研究背景：嵌合抗原受体（chimeric antigen receptor，CAR）T细胞疗法是肿瘤免疫治疗的代表性疗法之一，通过基因编辑使T细胞表达特异性识别肿瘤抗原的受体，从而高效靶向并清除肿瘤细胞。该疗法在治疗血液系统恶性肿瘤方面取得了重大突破。目前，CAR T细胞疗法已获批用于治疗复发难治性大B细胞淋巴瘤（large B-cell lymphoma，LBCL）、B细胞急性淋巴细胞白血病（B-cell acute lymphoblastic leukemia，B-ALL）和多发性骨髓瘤（multiple myeloma，MM），其完全缓解率分别为40%-66%、56%-81%和67%-79%，显著优于标准疗法（高剂量化疗联合自体干细胞移植）。然而，CAR T治疗后长期缓解率仍低于50%，有待进一步提高。

研究目的：CAR T细胞的激活信号受到靶抗原密度和CAR分子表达水平的双重调控，任何一方的低表达都会通过降低CAR信号强度来限制其抗肿瘤效能。目前，对于抗原密度降低引起的疾病复发研究较多。CAR T治疗B细胞恶性肿瘤后，靶抗原表达下调引发的抗原逃逸是耐药和疾病复发的常见机制；此外，实体瘤中抗原表达的异质性较大，低抗原逃逸成为CAR T治疗实体瘤的主要障碍之一。除了靶抗原表达下调，研究表明，CAR T细胞在与靶抗原接触后T细胞表面的CAR分子也迅速下调。然而，目前对CAR分子表达水平如何影响CAR T细胞命运及抗肿瘤能力的理解仍不充分，相关调控机制尚需进一步研究。为此，本研究深入分析接受CAR T治疗患者的外周血样本，并结合具有完整免疫系统的抗人CD19 CAR转基因小鼠及相应的血液和实体肿瘤治疗模型，系统探讨了CAR信号强度对CAR T细胞的调控机制，揭示了逆转CAR和抗原低表达引起疾病复发的潜在靶点，为减少肿瘤细胞免疫逃逸提供了重要指导意义。

研究方法：（1）联合CAR T治疗的患者样本及小鼠体内外模型，探究CAR表达水平如何影响CAR T细胞的激活程度，以及与治疗效果之间的关系。（2）构建CAR激活程度不足的细胞和小鼠模型。首先，构建不同CAR表达水平的抗人CD19转基因小鼠，检测CAR表达对CAR下游信号强度的影响。其次，构建不同人CD19抗原表达水平的肿瘤细胞系。检测抗原表达如何影响CAR T细胞激活程度。（3）CAR和抗原低表达导致的CAR信号强度不足如何影响CAR T细胞的体外功能和体内疗效。（4）探究CAR信号强度对LBCL患者CD8 T细胞的调控。随后，在血液和实体荷瘤小鼠中进行low、medium、high CAR T细胞共转移实验，以研究CAR信号强度调控CD8 T细胞的细胞和分子机制。对共转移实验中的三种CAR T细胞进行转录组测序，从基因表达层面探究CAR信号强度调控CAR T细胞命运的机制。（5）首先，在体外敲除Regnase-1对low CD8 CAR T细胞CAR下游信号传导事件及细胞功能的影响。其次，进行体内治疗实验以检测Regnase-1缺失对low CD8 CAR T细胞体内疗效的影响。再次，将WT和Regnase-1敲除的low CD8 CAR T细胞共转移至实体和血液瘤荷瘤小鼠中，分析肿瘤浸润CAR T细胞以探究Regnase-1如何调控CAR信号强度不足的T细胞。最后，在体内外探究敲除Regnase-1是否能够减轻抗原低表达引起的CAR激活程度不足和CAR T细胞功能障碍。

研究结果：（1）在LBCL患者中，CD8 CAR T细胞输注后7天即检测到CAR表达水平显著下降，并维持低水平表达。我们发现，在输注产品及输注后CAR T细胞扩增阶段，完全缓解(Complete response, CR)/ 部分缓解(Partial response, PR)患者的CAR表达水平显著高于疾病进展(Progressive disease，PD)患者。CD8 CAR T治疗的B-ALL-hCD19荷瘤小鼠后我们也检测到相似的现象。随后构建体外抗原刺激诱导的CAR分子下调模型，发现CAR下调后T细胞激活程度降低，效应功能减弱。（2）我们成功构建了CAR表达水平不同的三种转基因小鼠模型，和不同抗原表达水平的肿瘤细胞系。发现CAR低表达影响CAR下游信号分子活化强度及CAR T细胞激活程度，抗原低表达也影响CAR T细胞的激活程度。（3）首先，对于CAR低表达的T细胞：low CD8 CAR T细胞的体外激活程度和效应功能远不如medium和high CD8 CAR T细胞。在小鼠模型中，low CD8 CAR T细胞既无法控制肿瘤负荷，也未能延长小鼠的生存时间；medium和high CD8 CAR T细胞显著抑制了肿瘤生长，并延长了小鼠的生存期。其次，对于抗原低表达的肿瘤细胞：medium CD8 CAR T细胞在体外无法有效杀伤肿瘤细胞，在体内治疗效果显著减弱或无治疗效果。（4）在患者体内，高水平的CAR表达显著促进了CD8 T细胞的增殖，并推动了效应记忆T细胞的分化，介导更强的抗肿瘤功能。在小鼠模型中，输注后high CD8 CAR T细胞在肿瘤中的浸润数量、扩增能力、效应分化程度和细胞因子分泌水平均为最高，medium CAR T细胞次之，而low CAR T细胞表现最弱。在实体瘤模型中，high CD8 CAR T细胞耗竭程度最高，medium CAR T次之；与之不同的是，在血液瘤模型中，medium CAR T细胞的耗竭程度最低，耗竭前体细胞比例最高。转录组测序结果显示，low CD8 CAR T效应分化程度低，而medium和high CD8 CAR T细胞则高表达效应功能相关的基因和通路。（5）首先，对于CAR低表达的T细胞：敲除Regnase-1显著增强low CD8 CAR T细胞CAR下游信号强度。Regnase-1缺失的low CD8 CAR T 细胞的体内疗效与medium CAR T细胞相似。敲除Regnase-1后，low CD8 CAR T细胞在实体肿瘤和血液肿瘤中的浸润分别增强了约10-30倍和100-600倍左右，其激活程度、效应分化和分泌细胞因子水平也显著升高。其次，敲除Regnase-1显著促进medium CD8 CAR T细胞对抗原低表达肿瘤细胞的体外杀伤，提升medium CD8 CAR T细胞的激活程度。在抗原低表达的荷瘤小鼠体内，敲除Regnase-1后CD8 CAR T细胞浸润程度增加了60-100倍左右。

研究结论：在CAR结构不变的情况下，CAR表达水平和肿瘤抗原密度是决定CAR T细胞激活强度的两个关键因素。当CAR表达水平或抗原密度过低时，CAR分子下游信号强度不足，导致T细胞无法有效激活，进而影响CAR T细胞的功能和治疗疗效。敲除Regnase-1显著降低CAR T细胞激活所需的CAR和抗原表达阈值，有望减少靶抗原降低或CAR表达水平下降引起的免疫逃逸，逆转CAR信号强度不足引起的治疗抵抗。

Background: Chimeric antigen receptor (CAR) T-cell therapy significantly advances cancer immunotherapy. This approach involves the genetic modification of T cells to express receptors that target specific tumor antigens, facilitating T cell recognition and subsequent destruction of the tumor cells. Currently approved CAR-T cell therapies have achieved complete response rates of 40%-66%, 56%-81%, and 67%-79% for relapsed/refractory diffuse large B-cell lymphoma, B-cell acute lymphoblastic leukemia, and multiple myeloma, respectively. These response rates exceed traditional standard therapy (high-dose chemotherapy combined with autologous stem cell transplantation). However, the sustained remission rate of B-cell malignancies following commercial CAR-T therapy is less than 50%.

Objectives: CAR signaling strength is regulated by both antigen density and CAR molecule expression levels, and low expression of either factor can reduce signaling strength, thereby limiting the anti-tumor efficacy of CAR T cells. Currently, much research focuses on disease relapse caused by reduced antigen density. Following CAR T therapy for B-cell malignancies, antigen escape due to downregulation of target antigen expression is a common mechanism leading to treatment resistance and relapse. In solid tumors, the heterogeneity of antigen expression is even greater, and low-antigen escape has become one of the primary barriers to the success of CAR T therapy in treating solid tumors. Growing evidence indicates that CARs experience swift downmodulation following their interaction with tumor antigens. However, how CAR expression levels influence the fate and anti-tumor capabilities of CAR T cells remains inadequately explored, and the regulatory mechanisms are not fully understood. To address this, a comprehensive analysis of samples from patients undergoing CAR T therapy was conducted, alongside the use of anti-human CD19 CAR transgenic mice with intact immune systems and corresponding hematological and solid tumor models. This research explores the regulatory mechanisms of CAR signal strength on CAR T cell functionality, identifying potential targets to counteract disease relapse caused by inefficient CAR signal strength, and provides critical insights into reducing immune escape by tumor cells.

Methods: (1) Using peripheral blood samples from CAR T-treated patients and both in vitro and in vivo mouse models, this study investigates how CAR expression levels influence the activation of CAR T cells and their correlation with patient prognosis. (2) We established insufficient CAR activation signaling models. First, transgenic mice expressing different levels of anti-human CD19 CAR were generated to assess how CAR expression levels influenced downstream signaling strength. Second, tumor cell lines with varying human CD19 antigen expression levels were constructed to examine how antigen expression influences the degree of CAR T cell activation. (3) The effect of insufficient CAR signaling strength, caused by reduced CAR and antigen expression, on the in vitro performance and in vivo therapeutic potency of CAR T cells was evaluated. (4) We investigated how CAR signaling strength regulated CD8 T cells in LBCL patients. Subsequently, co-transfer experiments with low, medium, and high CAR T cells are performed in murine models of hematologic and solid tumors to investigate the cellular and molecular mechanisms through which CAR signaling intensity modulates CD8 T cell behavior. Then, we conducted transcriptomic sequencing with three types of CAR T cells from the co-transfer experiments to investigate the mechanisms by which CAR activation strength regulates the fate of CAR T cells. (5) First, the effects of Regnase-1 deletion on CAR downstream signaling events and cellular functions in low CD8 CAR T cells were investigated. Second, we used in vivo therapeutic experiments to assess the impact of Regnase-1 deficiency on the efficacy of low CD8 CAR T cells. Third, both wild-type and Regnase-1-deficient low CD8 CAR T cells were co-transferred into tumor-bearing mice. Then we analyzed tumor-infiltrating CAR T cells, investigating how Regnase-1 regulates T cells with insufficient CAR signaling. Finally, the effects of Regnase-1 on T cells with suboptimal CAR activation due to low antigen expression were examined in both in vitro and in vivo models.

Results: (1) In patients with LBCL, a significant reduction in CAR expression levels was observed as early as 7 days post infusion, and they remained low thereafter. Our findings revealed that both in the infusion product and during the expansion phase post-infusion, CAR expression levels in patients who achieved complete or partial remission (CR/PR) were significantly higher than those with progressive disease (PD). A similar trend was observed in B-ALL-hCD19 tumor-bearing mice treated with CD8 CAR T cells. Subsequently, an in vitro model of antigen stimulation-induced CAR downregulation was established, demonstrating that after CAR downregulation, T cell activation diminished and effector functions weakened. (2) We successfully established three transgenic mouse models with varying levels of CAR expression, along with tumor cell lines exhibiting different levels of antigen expression. Our findings demonstrated that low CAR expression impaired the activation of downstream signaling molecules, while low antigen expression lowered the activation level of CAR T cells. (3) First, for CAR low-expressing T cells: low CD8 and CD4 CAR T cells exhibited significantly lower activation and effector functions. low CD8 CAR T cells were unable to control tumor burden or extend the survival of the mice. In contrast, medium and high CD8 CAR T cells effectively suppressed tumor growth and prolonged survival. Second, for tumor cells with low antigen expression: medium CD8 CAR T cells were ineffective in killing tumor cells in vitro, and their therapeutic efficacy in vivo was either significantly diminished or absent. (4) In patients, high levels of CAR expression significantly enhanced CD8 T cell proliferation and promoted the differentiation into effector memory T cells, leading to stronger antitumor activity. In mouse models, high CD8 CAR T cells exhibited the greatest tumor infiltration, expansion capacity, effector differentiation, and cytokine secretion, followed by medium CAR T cells, while low CAR T cells demonstrated the weakest performance. Transcriptomic sequencing results revealed that low CD8 CAR T cells displayed limited effector differentiation, while medium and high CD8 CAR T cells showed elevated expression of genes and pathways associated with effector functions. (5) First, the deficiency of Regnase-1 significantly enhanced downstream CAR signaling strength in low CD8 CAR T cells. The in vivo efficacy of Regnase-1-deficient low CD8 CAR T cells was comparable to that of medium CAR T cells. Regnase-1 deletion significantly improved the infiltration, activation, effector differentiation, and cytokine secretion levels of low CD8 CAR T cells. Second, Regnase-1 deletion markedly enhanced the in vitro cytotoxicity of CAR T cells against low-antigen-density tumor cells and elevated the activation level of CAR T cells. In antigen low-expressing tumor-bearing mice, Regnase-1-deficient CAR T cells were present at approximately 60-100 times the frequency of control T cells.

Conclusion：The level of CAR expression and density of tumor antigens are two determinants influencing the strength of CAR signaling. When either CAR expression or antigen density is too low, CAR signal strength is insufficient, leading to inadequate T-cell activation. This, in turn, compromises the function and efficacy of CAR T cells. Knocking out Regnase-1 significantly lowers the threshold of CAR and antigen expression required for CAR T cell activation. This approach holds promise for mitigating immune escape caused by downregulation of the target antigen or CAR expression levels and for overcoming treatment resistance due to insufficient CAR signaling.