Dual-targeting CD133/PD-L1 CAR-T plus αPD-1 overcomes immunosuppressive microenvironment and enhanced by radiation pre-conditioning

Zhuoran Yao; Kai Kang; Pei-Heng Li; Limei Yin; Ruizhan Tong; Linglu Yi; Yonghong Song; Ren Luo; Yijun Wu; Shanghai Liu; Zichong Peng; Xianming Mo; Wenbo Wang; Jianxin Xue; You Lu

doi:10.1016/j.ymthe.2025.12.068

. 2026 Jan 10;34(3):1483–1502. doi: 10.1016/j.ymthe.2025.12.068

Dual-targeting CD133/PD-L1 CAR-T plus αPD-1 overcomes immunosuppressive microenvironment and enhanced by radiation pre-conditioning

Zhuoran Yao ^1,^2,⁸, Kai Kang ^1,^2,⁸, Pei-Heng Li ^3,⁸, Limei Yin ^1,⁴, Ruizhan Tong ^1,², Linglu Yi ², Yonghong Song ⁵, Ren Luo ^1,², Yijun Wu ^1,², Shanghai Liu ^1,², Zichong Peng ^1,², Xianming Mo ⁶, Wenbo Wang ⁷, Jianxin Xue ^1,^2,^∗, You Lu ^1,^2,^∗∗

PMCID: PMC12974209 PMID: 41521560

Abstract

Innovative chimeric antigen receptor (CAR) T cell designs and combinational approaches are needed for enhancing therapeutic effectiveness in solid tumors. We developed and assessed a novel dual-targeting CAR-T therapy that combines an αPDL1.CD28 chimeric receptor with a second-generation αCD133 CAR to target CD133⁺ tumors. The αPDL1.CD28 structure activated the CD3ζ signaling in cis by clustering with αCD133 CAR via CD28 dimerization. Binding to programmed cell death ligand-1 (PD-L1) through αPD-L1 CAR improved the CD133-targeted cytotoxic function of T cells by enhancing activation signals and countering inhibitory signals. Combination with programmed cell death receptor-1 (PD-1) blockade further disrupted the PD-L1/PD-1 inhibitory signal, achieving prolonged therapeutic efficacy. Moreover, radiation pre-conditioning (10 Gy/1 fraction or 4 Gy/2 fractions) maximized the antitumor effects of CAR-T plus PD-1 blockade, inducing complete tumor regression in mice. Radiation induced a unique tissue-resident memory CAR-T cell phenotype with high CXCR6 and CD103 expression. As the ligand of CD103, E-cadherin expression increased in tumor cells after irradiation, potentially mediating E-cadherin-CD103 interactions between tumor cells and tissue-resident memory T cells. Our study introduces a novel dual-targeting CD133/PD-L1 CAR-T cell and further demonstrates the efficacy and rationale of the triple-combination approach in solid tumors.

Keywords: chimeric antigen T cell receptor, PD-L1/PD-1 pathway, radiotherapy, tissue-resident memory T cells, E-cadherin

Graphical abstract

Yao and colleagues designed dual-targeting CD133/PD-L1 CAR-T cells exploiting CD28-driven co-clustering to share CD3ζ signaling in cis, boosting antitumor activity. Adding PD-1 blockade and radiation pre-conditioning establishes tissue-resident memory-like CAR-T cells and durable control, offering a practical framework for solid tumors.

Introduction

Adoptive cell transfer (ACT) therapies, such as chimeric antigen receptor (CAR)-T cell therapy, represent a breakthrough in treating hematologic malignancies. However, their application in solid tumors poses significant challenges. Our previous study demonstrated the safety and feasibility of ACT in refractory non-small cell lung cancer, but none of the patients achieved a clinical response.¹ While CAR-T cells exhibit superior tumor recognition ability over regular T cell transfer, their efficacy is compromised within the solid tumor microenvironment (TME).²^,³ Selecting target antigens, immunosuppressive microenvironment, and poor persistence are currently the major obstacles hindering the translation of CAR-T therapy in solid cancers.⁴^,⁵

CD133 serves as a cancer stem cell (CSC) marker highly expressed in various solid malignancies, including colorectal cancer (CRC), hepatocellular carcinoma (HCC), and gastric cancer.⁶^,⁷ However, CAR-T designs targeting a single tumor antigen may not be sufficiently effective in solid tumors.⁸ The programmed cell death receptor-1 (PD-1)/programmed cell death ligand-1 (PD-L1) axis plays a pivotal role in immunosuppressive signaling pathways involved in tumor immune evasion.⁹ Recently, chimeric switch receptors (CSRs) targeting this axis have emerged as enhancements to CAR-T cells.¹⁰^,¹¹^,¹²^,¹³ By fusing PD-L1 single-chain variable fragment (scFv) or PD-1 with an intracellular CD28 or 41-BB domain, CSR can reverse the inhibitory signals to positive co-stimulation. Additionally, targeting PD-L1 shows promise in directly eliminating PD-L1⁺ tumor cells. Dual-targeting CAR-T cells have been constructed against PD-L1 and other tumor antigens in several studies, demonstrating enhanced efficacy in treating solid tumors.¹⁴^,¹⁵ Collectively, these findings highlight the significance of optimizing PD-L1/PD-1 signals of CAR-T to overcome the suppressive tumor immune microenvironment.

Insufficient infiltration of CAR-T cells poses a significant limitation to the efficacy of CAR-T cell therapy for solid tumors. Factors such as inadequate chemotaxis, abnormal tumor vasculature, and restricted adhesion of CAR-T cells could limit T cells’ ability to reach and eliminate tumor cells.³ Radiotherapy can enhance immunotherapy by increasing tumor antigenicity, stimulating immune activation, modulating the vasculature, and altering chemokine and cytokine profiles.¹⁶ Numerous studies and clinical trials have demonstrated the synergetic effect of radiation and PD-1/PD-L1 blockade.¹⁷^,¹⁸^,¹⁹^,²⁰^,²¹ Combination therapy involving radiotherapy offers a promising approach to enhance CAR-T cell infiltration. However, the application of CAR-T therapy in treating solid tumors is limited.

Therefore, this study aimed to develop and evaluate a novel dual-targeting CAR-T therapy, combining an αPD-L1.28 CSR with a second-generation anti-CD133 CAR to target CD133⁺ tumors. We accidentally discovered that the αPDL1.CD28 structure induced CD3ζ signaling in cis by forming clusters with CD133 CAR through CD28 dimerization. This represents a form of bystander activation whereby, within the same CAR-T cell, one CAR engages its target and activates the signaling domain of another CAR without direct antigen binding by the latter.²²^,²³ This novel dual-targeting CD133/PD-L1 CAR-T (αPDL1.28/αCD133.28ζ) exhibited enhanced cytotoxicity against CD133⁺ tumor cells, regardless of their basal PD-L1 expression. Moreover, a triple-modality therapy involving dual-targeting CAR-T cells, anti-PD-1, alongside 10 Gy/1 f radiation, maximized antitumor responses and induced complete tumor eradication in mouse models. Our findings could shed light on a multifaceted strategy for overcoming the inhibitory microenvironment, poor infiltration, and limited persistence associated with CAR-T therapy in solid tumors.

Results

Dual CD133/PD-L1 targeting CAR-T shared one single CD3ζ chain to transduce activation signals

We confirmed CD133 expression in CRC and HCC microarrays (Figure S1) and developed second-generation CAR-T cells targeting CD133. Previous studies have demonstrated that co-expressing a CSR targeting PD-L1 (PD1.28) could augment efficacy in solid tumors.¹⁰ We designed a series of bicistronic vectors by adding a PD-L1 CSR comprising a high-affinity anti-PDL1 VHH domain and an intracellular co-stimulatory domain without CD3ζ chain (Figure 1A).¹⁰ Flow cytometry analysis revealed CAR expression in T cells after lentiviral transduction (Figure 1B). Unexpectedly, we found that αPDL1.28/αCD133.28ζ CAR-T cells demonstrated upregulated activation markers and released interferon (IFN)-γ cytokine upon engagement with PD-L1 alone (Figures 1C and 1D). This indicates that the incomplete αPDL1.28 CAR when engaging the antigen can use the CD3ζ expressed in cis provided by another moiety that does not directly recognize the antigen. This CD3ζ-sharing phenomenon was only observed in dual CARs containing identical CD28 hinges, transmembrane, and intracellular domains. Conversely, αPDL1.BB/αCD133.28ζ and αPDL1.28/αCD133.BBζ did not exhibit activation signatures or release higher cytokines upon PD-L1 engagement (Figures 1C and 1D). Furthermore, αPDL1.28/αCD133.28ζ CAR-T cells showed cytotoxic function against CD133^-/PD-L1+ tumor cells, unlike other CSR-modified CAR-T cells (Figure 1E). To explore whether the two CARs in αPDL1.28/αCD133.28ζCAR-T cells clustered in the same immune synapse upon engaging PD-L1 and subsequently triggered an activation signal through the shared CD3ζ chain, we fused green fluorescent protein (GFP) with αPDL1.28 and mCherry with αCD133.28ζ (Figure 1F). We stimulated the cells with PD-L1-Fc fusion protein or PD-L1 fusion protein and CD3 antibody. Using confocal microscopy imaging, we observed the formation of membrane clusters and colocalization between αPDL1.28 and αCD133.28ζ CARs upon CAR cross-linking (Figure 1G). Additionally, the bystander activation through crosslink with CAR-CD28-CD3ζ structure was also observed in αPDL1.28/αDLL3.28ζ CAR-T cells upon PD-L1 antigen engagement (Figures S2A–S2C) along with αCD133.28ζ CAR-T cells clustering with endogenous CD28 receptor upon anti-CD28 stimulation (Figures S2D and S2E).

Given that PD-L1 is detectable on T cells (Figure S3A), we explored whether bicistronic αPD-L1.28/αCD133.28ζ CAR-T cells become activated when cultured alone in the absence of target cells. In the resting state, cytokine secretion by αPD-L1.28/αCD133.28ζ CAR-T cells remained low (Figure S3B). During ex vivo expansion, αPD-L1.28/αCD133.28ζ cells showed slower expansion and higher IFN-γ release than αCD133.28ζ and vector T cells (Figures S3C and S3D). However, we observed no evidence of T cell fratricide, and bystander T cell proliferation remained unaffected (Figure S3E).

Dual CAR-T showed improved cytotoxic function against CD133⁺ tumor cells regardless of basal PD-L1 expression

To evaluate whether the αPDL1.28 engagement enhanced the cytotoxic function of αCD133.28ζ CAR-T cells to CD133⁺ target cells, we constructed tumor cells with diverse basal PD-L1 expressions. The surface expression of both CD133 and PD-L1 was detected by flow cytometry (Figure S4). Then αPDL1.28/αCD133.28ζ, αCD133.28ζ CAR-T cells, or vector T cells were co-cultured with tumor cells at 1:1 E:T ratios for 24 h in a cytokine-free medium (culture medium). As compared to αCD133.28ζ CAR-T cells, αPDL1.28/αCD133.28ζ CAR-T cells exhibited higher levels of activation markers, 4-1BB, CD69, and CD25 (Figure 2A), and the higher release of IFN-γ, tumor necrosis factor (TNF)-α, and IL-2 cytokines (Figure 2B). Accordingly, αPDL1.28/αCD133.28ζ CAR-T cells demonstrated superior antitumor cytotoxicity in a luciferase-based killing assay, regardless of basal PD-L1 expression (Figure 2C). Real-time fluorescence imaging of CAR-T and tumor cells revealed that αPDL1.28/αCD133.28ζ CAR-T cells formed significantly more and larger clonal clusters against tumor cells, with a faster killing effect of GFP⁺ tumor cells (Figures S5A–S5C and Videos S1, S2, and S3). Moreover, we tested the antitumor effects of αPDL1.28/αCD133.28ζ CAR-T cells in a patient-derived CD133⁺ PD-L1^- lung cancer organoid. The dual-targeting CAR-T cells eliminated the tumor cells more prominently, accompanied by higher secretion of IFN-γ and TNF-α (Figures 2D and 2E).

Video S1. Long-term cytotoxic interaction between αPDL1.28/αCD133.28ζ CAR-T cells and tumor cells captured by fluorescence microscopy

Download video file^{(4.8MB, mp4)}

Video S2. Long-term cytotoxic interaction between αCD133.28ζ CAR-T cells and tumor cells captured by fluorescence microscopy

Download video file^{(4.8MB, mp4)}

Video S3. Long-term cytotoxic interaction between vector T cells and tumor cells captured by fluorescence microscopy

Download video file^{(4.8MB, mp4)}

Based on the in vitro data, we assessed the antitumor activity of dual-targeting CAR-T cells in immunodeficient mice bearing tumor xenografts. We subcutaneously implanted NCG mice contralaterally with PD-L1 wild type (WT), CD133-positive tumor cells, or PD-L1-overexpressed CD133-positive tumor cells. Ten days later, the mice were infused with various T cell groups (Figures 2F and S5D). αPDL1.28/αCD133.28ζ CAR-T cells showed a significant delay of tumor growth in both PD-L1 WT and overexpressed models (Figures 2F and S5D). Moreover, tumor tissue treated with αPDL1.28/αCD133.28ζ CAR-T cells contained a higher proportion of T cells than αCD133.28ζ CAR-T and vector T cells (Figures 2G and S5E). However, αPDL1.28/αCD133.28ζ CAR-T cells were insufficient to fully eliminate CD133-positive tumors, especially in the PD-L1-overexpression model.

PD-1 blockade enhanced antitumor immunity of dual-targeting CD133/PD-L1 CAR-T

The in vivo and in vitro assessment of αPDL1.28/αCD133.28ζ CAR-T showed that simultaneous targeting of CD133 and PD-L1 could enhance cytotoxic function against CD133⁺ tumors, even in HCT116, a colorectal cell line with minimal baseline PD-L1 expression. We observed that incubating tumor cells with CAR-T cells resulted in elevated PD-L1 expression (Figure 3A), thereby enhancing antigen recognition via αPDL1.28 and suppressing PD-L1/PD-1 immunosuppressive signals. Following incubation, PD-L1 expression in HCT116 and 231-CD133 was lower in the αPDL1.28/αCD133.28ζ CAR-T group, suggesting that the dual-targeting CAR-T may eliminate CD133⁺ and PD-L1⁺ tumor subclones (Figure 3A). Conversely, the enhanced antitumor activity of αPDL1.28/αCD133.28ζ CAR-T was diminished in PD-L1-knockout tumor models (Figure 3B). In PD-L1-overexpression tumor models, surface PD-L1 levels remained high following incubation, suggesting excessive PD-L1 suppression through the PD-1 receptor on CAR-T cells (Figure 3C). Therefore, we further combined CAR-T cells with tislelizumab, an anti-PD-1 antibody (αPD-1), in multi-round co-culture experiments (Figure 3D). The PD-1 blockade significantly augmented the antitumor immunity of CAR-T cells, particularly in the PD-L1-overexpression tumor model (Figures 3E. When HCT116 and HCT116-PDL1 tumors were implanted contralaterally in NCG mice, a complete response was achieved on both sides (Figures 3F–3H).

Radiotherapy further synergized with CAR-T therapy and triple combination reached maximal tumor control

Intense discussions have centered on implementing CAR-T cell therapy into conventional treatment regimens. Radiotherapy (RT) plays a pivotal role in treating many solid tumors and is considered an adjunct to immunotherapy.¹⁶ Upon irradiation, in vitro assays showed an increase in CD133 and PD-L1 expression on the tumor cell surface, whereas in vivo analyses revealed elevated total protein levels of both markers within tumor tissues, potentially enhancing recognition by dual-targeting αPDL1.28/αCD133.28ζ CAR-T cells (Figure S6). In the Transwell co-culture system, tumor cells were seeded onto the lower compartment, while the upper insert contained immune cells (Figure 4A). Transwell assay results indicated significantly higher migration of CAR-T cells toward irradiated tumor cells than untreated tumor cells (Figure 4A). We further examined the synergistic efficacy of RT and CAR-T cells in vivo under stress conditions, initiating treatment 20 days after tumor inoculation, resulting in a relatively larger tumor burden (approximately 200 mm³). Initially, a single dose of 10 Gy/1 f irradiation was administered, followed by sequential infusion of αPDL1.28/αCD133.28ζ CAR-Ts after 5 days (Figures S7A and S7C). In mice bearing HCT116 or HCT116-PDL1, combination treatment of 10 Gy/f radiotherapy and αPDL1.28/αCD133.28ζ CAR-T cells showed an enhanced tumor control and prolonged survival compared with CAR-T monotherapy alone (Figures S7A–S7D). Tumor bioluminescence (BLI) images of mice treated with radiotherapy alone showed significant distant metastasis as early as 57 days after treatment initiation (Figures S7B and S7D). This underscores the limitations of locoregional therapy and emphasizes the necessity for combinations in treating solid tumors.

Considering the limited antitumor activity in PD-L1-overexpression tumor models, we proceeded to employ tislelizumab to further block the excessive PD-L1 inhibitory signals to αPDL1.28/αCD133.28ζ CAR-T. A single dose of 10 Gy/1 f irradiation was administered 5 days before CAR-T cell infusion, along with weekly injections of a PD-1 antibody (intraperitoneally [i.p.]) (Figure 4B). The triple combination achieved greater tumor control than the combination of dual or single agents. Particularly, 10 Gy/1 f + CAR-T + αPD-1 led to complete regression of the tumor and showed persistent antitumor immune memory (Figures 4C and 4D). The durable tumor regression was maintained over 100 days (Figure S8A). The triple combination was well tolerated, as shown by the absence of off-tumor organ cytotoxicity (Figure S9). Additionally, we further evaluated a de-escalated regimen (4 Gy in two consecutive fractions). The 4 Gy/2 f + CAR-T + αPD-1 group recapitulated the superiority of the triple combination over the corresponding dual- and single-agent groups (Figure S10).

Moreover, we longitudinally collected peripheral blood samples from all groups of mice at days 10, 20, 30, 40, 50, 60, and 79 after treatment initiation. The CAR-T copy number peaks at day 10 in all group, while the triple-therapy group consistently had higher CAR-T cell copy numbers in peripheral blood at all time points compared to the other groups, followed by the CAR-T + αPD-1 groups (Figures 4E and S8B). This indicates that the addition of PD-1 inhibitors is crucial for sustained presence of CAR-T cells in peripheral blood. The serum cytokine level was also measured to monitor potential cytokine release syndrome (CRS). The CAR-T + αPD-1 group consistently exhibited higher IFN-γ levels than the other groups at all time points (Figure 4F). Notably, the pre-administration of radiotherapy lower the peripheral cytokine level, implying that pre-administration of radiation might lower the risk of CRS in treating solid tumors.

Triple-therapy-induced intratumoral CD103⁺ CXCR6⁺ tissue-resident memory CD8⁺ CAR-T cell expansion

We then investigated whether, apart from the cytotoxic effect of irradiation, the curative effects of the triple-combination therapy were linked to the distinct characteristics of CAR-T cells. We analyzed the TME 17 days after treatment initiation using single-cell RNA sequencing (scRNA-seq) and employed the Xenocell algorithm to classify human-derived graft cells (tumors and T cells) in the xenograft tumor model (Figure 5A).²⁴ Overall, 71,681 human-derived cells were grouped into tumor and T cells (Figures 5B and S11A). The CAR-T/tumor ratio was significantly increased in the 10 Gy/1 f + CAR-T (RT + T) group and 10 Gy/1 f + CAR-T + αPD-1 (RT + T + ICI (Immune checkpoint inhibitor)) group (Figures 5B and S11B). Subsequently, 10,502 CAR-T cells were clustered and visualized (Figure 5C). Six subgroups with different transcriptional profiles were identified: CD8⁺CXCR6⁺ T cells, CD8⁺ AREG⁺ T cells, CD8⁺ cycling T cells, CD8⁺ FOXP3⁺ T cells, CD4⁺ CCR7⁺ T cells, and CD4⁺ CXCR6⁺ T cells (Figure 5D). Upon normalizing the proportions of sample sites within cellular components, CD8⁺CXCR6⁺ T cells were found to be predominantly derived from the RT + T and RT + T + ICI groups (Figure 5E). The CD8⁺CXCR6⁺ T cell subgroup was characterized by CXCR6 and ITGAE high expression (CD103), while gene set enrichment analysis (GSEA) exhibited high similarities with tissue-resident T (Trm) cell signatures as reported by Savas et al. (Figure 5F).²⁵ Compared with other T cell subgroups, CD8⁺CXCR6⁺ T cells exhibited a higher Trm score than other T cell subgroups (Figure S11C). Therefore, we define CD8⁺ CXCR6⁺ T cells as CD8⁺ Trm cells.

We conducted a pseudotime analysis to illustrate the dynamic evolution of CAR-T cells within major subsets of CD8⁺CAR-T cells. Two distinct trajectory paths emerged: one transition from cycling CD8⁺ T cells to CXCR6⁺CD8⁺T cells (path 1) and the other from cycling CD8⁺ T cells to AREG⁺CD8⁺T cells (path 2). Path 2 was mainly distributed in CAR-T (T) and CAR-T + αPD-1 (T + ICI) groups (Figures 5G and S11D). The upregulation of classic Trm cell marker genes, such as ITGA1, ZNF683, and CXCR6, occurred along path 1 as inferred from pseudotime (Figure 5G).²⁶ To verify the scRNA-seq findings, we performed flow cytometry and multiple-immunofluorescence (mIF) staining on the tumor tissues. We utilized CXCR6 and the canonical Trm marker CD103 to identify tissue-resident CAR-T cells (Trm). The proportion of CD103⁺ CXCR6⁺CD8⁺Trm cells was higher in the RT + T + ICI group than in the non-irradiated control group (Figure 5H). Additionally, CD8⁺Trm cells exhibited robust infiltration into the TME of the RT + T + ICI and T + ICI groups, as indicated via mIF (Figure 5I).

E-cadherin-CD103 appears as a vital ligand receptor between tumor cells and CD103⁺ CXCR6⁺ CD8⁺ CAR-Trm after irradiation

scRNA-seq and mIF analysis revealed that CD103⁺CXCR6⁺CD8⁺Trm cells were predominantly observed in irradiated groups, suggesting their crucial role following radiotherapy. Subsequently, we examined the effects of radiation on tumor cells using transcriptomic analysis. To standardize the variables, we conducted pairwise comparisons among the six groups (RT vs. CON, RT + T vs. T, and RT + T + ICI vs. T + ICI) to ensure that the sole variable between each pair was the receipt of radiotherapy. Gene Ontology (GO) analysis revealed that the upregulated genes in the radiotherapy group were consistently enriched in the cadherin-binding signaling pathway (Figure 6A). Similar to the GO analysis, the GSEA signature scores for E-cadherin related pathway in the RT, RT + T, and RT + T + ICI groups were also higher than those in the unirradiated controls (Figure S11E). The Venn diagram of the DEGs in the three pairs illustrated that CDH1 was consistently upregulated in the irradiated groups (Figures 6B and S11F). CDH1 encoded the E-cadherin, an adherens junction protein that plays a crucial role in tumorigenesis, invasion, and metastasis.²⁷ Interestingly, E-cadherin serves as a canonical ligand for integrin αEβ7, also known as CD103.²⁸ This interaction between E-cadherin and CD103 helps anchor Trm cells and strengthens the Trm immune response. Therefore, the upregulation of E-cadherin after radiation may facilitate the differentiation and maintenance of CD103⁺ CXCR6⁺CD8⁺ CAR-Trm cells. To validate this hypothesis, we analyzed mIF samples to assess the spatial distribution of E-cadherin⁺ tumor cells and CD103⁺ CXCR6⁺CD8⁺Trm cells. CD103⁺ CXCR6⁺CD8⁺ Trm cells were more proximal to E-cadherin⁺ tumor cells than to other non-Trm CD8⁺T cells (Figure 6C). Furthermore, in vitro validation corroborated our in vivo findings that demonstrate that radiation upregulates surface E-cadherin expression through flow cytometry (Figure 6D). Moreover, tumor cells cultured for 24 h after radiation induced a higher proportion of CD103⁺ CXCR6⁺CD8⁺ Trm cells in the co-culture system (Figure 6E).

Irradiation enhances the E-cadherin-CD103 interaction between tumor cells and CD8+ T cells in clinical samples

To further substantiate our findings, we conducted scRNA-seq on clinical samples from four patients diagnosed with small cell lung cancer (SCLC). Two patients had received first-line therapy without irradiation, while the other two had received first-line therapy that included irradiation at 15 Gy over five fractions. This irradiation regimen corresponds to a biological effective dose (BED) of 19.5 Gy (α/β = 10), comparable to 10 Gy delivered in a single fraction (BED_α/β=10 = 20 Gy) (Figure 7A). After quality control, a total of 33,078 cells were retained and classified into different cell types using typical marker genes (Figure 7B and Figure S11G).

Based on typical marker genes, T/natural killer (NK) cells were further divided into eight subtypes, namely CD8_ITGAE, CD8_GZMK, CD8_FGFBP2, CD4_FOXP3, CD4_ICOS, T_CCR7, T_MKI67, and NK_FCER1G (Figure 7C). Notably, irradiation led to an overall increase in the proportion of each subtype (Figure 7C). CD8_ITGAE was characterized by higher expression levels of ITGAE (coding CD103) and CXCR6, both of which are typical marker genes of Trm (Figure 7C). Based on the results of the differential gene expression (DGE) analysis, GSEA on signatures previously reported by Savas et al. further supported the similarity of CD8_ITGAE to Trm (Figure 7D).²⁵

In the RT group, SCLC cells expressed elevated levels of CDH1 (Figure 7E). Similar trends were also observed across individual samples (Figure 7F). Enrichment analysis exhibited that upregulated genes in SCLC cells of the RT group were significantly enriched in the cadherin-binding pathway, indicating that the influence of irradiation on SCLC cells is not limited to the CDH1 gene alone but rather at the pathway level (Figure 7G). Cell-cell communication analysis revealed that irradiation enhanced the E-cadherin-CD103 interaction between SCLC cells and CD8⁺ T cells, and this promotion was not limited to CD8_ITGAE but was common to all CD8⁺ T cells (Figure 7H). In addition, we extended our analysis to two independent, publicly available datasets, GSE246613 (breast cancer samples after immunotherapy or immunotherapy combined with RT) and PRJNA818695 (paired pre- and post-RT samples).²⁹^,³⁰ We also discovered that the proportion of CD8_CXCR6 T cell subset possessing Trm characteristics (Figures S12D and S12J) was increased after irradiation in breast cancer samples (Figures S12F and S12L). Besides, a consistent post-RT upregulation of epithelial CDH1 expression was observed across samples (Figures S12E and 12K).

Discussion

In this study, we co-expressed an αPDL1.CD28 chimeric receptor with a second-generation αCD133 CAR-T using a bicistronic vector. The αPDL1.CD28 structure activates the CD3ζ signaling in cis by clustering with αCD133 CAR via CD28 dimerization. These dual-targeting CAR-T cells showed enhanced cytotoxicity against CD133⁺ tumor cells in vitro and in vivo, irrespective of their basal PD-L1 expression. αPDL1.28/αCD133.28ζ CAR-T effectively eliminates PD-L1-positive tumor subclones, evading PD-L1 inhibitory signals. Combining it with tislelizumab, an anti-PD-1 antibody, may further block excessive PD-L1 and enhance the antitumor efficacy of αPDL1.28/αCD133.28ζ CAR-T, especially in PD-L1-overexpression tumor models. Given recent advancements in immune-radiotherapy, we evaluated the synergistic effects of radiotherapy and CAR-T cells. αPDL1.28/αCD133.28ζ CAR-T cells were challenged with a larger tumor burden. Pre-administration of 10 Gy/1 f radiation effectively improved tumor control, with the triple combination of 10 Gy/1 f + CAR-T cells + tislelizumab achieving maximal antitumor efficacy. Mechanistically, radiation-induced unique Trm CAR-T phenotype highly expresses levels of CD103 and CXCR6. E-cadherin and CD103 may form a potentially crucial receptor-ligand interaction as an underlying mechanism.

Despite their efficacy and significant potential in hematologic malignancies, most trials testing CAR-T cells in solid tumors have shown minimal clinical activity.³¹^,³²^,³³^,³⁴ Studies have explored the potential mechanism underlying the restricted efficacy of CAR-T cells in solid tumors,³^,⁴^,⁵ emphasizing several challenging factors, including insufficient specific target antigens, poor trafficking, short persistence, loss of effector function, and tumor antigen heterogeneity. In this study, we implemented several strategies to address existing limitations that hold promise for future application of CAR-T therapy in patients with solid tumors.

Emerging evidence suggests that cancer stem cells possess intrinsic mechanisms to resist traditional or novel therapies, driving tumor initiation, recurrence, and metastasis.³⁵^,³⁶ CD133 has been postulated to identify CSC populations in diverse solid tumor types, such as colorectal, hepatocellular, gastric, and lung cancer.⁷^,³⁷^,³⁸ Our tissue microarray analysis confirmed CD133 expression in several malignancy types, suggesting that CD133 is an ideal target for CAR-T therapy. However, canonical CD133-specific second-generation CAR-T cells show only moderate efficacy in clinical settings, demanding a therapeutic upgrade.³³^,³⁹

The immune checkpoint pathway has been recognized as a cornerstone of cancer immunology, and the development of PD-1/PD-L1 antibodies led to the renaissance of immunotherapy. We generated CAR-T cells to target the tumor antigen CD133 and immunosuppressive ligand PD-L1. Dual-targeting CD133/PD-L1 CAR-T cells showed superior antitumor efficacy, even in tumor models with low PD-L1 expression. The advantages of incorporating the PD-L1 CAR structure may be multifactorial. The high-affinity anti-PD-L1 receptor binds to PD-L1 on tumor cells, potentially countering inhibitory signals. Conversely, antigen recognition by PD-L1 through chimeric receptors transmits stronger T cell activation signals, enhancing T cell cytotoxic function. Here, we report the induction of PD-L1 expression following CAR activation by the first tumor antigen, suggesting that targeting PD-L1 is adaptable to various solid tumors regardless of their basal PD-L1 expression. The efficacy of dual targeting of CD133/PD-L1 CAR-T cells was further prolonged by combining them with PD-1 blockade. While targeting the same pathway at two points might appear redundant or overlapping, the anti-PD-L1 CAR on T cells alone is likely insufficient to block all PD-L1 molecules. Therefore, anti-PD-1 blockade is necessary to disrupt the excessive PD-L1/PD-1 inhibitory interaction.

Radiotherapy is acknowledged as an adjunct to immunotherapy, known to enhance the antitumor immune response.¹⁷^,¹⁸ Previous studies have highlighted the synergistic effects of RT and CAR-T cells in various tumor models.⁴⁰^,⁴¹^,⁴² Clinical data also support the feasibility of RT as a bridging strategy prior to CAR-T infusion, contributing to disease control and improved patient condition before leukapheresis.⁴³ In preclinical models, RT has been shown to enhance tumor antigen cross-presentation and facilitate CAR-T infiltration into tumors.⁴⁴ Moreover, low-dose RT can render antigen-negative tumor cells susceptible to CAR-T-mediated killing by inducing TRAIL sensitivity, thereby mitigating antigen escape.⁴⁵ Notably, myeloid cells play a complex role in modulating CAR-T cell function in response to RT, potentially influenced by varying radiation doses and fractionation regimens. FLASH radiotherapy, an ultra-high dose-rate irradiation modality, has been shown to reduce peroxisome proliferator-activated receptor-γ (PPARγ) and arginase 1 expression, thereby inhibiting immunosuppressive macrophage polarization and enhancing CAR-T cell infiltration and activation, while conventional fractionated radiation may exert opposing effects.⁴⁶ Additionally, hypofractionated RT has been found to impair CAR-T cell infiltration and efficacy by inducing the recruitment of myeloid-derived suppressor cells (MDSCs).⁴⁷

Here, we propose a sandwich therapy approach, combining RT + T + ICI to treat solid tumors, especially those with larger volumes. Pre-conditioning of RT reduces the tumor volume, mitigates the risk of CRS, upregulates antigen expression, and induces a Trm phenotype, as evidenced by our single-cell transcriptome analysis. For the first time, we identified a distinct subset of CAR-Trm cells in our triple-combination therapies. Trm cells are pivotal in tumor immune surveillance.²⁶ Studies have indicated that increased CD8⁺ tumor-infiltrating lymphocytes (TILs) with Trm features correlate with positive outcomes after immunotherapy, thus highlighting their significance as reservoirs of antitumor specificities.⁴⁸ Inducing CAR-Trm cells through ex vivo treatment with transforming growth factor β (TGF-β) exhibits enhanced stemness, residency, and antitumor persistence.⁴⁹ In parallel with the ex vivo TGF-β treatment, we successfully validated that pre-administration of RT could also provoke Trm differentiation in the TME. More importantly, a chemokine receptor, CXCR6, has been identified as a specific marker in the RT combination groups (RT + T and RT + T + ICI). CXCR6-armed CAR-T has shown sustained antitumor activity in pancreatic cancer.⁵⁰ Therefore, arming CAR-T cells with chemokine receptors might represent a promising strategy for next-generation CAR-Ts.

The mechanism beyond the synergism of RT and CAR-T remains unknown. Pairwise analysis revealed a consistent upregulation of the cadherin-binding pathway. Previous studies have highlighted the elevation of E-cadherin expression in tumor cells upon irradiation, which has been recognized as a radioresistance marker.⁵¹^,⁵² E-cadherin/integrin αEβ7 (CD103) constitutes a canonical ligand-receptor pair for T cell retention, potentially fostering Trm differentiation.²⁸^,⁵³ The spatial distance between E-cadherin⁺ tumor cells and CD103⁺ CXCR6⁺CD8⁺ Trm cells was shorter than that between non-Trm cells. Therefore, we hypothesize that E-cadherin upregulated by radiation serves as an anchor for CAR-T retention, which induces differentiation of Trm cells and prolongs the antitumor persistence. Therefore, we hypothesized that E-cadherin upregulated by radiation serves as an anchor for CAR-T cell retention, leading to Trm cell differentiation and prolonged antitumor persistence. Nevertheless, while we explored the E-cadherin/CD103 interaction, the enhanced antitumor efficacy might be a result of various factors such as chemokine gradient, radiation-induced tumor vascular changes, or upregulation of the transmigration family, which warrants further investigation.

Finally, one of the last observations we made is that the intracellular CD28 domain exhibited crosstalk between different CAR molecules. CD28 formed heterodimers on the T cell surface, allowing a “bystander” CAR to hijack this structure to share the intracellular CD3ζ chain.⁵⁴ This discovery hints at potential advancements in dual CAR-T cell designs. Currently, most dual CAR-T designs are bicistronic with a double CD3ζ chain.⁵⁵ The sharing of the CD3ζ chain can potentially reduce vector size, which could raise the transduction efficiency of CAR-Ts during manufacturing.

In addition to outlining the antitumor effects and mechanisms, acknowledging the limitations of the present study is essential. First, we did not use a mouse CAR in an immunocompetent model, which would have allowed for a more comprehensive analysis of TME characteristics in the RT + CAR-T + ICI combinations. The interaction between myeloid cells and CAR-Ts after radiation was not explored. Furthermore, given the presence of PD-L1 expression on normal cells to some extent, dual-targeting CD133/PD-L1 CAR-T could potentially pose “on-target/off-tumor” toxicity against normal organs. Optimizing the affinity of the PD-L1 CAR in the future could help achieve the ideal therapeutic window between normal and tumor cells. Finally, the adopted tumor models were primarily human-derived cell lines with artificially engineered CD133 and PD-L1 expression. To further validate our findings, future studies can utilize patient-derived models that are similar to the clinical scenario.

Conclusions

Our study, for the first time, introduced a novel CAR-T design featuring two CAR structures sharing a single CD3ζ through membrane clustering. Dual-targeting CD133/PD-L1 CAR-T cells combined with αPD-1 effectively enhanced antitumor cytotoxicity and overcame inhibitory signals from the TME. We demonstrated the rationale and efficacy of radiation pre-conditioning plus CAR-T and PD-1 blockade in solid tumors, which is a paradigm-shifting finding and has implications for future clinical use.

Materials and methods

Ethics statement

This research complies with all relevant ethical regulations. All mouse experiments were performed in compliance with the Guide for the Care and Use of Laboratory Animals of Sichuan University and were approved by the Institutional Animal Care and Use Committee of Sichuan University. Clinical samples were collected with written informed consent from patients, adhering to the Declaration of Helsinki and approved by the Biomedical Research Ethics Committee of West China Hospital of Sichuan University (no. 2021-264).

Plasmid construction and lentiviral production

CD133-specific second-generation CARs (αCD133.28ζ) were assembled using CD133-targeting scFv (patent number: 2013/0224202 A1), CD28 hinge, transmembrane and intracellular co-stimulatory domain, and CD3ζ domain. The PD-L1 specific VHH nanoantibody sequence was generated by NBbiolab as described before (patent number: ZL 2019 1 0788886. 3).⁵⁶ The αPDL1 CSR contains the PD-L1 targeting VHH, with CD8α hinge/transmembrane region, 4-1BB co-stimulatory domain, or CD28 hinge/transmembrane/intracellular domain. The dual CAR (αPDL1.28/αCD133.28ζ, αPDL1.BB/αCD133.28ζ, and αPDL1.28/αCD133.BBζ) contains two individual CARs that were linked by the T2A self-cleaving peptide sequence. DLL3-based dual-targeting CAR-T was generated similarly, with replacement of a DLL3-targeted VHH based CAR (patent number: ZL 2021 1 1327597.7) to CD133 sfFv domain. All CAR constructs were synthesized and cloned into a lentivirus vector backbone pHBLV (Hanbio) by Genescript.

Lentiviral vectors were produced after transfection of 293T and titered in 293T cells. 293T cells were seeded at 3 × 10⁶ in a total volume of 10 mL of medium in a 100-mm culture plate. Then 293T cells were transfected with 48 μg of target plasmid pHBLV containing CAR, 24 μg of pMD2.G, and 48 μg of psPAX (Hanbio). The viral supernatant was harvested at 48 and 72 h post transfection, 0.45-μm filtered, concentrated by PEG Virus Precipitation Kit per manufacturer’s protocol (BioVision), and frozen at −80°C until use.

CAR-T manufacturing

Human peripheral blood mononuclear cells from healthy donors were purchased from Milestone Biotechnologies (Beijing, China). Informed consent was obtained from all donors before specimen collection, following the principles of the Declaration of Helsinki. CD3+ T cells were isolated using CD3 microbeads (Miltenyi Biotec, Germany) according to the manufacturer’s instructions. Dynabeads human T-activator CD3/CD28 (Gibco) was used to activate the T cells. After stimulation for 24 h, the virus was added to activated T cells in a RetroNectin-coated plate and centrifuged for 2 h at 1,000 × g. 48 h after transfection, T cells were replaced with fresh medium and beads were removed. The T cell expansion culture medium used is ImmunoCult-XF T Cell Expansion Medium (Stemcell), supplemented with IL-2 (10 ng/mL).

Generation of cancer cell lines

Human colorectal cancer cell line HCT116, breast cancer cell lines MDA-MB-231, and human embryonic kidney epithelial HEK-293T cells were stored in the State Key Laboratory of Biotherapy of Sichuan University. Tumor cells negative for CD133 and PD-L1 expression were transduced with lentiviral vectors expressing CD133 or PD-L1 to construct tumor cell lines overexpressing CD133 and PD-L1. Additionally, luciferase-overexpressing tumor cell lines for co-culture cytotoxicity experiments were established in the same way. MDA-MB-231 and HCT116 cells were genome edited to delete CD274 (PD-L1) using the CRISPR-Cas9 system. 231 cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS), and 1% penicillin-streptomycin; HCT116 cells were cultured in RPMI. All cell lines were grown at 37°C and 5% CO₂ and were regularly validated to be Mycoplasma free and authenticated by short tandem repeat (STR) profiling.

In vitro cytolysis and activation experiments

The cytotoxicity of CAR-T cells was determined by a luciferase-based assay. In brief, the luciferase-expressing tumor cells were used to establish a cytolytic assay. Cytolysis of CAR-T cells was detected by co-culturing with luciferase-expressing tumor cells (HCT116 and MDA-MB-231) at various E:T ratios for 24 hours followed by measurement of the luciferase activity using a microplate reader (Biotek). For multi-round co-culture experiments, tumor cells were seeded in 24-well plates at a concentration of 1 × 10⁵ cells per well 1 day before addition of T cells. T cells were added at an E:T ratio of 2 to 1 without addition of exogenous cytokines. Every 24 h, T cells were collected and transferred into a new well in which tumor cells had been seeded before. The residual tumor cells on previous wells were measured for cytolysis. Anti-PD-1 antibody tislelizumab (BGB-A317, BeiGene) was added in each round at a final concentration of 25 μg/mL.

To perform an antigen stimulation assay, functional αCD3 (1μg/mL), αCD28 (1-4μg/mL) or PDL1-Fc (5-100μg/mL) proteins of different concentration were precoated on 96-well plate in a 60 μL phosphate-buffered saline (PBS) for 12 h. T cells were seeded at a concentration of 2 × 10⁵ cells per well. After 24-h stimulation, T cells were collected for activation marker detection.

The supernatants were collected for TNF-α, IL-2, and IFN-γ detection using a BD Cytometric Bead Array (BD Biosciences). The tests were carried out following the manufacturer’s instructions. Data were obtained using a BD FACSVerse Bioanalyzer. For data analysis, the FCAP Array software (v3.0.1; BD Biosciences) was used.

Flow cytometry assay

In general, single-cell suspensions were harvested and prepared. 2 × 10⁵ to 1 × 10⁶ cells/100 μL were stained with a conjugated antibody cocktail for 30 min on ice according to the manufacturer’s protocol. The cells were washed twice with PBS and resuspended for flow cytometry. APC-conjugated anti-human CD133 antibody (Biolegend, 372806) and PE-conjugated anti-human PD-L1 antibody (Biolegend, 329706) were used to determine the expression levels of CD133 and PD-L1 on different cancer cell lines. AF647-conjugated anti-human CD324 (E-cadherin) (Biolegend, 147308) was used to determine expression level of E-cadherin on HCT116. To detect CAR expression, cells were stained using CD133-his (Acro, CD3-H52H4) followed by PE-CY7 anti-His Tag antibodies (Biolegend, 362620) and PD-L1-biotin (Acro, PD1-H82F3) followed by streptavidin-APC Conjugate (Thermo Fisher, SA1005). For activation assay after co-culture experiments, T cells were stained with anti-CD3(Biolegend, 300448), anti-CD137 (4-1BB) (Biolegend, 309809), anti-CD69 (BD, 557745), and anti-CD25 (BD, 612806) antibodies.

For flow cytometry detection of tissue samples, tumor samples were dissected and digested thoroughly with RPMI 1640 containing 1.0 mg/mL collagenase I (Gibco, 17100-017), 0.5 mg/mL collagenase Ⅳ (Gibco, 17104-019), 2% FBS (Gibco, 10099-141), and 50 μg/mL deoxyribonuclease I (Sigma, D5025-15KU) at 37°C for 30 min and then filtered through a 70-μm mesh (Biofil). Red blood cells were lysed using ACK lysis buffer (Leagene, CS0003) for 5 min at 4°C. The cells were then resuspended in PBS for antibodies staining including anti-CD45 (Biolegend, 3304042), anti-CD3 (Biolegend, 300448), anti-CD8 (BD, 557834), anti-CD103 (Biolegend, 350214), and anti-CXCR6 (Biolegend, 356006). All experiments were performed on a FacsCanto 3 L, Fortessa 4 L HT, Symphony A5, and Fortessa 5 L (BD Biosciences) and the data were analyzed with FlowJo software (V.10, TreeStar).

Cell Trace Violet proliferation assay

Untransduced T cells (bystander population) were labeled with Cell Trace Violet (CTV; Thermo Fisher, catalog no. C34557). CTV-labeled bystander T cells were co-cultured with unlabeled effector cells (vector T or αPD-L1.28/αCD133.28ζ CAR-T) at an E:T ratio of 1:1 in 96-well U-bottom plates (final volume, 200 μL) in RMPI 1640 medium. Cultures were harvested at 24 h and day 4. For flow cytometry, cells were stained with a viability dye and acquired on a BD Fortessa 4 L HT.

Western blotting analysis

For western blotting analysis, the total tumor lysates were electrophoresed and transferred to polyvinylidene fluoride (PVDF) membranes. Then, the membranes were probed with antibodies against PD-L1 (CST, E1L3N, 1:1,000), CD133 (CST, D2V8Q, 1:1,000), and GAPDH (Abcam, ab68481, 1:1,000). Anti-rabbit antibodies (ProteinSimple) were used as secondary antibodies. Protein samples were mixed with 0.1× sample buffer and 5× master mix (ProteinSimple) to achieve a final protein concentration of 1 μg/μL and then denatured at 95°C for 5 min. The denatured protein samples, blocking buffer, primary antibodies, horseradish peroxidase-conjugated secondary antibodies, wash buffer, and chemiluminescent substrate (1:1 luminol-peroxidase mixture) were then added to specific wells of the assay plate. After loading the plate, separation electrophoresis was performed using a capillary system, and immunodetection was fully automated.

In vitro imaging of CAR-T cytolysis and confocal imaging

A real-time cell analyzer system (RTCA) (Opera Phenix Plus, PerkinElmer) and cancer cell fluorescence measurements were used to assess the cytotoxicity. In the RTCA system, GFP-labeled HCT116-PDL1 and 231-CD133 cells were seeded in 96-well plates with an optically clear bottom (PerkinElmer, PDL-coated CellCarrier-96 Ultra) and cultured for 24 h. Then, T cells were added into the unit at 5:1 E:T ratio. The fluorescence images were recorded at 15-min intervals. The signal-time curves were drawn to display the cytotoxicity. The fluorescence values were analyzed using high-intension confocal microscopy (Opera Phenix Plus, PerkinElmer) at different time points after co-culturation.

T cells expressing GFP-tagged PD-L1.28 and Mcherry-tagged CD133.28ζ were stimulated with either PDL1-Fc protein (5μg/mL)or PDL1-Fc protein plus anti-CD3 functional antibodies (1μg/mL), precoated on 96-well plates with an optically clear bottom (PerkinElmer, PDL-coated CellCarrier-96 Ultra) 24 h earlier. The subcellular imaging of CAR molecules was captured by high-intensity confocal microscopy (Opera Phenix Plus, PerkinElmer). All groups of images were acquired with the same settings.

Organoid construct and killing assay

Fresh tumor tissues collected from lung cancer patients with informed consent were placed in preservation solution (AimingMed, 100-032) and transported to the laboratory at 4°C. After washing three times with DPBS (Dulbecco’s phosphate-buffered saline) and mincing into 1-mm³ pieces, the tissues were digested with Digestion Solution I (AimingMed, 100-047) at 37°C for 30 min, followed by termination with DPBS and centrifugation. The cell pellet was then digested with Digestion Solution II (AimingMed, 100-048) at 37°C for 10 min, terminated, and centrifuged again. After red blood cell lysis and counting, cells were resuspended in Matrigel (Corning, 356231) to a density of 4,000–6,000 cells/μL. A 25-μL cell suspension was seeded into each well of a 48-well plate, solidified at 37°C with 5% CO₂ for 30 min, and cultured in SCLC complete medium (AimingMed, 10-100-316).

For organoid co-culture experiment, 1 day ahead, disperse the cultured organoids with cold DPBS-PS (penicillin–streptomycin). Try to be gentle to maintain organoid integrity. Typically, resuspend 3–4 50-μL domes of lung cancer organoids in 4 mL of DPBS-PS, approximately 800 organoids, which can be used for 1 mL of co-culture cell suspension. After centrifugation at 300 × g for 5 min, resuspend the organoids in 1 mL of MasterAim lung cancer organoid complete culture medium and culture in a 96-well clear Ultra-Low Attachment plate. CAR-T cells were added later at an E:T ratio of 5:1 in 1 mL of RPMI plus 10% heat-inactivated FBS for co-culture. Images of each condition were taken on day 1, day 2, and day 3, and supernatant was harvested for cytokine detection on each day.

Transwell study

Tumor cells, HCT116 and 231-CD133, were seeded into the lower chamber of a Transwell plate at a density of 1 × 10⁵ cells per well, with three replicates per group. After overnight adherence of the tumor cells in the lower chamber, they were either irradiated with 10 Gy/1 fraction or left untreated and then incubated at 37°C for 24 h. CAR-T cells were then seeded into the upper chamber of the Transwell plate at a density of 3 × 10⁵ cells per well, followed by incubation at 37°C for another 24 h. After 24 h, cells from the lower chamber were collected, washed twice with PBS, and centrifuged to obtain cell pellets. The cells were stained with CD3 flow cytometry antibodies, and absolute counts of CD3+ cells were determined using flow cytometry absolute counting beads. The migration index was calculated as follows: migration index = number of T cells in the lower chamber/total number of original T cells × 100%. Images of the T cells in the lower chamber were captured using a fluorescence microscope (IX83, Olympus).

Xenograft mouse models

Six- to 8-week-old male NCG (NOD/ShiLtJGpt-Prkdc^em26Cd52Il2rg^em26Cd22/Gpt) mice were purchased from Gempharmatech (JCYK Bioscience, Jiangsu, China). These mice were maintained and bred under specific pathogen-free conditions at Sichuan University with autoclaved food, bedding, and water. Animals were housed at room temperature (23°C ± 2°C) at a humidity of 30%–70% on a 12-h light/12-h dark cycle (6:00–18:00). NCG mice were inoculated with 1 × 10⁶ HCT116 or MDA-MB-231 tumor cells (90 μL PBS and mixed with 10 μL of Matrigel [Corning]) subcutaneously (s.c.). 10 days after tumor engraftment, when the tumor size reached approximately 100 mm³, the mice were randomly assigned to different groups (Figures 2F, 2G, and S4D). For radiation treatment study, tumor cells were inoculated in the right hindlimb. When the tumor size reached approximately 200 mm³, the mice were randomly assigned to different groups (Figures 4C and S6). In our mouse experiments, the tumor volumes were measured with calipers using the formula volume = (a × b²)/2, in which a and b are the largest and smallest perpendicular diameters, respectively.

Tumor growth and treatment

The mice were randomized into different groups as described above. T cells were harvested 12–15 days post transduction and washed once with PBS. The cell pellet was collected and resuspended in PBS. Each mouse received an intravenous injection of 5 × 10⁶ T cells via the tail vein. Before radiation treatment, each mouse was anesthetized and shielded using a lead box such that only the irradiated tumor was exposed. Radiation was delivered using an RS-2000 X-ray irradiator (Rad Source Technologies, Alpharetta, GA, USA) operated at 160 kV, 25 mA, and a dose rate of 1.836 Gy/min (0.3 mm copper filtration; the distance from the X-ray source to the target was 30 cm). From the day of CAR-T infusion, the PD-1 inhibitor tislelizumab was administered intraperitoneally (10 mg/kg) every 7 days. Tumor growth was monitored by bioluminescence imaging using the IVIS lumina II in vivo imaging system (PerkinElmer). Mice were euthanized when the tumor volume reached 2,000 mm³. Survival was calculated from the date of treatment initiation to death.

Copy number of CAR-T cells and cytokine detection in mouse peripheral blood

Perform retro-orbital blood collection of the mice using a disposable blood collection needle. Centrifuge the blood at 300 × g for 5 min. The supernatants are harvested for cytokine detection. Extract genomic DNA from mouse peripheral blood using the EasyPure Blood Genomic DNA kit. Design PCR primers based on the CD133 scFv gene sequence (Forward primer: 5′-CCTATTTTTGCGCCACCGAC-3′. reverse primer: 5′-GCACTTGAAACGGTCAGCG-3′). Establish a Cq value-log copy number standard curve using the αPDL1.28/αCD133.28ζ plasmid. Dilute the plasmid to create six gradients containing 10⁶, 10⁵, 10⁴, 10³, 10², and 10 copies per 100 ng of DNA. The PCR amplification reaction steps have been described before.⁵⁷ Use the standard curve to calculate the CAR sequence copy number in each sample by substituting the Cq values of the samples into the standard curve equation.

Histological staining and data analysis

The tissue samples from tumors and vital organs were processed into paraffin blocks and sectioned into 5-μm-thick FFPE (formalin-fixed paraffin-embedded) sections. Organ samples were stained with hematoxylin and eosin (Thermo Fisher) according to the manufacturer’s instructions. Multiplex immunohistochemistry (IHC) staining was performed using an Opal 7-color kit (Akoya Bioscience, NEL861001KT). The markers E-cadherin (ZA-0565, ZSGB-Bio, 1:200, Opal 570), CD3 (MAB-0740, Maxim Biological Technology, 1:1, Opal 620), CD8 (85736S, Cell Signaling Technology, 1:200, Opal 520), CD103 (ZA-0667, ZSGB-Bio, 1:50, Opal 690), and CXCR6 (ab273116, abcam, 1:200, Opal 480) were elevated via IHC. The sample sections were dewaxed using xylene for 20 min and then rehydrated with ethanol. Antigen retrieval was performed using microwave treatment with a buffer (pH 9.0) followed by cooling to room temperature. An antibody diluent/block (72424205, Akoya Bioscience) was applied at room temperature for 10 min. Sections were then incubated with primary antibodies, followed by secondary reagents at 37°C for 10 min and tyramide signal amplification reagents at room temperature for 10 min (Opal 480, Opal 520, Opal 570, Opal 620, Opal 690, Akoya Bioscience, 1:100). After staining all markers, a second round of antigen retrieval was performed using microwave treatment. Nuclear staining was done using DAPI (Akoya Bioscience, 1:5) at room temperature for 5 min. The slides were scanned using a confocal microscope with the Vectra Polaris (PerkinElmer) system. The percentage of positively stained cells among all nucleated cells was determined. Multispectral image unmixing was performed using QuPath software (Version 3.0).⁵⁸ DAPI-positive cells were identified through the “cell detection” command, and a single-channel threshold was selected for each marker. As a result, all detected cells were divided into different subgroups for further analysis. Defective samples or areas with staining artifacts were reanalyzed or excluded.

Tissue microarray and immunohistochemical staining

Human colon cancer, rectal cancer, and liver cancer tissue microarray, which contained 48 paired cases of cancer tissues and adjacent non-cancerous ones each, were purchased from Shanghai Zhuoli Bio (Shanghai, China). IHC was performed. The sections were incubated overnight at 4°C with an anti-CD133 monoclonal antibody (1:50) (64326s, Cell Signaling Technology) followed by an incubation with a secondary antibody (goat anti-rabbit/horseradish peroxidase, Santa Cruz Biotechnology, 1:200 dilution) at room temperature for 30 min. Finally, the sections were visualized with diaminobenzidine for ∼2 min and counterstained with hematoxylin. The assessment of the IHC was employed by a semi-quantitative immunoreactivity score as previously reported.⁵⁹

scRNA-seq and analysis of mouse samples

Single-cell suspensions were prepared from tumors subjected to different treatments. Following the manufacturer’s protocol in the Chromium Single Cell 3′ Reagents Kits v2 User Guide, we used the Chromium Single Cell 3′ Library & Gel Bead Kit v2 (PN-120237), Chromium Single Cell 3′ Chip Kit v2 (PN-120236), and Chromium i7 Multiplex Kit (PN-120262). The single-cell suspension was washed twice with 1× PBS containing 0.04% BSA. Cell number and concentration were confirmed using the TC20 Automated Cell Counter. The cells were immediately processed on the 10× Genomics Chromium Controller for gel bead-in-emulsion (GEM) generation. Using the 10× Genomics Chromium Single Cell 3′ Reagent Kit (V2 chemistry), we prepared barcoded complementary DNAs (cDNAs), which were subsequently recovered, purified, and amplified to generate sufficient material for library preparation. Library quality and concentration were assessed with the Agilent Bioanalyzer 2100. The libraries were then sequenced on the Illumina NovaSeq6000 platform using PE150 sequencing. The Xenocell algorithm was applied to classify graft-derived human and mouse cells.²⁴

We followed the scRNA-seq data analysis pipeline as we described previously.⁶⁰^,⁶¹^,⁶² We used the Read10X function from the Seurat package (version 4.0.4) to load the Cell Ranger output. Potential doublets were removed using Scrublet. Quality control was applied to the aggregated data: cells with fewer than 200 detected genes or more than 15% mitochondrial genes were excluded, as were cells with over 6,000 detected genes to further avoid doublets. After filtering, 71,681 cells were included for further analysis. We applied SCTransform for normalization, followed by principal-component analysis (PCA) to incorporate highly variable features and reduce dimensionality. The first 30 principal components (PCs) were selected for further analysis. To correct for batch effects, we used FindIntegrationAnchors in Seurat. Clustering analysis was performed with FindNeighbors and FindClusters, based on the edge weights between cells, producing a shared nearest-neighbor graph via the Louvain algorithm. The uniform manifold approximation and projection (UMAP) method was then used to visualize clusters. For subclustering, a similar procedure was followed. Differentially expressed features for each cluster were identified using the FindAllMarkers function, applying the default nonparametric Wilcoxon rank-sum test with Bonferroni correction.

To investigate the dynamic biological processes of CAR-T cells, we applied the Monocle algorithm (version 2.22.0). Using the NewCellDataset function, we created a new object from the transcript count data of the included cells. The estimateSizeFactors and estimateDispersions functions were subsequently used to prepare the data for further analysis. Genes were selected for ordering if they were expressed in at least 10% of the cells in the dataset and had a p value <0.01 as determined by the differentialGeneTest function. To reduce the dimensionality, we applied the ReduceDimension function to map the data to two dimensions. The orderCells function was then used to order the cells based on the selected order genes. Pseudotime-dependent genes were identified using the differentialGeneTest function, and smooth expression curves were visualized using the plot_genes_branched_pseudotime function.

GO enrichment analysis and GSEA were performed using the clusterProfiler package. Additionally, a nonparametric and unsupervised algorithm from the Gene Set Variation Analysis (GSVA) package (version 1.14.1) was also applied to assess the scoring of individual cells.

scRNA-seq and analysis of clinical samples

Four patients with SCLC underwent first-line chemotherapy ± ICI treatment or RT + chemotherapy + ICI at West China Hospital of Sichuan University. Clinical samples were collected with written informed consent from patients, adhering to the Declaration of Helsinki and approved by the Biomedical Research Ethics Committee of West China Hospital of Sichuan University (no. 2021-264). Sample characteristics, library construction, and sequencing protocols are detailed in our previous publication.⁵⁷

The sequencing data were merged and converted into Seurat objects based on Seurat (v.4.1.0). Cells with 500–10,000 detected genes and 1,000–50,000 detected unique molecular identifiers were retained, while cells with mitochondrial content exceeding 10% were excluded. Following quality control, the expression matrix was normalized and scaled, and 2,000 highly variable genes were identified using the FindVariableFeatures function for subsequent PCA, retaining the top 20 PCs. To correct batch effects across samples, Harmony (v0.1.0) was applied, followed by the RunUMAP function for visualization and the FindNeighbors and FindClusters functions for unsupervised clustering. Further subdivision of T/NK cells, identification of variable genes, and reduction of dimensions were repeated to better distinguish the differences between cells. Cell types were manually annotated based on typical marker genes. Cells expressing marker genes for two cell types, including cells expressing both typical marker genes for T cells and SCLC cells, and cells expressing both typical marker genes for T cells and erythrocytes were considered doublets and removed from subsequent analyses.

The DGE analysis of each cell type was performed using MAST (v.1.20.0). For CD8_ITGAE cells versus other CD8⁺ T cells, DGE results were ranked by log2(fold change) in descending order, and GSEA was performed using clusterProfiler (v.4.2.2) with previously established signatures of Trm cells. For DGE analysis between two SCLC cell groups, upregulated genes were filtered with adjusted p values <0.05, log2(fold change) > 0.25, and the proportion of genes expressed in cells >0.25. Enrichment analysis of upregulated genes based on molecular function terms in the GO database was also conducted using clusterProfiler (v.4.2.2).

The cell-cell communication analysis between SCLC cells and each subtype of CD8+ T cells was performed using CellPhoneDB (v.5.0.0). The significance of each ligand-receptor pair was assessed based on expression levels of ligands and receptors in each cell type.

Statistical analysis

Statistical analyses were performed using Prism v.9.5.0 software (GraphPad) and R. The details of the specific statistical analyses used in the individual experiments can be found in the appropriate figure legends. For the in vivo treatment experiments, all groups were randomized according to their tumor volumes before treatment. For the in vitro experiments, all samples were randomly assigned to untreated or treatment groups. To compare tumor growth curves, we used two-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test if more than two treatment groups were compared at a given time point, or Sidak’s multiple comparison test if only two treatment groups were compared. For comparing more than two groups, first, a one-way ANOVA was performed, followed by Tukey’s multiple comparison test. Survival was measured using the Kaplan-Meier method, and survival curves were analyzed using the log rank test. Data are represented as the mean ± SEM (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns, not significant).

Data and code availability

The raw sequencing data for this study have been deposited at the National Genomics Data Center (NGDC; https://ngdc.cncb.ac.cn), data for tumor models under GSA: CRA020795, and data for clinical samples under GSA-human: HRA009487. Other scRNA-seq data were retrieved from Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/). The scRNA-seq data are available under accession number (GSE246613 and PRJNA818695). All other data supporting the findings of this study are available from the corresponding author upon reasonable request.

Acknowledgments

We thank Li Wang and Fang Chen of Laboratory of Clinical Cell Therapy, West China Hospital, Sichuan University (for assistance with cell line and mouse experiments); Li Li, Fei Chen, Chunjuan Bao, and Yang Deng of the Institute of Clinical Pathology, West China Hospital, Sichuan University (for processing histological staining); Wentong Meng, Qiaorong Huang, and Xue Li of the Laboratory of Stem Cell Biology, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University; Ting Cao of Laboratory of Infectious Diseases and Vaccine, West China Hospital, Sichuan University; and Yan Wang, Huifang Li, and Mengli Zhu of Core Facilities of West China Hospital, Sichuan University (for assistance with flow cytometry and cytometric bead array); and Li Chai, Yi Li, Xing Xu, and Xiangyi Ren of Core Facilities of West China Hospital, Sichuan University (for assistance with mIF and confocal imaging). This study was supported by the Original Exploration Program of National Natural Science Foundation of China (No. 82350128); 1·3·5 project for disciplines of excellence, West China Hospital, Sichuan University (No. ZYJC21003 and ZYYC23010), the National Natural Science Foundation of China (No. 82203022 and 82303772), China Postdoctoral Science Foundation (No. 2022M722274), and Sichuan Science and Technology Program (2024NSFSC1914).

Author contributions

Y.L., J.X., L. Yin, and Z.Y. contributed to the concept. Y.L., J.X., L. Yin, Z.Y., K.K., P.-H.L., and L. Yi designed the experiment research. Z.Y., L. Yin, K.K., and P.-H.L. executed, interpreted, and analyzed the experiment data. Z.Y., L. Yin, K.K., and P.-H.L. were involved in the analysis and assembly of data. Z.Y., K.K., and P.-H.L. wrote the original draft. All authors revised the manuscript and approved the final version submitted for publication.

Declaration of interests

Y.L. has an invited speaker/project lead/principal investigator role with Roche, AstraZeneca, BeiGene, and Hengrui Pharmaceuticals and is an invited speaker role with Pfizer, Merck Sharp & Dohme, and Bristol-Myers Squibb.

Footnotes

Supplemental information can be found online at https://doi.org/10.1016/j.ymthe.2025.12.068.

Contributor Information

Jianxin Xue, Email: radjianxin@163.com.

You Lu, Email: radyoulu@hotmail.com.

Supplemental information

Document S1. Figures S1–S12

mmc1.pdf^{(73.5MB, pdf)}

Document S2. Article plus supplemental information

mmc5.pdf^{(96.2MB, pdf)}

References

1.Lu Y., Xue J., Deng T., Zhou X., Yu K., Deng L., Huang M., Yi X., Liang M., Wang Y., et al. Safety and feasibility of CRISPR-edited T cells in patients with refractory non-small-cell lung cancer. Nat. Med. 2020;26:732–740. doi: 10.1038/s41591-020-0840-5. [DOI] [PubMed] [Google Scholar]
2.June C.H., Sadelain M. Chimeric Antigen Receptor Therapy. N. Engl. J. Med. 2018;379:64–73. doi: 10.1056/NEJMra1706169. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Albelda S.M. CAR T cell therapy for patients with solid tumours: key lessons to learn and unlearn. Nat. Rev. Clin. Oncol. 2024;21:47–66. doi: 10.1038/s41571-023-00832-4. [DOI] [PubMed] [Google Scholar]
4.Wagner J., Wickman E., DeRenzo C., Gottschalk S. CAR T Cell Therapy for Solid Tumors: Bright Future or Dark Reality? Mol. Ther. 2020;28:2320–2339. doi: 10.1016/j.ymthe.2020.09.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Hou A.J., Chen L.C., Chen Y.Y. Navigating CAR-T cells through the solid-tumour microenvironment. Nat. Rev. Drug Discov. 2021;20:531–550. doi: 10.1038/s41573-021-00189-2. [DOI] [PubMed] [Google Scholar]
6.Glumac p.m., LeBeau a.m. The role of CD133 in cancer: a concise review. Clin. Transl. Med. 2018;7:18. doi: 10.1186/s40169-018-0198-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Smith L.M., Nesterova A., Ryan M.C., Duniho S., Jonas M., Anderson M., Zabinski R.F., Sutherland M.K., Gerber H.P., Van Orden K.L., et al. CD133/prominin-1 is a potential therapeutic target for antibody-drug conjugates in hepatocellular and gastric cancers. Br. J. Cancer. 2008;99:100–109. doi: 10.1038/sj.bjc.6604437. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Brudno J.N., Maus M.V., Hinrichs C.S. CAR T Cells and T-Cell Therapies for Cancer: A Translational Science Review. JAMA. 2024;332:1924–1935. doi: 10.1001/jama.2024.19462. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Sanmamed M.F., Chen L. A Paradigm Shift in Cancer Immunotherapy: From Enhancement to Normalization. Cell. 2018;175:313–326. doi: 10.1016/j.cell.2018.09.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Liu X., Ranganathan R., Jiang S., Fang C., Sun J., Kim S., Newick K., Lo A., June C.H., Zhao Y., Moon E.K. A Chimeric Switch-Receptor Targeting PD1 Augments the Efficacy of Second-Generation CAR T Cells in Advanced Solid Tumors. Cancer Res. 2016;76:1578–1590. doi: 10.1158/0008-5472.CAN-15-2524. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Liang Y., Liu H., Lu Z., Lei W., Zhang C., Li P., Liang A., Young K.H., Qian W. CD19 CAR-T expressing PD-1/CD28 chimeric switch receptor as a salvage therapy for DLBCL patients treated with different CD19-directed CAR T cell therapies. J. Hematol. Oncol. 2021;14:26. doi: 10.1186/s13045-021-01044-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Liu H., Lei W., Zhang C., Yang C., Wei J., Guo Q., Guo X., Chen Z., Lu Y., Young K.H., et al. CD19-specific CAR T Cells that Express a PD-1/CD28 Chimeric Switch-Receptor are Effective in Patients with PD-L1-positive B-Cell Lymphoma. Clin. Cancer Res. 2021;27:473–484. doi: 10.1158/1078-0432.CCR-20-1457. [DOI] [PubMed] [Google Scholar]
13.Qin L., Cui Y., Yuan T., Chen D., Zhao R., Li S., Jiang Z., Wu Q., Long Y., Wang S., et al. Co-expression of a PD-L1-specific chimeric switch receptor augments the efficacy and persistence of CAR T cells via the CD70-CD27 axis. Nat. Commun. 2022;13:6051. doi: 10.1038/s41467-022-33793-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Zhao W., Jia L., Zhang M., Huang X., Qian P., Tang Q., Zhu J., Feng Z. The killing effect of novel bi-specific Trop2/PD-L1 CAR-T cell targeted gastric cancer. Am. J. Cancer Res. 2019;9:1846–1856. [PMC free article] [PubMed] [Google Scholar]
15.Jiang W., Li T., Guo J., Wang J., Jia L., Shi X., Yang T., Jiao R., Wei X., Feng Z., et al. Bispecific c-Met/PD-L1 CAR-T Cells Have Enhanced Therapeutic Effects on Hepatocellular Carcinoma. Front. Oncol. 2021;11 doi: 10.3389/fonc.2021.546586. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Galluzzi L., Aryankalayil M.J., Coleman C.N., Formenti S.C. Emerging evidence for adapting radiotherapy to immunotherapy. Nat. Rev. Clin. Oncol. 2023;20:543–557. doi: 10.1038/s41571-023-00782-x. [DOI] [PubMed] [Google Scholar]
17.Herrera F.G., Ronet C., Ochoa de Olza M., Barras D., Crespo I., Andreatta M., Corria-Osorio J., Spill A., Benedetti F., Genolet R., et al. Low-Dose Radiotherapy Reverses Tumor Immune Desertification and Resistance to Immunotherapy. Cancer Discov. 2022;12:108–133. doi: 10.1158/2159-8290.CD-21-0003. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Yin L., Xue J., Li R., Zhou L., Deng L., Chen L., Zhang Y., Li Y., Zhang X., Xiu W., et al. Effect of Low-Dose Radiation Therapy on Abscopal Responses to Hypofractionated Radiation Therapy and Anti-PD1 in Mice and Patients With Non-Small Cell Lung Cancer. Int. J. Radiat. Oncol. Biol. Phys. 2020;108:212–224. doi: 10.1016/j.ijrobp.2020.05.002. [DOI] [PubMed] [Google Scholar]
19.Zhou X., Zhou L., Yao Z., Huang M., Gong Y., Zou B., Zhu J., Liu Y., Peng F., Zhang Y., et al. Safety and Tolerability of Low-Dose Radiation and Stereotactic Body Radiotherapy + Sintilimab for Treatment-Naive Stage IV PD-L1+ Non-Small Cell Lung Cancer Patients. Clin. Cancer Res. 2023;29:4098–4108. doi: 10.1158/1078-0432.CCR-23-0315. [DOI] [PubMed] [Google Scholar]
20.Antonia S.J., Villegas A., Daniel D., Vicente D., Murakami S., Hui R., Kurata T., Chiappori A., Lee K.H., de Wit M., et al. Overall Survival with Durvalumab after Chemoradiotherapy in Stage III NSCLC. N. Engl. J. Med. 2018;379:2342–2350. doi: 10.1056/NEJMoa1809697. [DOI] [PubMed] [Google Scholar]
21.Chang J.Y., Lin S.H., Dong W., Liao Z., Gandhi S.J., Gay C.M., Zhang J., Chun S.G., Elamin Y.Y., Fossella F.V., et al. Stereotactic ablative radiotherapy with or without immunotherapy for early-stage or isolated lung parenchymal recurrent node-negative non-small-cell lung cancer: an open-label, randomised, phase 2 trial. Lancet. 2023;402:871–881. doi: 10.1016/S0140-6736(23)01384-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Zhao Z., Condomines M., van der Stegen S.J.C., Perna F., Kloss C.C., Gunset G., Plotkin J., Sadelain M. Structural Design of Engineered Costimulation Determines Tumor Rejection Kinetics and Persistence of CAR T Cells. Cancer Cell. 2015;28:415–428. doi: 10.1016/j.ccell.2015.09.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Srivastava S., Riddell S.R. Engineering CAR-T cells: Design concepts. Trends Immunol. 2015;36:494–502. doi: 10.1016/j.it.2015.06.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Cheloni S., Hillje R., Luzi L., Pelicci P.G., Gatti E. XenoCell: classification of cellular barcodes in single cell experiments from xenograft samples. BMC Med. Genomics. 2021;14:34. doi: 10.1186/s12920-021-00872-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Savas P., Virassamy B., Ye C., Salim A., Mintoff C.P., Caramia F., Salgado R., Byrne D.J., Teo Z.L., Dushyanthen S., et al. Single-cell profiling of breast cancer T cells reveals a tissue-resident memory subset associated with improved prognosis. Nat. Med. 2018;24:986–993. doi: 10.1038/s41591-018-0078-7. [DOI] [PubMed] [Google Scholar]
26.Okla K., Farber D.L., Zou W. Tissue-resident memory T cells in tumor immunity and immunotherapy. J. Exp. Med. 2021;218 doi: 10.1084/jem.20201605. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Mendonsa a.m., Na T.Y., Gumbiner B.M. E-cadherin in contact inhibition and cancer. Oncogene. 2018;37:4769–4780. doi: 10.1038/s41388-018-0304-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Melssen M.M., Sheybani N.D., Leick K.M., Slingluff C.L., Jr. Barriers to immune cell infiltration in tumors. J. Immunother. Cancer. 2023;11 doi: 10.1136/jitc-2022-006401. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Shiao S.L., Gouin K.H., third, Ing N., Ho A., Basho R., Shah A., Mebane R.H., Zitser D., Martinez A., Mevises N.Y., et al. Single-cell and spatial profiling identify three response trajectories to pembrolizumab and radiation therapy in triple negative breast cancer. Cancer Cell. 2024;42:70–84.e8. doi: 10.1016/j.ccell.2023.12.012. [DOI] [PubMed] [Google Scholar]
30.Schalck A., Tran T., Li J., Sei E., Bai S., Hu M., Lin J., Bright S.J., Reddick S., Yang F., et al. The impact of breast radiotherapy on the tumor genome and immune ecosystem. Cell Rep. 2025;44 doi: 10.1016/j.celrep.2025.115703. [DOI] [PubMed] [Google Scholar]
31.Qi C., Gong J., Li J., Liu D., Qin Y., Ge S., Zhang M., Peng Z., Zhou J., Cao Y., et al. Claudin18.2-specific CAR T cells in gastrointestinal cancers: phase 1 trial interim results. Nat. Med. 2022;28:1189–1198. doi: 10.1038/s41591-022-01800-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Narayan V., Barber-Rotenberg J.S., Jung I.Y., Lacey S.F., Rech A.J., Davis M.M., Hwang W.T., Lal P., Carpenter E.L., Maude S.L., et al. PSMA-targeting TGFbeta-insensitive armored CAR T cells in metastatic castration-resistant prostate cancer: a phase 1 trial. Nat. Med. 2022;28:724–734. doi: 10.1038/s41591-022-01726-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Dai H., Tong C., Shi D., Chen M., Guo Y., Chen D., Han X., Wang H., Wang Y., Shen P. Efficacy and biomarker analysis of CD133-directed CAR T cells in advanced hepatocellular carcinoma: a single-arm, open-label, phase II trial. Oncoimmunology. 2020;9 doi: 10.1080/2162402X.2020.1846926. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Fu Q., Zheng Y., Fang W., Zhao Q., Zhao P., Liu L., Zhai Y., Tong Z., Zhang H., Lin M., et al. RUNX-3-expressing CAR T cells targeting glypican-3 in patients with heavily pretreated advanced hepatocellular carcinoma: a phase I trial. EClinicalMedicine. 2023;63 doi: 10.1016/j.eclinm.2023.102175. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Mao Q., Zhang Y., Fu X., Xue J., Guo W., Meng M., Zhou Z., Mo X., Lu Y. A tumor hypoxic niche protects human colon cancer stem cells from chemotherapy. J. Cancer Res. Clin. Oncol. 2013;139:211–222. doi: 10.1007/s00432-012-1310-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Jang J.W., Song Y., Kim S.H., Kim J., Seo H.R. Potential mechanisms of CD133 in cancer stem cells. Life Sci. 2017;184:25–29. doi: 10.1016/j.lfs.2017.07.008. [DOI] [PubMed] [Google Scholar]
37.Shmelkov S.V., Butler J.M., Hooper A.T., Hormigo A., Kushner J., Milde T., St Clair R., Baljevic M., White I., Jin D.K., et al. CD133 expression is not restricted to stem cells, and both CD133+ and CD133- metastatic colon cancer cells initiate tumors. J. Clin. Invest. 2008;118:2111–2120. doi: 10.1172/JCI34401. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Zheng Y., Wang L., Yin L., Yao Z., Tong R., Xue J., Lu Y. Lung Cancer Stem Cell Markers as Therapeutic Targets: An Update on Signaling Pathways and Therapies. Front. Oncol. 2022;12 doi: 10.3389/fonc.2022.873994. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Wang Y., Chen M., Wu Z., Tong C., Dai H., Guo Y., Liu Y., Huang J., Lv H., Luo C., et al. CD133-directed CAR T cells for advanced metastasis malignancies: A phase I trial. Oncoimmunology. 2018;7 doi: 10.1080/2162402X.2018.1440169. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Zhou M., Chen M., Shi B., Di S., Sun R., Jiang H., Li Z. Radiation enhances the efficacy of EGFR-targeted CAR-T cells against triple-negative breast cancer by activating NF-kappaB/Icam1 signaling. Mol. Ther. 2022;30:3379–3393. doi: 10.1016/j.ymthe.2022.07.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Weiss T., Weller M., Guckenberger M., Sentman C.L., Roth P. NKG2D-Based CAR T Cells and Radiotherapy Exert Synergistic Efficacy in Glioblastoma. Cancer Res. 2018;78:1031–1043. doi: 10.1158/0008-5472.CAN-17-1788. [DOI] [PubMed] [Google Scholar]
42.Quach H.T., Skovgard M.S., Villena-Vargas J., Bellis R.Y., Chintala N.K., Amador-Molina A., Bai Y., Banerjee S., Saini J., Xiong Y., et al. Tumor-Targeted Nonablative Radiation Promotes Solid Tumor CAR T cell Therapy Efficacy. Cancer Immunol. Res. 2023;11:1314–1331. doi: 10.1158/2326-6066.CIR-22-0840. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Ababneh H.S., Abramson J.S., Johnson P.C., Patel C.G. Assessing the role of radiotherapy in patients with refractory or relapsed high-grade B cell lymphomas treated with CAR T cell therapy. Radiother. Oncol. 2022;175:65–72. doi: 10.1016/j.radonc.2022.08.007. [DOI] [PubMed] [Google Scholar]
44.Kostopoulos N., Costabile F., Krimitza E., Beghi S., Goia D., Perales-Linares R., Thyfronitis G., LaRiviere M.J., Chong E.A., Schuster S.J., et al. Local radiation enhances systemic CAR T cell efficacy by augmenting antigen crosspresentation and T cell infiltration. Blood Adv. 2024;8:6308–6320. doi: 10.1182/bloodadvances.2024012599. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.DeSelm C., Palomba M.L., Yahalom J., Hamieh M., Eyquem J., Rajasekhar V.K., Sadelain M. Low-Dose Radiation Conditioning Enables CAR T Cells to Mitigate Antigen Escape. Mol. Ther. 2018;26:2542–2552. doi: 10.1016/j.ymthe.2018.09.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Ni H., Reitman Z.J., Zou W., Akhtar M.N., Paul R., Huang M., Zhang D., Zheng H., Zhang R., Ma R., et al. FLASH radiation reprograms lipid metabolism and macrophage immunity and sensitizes medulloblastoma to CAR-T cell therapy. Nat. Cancer. 2025;6:460–473. doi: 10.1038/s43018-025-00905-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Zhang B., Hu M., Ma Q., Li K., Li X., He X., Shu P., Chen Y., Gao G., Qin D., et al. Optimized CAR-T therapy based on spatiotemporal changes and chemotactic mechanisms of MDSCs induced by hypofractionated radiotherapy. Mol. Ther. 2023;31:2105–2119. doi: 10.1016/j.ymthe.2023.04.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Oliveira G., Egloff a.m., Afeyan A.B., Wolff J.O., Zeng Z., Chernock R.D., Zhou L., Messier C., Lizotte P., Pfaff K.L., et al. Preexisting tumor-resident T cells with cytotoxic potential associate with response to neoadjuvant anti-PD-1 in head and neck cancer. Sci. Immunol. 2023;8 doi: 10.1126/sciimmunol.adf4968. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Jung I.Y., Noguera-Ortega E., Bartoszek R., Collins S.M., Williams E., Davis M., Jadlowsky J.K., Plesa G., Siegel D.L., Chew A., et al. Tissue-resident memory CAR T cells with stem-like characteristics display enhanced efficacy against solid and liquid tumors. Cell Rep. Med. 2023;4 doi: 10.1016/j.xcrm.2023.101053. [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Lesch S., Blumenberg V., Stoiber S., Gottschlich A., Ogonek J., Cadilha B.L., Dantes Z., Rataj F., Dorman K., Lutz J., et al. T cells armed with C-X-C chemokine receptor type 6 enhance adoptive cell therapy for pancreatic tumours. Nat. Biomed. Eng. 2021;5:1246–1260. doi: 10.1038/s41551-021-00737-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Ray M.E., Mehra R., Sandler H.M., Daignault S., Shah R.B. E-cadherin protein expression predicts prostate cancer salvage radiotherapy outcomes. J. Urol. 2006;176:1409–1414. doi: 10.1016/j.juro.2006.06.014. [DOI] [PubMed] [Google Scholar]
52.Krischak A., Kowaliuk J., Sarsarshahi S., Dörr W., Kleiter M. Effect of irradiation on the expression of E-cadherin and beta-catenin in early and late radiation sequelae of the urinary bladder and its modulation by NF-kappaB inhibitor thalidomide. Strahlenther. Onkol. 2021;197:537–546. doi: 10.1007/s00066-021-01751-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Corgnac S., Boutet M., Kfoury M., Naltet C., Mami-Chouaib F. The Emerging Role of CD8(+) Tissue Resident Memory T (T(RM)) Cells in Antitumor Immunity: A Unique Functional Contribution of the CD103 Integrin. Front. Immunol. 2018;9:1904. doi: 10.3389/fimmu.2018.01904. [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Muller Y.D., Nguyen D.P., Ferreira L.M.R., Ho P., Raffin C., Valencia R.V.B., Congrave-Wilson Z., Roth T.L., Eyquem J., Van Gool F., et al. The CD28-Transmembrane Domain Mediates Chimeric Antigen Receptor Heterodimerization With CD28. Front. Immunol. 2021;12 doi: 10.3389/fimmu.2021.639818. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Xie B., Li Z., Zhou J., Wang W. Current Status and Perspectives of Dual-Targeting Chimeric Antigen Receptor T-Cell Therapy for the Treatment of Hematological Malignancies. Cancers (Basel) 2022;14 doi: 10.3390/cancers14133230. [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Sun L., Chen B., Liu X., Zhu Y., Zhang G., Liang X., Xing L., Xu W., Jiang S., Wang X. Alpaca-derived nanobody targeting the hydrophobic pocket of the HIV-1 gp41 NHR broadly neutralizes HIV-1 by blocking six-helix bundle formation. Curr. Res. Microb. Sci. 2024;7 doi: 10.1016/j.crmicr.2024.100263. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Wang H., Yao Z., Kang K., Zhou L., Xiu W., Sun J., Xie C., Yu M., Li Y., Zhang Y., et al. Preclinical study and phase II trial of adapting low-dose radiotherapy to immunotherapy in small cell lung cancer. Med. 2024;5:1237–1254.e9. doi: 10.1016/j.medj.2024.06.002. [DOI] [PubMed] [Google Scholar]
58.Bankhead P., Loughrey M.B., Fernández J.A., Dombrowski Y., McArt D.G., Dunne P.D., McQuaid S., Gray R.T., Murray L.J., Coleman H.G., et al. QuPath: Open source software for digital pathology image analysis. Sci. Rep. 2017;7 doi: 10.1038/s41598-017-17204-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Brugmann A., Eld M., Lelkaitis G., Nielsen S., Grunkin M., Hansen J.D., Foged N.T., Vyberg M. Digital image analysis of membrane connectivity is a robust measure of HER2 immunostains. Breast Cancer Res. Treat. 2012;132:41–49. doi: 10.1007/s10549-011-1514-2. [DOI] [PubMed] [Google Scholar]
60.Luo H., Xia X., Huang L.B., An H., Cao M., Kim G.D., Chen H.N., Zhang W.H., Shu Y., Kong X., et al. Pan-cancer single-cell analysis reveals the heterogeneity and plasticity of cancer-associated fibroblasts in the tumor microenvironment. Nat. Commun. 2022;13:6619. doi: 10.1038/s41467-022-34395-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Luo H., Xia X., Kim G.D., Liu Y., Xue Z., Zhang L., Shu Y., Yang T., Chen Y., Zhang S., et al. Characterizing dedifferentiation of thyroid cancer by integrated analysis. Sci. Adv. 2021;7 doi: 10.1126/sciadv.abf3657. [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Cao M., Deng Y., Deng Y., Wu J., Yang C., Wang Z., Hou Q., Fu H., Ren Z., Xia X., et al. Characterization of immature ovarian teratomas through single-cell transcriptome. Front. Immunol. 2023;14 doi: 10.3389/fimmu.2023.1131814. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Video S1. Long-term cytotoxic interaction between αPDL1.28/αCD133.28ζ CAR-T cells and tumor cells captured by fluorescence microscopy

Download video file^{(4.8MB, mp4)}

Video S2. Long-term cytotoxic interaction between αCD133.28ζ CAR-T cells and tumor cells captured by fluorescence microscopy

Download video file^{(4.8MB, mp4)}

Video S3. Long-term cytotoxic interaction between vector T cells and tumor cells captured by fluorescence microscopy

Download video file^{(4.8MB, mp4)}

Document S1. Figures S1–S12

mmc1.pdf^{(73.5MB, pdf)}

Document S2. Article plus supplemental information

mmc5.pdf^{(96.2MB, pdf)}

Data Availability Statement

[bib1] 1.Lu Y., Xue J., Deng T., Zhou X., Yu K., Deng L., Huang M., Yi X., Liang M., Wang Y., et al. Safety and feasibility of CRISPR-edited T cells in patients with refractory non-small-cell lung cancer. Nat. Med. 2020;26:732–740. doi: 10.1038/s41591-020-0840-5. [DOI] [PubMed] [Google Scholar]

[bib2] 2.June C.H., Sadelain M. Chimeric Antigen Receptor Therapy. N. Engl. J. Med. 2018;379:64–73. doi: 10.1056/NEJMra1706169. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib3] 3.Albelda S.M. CAR T cell therapy for patients with solid tumours: key lessons to learn and unlearn. Nat. Rev. Clin. Oncol. 2024;21:47–66. doi: 10.1038/s41571-023-00832-4. [DOI] [PubMed] [Google Scholar]

[bib4] 4.Wagner J., Wickman E., DeRenzo C., Gottschalk S. CAR T Cell Therapy for Solid Tumors: Bright Future or Dark Reality? Mol. Ther. 2020;28:2320–2339. doi: 10.1016/j.ymthe.2020.09.015. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] 5.Hou A.J., Chen L.C., Chen Y.Y. Navigating CAR-T cells through the solid-tumour microenvironment. Nat. Rev. Drug Discov. 2021;20:531–550. doi: 10.1038/s41573-021-00189-2. [DOI] [PubMed] [Google Scholar]

[bib6] 6.Glumac p.m., LeBeau a.m. The role of CD133 in cancer: a concise review. Clin. Transl. Med. 2018;7:18. doi: 10.1186/s40169-018-0198-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib7] 7.Smith L.M., Nesterova A., Ryan M.C., Duniho S., Jonas M., Anderson M., Zabinski R.F., Sutherland M.K., Gerber H.P., Van Orden K.L., et al. CD133/prominin-1 is a potential therapeutic target for antibody-drug conjugates in hepatocellular and gastric cancers. Br. J. Cancer. 2008;99:100–109. doi: 10.1038/sj.bjc.6604437. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib8] 8.Brudno J.N., Maus M.V., Hinrichs C.S. CAR T Cells and T-Cell Therapies for Cancer: A Translational Science Review. JAMA. 2024;332:1924–1935. doi: 10.1001/jama.2024.19462. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib9] 9.Sanmamed M.F., Chen L. A Paradigm Shift in Cancer Immunotherapy: From Enhancement to Normalization. Cell. 2018;175:313–326. doi: 10.1016/j.cell.2018.09.035. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib10] 10.Liu X., Ranganathan R., Jiang S., Fang C., Sun J., Kim S., Newick K., Lo A., June C.H., Zhao Y., Moon E.K. A Chimeric Switch-Receptor Targeting PD1 Augments the Efficacy of Second-Generation CAR T Cells in Advanced Solid Tumors. Cancer Res. 2016;76:1578–1590. doi: 10.1158/0008-5472.CAN-15-2524. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] 11.Liang Y., Liu H., Lu Z., Lei W., Zhang C., Li P., Liang A., Young K.H., Qian W. CD19 CAR-T expressing PD-1/CD28 chimeric switch receptor as a salvage therapy for DLBCL patients treated with different CD19-directed CAR T cell therapies. J. Hematol. Oncol. 2021;14:26. doi: 10.1186/s13045-021-01044-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] 12.Liu H., Lei W., Zhang C., Yang C., Wei J., Guo Q., Guo X., Chen Z., Lu Y., Young K.H., et al. CD19-specific CAR T Cells that Express a PD-1/CD28 Chimeric Switch-Receptor are Effective in Patients with PD-L1-positive B-Cell Lymphoma. Clin. Cancer Res. 2021;27:473–484. doi: 10.1158/1078-0432.CCR-20-1457. [DOI] [PubMed] [Google Scholar]

[bib13] 13.Qin L., Cui Y., Yuan T., Chen D., Zhao R., Li S., Jiang Z., Wu Q., Long Y., Wang S., et al. Co-expression of a PD-L1-specific chimeric switch receptor augments the efficacy and persistence of CAR T cells via the CD70-CD27 axis. Nat. Commun. 2022;13:6051. doi: 10.1038/s41467-022-33793-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib14] 14.Zhao W., Jia L., Zhang M., Huang X., Qian P., Tang Q., Zhu J., Feng Z. The killing effect of novel bi-specific Trop2/PD-L1 CAR-T cell targeted gastric cancer. Am. J. Cancer Res. 2019;9:1846–1856. [PMC free article] [PubMed] [Google Scholar]

[bib15] 15.Jiang W., Li T., Guo J., Wang J., Jia L., Shi X., Yang T., Jiao R., Wei X., Feng Z., et al. Bispecific c-Met/PD-L1 CAR-T Cells Have Enhanced Therapeutic Effects on Hepatocellular Carcinoma. Front. Oncol. 2021;11 doi: 10.3389/fonc.2021.546586. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] 16.Galluzzi L., Aryankalayil M.J., Coleman C.N., Formenti S.C. Emerging evidence for adapting radiotherapy to immunotherapy. Nat. Rev. Clin. Oncol. 2023;20:543–557. doi: 10.1038/s41571-023-00782-x. [DOI] [PubMed] [Google Scholar]

[bib17] 17.Herrera F.G., Ronet C., Ochoa de Olza M., Barras D., Crespo I., Andreatta M., Corria-Osorio J., Spill A., Benedetti F., Genolet R., et al. Low-Dose Radiotherapy Reverses Tumor Immune Desertification and Resistance to Immunotherapy. Cancer Discov. 2022;12:108–133. doi: 10.1158/2159-8290.CD-21-0003. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib18] 18.Yin L., Xue J., Li R., Zhou L., Deng L., Chen L., Zhang Y., Li Y., Zhang X., Xiu W., et al. Effect of Low-Dose Radiation Therapy on Abscopal Responses to Hypofractionated Radiation Therapy and Anti-PD1 in Mice and Patients With Non-Small Cell Lung Cancer. Int. J. Radiat. Oncol. Biol. Phys. 2020;108:212–224. doi: 10.1016/j.ijrobp.2020.05.002. [DOI] [PubMed] [Google Scholar]

[bib19] 19.Zhou X., Zhou L., Yao Z., Huang M., Gong Y., Zou B., Zhu J., Liu Y., Peng F., Zhang Y., et al. Safety and Tolerability of Low-Dose Radiation and Stereotactic Body Radiotherapy + Sintilimab for Treatment-Naive Stage IV PD-L1+ Non-Small Cell Lung Cancer Patients. Clin. Cancer Res. 2023;29:4098–4108. doi: 10.1158/1078-0432.CCR-23-0315. [DOI] [PubMed] [Google Scholar]

[bib20] 20.Antonia S.J., Villegas A., Daniel D., Vicente D., Murakami S., Hui R., Kurata T., Chiappori A., Lee K.H., de Wit M., et al. Overall Survival with Durvalumab after Chemoradiotherapy in Stage III NSCLC. N. Engl. J. Med. 2018;379:2342–2350. doi: 10.1056/NEJMoa1809697. [DOI] [PubMed] [Google Scholar]

[bib21] 21.Chang J.Y., Lin S.H., Dong W., Liao Z., Gandhi S.J., Gay C.M., Zhang J., Chun S.G., Elamin Y.Y., Fossella F.V., et al. Stereotactic ablative radiotherapy with or without immunotherapy for early-stage or isolated lung parenchymal recurrent node-negative non-small-cell lung cancer: an open-label, randomised, phase 2 trial. Lancet. 2023;402:871–881. doi: 10.1016/S0140-6736(23)01384-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib22] 22.Zhao Z., Condomines M., van der Stegen S.J.C., Perna F., Kloss C.C., Gunset G., Plotkin J., Sadelain M. Structural Design of Engineered Costimulation Determines Tumor Rejection Kinetics and Persistence of CAR T Cells. Cancer Cell. 2015;28:415–428. doi: 10.1016/j.ccell.2015.09.004. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib23] 23.Srivastava S., Riddell S.R. Engineering CAR-T cells: Design concepts. Trends Immunol. 2015;36:494–502. doi: 10.1016/j.it.2015.06.004. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib24] 24.Cheloni S., Hillje R., Luzi L., Pelicci P.G., Gatti E. XenoCell: classification of cellular barcodes in single cell experiments from xenograft samples. BMC Med. Genomics. 2021;14:34. doi: 10.1186/s12920-021-00872-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib25] 25.Savas P., Virassamy B., Ye C., Salim A., Mintoff C.P., Caramia F., Salgado R., Byrne D.J., Teo Z.L., Dushyanthen S., et al. Single-cell profiling of breast cancer T cells reveals a tissue-resident memory subset associated with improved prognosis. Nat. Med. 2018;24:986–993. doi: 10.1038/s41591-018-0078-7. [DOI] [PubMed] [Google Scholar]

[bib26] 26.Okla K., Farber D.L., Zou W. Tissue-resident memory T cells in tumor immunity and immunotherapy. J. Exp. Med. 2021;218 doi: 10.1084/jem.20201605. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] 27.Mendonsa a.m., Na T.Y., Gumbiner B.M. E-cadherin in contact inhibition and cancer. Oncogene. 2018;37:4769–4780. doi: 10.1038/s41388-018-0304-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib28] 28.Melssen M.M., Sheybani N.D., Leick K.M., Slingluff C.L., Jr. Barriers to immune cell infiltration in tumors. J. Immunother. Cancer. 2023;11 doi: 10.1136/jitc-2022-006401. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib29] 29.Shiao S.L., Gouin K.H., third, Ing N., Ho A., Basho R., Shah A., Mebane R.H., Zitser D., Martinez A., Mevises N.Y., et al. Single-cell and spatial profiling identify three response trajectories to pembrolizumab and radiation therapy in triple negative breast cancer. Cancer Cell. 2024;42:70–84.e8. doi: 10.1016/j.ccell.2023.12.012. [DOI] [PubMed] [Google Scholar]

[bib30] 30.Schalck A., Tran T., Li J., Sei E., Bai S., Hu M., Lin J., Bright S.J., Reddick S., Yang F., et al. The impact of breast radiotherapy on the tumor genome and immune ecosystem. Cell Rep. 2025;44 doi: 10.1016/j.celrep.2025.115703. [DOI] [PubMed] [Google Scholar]

[bib31] 31.Qi C., Gong J., Li J., Liu D., Qin Y., Ge S., Zhang M., Peng Z., Zhou J., Cao Y., et al. Claudin18.2-specific CAR T cells in gastrointestinal cancers: phase 1 trial interim results. Nat. Med. 2022;28:1189–1198. doi: 10.1038/s41591-022-01800-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib32] 32.Narayan V., Barber-Rotenberg J.S., Jung I.Y., Lacey S.F., Rech A.J., Davis M.M., Hwang W.T., Lal P., Carpenter E.L., Maude S.L., et al. PSMA-targeting TGFbeta-insensitive armored CAR T cells in metastatic castration-resistant prostate cancer: a phase 1 trial. Nat. Med. 2022;28:724–734. doi: 10.1038/s41591-022-01726-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] 33.Dai H., Tong C., Shi D., Chen M., Guo Y., Chen D., Han X., Wang H., Wang Y., Shen P. Efficacy and biomarker analysis of CD133-directed CAR T cells in advanced hepatocellular carcinoma: a single-arm, open-label, phase II trial. Oncoimmunology. 2020;9 doi: 10.1080/2162402X.2020.1846926. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib34] 34.Fu Q., Zheng Y., Fang W., Zhao Q., Zhao P., Liu L., Zhai Y., Tong Z., Zhang H., Lin M., et al. RUNX-3-expressing CAR T cells targeting glypican-3 in patients with heavily pretreated advanced hepatocellular carcinoma: a phase I trial. EClinicalMedicine. 2023;63 doi: 10.1016/j.eclinm.2023.102175. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib35] 35.Mao Q., Zhang Y., Fu X., Xue J., Guo W., Meng M., Zhou Z., Mo X., Lu Y. A tumor hypoxic niche protects human colon cancer stem cells from chemotherapy. J. Cancer Res. Clin. Oncol. 2013;139:211–222. doi: 10.1007/s00432-012-1310-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib36] 36.Jang J.W., Song Y., Kim S.H., Kim J., Seo H.R. Potential mechanisms of CD133 in cancer stem cells. Life Sci. 2017;184:25–29. doi: 10.1016/j.lfs.2017.07.008. [DOI] [PubMed] [Google Scholar]

[bib37] 37.Shmelkov S.V., Butler J.M., Hooper A.T., Hormigo A., Kushner J., Milde T., St Clair R., Baljevic M., White I., Jin D.K., et al. CD133 expression is not restricted to stem cells, and both CD133+ and CD133- metastatic colon cancer cells initiate tumors. J. Clin. Invest. 2008;118:2111–2120. doi: 10.1172/JCI34401. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib38] 38.Zheng Y., Wang L., Yin L., Yao Z., Tong R., Xue J., Lu Y. Lung Cancer Stem Cell Markers as Therapeutic Targets: An Update on Signaling Pathways and Therapies. Front. Oncol. 2022;12 doi: 10.3389/fonc.2022.873994. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib39] 39.Wang Y., Chen M., Wu Z., Tong C., Dai H., Guo Y., Liu Y., Huang J., Lv H., Luo C., et al. CD133-directed CAR T cells for advanced metastasis malignancies: A phase I trial. Oncoimmunology. 2018;7 doi: 10.1080/2162402X.2018.1440169. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] 40.Zhou M., Chen M., Shi B., Di S., Sun R., Jiang H., Li Z. Radiation enhances the efficacy of EGFR-targeted CAR-T cells against triple-negative breast cancer by activating NF-kappaB/Icam1 signaling. Mol. Ther. 2022;30:3379–3393. doi: 10.1016/j.ymthe.2022.07.021. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] 41.Weiss T., Weller M., Guckenberger M., Sentman C.L., Roth P. NKG2D-Based CAR T Cells and Radiotherapy Exert Synergistic Efficacy in Glioblastoma. Cancer Res. 2018;78:1031–1043. doi: 10.1158/0008-5472.CAN-17-1788. [DOI] [PubMed] [Google Scholar]

[bib42] 42.Quach H.T., Skovgard M.S., Villena-Vargas J., Bellis R.Y., Chintala N.K., Amador-Molina A., Bai Y., Banerjee S., Saini J., Xiong Y., et al. Tumor-Targeted Nonablative Radiation Promotes Solid Tumor CAR T cell Therapy Efficacy. Cancer Immunol. Res. 2023;11:1314–1331. doi: 10.1158/2326-6066.CIR-22-0840. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib43] 43.Ababneh H.S., Abramson J.S., Johnson P.C., Patel C.G. Assessing the role of radiotherapy in patients with refractory or relapsed high-grade B cell lymphomas treated with CAR T cell therapy. Radiother. Oncol. 2022;175:65–72. doi: 10.1016/j.radonc.2022.08.007. [DOI] [PubMed] [Google Scholar]

[bib44] 44.Kostopoulos N., Costabile F., Krimitza E., Beghi S., Goia D., Perales-Linares R., Thyfronitis G., LaRiviere M.J., Chong E.A., Schuster S.J., et al. Local radiation enhances systemic CAR T cell efficacy by augmenting antigen crosspresentation and T cell infiltration. Blood Adv. 2024;8:6308–6320. doi: 10.1182/bloodadvances.2024012599. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib45] 45.DeSelm C., Palomba M.L., Yahalom J., Hamieh M., Eyquem J., Rajasekhar V.K., Sadelain M. Low-Dose Radiation Conditioning Enables CAR T Cells to Mitigate Antigen Escape. Mol. Ther. 2018;26:2542–2552. doi: 10.1016/j.ymthe.2018.09.008. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib46] 46.Ni H., Reitman Z.J., Zou W., Akhtar M.N., Paul R., Huang M., Zhang D., Zheng H., Zhang R., Ma R., et al. FLASH radiation reprograms lipid metabolism and macrophage immunity and sensitizes medulloblastoma to CAR-T cell therapy. Nat. Cancer. 2025;6:460–473. doi: 10.1038/s43018-025-00905-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib47] 47.Zhang B., Hu M., Ma Q., Li K., Li X., He X., Shu P., Chen Y., Gao G., Qin D., et al. Optimized CAR-T therapy based on spatiotemporal changes and chemotactic mechanisms of MDSCs induced by hypofractionated radiotherapy. Mol. Ther. 2023;31:2105–2119. doi: 10.1016/j.ymthe.2023.04.008. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib48] 48.Oliveira G., Egloff a.m., Afeyan A.B., Wolff J.O., Zeng Z., Chernock R.D., Zhou L., Messier C., Lizotte P., Pfaff K.L., et al. Preexisting tumor-resident T cells with cytotoxic potential associate with response to neoadjuvant anti-PD-1 in head and neck cancer. Sci. Immunol. 2023;8 doi: 10.1126/sciimmunol.adf4968. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib49] 49.Jung I.Y., Noguera-Ortega E., Bartoszek R., Collins S.M., Williams E., Davis M., Jadlowsky J.K., Plesa G., Siegel D.L., Chew A., et al. Tissue-resident memory CAR T cells with stem-like characteristics display enhanced efficacy against solid and liquid tumors. Cell Rep. Med. 2023;4 doi: 10.1016/j.xcrm.2023.101053. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib50] 50.Lesch S., Blumenberg V., Stoiber S., Gottschlich A., Ogonek J., Cadilha B.L., Dantes Z., Rataj F., Dorman K., Lutz J., et al. T cells armed with C-X-C chemokine receptor type 6 enhance adoptive cell therapy for pancreatic tumours. Nat. Biomed. Eng. 2021;5:1246–1260. doi: 10.1038/s41551-021-00737-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib51] 51.Ray M.E., Mehra R., Sandler H.M., Daignault S., Shah R.B. E-cadherin protein expression predicts prostate cancer salvage radiotherapy outcomes. J. Urol. 2006;176:1409–1414. doi: 10.1016/j.juro.2006.06.014. [DOI] [PubMed] [Google Scholar]

[bib52] 52.Krischak A., Kowaliuk J., Sarsarshahi S., Dörr W., Kleiter M. Effect of irradiation on the expression of E-cadherin and beta-catenin in early and late radiation sequelae of the urinary bladder and its modulation by NF-kappaB inhibitor thalidomide. Strahlenther. Onkol. 2021;197:537–546. doi: 10.1007/s00066-021-01751-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib53] 53.Corgnac S., Boutet M., Kfoury M., Naltet C., Mami-Chouaib F. The Emerging Role of CD8(+) Tissue Resident Memory T (T(RM)) Cells in Antitumor Immunity: A Unique Functional Contribution of the CD103 Integrin. Front. Immunol. 2018;9:1904. doi: 10.3389/fimmu.2018.01904. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib54] 54.Muller Y.D., Nguyen D.P., Ferreira L.M.R., Ho P., Raffin C., Valencia R.V.B., Congrave-Wilson Z., Roth T.L., Eyquem J., Van Gool F., et al. The CD28-Transmembrane Domain Mediates Chimeric Antigen Receptor Heterodimerization With CD28. Front. Immunol. 2021;12 doi: 10.3389/fimmu.2021.639818. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib55] 55.Xie B., Li Z., Zhou J., Wang W. Current Status and Perspectives of Dual-Targeting Chimeric Antigen Receptor T-Cell Therapy for the Treatment of Hematological Malignancies. Cancers (Basel) 2022;14 doi: 10.3390/cancers14133230. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib56] 56.Sun L., Chen B., Liu X., Zhu Y., Zhang G., Liang X., Xing L., Xu W., Jiang S., Wang X. Alpaca-derived nanobody targeting the hydrophobic pocket of the HIV-1 gp41 NHR broadly neutralizes HIV-1 by blocking six-helix bundle formation. Curr. Res. Microb. Sci. 2024;7 doi: 10.1016/j.crmicr.2024.100263. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib57] 57.Wang H., Yao Z., Kang K., Zhou L., Xiu W., Sun J., Xie C., Yu M., Li Y., Zhang Y., et al. Preclinical study and phase II trial of adapting low-dose radiotherapy to immunotherapy in small cell lung cancer. Med. 2024;5:1237–1254.e9. doi: 10.1016/j.medj.2024.06.002. [DOI] [PubMed] [Google Scholar]

[bib58] 58.Bankhead P., Loughrey M.B., Fernández J.A., Dombrowski Y., McArt D.G., Dunne P.D., McQuaid S., Gray R.T., Murray L.J., Coleman H.G., et al. QuPath: Open source software for digital pathology image analysis. Sci. Rep. 2017;7 doi: 10.1038/s41598-017-17204-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib59] 59.Brugmann A., Eld M., Lelkaitis G., Nielsen S., Grunkin M., Hansen J.D., Foged N.T., Vyberg M. Digital image analysis of membrane connectivity is a robust measure of HER2 immunostains. Breast Cancer Res. Treat. 2012;132:41–49. doi: 10.1007/s10549-011-1514-2. [DOI] [PubMed] [Google Scholar]

[bib60] 60.Luo H., Xia X., Huang L.B., An H., Cao M., Kim G.D., Chen H.N., Zhang W.H., Shu Y., Kong X., et al. Pan-cancer single-cell analysis reveals the heterogeneity and plasticity of cancer-associated fibroblasts in the tumor microenvironment. Nat. Commun. 2022;13:6619. doi: 10.1038/s41467-022-34395-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib61] 61.Luo H., Xia X., Kim G.D., Liu Y., Xue Z., Zhang L., Shu Y., Yang T., Chen Y., Zhang S., et al. Characterizing dedifferentiation of thyroid cancer by integrated analysis. Sci. Adv. 2021;7 doi: 10.1126/sciadv.abf3657. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib62] 62.Cao M., Deng Y., Deng Y., Wu J., Yang C., Wang Z., Hou Q., Fu H., Ren Z., Xia X., et al. Characterization of immature ovarian teratomas through single-cell transcriptome. Front. Immunol. 2023;14 doi: 10.3389/fimmu.2023.1131814. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Dual-targeting CD133/PD-L1 CAR-T plus αPD-1 overcomes immunosuppressive microenvironment and enhanced by radiation pre-conditioning

Zhuoran Yao

Kai Kang

Pei-Heng Li

Limei Yin

Ruizhan Tong

Linglu Yi

Yonghong Song

Ren Luo

Yijun Wu

Shanghai Liu

Zichong Peng

Xianming Mo

Wenbo Wang

Jianxin Xue

You Lu

Abstract

Graphical abstract

Introduction

Results

Dual CD133/PD-L1 targeting CAR-T shared one single CD3ζ chain to transduce activation signals

Figure 1.

Dual CAR-T showed improved cytotoxic function against CD133+ tumor cells regardless of basal PD-L1 expression

Figure 2.

PD-1 blockade enhanced antitumor immunity of dual-targeting CD133/PD-L1 CAR-T

Figure 3.

Radiotherapy further synergized with CAR-T therapy and triple combination reached maximal tumor control

Figure 4.

Triple-therapy-induced intratumoral CD103+ CXCR6+ tissue-resident memory CD8+ CAR-T cell expansion

Figure 5.

E-cadherin-CD103 appears as a vital ligand receptor between tumor cells and CD103+ CXCR6+ CD8+ CAR-Trm after irradiation

Figure 6.

Irradiation enhances the E-cadherin-CD103 interaction between tumor cells and CD8+ T cells in clinical samples

Figure 7.

Discussion

Conclusions

Materials and methods

Ethics statement

Plasmid construction and lentiviral production

CAR-T manufacturing

Generation of cancer cell lines

In vitro cytolysis and activation experiments

Flow cytometry assay

Cell Trace Violet proliferation assay

Western blotting analysis

In vitro imaging of CAR-T cytolysis and confocal imaging

Organoid construct and killing assay

Transwell study

Xenograft mouse models

Tumor growth and treatment

Copy number of CAR-T cells and cytokine detection in mouse peripheral blood

Histological staining and data analysis

Tissue microarray and immunohistochemical staining

scRNA-seq and analysis of mouse samples

scRNA-seq and analysis of clinical samples

Statistical analysis

Data and code availability

Acknowledgments

Author contributions

Declaration of interests

Footnotes

Contributor Information

Supplemental information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Dual CAR-T showed improved cytotoxic function against CD133⁺ tumor cells regardless of basal PD-L1 expression

Triple-therapy-induced intratumoral CD103⁺ CXCR6⁺ tissue-resident memory CD8⁺ CAR-T cell expansion

E-cadherin-CD103 appears as a vital ligand receptor between tumor cells and CD103⁺ CXCR6⁺ CD8⁺ CAR-Trm after irradiation