CAR-T Therapy Targets Extra Domain B of Fibronectin Positive Solid Tumor Cells

ABSTRACT Background CAR-T cell immunotherapy has achieved remarkable success in malignant B-cell malignancies, but progress in solid tumors is slow, and one of the key reasons is the lack of ideal targets. Cancer-specific extra domain B of fibronectin (EDB-FN) is widely upregulated in solid tumors and expressed at low levels in normal tissues. Many imaging and targeted cancer therapies based on EDB-FN targets have been developed and tested in clinical trials, making EDB-FN an ideal target for immunotherapy. Methods We constructed two EDB-FN-targeted CAR-Ts based on the peptide APT0 and the single-chain antibody CGS2 in a lentiviral infection manner for the first time. Luciferase cytotoxicity assay to assess CAR-T killing of tumor cells. An enzyme-linked immunosorbent assay was used to detect the release of the cytokine IFN-γ. Fluorescence imaging to evaluate the dynamics of CAR-T cell and tumor cell coculture. Knockdown assays were used to validate the target specificity of CAR-T cells. Results In this research, two CAR-Ts targeting EDB-FN, APT0 CAR-T, and CGS2 CAR-T, were constructed. In vitro, both CAR-T cells produced broad-spectrum killing of multiple EDB-FN-positive solid tumor cell lines and were accompanied by cytokine IFN-γ release. Regarding safety, the two CAR-T cells did not affect T cells’ normal growth and proliferation and were not toxic to HEK-293T human embryonic kidney epithelial cells. Conclusion APT0 CAR-T and CGS2 CAR-T cells are two new CAR-Ts targeting EDB-FN. Both CAR-T cells can successfully identify and specifically kill various EDB-FN-positive solid tumor cells with potential clinical applications.


Introduction
Chimeric antigen receptor T cells (CAR-T) refer to the use of genetic engineering techniques to express a synthetic receptor on the surface of T cells that recognizes tumor antigens in a manner that does not depend on the major histocompatibility complex (MHC) to identify and kill tumor cells (June et al., 2018).Among the emerging immunotherapies for various tumors, CAR-T immunotherapy has achieved great success in treating malignant B-cell malignancies, and several CD19 CAR-T and BCMA CAR-T products have been approved by the FAD for clinical treatment (Lu & Jiang, 2022).However, the application of CAR-T therapy in solid tumors is hampered by the tumor immunosuppressive microenvironment, the difficulty of CAR-T cell transport and infiltration into cancer, and the selection of CAR-T-specific targets, of which the application of a particular target is critical.On the one hand, it increases the specificity of CAR-T to kill tumor cells, and on the other hand, it reduces the toxic side effects on nontumor tissues (Flugel et al., 2023).
Fibronectin (FN), a widely studied dimeric glycoprotein, is a precursor mRNA produced by the FN gene that is selectively sheared at three sites in the EDA, EDB, and IIICS structural domains to produce a variety of isoforms (Kumra & Reinhardt, 2016).EDB-FN is a secreted stromal oncoprotein.In solid tumors, EDB-FN is present in the extracellular matrix and adheres to the cell surface to promote tumor angiogenesis.EDB-FN is positively correlated with the malignancy of cancer (Castellani et al., 1994(Castellani et al., , 2002;;Khan et al., 2005).EDB-FN is widely found in various solid tumors, such as breast, pancreatic, liver, glioblastoma, and colorectal cancer (Lieverse et al., 2020).Under physiological conditions, EDB-FN is involved in neovascularization and is found in embryonic tissue, adult ovaries, and the uterus but is rare in other normal tissues (De Candia & Rodgers, 1999).There are many immunotherapies based on the EDB-FN target.TRX-EDB is an EDB-FN immune vaccine injected into mice, producing good efficacy without interfering with wound healing (Huijbers et al., 2010).In addition, L19, an antibody that specifically binds EDB-FN to different therapeutic agents, has been used in clinical trials to treat various solid tumor types without off-target effects, producing severe organ toxicity (Lieverse et al., 2020).
In recent years, preclinical studies on EDB-FN-targeted CAR T cells have emerged, showing that EDB-FN can induce solid tumor cell line killing and inhibit growth in mouse tumor models (Wagner et al., 2021;Xie et al., 2019;Zhang et al., 2022).Given that the existing EDB CAR-T has been validated on a relatively small number of solid tumors and a large number of solid tumor types remain to be validated, we constructed CAR-T cells targeting EDB-FN based on peptide APT0 and antibody CGS2 for the first time (Carnemolla et al., 1996;Saw et al., 2021).CAR-T cell function was validated in EDB-FN-positive hepatocellular carcinoma, pancreatic cancer, breast cancer, and malignant glioma cell lines, and the results of the research showed that APT0 CAR-T and CGS2 CAR-T cells had sound antitumor effects.Therefore, APT0 CAR-T and CGS2 CAR-T cells can target EDB-FN-positive solid tumors, enriching CAR-T therapies for solid tumors and providing new options for treating solid tumors.

Plasmid construction and lentiviral packaging
The sequences of the CGS2 scFv, CD19 scFv, L19 scFv, and APT0 peptide sequences used in the experiments were obtained from published articles (Supplementary Table S1) (Carnemolla et al., 1996;Pini et al., 1998;Saw et al., 2021;Yang et al., 2019), and sequence synthesis was performed by a company (Youkang, China).The synthetic sequence was then cloned into the CAR-encoding lentivirus backbone containing the CD8 hinge spacer and transmembrane domain, 4-1BB, and CD3ζ endo-domains, which was linked to a fluorescent tag mKATE2 protein via T2A, with all elements expressed under the control of the CMV promoter.
For lentiviral packaging, lentiviral DNA vectors were transfected with Lipo8000™ Transfection Reagent (Beyotime, China) according to the manufacturer's protocol: the core plasmid and two packaging plasmids, pCMVΔ8.9 and pMD2.G were co-transfected into HEK-293T cells at a 10:5:2 ratio, and the lentivirus was obtained as previously described (Yang et al., 2019).

Cell lines and culture
Tumor cell lines (Bxpc3, MIA-PACA2, HEPG2, MDA-MB-231, U251 MG, U87 MG) were all purchased from the China Infrastructure of Cell Line Resources (Kunming, Beijing, or Shanghai, China).SK-HEP1 and MDA-MB-468 were obtained from Sichuan Bio Technology and West China Hospital, respectively.HEK-293T cells were obtained from A. Lasorella (The Institute for Cancer Genetics, Columbia University Medical Center, New York, NY).Tumor cell lines MIA-PACA2, HEPG2, and MDA-MB-468 were infected with pTOMO-CMV-Luciferase-EGFP-IRES-puro lentivirus to generate MIA-PACA2-Luciferase-EGFP, HEPG2-Luciferase-EGFP, and MDA-MB-468-Luciferase-EGFP cell lines.Other cell lines were infected with pTOMO-CMV-Luciferase-IRES-puro lentivirus to obtain cell-Luciferase.SK-HEP1-shFN cell lines were acquired with lentiviral infection.All cells were cultured in DMEM (Life Technologies) supplemented with 10% fetal bovine serum (Life Technologies), 100 U/mL penicillin (Life Technologies), and 100 mg/mL streptomycin sulfate (Life Technologies).All the cells were cultured at 37°C in a humidified incubator with 5% CO2 and routinely confirmed to be mycoplasma free by PCR.Cells were regularly passaged when they reached approximately 80% confluence.

CAR-T cell preparation
T cells were isolated from the peripheral blood of healthy donors with RosetteSep™ Human T Cell Enrichment Cocktail (STEMCELL, Canada).T cells were then activated by CD3/ CD28 Dynabeads (Life Technologies, USA) and cultured in RPMI 1640 medium (Life Technologies, USA) containing 10% FBS (Life Technologies, USA) with 200 U/ml IL-2 (PeproTech, USA).After 72 h, lentiviral particles were added to the cultures containing T cells at a multiplicity of infection (MOI) of 100 in the presence of Lentibost (Sirion Biotech).T cells were infected with CAR lentivirus for three days and used for in vitro experiments.

Cytotoxicity assay and cytokine release detection
The specific cytotoxicity of CAR-T cells against various tumor cell lines was tested at a defined effector target ratio (E: T).CAR-T cells were co-cultured with tumor cells in RPMI 1640 (Life Technologies, USA) culture medium supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 mg/mL streptomycin sulfate.After 24 h, the supernatant was collected to determine the concentration of cytokine IFN-γ (Invitrogen, KHC4021), and the concentration was determined using a specific enzyme-linked immunosorbent assay (ELISA) kit according to the manufacturer's instructions.Cells were also lysed for luciferase assay according to the manufacturer's instructions, and cytotoxicity was standardized using a Cytotoxicity Fluorometric Assay Kit (Promega).

RNA isolation and real-time PCR
Total RNA was isolated from the cell lines using TRIzol (Sigma-Aldrich, USA) as described previously (Yang et al., 2018).Then, the TURBO DNA-free TM Kit (Life Technologies, USA) and RevertAid First Strand cDNA Synthesis kit (Thermo Scientific, USA) were used to obtain cDNA without DNA contamination.The relative expression of EDB-FN in tumor cells was detected by a SYBR Green qPCR Kit (Life Technologies, USA).EDB F-primer: 5-AACTCACTGACCTAAGCTTT-3 and R-primer: 5-CGTTTGTTGTGTCAGTGTAG-3;18S F-primer: 5-GTAACCCGTTGAACCCCATT-3 and R-primer: 5-CCATCCAATCGGT AGTAGCG-3.

Flow cytometry
The harvested cells were washed twice with 1 × PBS and then suspended in 4°C 1 × PBS containing 2% FBS (density of 1 × 10 6 cells/ml).CD4 (BioLegend, 357408) and CD8 (BioLegend, 344704) antibodies were used to assess the T cell phenotype.The intensity of mKATE2 fluorescence represents CAR expression.Samples were analyzed on a BD LSR Fortessa system (BD Biosciences, USA) using the FlowJo_V10 analysis software package (TreeStar, USA).

Statistical analysis
All statistical analyses were performed using GraphPad Prism 8.0 statistical software (Dotmatics, USA).Statistical data between the two groups were analyzed using the unpaired two-tailed Student's t-test.Statistical data for three or more groups were analyzed with oneway ANOVA.All data are presented as the mean ± SD, and p values were considered statistically significant: ns: not significant, *p < .05,**p < .01,***p < .001,and ****p < .0001.

Construction, screening, and characterization of CAR-T cells
To generate CAR-T cells targeting EDB-FN, we synthesized the APT0 peptide sequence and the scFv sequence of CGS2, which can bind EDB-FN (Carnemolla et al., 1996;Saw et al., 2021).We then cloned the synthesized sequence into a lentiviral expression vector.The vector connects the extracellular antigen recognition domain to the intracellular signaling structural domain with the CD8 hinge region and CD28 transmembrane structural domain.We also added the mKATE2 element to the intracellular conductance structural domain.We named the constructed CARs APT0 CAR-T and CGS2 CAR-T.In addition, we used nontransduced T cells (NTD) and CD19 CAR-T as negative controls in our experiments and L19 CAR-T, which was reported to specifically kill EDB-FN-positive tumor cells, as a positive control (Wagner et al., 2021;Yang et al., 2019;Zhang et al., 2022)(Figure 1a).To visually characterize the ability of lentiviral particles to infect T cells, we observed the expression of CAR molecules on T cells under inversion fluorescence microscopy.We found that APT0 CAR-T and CGS2 CAR-T cells grew normally, CAR-T cells aggregated into apparent clusters, and both APT0 CAR-T and CGS2 CAR-T cells emitted bright red light under laser excitation (Supplementary Figure S1).To further quantify the positive rate of CAR-T cells, we detected mKATE2 by flow cytometry.We found that all of the constructed CAR-T cells showed a positive rate of more than 40% compared to the NTD, demonstrating that all of our constructed CARs have high infectivity (Figure 1b).
EDB-FN is a precancerous tumor target expressed only in human embryonic, ovarian, and uterine normal tissues and is widely expressed in various solid tumors (De Candia & Rodgers, 1999;Menrad & Menssen, 2005).EDB-FN targets were detected by real-time quantitative PCR in hepatocellular carcinoma, pancreatic cancer, breast cancer, and glioblastoma cell lines.We found that the tumor cell lines tested showed a high level of EDB expression compared to human embryonic kidney cells HEK-293T (Figure 1c).Subsequently, we used glioblastoma U251 MG cells to validate the tumor cell lysis ability of APT0 CAR-T and CGS2 CAR-T cells.The results showed that both APT0 CAR-T and CGS2 CAR-T cells exhibited the same potent cytotoxicity as L19 CAR-T cells compared to the CD19 CAR-T group (Figure 1d).Meanwhile, we performed an IFN-γ cytokine assay on the supernatant of cocultured cells to verify the specific cytokine-dependent cytotoxicity of CAR-T cells.The results showed that both APT0 CAR-T and CGS2 CAR-T cells exhibited higher levels of IFN-γ release than the L19 CAR-T cells (Figure 1e).The above results show that we successfully constructed APT0 CAR-T and CGS2 CAR-T cells with cytotoxicity targeting EBD-FN.
In addition, we performed cell counts on APT0 CAR-T and CGS2 CAR-T cells starting from day 0 and every 3 days until day 12.The counting results showed that both the APT0 CAR-T and CGS2 CAR-T groups showed similar proliferative efficiency as the control group, indicating that the insertion of our APT0 sequence and CGS2 sequence did not affect the usual proliferation of T cells and did not have cytotoxic effects on T cells (Figure 1f).Meanwhile, we found that APT0 CAR-T and CGS2 CAR-T cells were able to differentiate normally with CD4 and CD8 phenotypes (Supplementary Figure S2).

APT0 CAR-T and CGS2 CAR-T cells can widely lyse solid tumor cells
To further evaluate the extensive tumor cytotoxicity of APT0 CAR-T and CGS2 CAR-T cells, We selected new solid tumor cell lines for research that L19 CAR-T cells do not use.We cocultured APT0 CAR-T and CGS2 CAR-T cells with tumor cells at an E: T ratio of 5:1 and then determined the ability of CAR-T cells to lyse tumor cells by analyzing the luciferase signal in the samples.APT0 CAR-T and CGS2 CAR-T cells were cocultured with hepatocellular carcinoma cell lines (SK-HEP1, HEPG2), pancreatic cancer cell lines (MIA-PACA2, Bxpc3), breast cancer cell lines (MDA-MB-231, MDA-MB-468) and malignant glioblastoma cell lines (U87 MG), and the human embryonic kidney cell line HEK-293 T cells were used as control cells.Our results demonstrate that APT0 CAR-T and CGS2 CAR-T cells can identify and effectively lyse a variety of solid tumor cell lines (Figure 2a).The supernatants of the cocultures were also assayed for IFN-γ cytokine levels, which represent indicators of T cell activation and specific cytotoxicity.The results showed that APT0 CAR-T and CGS2 CAR-T cells had higher levels in each solid tumor cell line than CD19 CAR-T cells (Figure 2b).
In conclusion, cytotoxicity and cytokine assays provide convincing evidence that APT0 CAR-T and CGS2 CAR-T cells can induce cytokine production and broadly identify and effectively kill a wide range of solid tumor cells in vitro.

APT0 CAR-T and CGS2 CAR-T cells lyse tumor cells in a time-dependent and gradient-dependent manner
Next, we constructed EGFP elements in MDA-MB-468 tumor cells to further observe the changes in tumor cells under the influence of APT0 CAR-T and CGS2 CAR-T cells in realtime.The APT0 CAR-T and CGS2 CAR-T cells cocultured with tumor cell lines were photographed at 0 h, 12 h, 24 h, and 48 h by inverted fluorescence microscopy.Taking MDA-MB-468 tumor cells as an example, with the extension of coculture time, APT0 CAR-T and CGS2 CAR-T cells showed a gradually enhanced killing effect on HEPG2 tumor cells.Among them, APT0 CAR-T cells had a significant ability to lyse MDA-MB-468 cells, and almost all the tumor cells were killed at 24 hours.In the presence of APT0 CAR-T and CGS2 CAR-T cells, intact tumor cells were lysed into fragments, and the fluorescence of tumor cells became dim.At 24 h, APT0 CAR-T and CGS2 CAR-T cells had significant aggregation and proliferation of cells compared to the CD19 CAR-T group, and they both surrounded the tumor cells, implying an expanded expansion and specific killing of APT0 CAR-T and CGS2 CAR-T cells under the stimulation of EDB-FN targets (Figure 3a).At the same time, we counted the surviving MDA-MB-468 tumor cells at 48 h, and the quantitative results more directly showed the powerful tumor lysis ability of APT0 CAR-T and CGS2 CAR-T cells (Figure 3b).Thus, APT0 CAR-T and CGS2 CAR-T cells can exert practical cytotoxic effects on tumor cell lines in a time-dependent manner in vitro.
Our results also show that APT0 CAR-T and CGS2 CAR-T cells can specifically kill tumor cells in a gradient-dependent manner.We cocultured APT0 CAR-T and CGS2 CAR-T cells with MDA-MB-468 and Bxpc3 cell lines at different E: T ratios (1:1, 2.5:1, and 5:1).The results showed that as the gradient rose, APT0 CAR-T and CGS2 CAR-T cells also rose in their ability to lyse tumor cells, and CD19 CAR-T cells showed a weaker change (Figure 3c).

APT0 CAR-T and CGS2 CAR-T cells lyse tumor cells in an EDB-FN-dependent manner with particular safety
The peptide APT0 and the antibody CGS2 have long been shown to bind EDB-FN specifically, and we next verified whether CAR-T cells constructed from their sequences were dependent on the specificity of EDB-FN.We packaged knockdown lentiviral particles with fibronectin.We then infected SK-HEP1 cells with the above lentiviral particles to obtain stably transfected cells.The knockdown efficiency was verified by real-time PCR.The results demonstrated that the shFN knockdown efficiency on SK-HEP1 cell fibronectin was significantly higher compared with the shCOO2 control group (Figure 4a).As expected, APT0 CAR-T and CGS2 CAR-T cells exhibited low cytotoxicity against cells after FN knockdown (Figure 4b).These results suggest that the cytotoxicity of APT0 CAR-T and CGS2 CAR-T cells against tumor cells is EDB-FN dependent.Meanwhile, we initially validated the safety of APT0 CAR-T and CGS2 CAR-T cells in vitro using HEK-293T cells, which showed no significant cytotoxicity to human embryo kidney cells HEK-293T compared to CD19 CAR-T cells (Figure 4c).

Discussion
A specific target is critical for successfully translating CAR-T immunotherapy to solid tumors.Most of the typical clinical tumor-associated antigens, such as MSLN, HER2, B7-H3, and EpCAM, are expressed in other normal tissues of the body, and their CAR-T treatment faces the toxic side effects of targeting nontumor tissues, which may lead to serious adverse events (Flugel et al., 2023;Richman et al., 2018).
EDB-FN, a secreted stromal oncoprotein, is produced by selective shearing of fibronectin.EDB-FN is present in the tumor extracellular matrix, adheres to the cell surface, is involved in tumor angiogenesis, and correlates with tumor malignancy (Neri & Bicknell, 2005).EDB-FN is widely present in solid tumors, and analysis of the TCGA database has been performed to confirm the expression of EDB-FN in various solid tumors (Wagner et al., 2021).EDB-FN is present only in the ovary and uterus in adult tissues and plays a vital role in angiogenesis (De Candia & Rodgers, 1999).Preclinical and clinical studies have been conducted to design drugs for EDB-FN to trace and treat solid tumors without serious adverse effects on the organism or interference with wound healing.Therefore, treatment targeting EDB-FN has a good safety profile (Huijbers et al., 2010;Johannsen et al., 2010;Spitaleri et al., 2013).
The application of the EDB-FN target to CAR-T immunotherapy is still in its initial stage.In the past two years, CAR-T cells designed to target EDB-FN using the NJB2 nanoantibody and L19 antibody have shown antitumor effects in melanoma, lung cancer, rhabdomyosarcoma, and glioma (Wagner et al., 2021;Xie et al., 2019;Zhang et al., 2022).ATP0 is a highaffinity peptide that recognizes EDB-FN, and peptide-drug conjugates based on ATP0 can produce promising antitumor effects.CGS2 is another monoclonal antibody targeting EDB-FN.APT0 and CGS2 have a nanomolar affinity for EDB-FN and exhibit high specificity (Carnemolla et al., 1996;Saw et al., 2021).In this study, we constructed two novel CAR-T cells (APT0 CAR-T and CGS2 CAR-T) targeting EDB-FN based on the APT0 peptide and monoclonal antibody CGS2 for the first time.Compared to NJB2 CAR-T or L19 CAR-T cells, our CAR-T cells showed similar tumor cell lysis ability and IFN-γ cytokine release in the experimental cell line U87 MG.We also explored the efficacy, safety, and specificity of these two CAR-Ts targeting EDB-FN in new tumor types (liver, breast, pancreatic).We found that APT0 CAR-T and CGS2 CAR-T cells have strong tumor cell lysis ability, further enriching the study of CAR-T targeting EDB-FN on broad-spectrum solid tumor cell lines.This study used EDB-FN-positive tumor cells reported in the literature for CAR-T functional validation (Gao et al., 2019;Saw et al., 2021;Tavian et al., 1994;Wagner et al., 2008;Ye et al., 2017).The results of the CAR-T gradient killing assay and fluorescence imaging assay showed that APT0 CAR-T and CGS2 CAR-T had CAR-T-related gradient-dependent and time-dependent antitumor activity, and the process of tumor cell lysis was accompanied by a significant release of the cytokine IFN-γ, demonstrating that tumor cell lysis was caused by CAR-T specific killing.
Regarding safety and specificity, the peptide APT0 and the monoclonal antibody CGS2 have long been shown to bind specifically to EDN-FN and not to other tissues in the body, providing excellent safety and specificity (Carnemolla et al., 1996;Saw et al., 2021).Therefore, preparing CAR-T cells with APT0 and CGS2 sequences can guarantee the security and specificity of APT0 CAR-T and CGS2 CAR-T cells to a certain extent.Regarding CAR-T specificity, we performed a fibronectin knockdown assay to knock down the expression of the cell line EDB-FN to validate the specificity of both CARs and avoid other toxicities due to the broad-spectrum recognition of ATP0 and CGS2.We also conducted a preliminary exploration of safety.In CAR-T amplification experiments, the two new CAR-Ts showed similar amplification ability to the control, indicating that the insertion of the new sequence would not impair normal cell growth and development, and APT0 CAR-T and CGS2 CAR-T were not toxic to HEK-293T cells, which laterally reflected their good safety.In addition, when APT0 CAR-T enters the human body possibly due to its immunogenicity, it may induce B cells to produce anti-APT0 antibodies and lead to CAR-T inactivation.Although the low immunogenicity of the APT0 peptide itself reduces this interference, its potential for reaction in humans remains unavoidable (Saw et al., 2021).We can mitigate the immunogenicity of CAR-T cells using an enhanced lymphocyte clearance protocol to avoid triggering the immune response of B cells after APT0 CAR-T enters the body (Gargett et al., 2016;Haas et al., 2019;Junghans et al., 2016).However, this study is only a preliminary in vitro test of the antitumor capacity and safety of APT0 CAR-T and CGS2 CAR-T cells.In the future, we will further investigate the antitumor activity and safety in vivo using mouse tumor models to explore the effect pattern of APT0 CAR-T cells and CGS2 CAR-T cells in entering mouse models.
In summary, our study constructed new APT0 CAR-T and CGS2 CAR-T cells targeting EDB-FN and expanded the cell line killing assays for liver, breast, and pancreatic cancers, further enriching the EDB CAR-T category and expanding the types of solid tumors targeting EDB-FN, which is expected to provide new treatment options for patients with solid tumor types.

Figure 1 .
Figure 1.Construction, screening, and characterization of CAR-T cells.(a) Schematic presentation of CD19, CGS2, APT0, and L19 CAR domains.(b) CAR expression on the surface of human T cells was characterized by detecting mKATE2 protein by flow cytometry on day 3. Representative results for three independent experiments.(c) Analysis of EDB-FN expression in HEK-293T and solid tumor cells.(d) Cytotoxicity of APT0 CAR-T and CGS2 CAR-T cells compared with negative control CD19 CAR-T cells and positive control L19 CAR-T cells at an effector cell: target cell ratio of 5:1 against U251 MG cells.Cytotoxicity was determined by luciferase cytotoxicity assay.(e) The levels of IFN-γ released from CD19 CAR-T, L19 CAR-T, APT0 CAR-T, and CGS2 CAR-T cells were measured by ELISA after coculture of target cells and CAR-T cells for 24 hours at an E: T ratio of 5:1.(f) Curves for accumulative proliferation generated by cell counting every three days of CAR-T cell in vitro culture.In vitro, experiments were repeated at least three times independently, and data are expressed as the mean ± SD of triplicate samples.One-way ANOVA and two-tailed student's t-test were performed for analysis (* p < .05;** p < .01;*** p < .001;**** p < .0001).

Figure 2 .
Figure 2. APT0 CAR-T and CGS2 CAR-T cells can widely lyse solid tumor cells.(a) Extensive tumor cytotoxicity of APT0 CAR-T and CGS2 CAR-T cells.APT0 CAR-T and CGS2 CAR-T cells were cocultured with multiple EDB-FN-positive tumor cells at an E: T ratio of 5:1.Cytotoxicity was determined by luciferase cytotoxicity assay.(b) The levels of IFN-γ released from CD19 CAR-T, APT0 CAR-T, and CGS2 CAR-T cells were measured by ELISA after coculture of multiple tumor cells and CAR-T cells at a 5:1 ratio for 24 hours.In vitro, experiments were repeated at least three times independently, and data are expressed as the mean ± SD of triplicate samples.One-way ANOVA was performed for analysis (**p < .01;***p < .001;****p < .0001).

Figure 3 .
Figure 3. APT0 CAR-T and CGS2 CAR-T cells lyse tumor cells in a time-dependent and gradientdependent manner.(a) APT0 CAR-T and CGS2 CAR-T (mKATE2) cells were cocultured with MDA-MB -468 (EGFP) cells (E: T ratio of 5:1) for 48 h.Cell changes were observed under inversion fluorescence microscopy.(b) Cell counting of surviving MDA-MB-468 tumor cells at 48 h of coculture.(c) APT0 CAR-T and CGS2 CAR-T cells were cultured with the indicated cell lines at different E: T ratios (1:1, 2.5:1, and 5:1) for 24 h and compared to CD19 CAR-T cells.Cytotoxicity was determined by luciferase cytotoxicity assay.Three independent experiments were performed.Data are presented as the mean ± SD of triplicates.Statistical significance was calculated using one-way ANOVA (**p < .01;***p < .001;****p < .0001).

Figure 4 .
Figure 4. APT0 CAR-T and CGS2 CAR-T cells lyse tumor cells in an EDB-FN-dependent manner with particular safety.(a) FN expression in FN-knockdown and shRNA control SK-HEP1 cells was detected by real-time PCR.(b) Cytotoxicity of APT0 CAR-T and CGS2 CAR-T cells at an E: T ratio of 5:1 cocultured with SK-HEP1-shCOO2 cells and SK-HEP1-shFN cells.(c) Cytotoxicity of APT0 CAR-T and CGS2 CAR-T cells at an E: T ratio of 5:1 cocultured with HEK-293T cells for 24 h.Cytotoxicity was determined by luciferase cytotoxicity assay.Two independent experiments were performed.Data are presented as the mean ± SD of triplicates.One-way ANOVA and two-tailed student's t-test were performed for analysis (ns: not significant; *p < .05;****p < .0001).