CD45RO + CD62L − T cell (14.76 ± 5.44% vs 15.77 ± 2.58%, p = 0.7858) and naïve T cell (44.1 ± 5.79% vs 42.1 ± 8.08%, p = 0.7451). No difference of the activation marker CD28, CD27, CD69 and HLA-DR on CD3 + T cells was detected between the sgRNAhPD-1 T cells and control T cells (p > 0.01) (Fig. 5d). These results indicated that the CD4 or CD8 subsets constitution or the memory and activation status of the T cells is stable with sgRNA:Cas9-mediated PD-1 disruption.

Discussion
The effective activation of the tumor reactive T cells and the suppression of checkpoint inhibitor has long been the key problem of immunotherapy. Recently, the utilization of checkpoint blockade targeting the PD-1/ PD-L1 pathway has shown remarkable antitumor responses in patients with advanced melanoma, lung cancer and against other cancers with durable clinical responses [23][24][25][26] . T cells activated in the absence of PD-L1/PD-1 co-stimulation are functional activated, exhibiting increased proliferation by the stimulation of DCs or tumors and produce higher levels of Th-1 cytokines, in particular IFN-γ , IL-2 and TNF-α and enhanced lytic activities 27 . However, most of these studies have been performed using either blocking antibodies or RNA interference with Adoptive cell therapy using autologous gene editing T cells such as TCR-T or CAR-T has emerged as promising approach for the treatment of cancers 30 . There is demonstration that blockade of the PD-1 immunosuppressive pathway using an anti-PD-1 antibody significantly enhance the anti-tumor efficacy of genetically modified T cells expressing a chimeric antigen receptor (CAR) 31 . Despite the dramatic benefit achieved by these strategies, it has to be noted that in one hand, sustained expansion of PD-1-expressing CTLs by the stimulation of tumor antigens in vitro and in vivo may require continuous treatment with anti-PD-1 antibody which is costly, on the other hand, the long term systematic administration of the blocking antibody carries the risk of breaking immune tolerance may cause immune attack of normal tissues. In addition, RNA interference with siRNAs on T cell may be temporary and less efficient.
In our study, we described, for the first time, a new approach of inhibiting PD-1/PD-L1 co-stimulation by directly disrupting genome PD-1 expression on human primary T cells through the Cas9:sgRNA gene knock out system. This was achieved by electro-co-transfer of two DNA plasmids into cultured human T cells. A major obstacle of editing primary T cell is the low transfection efficacy. Here, we made use of nucleofection to achieve higher transfection efficacy and better cell viability. This non-viral mediated gene disruption method has the advantage of clinical application at affordable costs 32 . To optimize the established protocol for co-transfection of primary T cell, by using a GFP reporting plasmid, we observed significantly improved cell viability with program T-007 than with U-014 and V-024 which are mostly used in the studies of T cell editing 33 . In addition, the ratio of co-transfection of two plasmid were vital for the efficient gene disruption of sgRNA:Cas 9 system. Interestingly, our approach to disrupt PD-1 expression on human T cells was successfully utilized on several cancer patients and healthy donors as the requirement of T cell editing for adoptive transfer of patient's autologous lymphocytes or in some cases allogenetic lymphocytes of healthy donors. We found that PBMC-derived T cells from patients achieved less transfection efficacy comparing with T cells from healthy donors. Those functional impaired T cells may account for the low sensitivity to electroporation. The phenomenon inspired us to testify this approach firstly in the allogenetic adoptive therapy of cancer treatment using PBMC from healthy donors. Moreover, we observed for over 3 w after PD-1 gene disruption, and found no significant differences in the cell expansion rate between sgRNA:Cas9-treated cells and control cells, which encouraged us to further apply this approach into clinical application in future. In addition, the activation induced up-regulation of PD-1 was disrupted through the in vitro expansion stimulated by peptides pulsed DCs or irradiated whole tumor cells. As we know that the most popular schedule for T cell in vitro culture would not surpass 3 ~ 4 w.
Previous studies already demonstrated that blockade of PD-1/PDL1 by mAb improved IFN-γ production as well as cytotoxicity both in vitro and in vivo 29,31,34 . Here we also demonstrated that these sgRNA hPD-1:Cas9 modified primary T cells from healthy donors or late stage cancer patients exhibited enhanced IFN-γ production by stimulating with the related peptide antigens and at the same time we found improved tumor cells lysis by the disruption of PD-1 which may due to the reversed immune resistance mediated by PD-1/PDL1 interaction. IFN-γ is one of Th1 cytokines which mediated cellular immune response and activate cytotoxic T cells, indirectly regulate tumors lysis by several mechanisms 35 . Therefore, we believe in our case, IFN-γ activates cytotoxicity indirectly. Additionally, in our system using Cas9:sgRNA mediated gene editing of T cells from both patients and healthy donors, improved cytotoxicity on tumor cell lines was clarified on two PD-L1 positive target cell lines and further testified by induction of PD-L1 expression on the target cell. The expression of its receptor PD-L1 should be taken into consideration for good outcome of utilizing PD-1/PD-L1 inhibiting strategy 36 .
Nevertheless, our study did not precisely focus on a specific antigen and our exploration was nor clearly clarified in an antigen dependent manner. We are currently explore PD-1 disruption of T cells by generating CTL targeting on specific tumor antigen, as previous studies reported that sustained expansion of PD-1-expressing CTLs may require continuous treatment to interrupt PD-1/PD-L1 co-stimulation 27 . In addition, as it is difficult to obtain large number of PD-1 high expression T cells from periphery lymphocytes, the difference of the functional analyzes between our "PD-1 KO cells" and "control cells" need further study. We sought to use the approach in the following research by changing the model into PD-1 high expression tumor infiltrating lymphocytes in solid tumors or tumor associated lymphocytes in ascites or pleural effusion of cancer patients, as they expose intensively to tumor antigens and acquired adaptive immune resistance that mediates resistance to immunotherapies 37 .
In conclusion, we established an sgRNA:Cas9-basd effective gene disruption method for the highly efficient disruption of PD-1 on primary human T cells. This easy handling technique has great potential to achieve nice effect. And this electroporation mediated approach provides an alternative to labor-intensive and time-consuming viral-mediated gene transfer methods. Therefore, our gene editing method might be suited for both research and clinical applications and we are confident that it will be beneficial to cancer adoptive cell transfer treatments using tumor-specific lymphocytes in the near future.

Materials and Methods
Ethics statement. All the experimental methods were carried out in accordance with the approved guidelines. The blood collection procedure was carried out in accordance with the guidelines verified and approved by the Ethics Committee of Drum Tower Hospital. All donors signed an informed consent for scientific research statement.
Plasmid expression vectors. The Cas9 expression construct pST1374-Cas9-N-NLS-Flag-linker (Addgene 44758) was modified by adding an EGFP sequence to its C terminal. Oligos (Supplementary Table 1) for generation of sgRNA expression plasmids were annealed and cloned into the BsaI sites of pGL3-U6 sgRNA-PGK-Puro vector. The pGL3-dual U6-sgRNAs-PGK-Puro vector was produced based on pGL3-U6 sgRNA-PGK-Puro vector (Addgene 51133), which was modified by removal the small fragment between two BsaI sites and replaced with a ccdb suicide gene. The BsaI sites were also substituted for Esp3I sites. Taken the origin U6 promoter and sgRNA scaffold on pGL3-U6 sgRNA-PGK-Puro as the 1 st sgRNA promoter and the 2 nd sgRNA structure, primers containing pairs of 20 bp targeted sequences (Supplementary Table 2) were used to PCR-amplify the 1 st sgRNA scaffold and the 2 nd U6 promoter from a pUC57-sgRNA-U6 plasmid. The PCR products were then digested with Esp3I and the ccdb gene was replaced, forming the tandem dual sgRNA expression vectors.
Preparation of primary human PBMCs. Apheresis specimens were collected from stage III/IV cancer patients or healthy donors. PBMCs were isolated by centrifugation on a Ficoll density gradient and suspended in AIM-V medium (Gibico, USA). Cells were frozen in 90% FBS serum (Gibico, USA), and 10% dimethyl sulfoxide (Sigma, USA). All PBMCs were used for experiments or stored in a secure liquid nitrogen freezer until use.
T cell activation and electroporation. PBMC were cultured by adherence for 1 ~ 2 h and non-adherent cells were moved and suspended in AIM-V medium supplemented with 1000 U/ml IFN-γ on day 1 and 50ng/ml OKT-3 (eBioscienc, USA) and 300 U/ml of human recombinant IL-2 (eBioscienc, USA) on day 2 for 2 ~ 3 d. Cells were transfected with the intended plasmids by Nucleo-fector 2B (Lonza, Germany) using the Amaxa Human T cells Nucleofector Kit, VPA-1002 (Lonza, Germany). 5 ~ 10 × 10 6 cells were washed twice with DPBS by centrifuging at 800 rpm for 5 m and resuspended in 100 μ l transfection buffer and then transferred into the electroporation cuvette. Program T-007 was selected for both high transfection and high efficiency. After electroporation, cells were resuspended in 500 μ l pre-warmed AIM-V medium containing 10% FBS and transferred into 6-well cell plate and incubated at 37 °C in 5% CO 2 . The transfection efficiency was evaluated by the fluorescent counts 24 h after electroporation. Cells culture medium was half replaced by fresh complete medium containing IL-2 (100 ~ 300 U/ml) every 2 ~ 3 d. Cell viability was evaluated by flow cytometry 24 h after electroporation. T7EN1 cleavage assay and sequencing. Cells were harvested and digested with 100 μ g/ml Proteinase K in lysis buffer (10 μ M Tris-HCl, 0.4 M NaCl, 2 μ M EDTA and 1% SDS). Genomic DNA was extracted by phenol-chloroform and alcohol precipitation. The T7EN cleavage assay was performed as follows: briefly, targeted regions of PD1 were PCR-amplified from genomic DNA using rTaq (Takara, DR001BM) and the products were purified with a PCR cleanup kit (Axygen, APPCR-50). Purified PCR product was denatured and re-annealed in NEBuffer 2 (NEB) using a thermocycler (ABI, Veriti9902). Hybridized PCR products were digested with T7EN1 (NEB, M0302L) for 30 m and separated by 2% agarose gel. Primers for PCR are listed in Supplementary Table  3. The purified PCR products were ligated with pMD19T vector (Takara, 6013) using DNA ligation Kit Ver. 2.1 (Takara, 6022). Ligation products were used for transformation and about 20 ~ 30 colonies per kind are sequenced by using universal primer M13F.
In vitro generation of autologous DC. DCs were generated from monocytes enriched by adherence for 1 ~ 2 h, and cultured in AIM-V medium containing 10% FBS together with human GM-CSF (500 U/ml, Peprotech) and IL-4 (500 U/ml, Peprotech) to obtain immature DCs. To obtain mature DCs (mDCs), fresh complete medium containing TNF-α (500 U/ml, Peprotech), IFN-α (500 U/ml, Peprotech) and PGE2 (50 ng/ml, Peprotech) was added to the culture on day 5. The culture was continued for an additional 2 d. On day 7, all DCs were harvested to be frozen or used for experiments. As described previously, these DCs possess the ability to present peptide antigen and express CD80, CD86, HLA-DR and CD11c.

In vitro expansion of PD-1 KO T cells. Control T cells or PD-1 KO T-cells post transfection were cultured
in AIM-V medium supplemented with IL-2 (300 U/ml, Peprotech) and half replaced by fresh complete medium containing IL-2 every 2 ~ 3 d until analyzes. For antigen stimulated T cells. Mature DCs were pulsed by peptide (25 μ g/mL) for 2 ~ 3 h at 37°C, washed with pre-warmed PBS and then incubated with control T cells or PD-1 KO T-cells at a ratio of 1:10 in complete AIM-V medium supplemented with IL-2 (100 U/ml, Peprotech) in 6-well-plates (5 × 10 6 cells/well) on day 7 post electroporation. IL-7 and IL-15 (5 ng/mL, Peprotech) were added with fresh medium. For re-stimulation, autologous DCs were pulsed with peptide (25 μ g/mL) for 2 h and added to the cultured cells for another 7 d. Fresh complete medium was added containing cytokines every 2 to 3 d until use for experiments.
ELISPOT assay. IFN-γ ELISPOT kit (Dakewei, China) was used to determine the frequency of cytokine-expressing T cells after overnight activation with peptides. Briefly, T cells (10 5 per well) and peptides (50 μ g/ml) were added to duplicate wells and DCs were added at ratio (DC:T) of 1:5 ~ 1:10 for 18 ~ 20 h. The plates were washed before the addition of the diluted detection antibody (1:100 dilution) and then incubated for 1 h in 37 °C. After washing the plates, streptavidin-AP (1:100 dilution) was added and incubated at 37 °C for another 1 h. AEC solution mix was then added to each well, and the plates were left in the dark for about 15 ~ 25 m at room temperature before deionized water was added to stop development. Plates were scanned by Elispot CTL Reader (Cell Technology Inc, Columbia, MD) and the results were analyzed with Elispot software (AID, Strassberg, Germany). Cytotoxicity assay. Transduced T cells were tested for lytic activities by CFSE/PI labeling cytotoxicity assay.
Target tumor cells were labeled with 4 μ M CFSE (Carboxyfluorescein succinimidyl ester) (Invitrogen) for 10 m at 37 °C in PBS. Labeling was stopped by adding 10 fold volume of PBS and extensively washed in PBS before seeding into the 24-well plates. CFSE-labeled cells were then incubated with T cells by different effector to target ratio for 6 ~ 16 h. Propidium iodide (PI) (Sigma) was added to determine the ratio of cell death. Samples were analyzed by flow cytometry. Statistical analysis. Graphpad Prism 5.0 (Graphpad software, San Diego, CA) was used for all statistical analysis. The mean ± S.E.M. was determined for each treatment group in the individual experiments. And the one-tailed Student t-test was used to determine the significances between treatment and control group. P-values < 0.05 were significant.