Dual targeting of CD155 augments the antitumor efficacy of ROR1-CAR-T cells in ovarian cancer

Abstract

Background

Exploring novel therapeutic targets and developing targeted therapies constitute an urgent clinical need for improving the prognosis of ovarian cancer (OC), particularly among patients with advanced stages. Currently, chimeric antigen receptor T (CAR-T) cell therapy has been demonstrated to have a remarkable therapeutic effect in hematological malignancies, while its application remains limited in OC due to the absence of appropriate target molecules and the complex immunosuppressive tumor microenvironment (TME). Poliovirus receptor (PVR, CD155) has been the subject of extensive research in the field of regulatory molecules within the immune microenvironment. However, there has been a paucity of research investigating its role in OC. Receptor tyrosine kinase-like orphan receptor 1 (ROR1) is barely expressed in normal tissues but widely expressed in tumor tissues, making it a promising target for CAR-T therapy. Nevertheless, the potential effectiveness of CAR-T cell targeting ROR1 in OC remains unknown. Therefore, the purpose of this study is twofold: The primary objective of this study is to investigate the potential efficacy of single-target ROR1-CAR-T cells on OC. The secondary objective is to examine the feasibility of CD155 as an immunotherapy target for OC and to determine whether combined targeting of CD155 can enhance the function of ROR1-CAR-T cells in OC.

Method

ROR1 and CD155 expression were detected via flow cytometry analysis. In vitro experiments were conducted to explore the regulatory effect of CD155 on OC proliferation, invasion, angiogenesis, and T cell function. ROR1-CAR, CD155-CAR, and ROR1/CD155 bispecific CAR constructs were designed and synthesized. Then, they were introduced into T cells using lentiviral particles to generate CAR-T cells. We subsequently validated the synergistic effects of CD155 in ROR1/CD155 bispecific CAR-T cells based on cytotoxic efficacy, activation, exhaustion, and differentiation status.

Results

ROR1-CAR-T cells exhibited tumoricidal activity in OC, but elevated tonic signaling was observed, resulting in rapid depletion. CD155 constitutes an ideal therapeutic target in OC: firstly, ubiquitous CD155 expression in OC cell lines. Secondly, CD155 promotes tumor proliferation, migration, and angiogenesis in OC cell lines, acting as an oncogenic driver. Thirdly, CD155 impairs T cell function and accelerates their depletion, contributing to an immunosuppressive TME. The bispecific CAR-T combined targeting CD155 and ROR1 demonstrated superior cytotoxicity compared to single-target ROR1-CAR-T or CD155-CAR-T. Co-targeting CD155 significantly attenuated tonic signaling and delayed CAR-T cell exhaustion.

Conclusion

CD155 emerges as a promising therapeutic target for CAR-T therapy in OC. The bispecific CAR-T construct that co-targets CD155 and ROR1 demonstrates superior and durable tumoricidal activity, offering new perspectives on OC targeted therapy.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00262-025-04283-x.

Introduction

Receptor tyrosine kinase-like orphan receptor 1 (ROR1) is a transmembrane protein that functions primarily through the atypical Wnt signaling pathway, thereby promoting the proliferation and invasion of tumor cells [1]. ROR1 is broadly expressed during embryonic development, while it exhibits minimal or undetectable expression in normal organs or tissues in adults [2]. Current research demonstrates it highly overexpressed across diverse malignancies including chronic lymphocytic leukemia, breast cancer, OC, melanoma, and lung adenocarcinoma [3, 4]. This tumor-associated expression profile establishes ROR1 as a promising therapeutic target. Zilovertamab vedotin (an antibody–drug conjugate targeting ROR1) showed antitumor effectiveness and safety in lymphoid cancer patients [5].

In solid tumors, Zilovertamab combined with paclitaxel has demonstrated safety and tolerability in a Phase Ⅰb clinical trial for advanced breast cancer [6]. ROR1-targeted therapy demonstrates considerable promise for clinical application. However, extant research on ROR1 in OC remains limited, primarily focusing on its biological behavior and expression patterns. Zhang et al. [7] demonstrated significantly elevated ROR1 expression in tumor tissues compared to normal ovarian tissues in patients, identifying ROR1 as an independent prognostic factor for overall survival (OS). Furthermore, ROR1 was validated to augment the tumorigenic capacity of cancer stem cells in OC, with elevated ROR1 expression correlating with increased recurrence rates compared to low-expression counterparts [8]. Collectively, these findings establish ROR1 as a potential immunotherapeutic target for OC. Nevertheless, ROR1-targeted therapeutic strategies remain underexplored. Investigating the therapeutic potential of ROR1 targeting in OC may offer novel perspectives for clinical management.

Chimeric antigen receptor T cell (CAR-T) therapy involves the ex vivo genetic engineering of T cells to integrate a synthetic receptor comprising a specific antigen-recognition domain and T cell activation motifs. Upon contact with tumor cells, these modified T cells specifically recognize and bind to surface tumor antigens, thereby triggering their activation. Subsequently, they mediate direct cytolysis through the release of effector molecules including IFN-γ, perforin, and GZMB [9]. Current clinical data demonstrate remarkable efficacy of CAR-T cell therapy in hematological malignancies, establishing it as a promising treatment for relapsed–refractory hematopoietic cancers [10–12]. This success has prompted investigations into CAR-T applications for solid tumors. Frustratingly, CAR-T cell therapy remains constrained in solid tumors due to multiple barriers including: (1) the complex immunosuppressive tumor microenvironment (TME) [13, 14], (2) inadequate CAR-T cell infiltration into tumor parenchyma [15], and (3) rapid T cell exhaustion upon repeated antigen exposure [16]. Consequently, an optimal target for solid tumor CAR-T therapy would exhibit tumor-restricted overexpression, absence in normal tissues, and active involvement in modulating the TME. Poliovirus receptor (PVR, CD155), a transmembrane glycoprotein member of the immunoglobulin superfamily (IgSF), has been demonstrated to regulate both immune cell functions [17, 18] and angiogenesis [19]. Notably, CD155 exhibits significantly elevated levels across multiple malignancies including lung cancer [20], digestive system malignancies [21–24], breast cancer [25], as well as head and neck squamous cell carcinomas [26]—while exhibiting minimal or undetectable expression in normal tissues. This expression profile positions CD155 as a promising target for cancer immunotherapy. However, the biological functions of CD155 in OC remain incompletely defined. Consequently, the objective of this study is twofold: first, to investigate the functional efficacy of ROR1-CAR-T cells in OC; and second, to determine whether CD155 constitutes an optimal immunotherapeutic target for OC and assess whether dual targeting of CD155 enhances the antitumor activity of single-target ROR1-CAR-T therapy.

This study utilizes in vitro assays to demonstrate the feasibility of ROR1-CAR-T therapy in OC, evidenced by its potent cytolytic efficacy, and its limitations, particularly the rapid exhaustion of cells due to elevated tonic signaling. Functional experiments established CD155 as an oncogenic driver in OC, mechanistically involved in promoting the proliferation of OC cells, suppressing T cell function, and promoting angiogenesis, thereby validating its potential as an immunotherapeutic target. Furthermore, the bispecific CAR-T co-targeting CD155 and ROR1 was constructed, and rigorous in vitro characterization confirmed that CD155 co-targeting enhances tumoricidal activity while mitigating T cell exhaustion. Collectively, our findings substantiate the therapeutic viability of dual ROR1/CD155 targeting in OC CAR-T therapy and demonstrate the superiority of bispecific CAR architecture, providing novel mechanistic insights for advancing CAR-T interventions in this malignancy.

Materials and methods

Cell lines and treatment

The human leukemia T lymphocyte cells (Jurkat) and human OC cell lines (SKOV3, HEY) were purchased from Procell Life Science & Technology Co., Ltd. Cell lines CaOV3, MDA-MB-231, HEK 293 T, HT-29, COLO-678 were all obtained from the Cell Bank of the Chinese Academy of Sciences. Cell lines SKOV3, HEY, CaOV3, and HEK 293 T were cultured in high-glucose Dulbecco’s modified Eagle medium (DMEM; Gibco, USA). Jurkat cells and COLO-678 cells were treated with RPMI-1640 medium (Gibco, USA). MDA-MB-231 cells were cultivated in Leibovitz's L-15 medium (Gibco, USA). HT-29 cells were cultured in McCoy's 5A medium (Gibco, USA). All culture medium was supplemented with 10% fetal bovine serum (FBS; Gibco, USA) and 1% penicillin–streptomycin (P/S; Servicebio, China). All cell lines were cultivated at 37℃ in a 5% CO₂ incubator. All target cells were modified to express the fusion protein firefly luciferase.

The SKOV3 and HEY cells in the overexpressing group were infected with pcDNA3.1-CD155-overexpressing plasmids (Public Protein/Plasmid Library), while the control group was treated with an equivalent dosage of Lipofectamine 2000 reagent (Invitrogen, USA). The lentiviral vector encoding ROR1 protein (pCMV3-ROR1-FLAG) (SinoBiological, HG13968-CF) was transduced to SKOV3 cells to generated SKOV3-ROR1 cells.

Flow cytometry

The flow cytometric analyses in this research were conducted using the Beckman Coulter flow cytometer, and the FlowJo software (version 10) was employed for data analysis. Prior to staining, cells were washed with PBS (Gibco, USA) containing 1% BSA (Servicebio, China). Subsequently, cell staining was carried out at a temperature of 4 °C for a duration of 30 min, without exposure to light. Next, cells were subjected to two rounds of washing, followed by resuspension in 100 μL of PBS containing 1% BSA. The samples were finally analyzed by the flow cytometer. Antibodies used in this research for flow cytometric analyses are listed in Table S1.

RNA Extraction and Real-Time qPCR

The total RNA was extracted using the total RNA extraction reagent (ABclonal, China). Subsequently, the concentration and purity of RNA were detected. The ABScript III RT Master Mix for qPCR (ABclonal, Cat#RK20428) was utilized for the reverse transcription process. The SYBR Green qPCR Mix (ABclonal, Cat#RK21219) was employed to conduct the RT-qPCR, with ACTB serving as an internal control. The relative mRNA expression levels were determined using the 2^−△Ct method [27]. Table S2 presents a comprehensive list of the primers employed in this research project.

Western blotting (WB)

The lysis of cells was achieved through the utilization of a RIPA buffer containing the protease and phosphatase inhibitor (NCM Biotech, China). The bicinchoninic acid method was used to detect the protein concentrations. Subsequent to undergoing a thermal treatment at 100 °C for a duration of 10 min, the extracted proteins were subjected to 10% SDS-PAGE (Servicebio, China) and transferred to polyvinylidene fluoride membranes (Servicebio, China). Then, the membranes were subjected to a blocking procedure involving the application of 5% lipid-free milk (Servicebio, China) for a duration of one and a half hours. Next, the samples were subjected to treatment with primary antibodies (Table S3) at 4 °C for 12 h. Subsequently, the membranes were washed and treated with secondary antibodies (ABclonal, China) for a period of 2 h. Finally, the protein index was visualized using enhanced chemiluminescence reagents (Epizyme, China).

ELISA assay

The concentration of GZMB, IFN-γ, and TNFα was detected using ELISA kits specific for GZMB (Abclonal, Cat#RK00089), IFN-γ (Abclonal, Cat#RK00015), and TNFα (Abclonal, Cat#RK00030), respectively. All procedures were carried out according to the manufacturer’s instructions.

Cell proliferation assays

Cell counting kit-8 (CCK-8) assays were employed to compare viability of cells in distinct groups. Prior to the detection, 100μL of cell suspension (with a density of 3 × 10⁴/mL) was added into a 96-well plate (Corning, NY, USA). Then, 10 µL of CCK-8 reagent (Vazyme, China) was added to each well and incubated for a period of 2 h. Ultimately, the optical density (OD) was detected at a wavelength of 450 nm.

The capacity for proliferation was assessed through 5-Ethynyl-20-deoxyuridine (EdU) assays, employing the Click-iT EdU-555 Kit (Servicebio, China). A total of 2 × 10⁴ cells per well were seeded into the 96-well plate and cultured for 24 h. Subsequently, the EdU solution was introduced into the culture medium, and the cells were incubated for two hours. Subsequently, 4% paraformaldehyde was utilized to fix the cells, and then the cells were permeabilized with 0.3% Triton X-100 (Servicebio, China). The Click solution was prepared according to the instructions outlined in the manual and utilized for the incubation of cells for a duration of 30 min. Next, the DAPI solution (Servicebio, China) was employed to counterstain the cell nuclei. The visualization of cells was accomplished through the utilization of an inverted fluorescence microscope (Carl Zeiss, Germany).

Cell migration and invasion assays

Transwell assays were applied for the evaluation of cell migration capacity. Initially, 150 µL of cells (with a density of 2 × 10⁵/mL) was added to the upper chamber of the Transwell plate (Corning, Cat#3422), while the lower chamber was filled with 800 µL of culture medium. Following a 24-h incubation period, the lower chamber and supernatant in the upper chamber were removed. The polycarbonate membrane was rinsed twice with PBS and fixed with 4% paraformaldehyde for 15 min. Next, the polycarbonate membrane was stained with crystal violet for 15 min and then rinsed twice with PBS. Subsequently, the moisture on the polycarbonate membrane was dried, and the upper charmer was observed under a microscope, with five fields chosen randomly for imaging.

Wound-healing assays were also employed to assess the invasion capacity of cells. Initially, cells were plated into six-well plates. Following the attainment of 95% confluence, a scratch was made using a 200 µL pipette tip. Subsequently, the adherent cells were cultivated in serum-free culture medium, and images were captured at 0, 24, and 48 h post-scratching. This procedure was performed to evaluate the wound area, which is an indicator of cell invasion capacity.

Plasmid and lentivirus production

DNA sequences of anti-ROR1 CAR, anti-CD155 CAR, and dual-targeting CD155/ROR1 CAR were synthesized by OriCell Therapeutics Co. Ltd. (Shanghai, China) and cloned into the lentiviral vector. Plasmids mentioned above were delivered into the HEK 293 T cells using the four-plasmid system for 48 h. Then, supernatant was collected and centrifuged at 3000 × g for 10 min. Then, lentivirus supernatant was sorted at − 80 °C.

Generation of CAR-T cells

Human peripheral blood mononuclear cells (PBMC) from healthy donors were purchased from Hycells Biotechnology Co., Ltd. (Shanghai, China) and isolated by density gradient centrifugation (STEMCELL Technologies, Canada). Subsequently, CD3⁺ T cells were separated using magnetic bead separation (Miltenyi, Germany). Then, CD3⁺ T cells were activated by CD3/CD28 dynabeads (Gibco, USA) and cultivated in X-VIVO¹⁵ culture medium (Lonza, Switzerland) supplemented with 5% FBS and 200U/mL IL-2. Next, CD3⁺ T cells were transduced with lentivirus for 24 h. After the activation of 5 days, dynabeads were extracted from the culture medium, and after the transduction of lentivirus for 7 days, the CAR expression was detected by flow cytometric analyses.

Cytotoxicity assay

Cytotoxicity assays based on bioluminescence were used to evaluate the tumoricidal capacity of CAR-T cells. Target cells were engineered to express luciferase ahead before. A co-culture system consisting of CAR-T/Mock-T cells and tumor cells was constructed and was designated as the test well. Firstly, 100 µL of tumor cell suspension (with the density of 1 × 10⁵/mL) was added into each well of the 96-well plate. Then, 100 µL of CAR-T cells or Mock-T cells was added into tumor cell suspension, respectively, according to effector-to-target (E:T) ratio. Furthermore, the establishment of control wells, which exclusively comprise target cells, has been implemented. Following a 24-h co-culture period, 100 µL of supernatant was extracted from each well, and 100 µL of luciferase reagent (Promega, USA) was added into each well without exposure to the light. Ultimately, the OD value was detected at a wavelength of 560 nm. The calculation of the cytotoxic effect was performed using the following formula:

$$\frac{OD\left(C,o,n,t,r,o,l,w,e,l,l\right)-OD(Testwell)}{OD(Controlwell)}\times 100\%$$

Cytokine release assay

Following a 24-h co-culture with an E:T ratio of 1:1, the supernatant was harvested to measure the concentrations of IFN-γ and IL-2 using a cytokine detection kit that is specific for IFN-γ (R&D, Cat#DY285B) and IL-2 (R&D, Cat#DY202-05). All procedures were performed in accordance with the manufacturer's instructions.

Repeated antigen stimulation assay

In order to emulate the in vivo context in which CAR-T cells encounter antigen stimulation, a series of in vitro repeated antigen stimulation assays were conducted. A co-culture system was established, consisting of CAR-T cells and SKOV3 cells at a ratio of 1:1. Subsequent to a 3-day co-culture period, the CAR-T cells were collected and washed with PBS. Then, the CAR-T cells were resuspended in X-VIVO¹⁵ medium and enumerated using an AO/PI reagent. Next, the collected CAR-T cells were plated into a new 6-well plate, while an equivalent number of target cells were added into the CAR-T cell suspension and co-cultured for an additional 3 days. At the conclusion of each round, the CAR-T cells were collected for subsequent enumeration and detection.

CIBERSORT analysis

Clinical information and expression profile of OC samples were extracted from TCGA database. CIBERSORT algorithm (https://cibersortx.stanford.edu/) [28] was utilized to assess the infiltration of immune cells of each OC sample. Initially, OC samples were stratified into CD155-high and CD155-low clusters, with the median value of CD155 expression as the cut off. Subsequently, infiltration differences between two clusters were evaluated using R studio software, and Wilcoxon rank-sum test was applied for the significance analyses.

Gene set enrichment analysis (GSEA)

The R packages “org.Hs.eg.db” and “clusterProfiler” [29] were applied for the GSEA analyses. The signal pathway that satisfied both the normalized enrichment score (NES) exceeding 1 and a p value of less than 0.05 was identified as enriched.

scRNA-seq data processing

The R software was used to analyze the 10 × scRNA-seq data from GSE184880 (including samples: GSM5599220, GSM5599221, GSM5599222, GSM5599223, and GSM5599224). Firstly, the “Seurat” R package [30] was used to transform the scRNA-seq data from GSE184880 into a Seurat object. Low quality samples (proportions of mitochondrial genes in excess of 25% and proportions of erythrocyte genes in excess of 1%) were removed. Besides, cells with total cellular gene transcripts (nCount) exceeding 25,000 and total cellular gene counts (nFeature) below 200 or above 6000 were excluded. Then, the “LogNormalize” method was applied to normalize the data, and the “FindVariableFeatures” was exploited to select 2000 highly variable genes. Next, the “RunPCA” method was utilized to reduce the dimensionality of the 2000 highly variable genes. Subsequently, the “ElbowPlot” method and the “t-SNE” method were applied for the purpose of dimensionality reduction as well as cluster identification. The latter process yielded 21 principal components (PCs) for subsequent analysis. Finally, the “FindAllMarkers” method was utilized to identify differentially expressed markers among the various clusters and the cell populations were annotated using the “SingleR” algorithm [31].

Statistics

Statistical analyses in this research were conducted using R studio (version 4.2.1) and GraphPad Prism 9. Differences between two groups were assessed using unpaired student’s t tests, and p value < 0.05 was considered statistically significant. Statistical significance was defined as follows: ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05; ns p > 0.05.

Results

ROR1-CAR-T cells efficiently target OC cells but exhibit excessive tonic signaling

To elucidate the application significance of targeted ROR1 in OC, we initially detected the expression of ROR1 in OC cell lines by means of flow cytometry. The results of the flow cytometric analyses revealed the expression of ROR1 in all the examined cell lines. Among these, SKOV3-ROR1 exhibited the highest expression levels, while CaOV3 displayed the lowest (Fig. 1A). The structure of lentiviral vector encoding ROR1-specific CAR is presented in Fig. 1B, and a ROR1-directed single chain variable fragment was fused to a Myc tag to allow for CAR detection. Following lentivirus transduction, CAR expression was detected by flow cytometric analyses utilizing a c-Myc tag antibody (Fig. 1C). According to the results of cytotoxicity assays, ROR1-CAR-T cells exhibited a notable tumoricidal capacity in OC cells with varying ROR1 expression levels, particularly when the effector-to-target ratio was maintained at 1:1 (Fig. 1D). However, cytokine release assays demonstrated ROR1-CAR-T cells secreted high levels of IFN-γ in the absence of target cell stimulation, indicating excessive tonic signaling (Fig. 1E). In general, CAR-T cells with excessive tonic signaling undergo rapid depletion, which hinders their ability to continuously exert their tumor-eliminating effects [32]. Consequently, modifications to the CAR structure are imperative to optimize the functionality of ROR1-CAR-T. Recent studies have indicated that patients diagnosed with hematological malignancies who have undergone single-target CAR-T cell therapy have demonstrated antigen escape [33]. The application of dual-target CAR-T has been demonstrated to effectively mitigate the occurrence of antigen escape [34, 35]. In consideration of the phenomenon of antigen escape, we hypothesize that the modification of the ROR1-CAR-T structure is attained through the incorporation of another target.

Fig. 1.

ROR1-CAR-T cells efficiently target OC cells but exhibit excessive tonic signaling. A Flow cytometric analyses revealed the expression level of ROR1 in OC cell lines. B Schematic illustration of a ROR1-CAR vector. C Transgene expression of the CAR constructs after retroviral transduction of primary human T cells. D The cytotoxic effect of ROR1-CAR-T cells was assessed on OC cell lines exhibiting varying effector-to-target (E:T) ratios through luciferase assays (p value: ***p < 0.001, **p < 0.01, *p < 0.05, ns p > 0.05). E ELISA results showed IFN-γ concentration of ROR1-CAR-T/Mock-T cells in the absence of target cell stimulation (p value: ***p < 0.001).

CD155 exerts positive regulatory effect on the proliferation and migration of OC

CD155 has been the focus of extensive research in the field of immune regulation. However, the biological behavior of CD155 in OC remains unclear. Firstly, survival analysis was employed to elucidate its effect on overall survival (OS) of OC. As illustrated in Fig. 2A, OC patients exhibiting low CD155 expression were more likely to possess prolonged OS, suggesting that the function of CD155 may be pro-cancerous. In order to elucidate the feasibility of targeting CD155 in OC, the expression of CD155 in OC cell lines was detected by flow cytometry, and the results demonstrated that CD155 was expressed in all the OC cell lines detected (Fig. S1A). Besides, given the potential for on-target/off-tumor effects, the expression of CD155 and ROR1 was evaluated through scRNA-seq data analysis. Initially, scRNA-seq data of five cases of non-tumor ovarian tissues were collected. Following the quality control, dimensionality reduction and cluster identification, a total of 26,841 cells were grouped into 8 distinct cell clusters, comprising smooth muscle cells, T cells, NK cells, endothelial cells, monocytes, mesenchymal stem cells, B cells, and fibroblasts (Fig. S1B). It can be observed that the expression of CD155 is minimal in the cells of normal tissues (Fig. S1C). Furthermore, ROR1 is almost undetectable in non-tumor cells (Fig. S1D). In addition to OC, differential analysis revealed upregulated CD155 expression in malignancies including cholangiocarcinoma and gastric carcinoma, implying its potential as multi-cancer target (Fig. S2). In vitro experiments were performed to investigate the regulation of CD155 on the proliferation and migration of OC cells. Subsequent to the transfection of CD155-overexpressing plasmid, the overexpression efficacy of OC cells was detected by WB assays (Fig. 2B). CCK-8 assays were conducted to assess the viability of OC cells in distinct groups. As demonstrated in Fig. 2C, the overexpression of CD155 was observed to markedly increase the viability of OC cells. Besides, the results of EdU assays indicated that the overexpression of CD155 enhanced the proliferation capacity of OC cells (Fig. 2E). The migratory capacity of OC cells within distinct groups was evaluated through the implementation of Transwell assays and wound-healing assays. The CD155-overexpression group exhibited an augmented migratory capacity as determined by the Transwell assay (Fig. 2D). Furthermore, the wound-healing assay revealed that the CD155-overexpression group exhibited a greater migratory capacity compared to the control group (Fig. 2F). All these results collectively indicated that CD155 is a pro-cancer molecule that plays a positive regulatory role in the proliferation and migration of OC cells. The potential for the treatment of OC through the inhibition of CD155 represents a promising avenue for further research. More importantly, CD155 is widely expressed in OC cell lines, suggesting that targeting of CD155 is feasible.

Fig. 2.

CD155 exerts positive regulatory effect on the proliferation and migration of OC. A Survival analysis for OC patients with different CD155 expression, p = 0.0063 by the log-rank test. B WB results showed the CD155 protein level in both control group and overexpressed group derived from HEY and SKOV3, respectively. C It is shown that the HEY cells (at times of 24, 48, 72, and 96 h) and SKOV3 cells (at times of 24, 48, and 72 h) living in the CD155-overexpressed group are always significantly greater than the normal control group by CCK-8 (p value: ****p < 0.0001, *p < 0.05). D Transwell assay conducted on HEY and SKOV3 cell lines; in each group, we observed the chamber bottom under 10 × stereomicroscope and randomly chose five visual fields and performed cell counting through Image J software. The differences in the invaded cell number between the two groups were taken into statistics (p value: ****p < 0.0001, **p < 0.01). E EdU assay was used to detect the proliferation rate in two hours of each group of HEY and SKOV3 cells. Images were taken under 10 × fluorescence microscope (p value:**p < 0.01, *p < 0.05). Scale bar, 100 µm. F Wound-healing assays accessed the differences in the migration ability of HEY cells and SKOV3 cells between the normal control group and the overexpression group. Scale bar, 100 µm.

CD155 impedes functions of T cells

In addition to its role in the regulation of OC cells, CD155 has also been implicated in the modulation of T cell function within the OC microenvironment. A co-culture system consisting of Jurkat cells and OC cells with a ratio of 5:1 was established to investigate its effect on T cells. Following a 24-h co-culture period, OC cells overexpressing CD155 significantly downregulated the activation level of Jurkat cells in comparison with the control group (Fig. 3A, B). Besides, Jurkat cells co-cultured with OC cells overexpressing CD155 exhibited a diminished secretion of cytotoxic effectors, including IFN-γ, TNFα, and perforin (Fig. 3E). More importantly, an elevated level of PD-1, TIM-3, LAG-3, and TIGIT was observed in Jurkat cells co-cultured with OC cells overexpressing CD155 after a 24-h co-culture period (Fig. 3C, D), indicating that CD155 contributed to the exhaustion of T cells. Following a prolonged co-culture period (48 h), the results of the WB assay revealed that the expression of CD69 in Jurkat cells from the overexpressing group was reduced, while the expression of PD-1 was augmented (Fig. 3F). The aforementioned results indicate that CD155 hinders T cell activation and accelerates T cell depletion, thereby participating in the regulation that impedes T cell function. Consequently, the efficacy of targeted CD155 therapy may be attributed to its capacity to restore the function of T cells within the TME.

Fig. 3.

CD155 impedes functions of T cells. A, B RT-qPCR tests showed the transcription level of T cell activation markers in Jurkat cells co-cultured with distinct groups of HEY cells (A) and SKOV3 cells (B), respectively (p value: ***p < 0.001, **p < 0.01, *p < 0.05, ns p > 0.05). C, D RT-qPCR tests showed the transcription level of T cell exhaustion markers in Jurkat cells co-cultured with distinct groups of HEY cells (C) and SKOV3 cells (D), respectively (p value: ***p < 0.001, **p < 0.01, *p < 0.05, ns p > 0.05). E Results of ELISA assays indicated a diminished secretion of IFN-γ, TNFα, and perforin of Jurkat cells co-cultured with SKOV3 cells overexpressing CD155 (p value: ***p < 0.001, *p < 0.05). F WB assays demonstrated the protein level of CD69 and PD-1 of Jurkat cells in distinct groups.

CD155 is involved in inducing angiogenesis in OC

The intricate tumor microenvironment comprises a variety of components, including immune cells, tumor cells, fibroblasts, blood vessels, and the extracellular matrix. Current research has revealed that the aberrant proliferation of blood vessels within the TME hinders the recruitment of T cells into the tumor parenchyma [36]. Thus, in addition to the regulation of tumor cells and T cell functions, bioinformatics analyses and in vitro experiments were conducted to evaluate the effect of CD155 on tumor angiogenesis in OC. The CIBERSORT analysis revealed a lower proportion of CD8⁺ T cells in the CD155-low cluster (Fig. 4A), which might be attributed to the upregulation of angiogenesis and the subsequent hindrance of T cell infiltration by CD155. Besides, results of the GSEA analysis illustrated that the “VEGF signaling pathway” was significantly enriched in the CD155-high cluster (Fig. 4B). Tumor cells continuously facilitate angiogenesis to meet their own proliferation and metastasis needs, manifested as the upregulation of VEGF family genes and the activation of related signaling pathways [37]. CD31 has been found to be involved in the processes of endothelial cell proliferation and migration, thereby promoting tumor angiogenesis [38]. In this research, results of WB assays demonstrated an elevated levels of VEGF and CD31 in OC cells overexpressing CD155 (Fig. 4C, D). All these results indicated CD155 facilitated angiogenesis in OC. Consequently, it can be deduced that CD155 targeted therapy exerts an anti tumor effect by down-regulating abnormal angiogenesis in the TME and subsequent facilitating T cells infiltration in the tumor parenchyma.

Fig. 4.

CD155 is involved in inducing angiogenesis in OC. A Violin plot showed the ratio differentiation of 22 types of immune cells between OC cases with high or low expression relative to the median of CD155 expression, and Wilcoxon rank-sum was applied for the significance test. B GSEA analysis was done under the CD155-high-expression group. C Results of WB assays showed the protein level of VEGF was stronger in HEY cells and SKOV3 cells overexpressing CD155, respectively. D Results of WB assays showed the protein level of CD31 was upregulated in HEY cells and SKOV3 cells overexpressing CD155, respectively.

Dual targeting of CD155 augments the tumoricidal capacity of ROR1-CAR-T cells

In light of the aforementioned results, it can be posited that CD155 is an optimal immunotherapy target for OC. Therefore, we seek to incorporate CD155 as a secondary target, thereby constructing a bispecific CAR-T dual-targeting ROR1 and CD155. The structures of lentiviral vector are shown in Fig. 5A. Following lentivirus transduction, CAR expression was detected using flow cytometric analyses (Fig. S3). Subsequent to the generation of CAR-T cells, a co-culture system was established comprising SKOV3 cells and CAR-T cells at an effector-to-target ratio of 1:1. Following a 24-h co-culture period, cytometric analyses were conducted to evaluate the activation level of CAR-T cells within distinct groups. An elevated expression level of CD69 and CD25 was observed in the bispecific CAR-T cells in comparison with the ROR1-CAR-T cells (Fig. 5B, C). Besides, results of ELISA assays revealed a trending increase in IFN-γ and IL-2 production by bispecific CAR-T cells versus single-target constructs at 1:1 E:T ratio following 24 h OC cells engagement (Fig. S4A, B). These findings suggest that the incorporation of CD155 as a secondary target enhances the activation of CAR-T cells. Cytotoxicity assays were conducted to evaluate the anti tumor effect of single-target CAR-T cells and bispecific CAR-T cells. As depicted in Fig. 5D, at a high E:T ratio (1:1), bispecific CAR-T cells exhibited tumor-killing efficacy comparable to single-target ROR1-CAR-T cells. However, at lower E:T ratios (1:3 and 1:9), bispecific CAR-T cells demonstrated a progressively enhanced tumor-killing trend relative to their single-target counterparts. In addition to OC cells, consistent CD155 surface expression was confirmed by flow cytometry in breast cancer cell line (MDA-MB-231), and colorectal cancer cell lines (HT-29, COLO-678) (Fig. S5A). Cytotoxicity assays demonstrated marked tumor-killing efficacy of bispecific CAR-T cells against these cell lines (Fig. S5B-D), suggesting broad therapeutic relevance of co-targeting ROR1 and CD155.

Fig. 5.

Dual targeting of CD155 augments the tumoricidal capacity of ROR1-CAR-T cells. A Schematic illustration of the ROR1-CAR, CD155-CAR, and bispecific CD155-ROR1-CAR vector. Dual CAR: dual anti-CD155/anti-ROR1 bispecific CAR construct. B The expression of CD69 in CAR-T cells was detected by flow cytometry. C The expression of CD25 in CAR-T cells was detected by flow cytometry. D The cytotoxic effect of CAR-T cells was assessed on OC cell lines exhibiting varying effector-to-target (E:T) ratios through luciferase assays (p value: ***p < 0.001, *p < 0.05, ns p > 0.05)

Co-targeting CD155 attenuated tonic signaling and reduced exhaustion

As mentioned above, single-target ROR1-CAR-T cells exhibited elevated tonic signaling, thereby expediting functional exhaustion. Subsequent in vitro experiments were utilized to investigate whether CD155 co-targeting effectively attenuates this tonic activation. Initially, IFN-γ concentration was quantified by ELISA following a 24-h monoculture of respective CAR-T cell groups. Results demonstrated that single-target ROR1-CAR-T cells exhibited elevated tonic signaling, whereas a remarkable diminished level was found in bispecific CAR-T cells co-targeting CD155 and ROR1 (Fig. 6A). On days 14 and 16 of CAR-T cell culture, we assessed the expression of T cell exhaustion markers (PD-1, TIM-3, and LAG-3) across respective CAR-T groups via flow cytometry to evaluate their exhaustion status. A notable increase in the expression of PD-1, TIM-3, and LAG-3 was observed in single-target ROR1-CAR-T cells on days 14 (Fig. 6B) and 16 (Fig. 6C), suggesting an elevated level of exhaustion, which may be attributable to heightened tonic signaling. Conversely, bispecific CAR-T cells that co-target CD155 and ROR1 exhibited a substantial decrease in exhaustion in comparison with single-target ROR1-CAR-T cells. This finding suggests that the concurrent targeting of CD155 has the potential to attenuate tonic signaling in ROR1-directed CAR-T cells, thereby contributing to the mitigation of T cell exhaustion. Subsequently, repetitive stimulation assays were utilized to emulate the in vivo context in which CAR-T cells encounter repeated antigen exposure (Fig. 6D). Following two stimulation cycles, flow cytometric analyses revealed that bispecific CAR-T cells exhibited significantly lower PD-1 and LAG-3 expression compared to single-target ROR1-CAR-T cells, indicating reduced exhaustion (Fig. 6E). Furthermore, after each stimulation cycle, CAR-T cell differentiation status was assessed via CD45RO and CCR7 co-staining. Results demonstrated that bispecific CAR-T cells maintained superior capacity for effector memory T cell differentiation (CD45RO⁺CCR7⁻), suggesting their enhanced ability to mount rapid effector responses upon recurrent antigen encounter (Fig. S6A, B). We also detected CAR expression after each stimulation round, greater enrichment of CAR-positive cells was found in bispecific CAR-T cells in comparison with single-target counterparts (Fig. 6F, G). The aforementioned results collectively suggest that co-targeting CD155 attenuated tonic signaling and reduced exhaustion.

Fig. 6.

Co-targeting CD155 attenuated tonic signaling and reduced exhaustion. A ELISA results showed IFN-γ concentration of CAR-T cells in the absence of target cell stimulation (p value: *p < 0.05). B The expression of PD-1, TIM-3, and LAG-3 in CAR-T cells was detected by flow cytometry on day 14 post-activation. C The expression of PD-1, TIM-3, and LAG-3 in CAR-T cells was detected by flow cytometry on day 16 post-activation. D Schema of the repeated antigen stimulation experiment. E Representative exhaustion marker expression following two stimulation cycles. F, G CAR expression in co-cultured T cells was detected by flow cytometry following one stimulation cycle (F) and two stimulation cycles (G).

Discussion

Due to its insidious onset and lack of effective early screening modalities, a majority of OC patients present with advanced-stage disease at first diagnosis [39]. These advanced cases exhibit poor prognoses, posing a substantial global health burden for women. Although contemporary clinical evidence suggests that maintenance therapy with PARP inhibitors (PARPi) alone or combined with anti-angiogenic agents (e.g., bevacizumab) following platinum-based chemotherapy prolongs progression-free survival [40, 41], the prognosis for advanced OC remains dismal, with 5-year and 10-year survival rates approximating merely 25% and 15%, respectively [42]. Consequently, developing novel therapeutic approaches and identifying new molecular targets constitute urgent unmet needs for improving OC outcomes.

The advent of CAR-T cell therapy has revolutionized the treatment paradigm for hematological malignancies. Kymriah (tisagenlecleucel), the first FDA-approved CAR-T cell product targeting CD19, induces potent and durable clinical responses in B cell acute lymphoblastic leukemia [43, 44]. Presently, six CAR-T cell therapies are FDA-approved for clinical use in hematopoietic cancers [45]. The established success of CAR-T in hematologic malignancies has catalyzed its investigation in solid tumors. Although numerous CAR-T cells targeting diverse antigens are in preclinical development or clinical trials for solid tumors [46], no CAR-T therapy has yet received FDA approval for solid tumor indications. The limited clinical translation of CAR-T cells in solid tumors is primarily attributable to target antigen-related challenges and the distinct, complex tumor microenvironment compared to hematological cancers [47].

Limitations at the target antigen level encompass both off-target toxicity and antigen escape. The majority of candidate target antigens in solid tumors are tumor-associated antigens (TAAs) co-expressed in malignant and normal tissues. CAR-T cells engineered to recognize such antigens may bind and activate against target-positive normal tissues, potentially causing fatal damage [48]. The restricted antigen expression, which is minimal or absent in normal tissues, is critical to avoid off-target toxicity. The tumor-restricted expression pattern of ROR1, characterized by high tumor and minimal normal tissue expression, is well-documented. Our study extends this by validating its nearly absent expression in non-neoplastic ovarian tissues at single-cell data analysis, solidifying its rationale as an immunotherapy target. More importantly, in vitro experiments in this study demonstrated the remarkable efficacy of ROR1-CAR-T cells in eradicating tumor cells, accompanied by their limitation of elevated tonic signaling. It is imperative to reduce such signaling through the modification of CAR structure in order to maintain the efficacy of the therapeutic treatment.

Antigen escape refers to target antigen downregulation or loss following single-target CAR-T therapy. Although single-target CD19-CAR-T achieves durable remissions [49], it has been observed that 30–70% of relapsed patients exhibit CD19 downregulation [50]. Antigen escape has also been observed in solid tumors subsequent to single-target CAR-T treatment. A diminished expression level of IL13Rα2 was found in recurrent glioblastoma lesions following IL13Rα2-CAR-T therapy [51]. Reintroduction of the same single-target CAR-T in antigen-escape relapses has been shown to yield diminished efficacy, thus prompting the development of dual/multi-target CAR-T strategies. For instance, CD19/CD22 co-targeting CAR-T partially overcomes CD19 escape in relapsed B cell leukemia [52]. In glioblastoma, HER2 and IL13Rα2 co-targeting CAR-T reduces antigen-escape incidence [53]. Consequently, we hypothesize that the modification of ROR1-CAR-T is achieved by the addition of a target. Preview research [54] has revealed high CD155 antigen coverage in OC through immunohistochemical analysis, and of particular significance is the observation that CD155 demonstrated minimal expression in normal fallopian tube epithelial tissues. In addition, clinical studies have shown that the concentration of soluble CD155 in the serum of patients with malignant tumors, including ovarian cancer, is higher than that of healthy people [55]. In our research, our analysis of single-cell transcriptomic profiles confirmed a minimal expression of CD155 in non-malignant ovarian tissues, supporting its favorable safety profile as a therapeutic target. More importantly, Xiong et al.’s thorough in vivo studies have revealed the efficacy and safety of CD155-CAR-T in the treatment of solid tumors [56]. Complementary in vitro experiments in our research demonstrated CD155’s role as an oncogenic driver that positively regulates OC cell proliferation and migration, suggesting that CD155 inhibition represents a potential therapeutic strategy for OC.

Furthermore, inadequate infiltration into the tumor parenchyma serves as a primary barrier limiting CAR-T efficacy in solid tumors. Unlike hematological malignancies, the penetration of CAR-T cells into the tumor parenchyma is imperative for target recognition and the initiation of cytolytic processes. However, tumor cells develop aberrant vasculature to support proliferation and metastasis [57]. As demonstrated by Huang et al. [36], the presence of abnormal vascular networks within solid tumors has been shown to impede CD8⁺ T cell infiltration, whereas vessel normalization increases T cell infiltration. In our in vitro experiments, CD155-overexpressing OC cells exhibited elevated levels of angiogenesis-related molecules including VEGF and CD31. Consequently, it could be inferred the prevention of CD155 expression may facilitate the infiltration of CAR-T cells by normalizing the tumor vasculature within the OC parenchyma.

Notably, the immunosuppressive TME, which impedes CAR-T cell recruitment and function, constitutes another major barrier to CAR-T efficacy in solid tumors. As an immunoregulatory molecule, CD155 extensively modulates both innate and adaptive immune cells within the TME. Zhu et al. [58] demonstrated significantly elevated CD155 expression on tumor-associated macrophages (TAMs) in colorectal carcinoma compared to macrophages in paracancerous and adjacent normal tissues. CD155⁺ TAMs exhibited M2-like polarization, promoting colorectal cancer cell migration, invasion, and tumor growth. Furthermore, TAMs have been shown to exert a negative regulatory effect on CD8⁺ T cell recruitment and function. This regulatory effect is mediated through the expression of immunosuppressive molecules, including programmed death-ligand 1 (PD-L1) and CD155 [13]. Furthermore, CD155 expression on tumor cells has been shown to downregulate IFN-γ secretion capacity in NK cells [18]. In regard to CD8⁺ T cells, which have been identified as pivotal antitumor effectors, He et al. [17] demonstrated that gastric cancer cells trigger TIGIT upregulation on CD8⁺ T cells. This prompts the initiation of CD155/TIGIT signaling, a process that has been shown to suppress T cell metabolism and compromise effector function. Notably, both the TIGIT blockade and the CD155 prevention reversed the inhibition of T cell IFN-γ secretion. In this study, co-culture assays revealed that CD155-expressing OC cells suppressed T cell activation, downregulated secretion of tumoricidal mediators (TNFα, IFN-γ, and perforin), and upregulated T cell exhaustion markers (PD-1, TIM-3, LAG-3, and TIGIT). Collectively, CD155 participates in both immunomodulation within the TME and angiogenesis regulation, thereby contributing to an immunosuppressive niche. Consequently, blocking CD155 may foster a “hot” immunophenotype by enhancing immune cell functionality and normalizing tumor vasculature. Based on this rationale, we engineered bispecific CAR-T cells co-targeting CD155 and ROR1 to investigate their synergistic efficacy versus single-target CAR-T. The present study demonstrated in vitro that bispecific CAR-T cells co-targeting CD155 and ROR1 exhibited enhanced cytotoxic activity and activation profiles. This enhancement is presumably attributable to the partial alleviation of CD155-mediated suppression on T cell activation and effector functions. More significantly, CD155 co-targeting effectively ameliorated the intrinsic limitation of elevated tonic signaling in single-target ROR1-CAR-T cells, reducing exhaustion levels during both monoculture and repetitive stimulation. The attenuation of exhaustion during repetitive stimulation may be mediated by blocking CD155's engagement with inhibitory receptors on CAR-T cells. Besides, it is noteworthy that single-target CD155-CAR-T cells exhibited superior exhaustion resistance during repetitive stimulation assays in comparison with both single-target ROR1-CAR-T and bispecific CAR-T cells. This finding suggests that the prevention of CD155 engagement between T cells and target cells—independent of ROR1 targeting—can mitigate T cell exhaustion, thereby corroborating our previously formulated mechanistic hypothesis.

Notably, our study has certain limitations. The therapeutic efficacy and safety profile of the bispecific CAR-T co-targeting CD155/ROR1 require further validation through in vitro and in vivo studies, which will be a primary focus of our future research. A key priority is to systematically validate the expression patterns of CD155 and ROR1, along with their antigen coverage in tumor samples, by analyzing clinic pathological specimens from ovarian cancer patients and non-neoplastic tissues from vital human organs. Besides, whether bispecific CAR-T strategy co-targeting CD155 can enhance T cell infiltration into the parenchyma of ovarian tumors and subsequently remodel the immunologically “cold” TME requires further investigation through well-designed in vivo studies.

Conclusion

In summary, our study developed a novel bispecific CAR-T and conducted an experimental investigation of its therapeutic potential in OC. The present study delineates the potent antitumor efficacy and inherent limitations of single-target ROR1-CAR-T therapy via experimental investigation. Of particular significance is the work’s establishment of CD155 as a promising immunotherapeutic target for OC and its demonstration that co-targeting CD155 enhances the functionality of ROR1-CAR-T cells while elucidating the underlying mechanisms. Collectively, these findings provide novel mechanistic and translational perspectives for target selection in OC CAR-T therapy.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

CAR-T

Chimeric antigen receptor T cell

CD155(PVR)

Poliovirus receptor

GZMB

Granzyme B; PD-1: Programmed cell death 1

IFN-γ

Interferon gamma

IL-2

Interleukin-2

KM

Kaplan–Meier

LAG-3

Lymphocyte-activation gene 3

NK

Natural killer

OC

Ovarian cancer

OS

Overall survival

PBS

Phosphate-buffered saline

ROR1

Receptor tyrosine kinase-like orphan receptor 1

TCGA

The Cancer Genome Atlas

TME

Tumor microenvironment

TNFα

Tumor necrosis factor alpha

TIM-3

T cell immunoglobulin domain and mucin domain-3

TIGIT

T cell immune receptor with Ig and ITIM domains

Author contributions

Shaohua Xu and Xiaowen He conceived and supervised the study. Yingjun Ye and Tingwei Liu designed and conducted the experiments, and Chao Cheng helped analyze the data. Yingjun Ye and Tingwei Liu drafted the original manuscript. Huajing Wang and Jiacheng Shen were responsible for the review and editing. All authors have read and approved the work content and the final version of this paper.

Funding

This work was supported by the following funding sources: the National Natural Science Foundation of China (Grant No. 81772762), Natural Science Foundation of Shanghai (Grant No. 21ZR1450900), and Shanghai Science and Technology Planning Project (Grant No. 21Y11907000).

Data availability

The datasets generated and analyzed in this study are available in the TCGA database (https://portal.gdc.cancer.gov), KM plotter database (https://kmplot.com/analysis/), and GEPIA 2 database (http://gepia2.cancer-pku.cn/). The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

The authors declare that they have no competing interests. All the authors approved the publication.