胃癌是一个全球性的健康问题。全世界每年新发胃癌人数约100万，死亡患者近70万，在癌症相关死亡中排名第三，且年轻人的胃癌发病率在逐年上升。由于早期缺乏明确的临床症状，大多数患者在晚期诊断，Ⅲ期后占比65%。临床上现有的针对胃癌的疗法包括一些获批的针对肿瘤靶点的靶向药物，但是治疗效果有限，预后不佳。

Claudin18.2（CLDN18.2）通常存在于正常胃黏膜细胞的紧密连接中，发挥维持胃黏膜屏障的功能。研究表明，当组织恶性转化时，癌变细胞极性紊乱导致CLDN18.2表位暴露，在胃癌和胃转移瘤中持续表达，可作为肿瘤免疫治疗的一个靶点。目前靶向CLDN18.2抗体药物的临床试验包括单克隆抗体、双特异性抗体和抗体偶联药物。靶向CLDN18.2的嵌合抗原受体T细胞（Chimeric Antigen Receptor T-Cell，CAR-T）也有多项临床试验，均展现出较好的治疗效果，表明CLDN18.2有望成为特异性针对胃癌的有效靶点。但靶向CLDN18.2的CAR-T细胞也存在一些缺陷，比如所有患者均出现了G3及以上的血液学毒性，94.6%的患者发生1级或2级细胞因子释放综合征(Cytokine release syndrome，CRS)。此外，CAR-T细胞对实体瘤的浸润性差，且仅限于自体治疗，无法用于开发同种异体的通用型产品，这些都大大制约了CAR-T细胞的临床应用和疗效。

γδ T细胞是一类表达γδ T细胞受体（γδ T cell receptor, γδ TCR）的固有免疫细胞，因其独特的优势，逐渐成为肿瘤过继细胞免疫治疗的研究热点。成熟的γδ T细胞优势定植于皮肤、呼吸道、消化道、生殖道黏膜等部位，对实体肿瘤具有更好的浸润性。γδ T细胞以主要组织相容性复合体（major histocompatibility complex，MHC）非限制性的方式直接识别肿瘤细胞，在体外对大多数实体肿瘤细胞系具有强大的细胞毒活性。此外，已发表的临床试验结果表明同种异体的γδ T细胞在体内安全性良好，没有出现血液毒性和CRS，且能在体内发挥较好的抗肿瘤疗效。γδ T细胞的这些特点正好弥补了CAR-T细胞的不足。

基于以上研究背景，本研究拟以原代人外周血γδ T细胞为受体细胞，以慢病毒感染的方式，构建一种新型的、通用型的靶向CLDN18.2的CAR-γδ T细胞，并在体内外验证其抗肿瘤活性。

首先，我们设计并成功构建了CAR-CLDN18.2慢病毒载体，将其瞬时转染至HEK293T细胞后，通过Western blot验证了EGFP和CD3ζ的表达，证明该慢病毒载体可以用于后续使用。将CAR-CLDN18.2慢病毒载体与慢病毒包装质粒psPAX2、pMD2.G同时转染至HEK293T细胞，48 h后收集细胞上清并通过低温超滤浓缩得到慢病毒，流式细胞术检测病毒滴度为4.90×10⁸ TU/mL。分离健康人外周血单个核细胞（Peripheral blood mononuclear cell,PBMC)，分别扩增γδ T细胞和αβ T细胞。将浓缩的慢病毒感染原代T细胞，第8天能在EVOS荧光显微镜下观察到较强的绿色荧光信号，该信号主要集中于T细胞克隆团中，流式检测CAR-CLDN18.2慢病毒感染原代γδ T细胞和αβ T细胞的效率分别为31.76 ± 4.122%和44.13 ± 4.436%，两组间没有统计学差异。在细胞扩增的过程中，CAR分子的表达水平整体上趋于平稳。通过台盼蓝染色检测慢病毒感染前后γδ T细胞和αβ T细胞的活性，结果表明CAR-CLDN18.2感染后第4天T细胞的存活率会有明显下降，但在之后会逐渐恢复正常；而Vector感染后对细胞的存活率影响较小。多种实体肿瘤细胞系流式检测结果显示，CLDN18.2只特异性的表达在胃癌细胞系SNU-601膜表面，阳性率为58.6%。通过流式分选得到CLDN18.2的表达率90%以上的细胞，记为SNU-601-CLDN18.2+并将其作为阳性细胞，将不表达CLDN18.2的胃癌细胞AGS作为阴性细胞用于后续体外毒活性的检测。结果表明，与经典的CAR-αβ T细胞相比，CAR-CLDN18.2-γδ T细胞通过分泌高水平的颗粒酶B（Granzyme-B，GZMB），穿孔素1（Perforin-1，PRF1）和干扰素γ（Interferon γ，IFN-γ），对CLDN18.2阳性的胃癌细胞表现出更强大的细胞毒活性。最后，构建皮下SNU-601-CLDN18.2+荷瘤的NSG小鼠模型，评价CAR-CLDN18.2-γδ T细胞的体内抗肿瘤疗效。结果显示，CAR-CLDN18.2-γδ T细胞能显著抑制CLDN18.2阳性肿瘤的生长，并延长小鼠的存活时间，与CAR-CLDN18.2-αβ T细胞的治疗效果无明显差异。

综上，我们用慢病毒感染原代T细胞，构建了靶向CLDN18.2的新型CAR-γδ T细胞。体内外实验结果表明，CAR-γδ T细胞通过分泌高水平的Perforin-1, Granzyme-B和IFN-γ，对CLDN18.2阳性实体肿瘤表现出很好的细胞毒活性。尽管体内实验中，CAR-CLDN18.2-αβ T和CAR-CLDN18.2-γδ T细胞治疗组之间的肿瘤体积变化无明显差异，但结合其对于实体肿瘤的浸润性、安全性以及能用于同种异体治疗等优势，该新型CAR-CLDN18.2-γδ T细胞有望成为一种针对CLDN18.2阳性实体肿瘤的同种异体通用型细胞治疗新策略。

Gastric cancer is a global health problem. Each year, there are approximately one million new cases of gastric cancer worldwide, and nearly 700,000 deaths, ranking third among cancer-related deaths. Moreover, the incidence of gastric cancer among young people is increasing year by year. Due to the lack of clear clinical symptoms in the early stage, most patients are diagnosed in the late stage, accounting for 65% after stage III. The existing clinical therapies for gastric cancer include some approved targeted drugs targeting tumor targets, but the therapeutic effect is limited and the prognosis is poor.

Claudin18.2 (CLDN18.2) is usually present in the tight junctions of normal gastric mucosal cells and plays a role in maintaining the gastric mucosal barrier. Studies have shown that when tissues undergo malignant transformation, the polarity disorder of cancerous cells leads to the exposure of the CLDN18.2 epitope, which is continuously expressed in gastric cancer and gastric metastases and can serve as a target for tumor immunotherapy. At present, clinical trials of drugs targeting CLDN18.2 antibodies include monoclonal antibodies, bispecific antibodies and antibody-drug conjugates. Chimeric Antigen Receptor T-cells (CAR-T) targeting CLDN18.2 have also undergone multiple clinical trials, all of which have shown good therapeutic effects, indicating that CLDN18.2 is expected to become an effective target specifically for gastric cancer. However, CAR-T cells targeting CLDN18.2 also have some problems. For instance, all patients presented with hematological toxicity of G3 or above, and 94.6% of the patients developed Cytokine release syndrome (CRS) of grade 1 or 2. Furthermore, CAR-T cells have poor invitability to solid tumors and are limited to autologous treatment, and cannot be used to develop allogeneic universal products. All these have greatly restricted the clinical application and efficacy of CAR-T cells.

γδ T cells are a type of innate immune cells expressing γδ T cell receptor (γδ TCR). Due to their unique advantages, they have gradually become a research hotspot in adoptive cell immunotherapy for tumors. Mature γδ T cells preferentially colonize in areas such as the skin, respiratory tract, digestive tract, and mucous membranes of the reproductive tract, and have better invasibility towards solid tumors. γδ T cells directly recognize tumor cells in a non-restrictive manner of the major histocompatibility complex (MHC) and have strong cytotoxic activity against most solid tumor cell lines in vitro. Furthermore, the published clinical trial results show that allogeneic γδ T cells have good safety in vivo, without hematotoxicity or CRS, and can exert good anti-tumor efficacy in vivo. These characteristics of γδ T cells precisely make up for the deficiencies of CAR-T cells.

Based on the above research background, this study intends to use primary human peripheral blood γδ T cells as recipient cells and construct a novel and universal CAR-γδ T cell targeting CLDN18.2 through lentiviral infection, and verify its anti-tumor activity in vivo and in vitro.

Firstly, we designed and successfully constructed the CAR-CLDN18.2 lentiviral vector. After transient transfection of it into HEK293T cells, the expressions of EGFP and CD3ζ were verified by Western blot, demonstrating that this lentiviral vector could be used for subsequent applications. The CAR-CLDN18.2 lentiviral plasmid, lentiviral packaging plasmids psPAX2 and pMD2.G were simultaneously transfected into HEK293T cells. After 48 hours, the cell supernatants were collected and concentrated by low-temperature ultrafiltration to obtain lentiviruses. The virus titer detected by flow cytometry was 4.90×10⁸ TU/mL. Peripheral blood mononuclear cells (PBMC) of healthy individuals were isolated and γδ T cells and αβ T cells were expanded respectively. The concentrated lentivirus infected the primary T cells. On the 8th day, a strong green fluorescence signal could be observed under the EVOS fluorescence microscope, and this signal was mainly concentrated in the T cell clone mass. The efficiency of CAR-CLDN18.2 lentivirus in infecting primary γδ T cells and αβ T cells detected by flow cytometry was 31.76 ± 4.122% and 44.13 ± 4.436%, respectively. There was no statistical difference between the two groups. During the process of cell expansion, the expression level of CAR molecules tends to be stable overall. The activities of γδ T cells and αβ T cells before and after lentivirus infection were detected by trypan blue staining. The results showed that the survival rate of T cells would significantly decrease on the 4th day after CAR-CLDN18.2 infection, but would gradually return to normal afterwards. However, Vector infection has a relatively small impact on the survival rate of cells. The flow cytometry detection results of multiple solid tumor cell lines showed that CLDN18.2 was specifically expressed only on the membrane surface of the gastric cancer cell line SNU-601, with a positive rate of 58.6%. Cells with an expression rate of CLDN18.2 of more than 90% were obtained through flow cytometry sorting and were recorded as SNU-601-CLDN18.2+ and used as positive cells. Gastric cancer cells AGS without CLDN18.2 expression were used as negative cells for the subsequent in vitro toxicity activity detection. The results indicated that, compared with the classic CAR-αβ T cells, CAR-CLDN18.2-γδ T cells secrete high levels of Granzyme-B (GZMB), Perforin-1 (PRF1), and Interferon γ (IFN-γ), It shows more powerful cytotoxic activity against gastric cancer cells positive for CLDN18.2. Finally, a subcutaneous NSG mouse model with tumor bearing of SNU-601-CLDN18.2+ was constructed to evaluate the in vivo anti-tumor efficacy of CAR-CLDN18.2-γδ T cells. The results showed that CAR-CLDN18.2-γδ T cells could significantly inhibit the growth of CLDN18.2-positive tumors and prolong the survival time of mice. There was no significant difference in the therapeutic effect compared with CAR-CLDN18.2-αβ T cells.

In conclusion, we infected primary T cells with lentiviruses and constructed a novel CAR-γδ T cell targeting CLDN18.2. The results of in vivo and in vitro experiments indicated that CAR-γδ T cells exhibited excellent cytotoxic activity against CLDN18.2-positive solid tumors by secreting high levels of Perforin-1, Granzyme-B and IFN-γ. Although there was no significant difference in tumor volume changes between the CAR-CLDN18.2-αβ T and CAR-CLDN18.2-γδ T cell treatment groups in vivo experiments, combined with its advantages such as invitability for solid tumors, safety, and the ability to be used for allogeneic therapy, this novel CAR-CLDN18.2-γδ T cell is expected to become a new allogeneic universal cell therapy strategy for CLDN18.2-positive solid tumors.