CNTN4/APP axis of cancer cells and T-cells

CNTN4/APP axis of cancer cells and T-cells | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article CNTN4/APP axis of cancer cells and T-cells Bu-Nam Jeon, Sujeong Kim, Yunjae Kim, Hyunkyung Yu, Hyunuk Kim, and 8 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-2979573/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Immune checkpoint inhibitors have significantly advanced tumor treatment, but their limited benefits and strong responses in only a subset of patients persist as challenges. CNTN4, a neuronal membrane protein involved in cell adhesion and synapse signaling, has unclear immunomodulatory functions. In this study, we reveal the immune checkpoint role of CNTN4 in T-cell proliferation and activation both in vitro and in vivo . We found that CNTN4, highly expressed in numerous tumor tissues, impedes T-cell proliferation, cytotoxicity, and the secretion of cytotoxic cytokines in vitro . On T cells, CNTN4 binds to two APP isoforms, APP770 and APP751, which results in attenuated TCR signaling and diminished cell adhesion capacity. To target this interaction, we developed GENA-104A16 against CNTN4 and an anti-APP antibody (5A7) that blocks the binding between CNTN4 and APP. Administering these two antibodies demonstrated anti-tumor effects in a syngeneic tumor mouse model and increased tumor-infiltrating lymphocytes within tumor tissues in vivo . Furthermore, elevated CNTN4 levels are associated with poor prognosis and negatively correlated with various cytotoxic immune-related markers. In conclusion, CNTN4 serves as a bona fide immune checkpoint protein and represents a promising therapeutic target for developing immunotherapeutic drugs. Biological sciences/Immunology/Immunotherapy/Immunosuppression Biological sciences/Cancer/Tumour immunology Biological sciences/Cancer/Cancer microenvironment Figures Figure 1 Figure 2 Figure 3 Figure 4 Main text The tumor microenvironment (TME) allows cancer cells to evade the immune system through interactions between T lymphocytes and immune checkpoints (ICPs) 1,2 . Immune checkpoint inhibitors (ICIs) have emerged as innovative cancer treatments, with the most notable targets being PD-1 (Programmed cell death-protein 1) and PD-L1 (Programmed cell death-ligand 1) 3,4 . These ICIs hold promise for activating therapeutic antitumor immunity and overcoming immune evasion. However, recent studies reveal that the response rate to ICIs remains low, and some patients exhibit resistance 5–9 . Numerous attempts have been made to address this resistance, but challenges persist 10–13 . CNTN4, a contactin family immunoglobulin member, functions as an axon-associated synaptic cell adhesion molecule in the brain as a GPI (Glycosylphosphatidylinositol)-anchored membrane protein 14 . Research has explored the relationship between CNTN4 and brain and nervous system development, finding Cntn4 expression in the olfactory bulb, thalamus, hippocampus, and cerebral cortex, suggesting that it plays an important role in neuropsychiatric phenotypes, similar to other contactins 15,16 . However, the immunomodulatory roles of CNTN4 within the TME remain unknown. In this study, we demonstrate that CNTN4 is highly expressed in various cancer tissues and suppresses T cell activation by binding to APP on T cells. Targeting the CNTN4/APP axis could enhance T-cell activation-related processes and increase anti-tumor effects in a syngeneic tumor mouse model. Furthermore, we assess the clinical implications of CNTN4 in immunotherapy-treated cancer cohorts (Fig. 1a). CNTN4 is overexpressed in tumor tissues We examined the expression of CNTN4 in normal human tissues and found that most normal tissues did not express CNTN4 (Fig. 1b; a z-score > 5 suggests gene expression in that tissue). In contrast, IHC analysis confirmed CNTN4 expression in 21 types of human cancer tissues. Notably, 13 types of human cancer tissues (lung, breast, thyroid, sarcoma, colon, bladder, prostate, skin, liver, endometrium, stomach, pancreas, and gallbladder) exhibited a high positivity (+1 or higher) rate of over 50% (Fig. 1c, d; Extended Data Fig. 1a). These data indicate that CNTN4 is expressed in various cancer tissues but not in normal tissues, potentially affecting tumor formation or growth. To investigate the characteristics of cancer cells in relation to CNTN4 expression, we utilized CNTN4 KD (Knockdown) U-2 OS cells and Cntn4 overexpression in CT26 cells (referred to as CT26/Cntn4). Both CNTN4 KD U-2 OS cells and CT26/Cntn4 cells displayed no significant differences in in vitro cellular proliferation, migration, and invasiveness according to CNTN4 expression (Extended Data Fig. 1b, c). Although CNTN4 expression did not alter cancerous characteristics in vitro , whole transcriptomic analysis of in vivo CT26 and CT26/Cntn4 tumors revealed 6,098 differentially expressed genes (FDR; False discovery rate 2). Downregulated hallmark pathways in CT26/Cntn4 tumors compared to CT26 tumors included TNF-α signaling via NF-kB (FDR = 0.034), apoptosis (FDR = 0.117), and inflammatory responses (FDR = 0.145) (Fig. 1e). These results suggest that CNTN4 in tumor tissues may reduce apoptosis and immune response sensitivity. Suppression of T cell activity by CNTN4 We next investigated whether CNTN4 could influence immune cell activity, given the reduced sensitivity of CNTN4-overexpressed tumors to immune response. While CNTN4 did not affect the differentiation or activity of B cells, NK (Natural killer) cells, DCs (Dendritic cells), or macrophages, it bound to human or mouse T-cells in a dose-dependent manner and constantly inhibited their proliferation and cytokine secretion (IFN-g and TNF-a) using plate-bound anti-CD3 antibody (Fig. 1f, Extended Data Fig. 1d). The inhibition of T-cells by CNTN4 was higher than that by PD-L1 at high concentrations, indicating that CNTN4 could be a novel immune checkpoint candidate that negatively regulates T-cell activity ure 2: GENA-104A16 Enhances T-Cell Mediated Cytotoxicity of CNTN4-Expressing Tumors in vivo. a, GENA-104A16's binding affinity to human or mouse CNTN4 measured using sandwich ELISA. EC50 values were obtained using a nonlinear four-parameter logistic curve via GraphPad PRISM. b, BioLayer Interferometry (BLI) sensorgrams demonstrating kinetic analysis of CNTN4's binding to GENA-104A16, with measurements taken during 600 seconds of association and dissociation. c, Evaluating the impact of GENA-104A16 (0.75 ~ 3.0 µM) on T-cell proliferation and cytokine secretion (IFN-γ and TNF-α) in the presence of CNTN4 on an anti-CD3 antibody pre-coated plate. d, Analysis of conjugation formation and cytotoxicity. The percentage of conjugates and apoptotic U-2 OS cells were determined using flow cytometry in the presence or absence of GENA-104A16. e, Confirmation of T-cell activation signaling pathway-related proteins' phospho-form expression through Western blotting after treatment with GENA-104A16. f, Validation of p65 or NFATc1 binding to the IFN-γ gene promoter via ChIP and qRT-PCR after an incubation period of 12 hours. g, h, i, Assessment of tumor growth kinetics in mice injected with MC38 (g), CT26 (h), or CT26/Cntn4 (i) cells, and treated with intraperitoneal injections of GENA-104A16. j, k, Analysis of tumor-infiltrating cell populations and marker gene expressions in CD8 + T-cells from CT26/Cntn4 tumor tissues treated with GENA-104A16, using single-cell RNA sequencing. l, Tumor growth kinetics in mice injected with B16F10 cells and treated with ICIs. All statistical analyses were conducted using GraphPad PRISM and involved one-way or two-way ANOVA, with Tukey's multiple tests further determining significance. *, p < 0.05; **, p < 0.01; ***, p < 0.005; ****, p < 0.001 and compared to the isotype antibody treatment #, p < 0.05; ##, p < 0.01; ###, p < 0.005; ####, p < 0.001. GENA-104 neutralizes CNTN4-induced suppression of T-cell To examine whether CNTN4 blockade could reverse its T-cell suppression, we generated a monoclonal antibody against CNTN4 (named GENA-104A16) and determined its binding affinity to human or mouse CNTN4. GENA-104A16 bound to both human and mouse CNTN4 (with over 95% homology between them), exhibiting EC 50 values of 0.166 nM and 0.167 nM, respectively (Fig. 2a). The bio-layer interferometry (BLI)-derived KD of 90 pM for the recombinant CNTN4 protein and GENA-104A16 indicated high-affinity antigen-binding capability (Fig. 2b). Testing GENA-104A16 binding to CNTN4-overexpressed HEK293FT cells yielded an EC 50 value of 0.316 nM (Extended Data Fig. 2a), and GENA-104A16 neutralized CNTN4-mediated T-cell inhibition in a dose-dependent manner (Fig. 2c). Additionally, GENA-104A16 treatment dose-dependently increased CD3 + T/U-2 OS conjugates and CD3 + T-cell cytotoxicity against U-2 OS (Fig. 2d). In the case of CNTN4 KD U-2 OS cells, conjugates and CD3 + T-cell cytotoxicity were higher than those of U-2 OS cells, regardless of GENA-104A16 treatment (Fig. 2d). These results demonstrate that GENA-104A16 is a potent inhibitor of CNTN4 binding to T-cells, with sub-nanomolar inhibitory activity. To elucidate the T-cell regulation mechanisms of CNTN4 and GENA-104A16, we assessed their effects on human TCR signaling. CNTN4 inhibited the phosphorylation of TCR signaling-related proteins (ZAP70, PI3K, LCK, ERK1/2, JNK, and p38), and GENA-104A16 neutralized this phenomenon (Fig. 2e and Extended Data Fig. 2b). CNTN4 also impeded the translocation of transcription factors (p65 and NFATc1) from the cytoplasm to the nucleus upon T-cell activation, which GENA-104A16 treatment counteracted (Extended Data Fig. 2c). ChIP assays revealed that CNTN4 inhibited the binding of transcription factors to target gene promoters (IFN-g and TNF-a) in human T-cells stimulated by anti-CD3 antibody, and GENA-104A16 neutralized the decrease in transcription factor binding to these sites (Fig. 2f, Extended Data Fig. 2d-g). These findings demonstrate that GENA-104A16 promotes T-cell activation and TCR signaling cascades which are inhibited by CNTN4. GENA-104A16 promotes killing of CNTN4-expressing tumors We subsequently examined the in vivo effectiveness of GENA-104A16. Mice received subcutaneous injections of MC38, CT26, or CT26/Cntn4 cells, and intraperitoneal injections of mIgG1 or GENA-104A16.mIgG1 (1 mg/kg or 3 mg/kg) five times at three-day intervals. In MC38 with no CNTN4 expression, GENA-104A16 had no effect on tumor reduction (Fig. 2g). The treatment of GENA-104A16 has led to a significant reduction in CT26 tumor size, a decrease that becomes even more pronounced when CNTN4 is overexpressed, as demonstrated in CT26/Cntn4 tumors (Fig. 2h, i). These results reveal that the efficacy of GENA-104A16 is dependent on CNTN4 expression. In light of Cntn4 expression modulating T-cell activation, we analyzed the immunological status associated with Cntn4 expression through FACS analysis and whole transcriptomic analysis in tumor tissues. GENA-104A16 treatment increased the population of CD4 + CD3 + /CD45 + T-cells and decreased the population of Tregs (Foxp3 + CD25 + /CD4 + ) in CT26 cancer tissues compared to mIgG1 treatment, with more pronounced changes in CT26/Cntn4 cancer tissues (Extended Data Fig. 3a, b). Intriguingly, CT26/Cntn4 tumors exhibited a significantly higher tumor-infiltrating Treg population than CT26, indicating that elevated CNTN4 expression promotes immune evasion (Extended Data Fig. 3a, b). Consistent with FACS analysis, whole transcriptomic analysis revealed that TNF-a signaling and IFN-g response pathway gene sets tended to increase in GENA-104A16-treated CT26/Cntn4 compared to control IgG1-treated CT26/Cntn4, while CT26/Cntn4 tumors reduced sensitivity to immune response compared to CT26 tumors (Extended Data Fig. 3c). We performed single-cell RNA (scRNA) sequencing and analysis to further assess the characteristics of TILs in CT26/Cntn4 tumors from mice treated with GENA-104A16 (Extended Data Fig. 4a–d). We obtained 17,563 cells and calculated gene expression counts for each single cell, then distinguished them into non-lymphocytes and lymphocytes based on Ptprc (Cd45) gene expression, a lymphocyte marker. The 15,082 cells with low Ptprc expression (count < 1) were classified as non-lymphocytes and divided into 12 clusters, while the remaining 6,098 cells expressing Ptprc (count ≥ 1) were classified as lymphocytes and divided into 11 clusters. Immune cells were classified into CD4 + T-cells, CD8 + T-cells, monocytes, macrophages, NK cells, pDC, and neutrophils by profiling the expression of variable genes in each cluster. Interestingly, a higher proportion of TILs was observed in GENA-104A16 treated tumors (18.63%) compared to mIgG1 treated tumors (9.57%) (Fig. 2j). Both CD4 + T-cells (0.34% vs. 1.24%) and CD8 + T-cells (2.81% vs. 4.18%) were more abundant in GENA-104A16-treated tumors (Fig. 2j), with most of these CD8 + cells expressing cytotoxic markers (Fig. 2k). These results suggest that GENA-104A16 can neutralize CNTN4-mediated T-cell suppression, enhance T-cell activity, and sequentially activate the cancer-immunity cycle, thereby increasing the number of TILs. Since the effectiveness of many immunotherapeutic agents as monotherapy options has been a recent major concern, we sought to investigate the potential for combining GENA-104A16 with traditional therapies or other immunotherapeutic agents to achieve maximum therapeutic benefit for a wide range of patients 17 . We examined the antitumor effect of GENA-104A16 in combination with anti-PD-L1 or anti-PD-1 antibodies. Since CT26 tumors exhibit a significant single effect of GENA-104A16, determining whether the combination with other ICIs is synergistic or additive is challenging. Therefore, B16F10 cells, which express minimal CNTN4, were subcutaneously injected into both flanks of C57BL/6 mice, and intraperitoneally injected with antibodies (mIgG1, GENA-104A16.mIgG1, anti-PD-L1 antibody, or a combination of GENA-104A16.mIgG1 and anti-mPD-L1 antibodies) four times at three-day intervals. While GENA-104A16 administration alone did not reduce B16F10 tumor volume, the combination of anti-mPD-L1 antibody and GENA-104A16 further increased the reduction in tumor volume (Fig. 2l, left). Similarly, the combination of anti-mPD-1 antibody and GENA-104A16 resulted in a significant decrease in tumor volume, whereas monotherapy did not have a significant effect (Fig. 2l, right). These findings suggest that GENA-104A16 may enhance the antitumor effects when combined with other immune checkpoint inhibitors. CNTN4 interaction with APP on T-cells We identified a receptor on T-cells that binds to CNTN4 through receptor screening, which revealed interactions with two APP isoforms (APP770 and APP751) (Fig. 3a). The CNTN4 protein specifically bound to APP-overexpressed HEK293FT cells, akin to PD-1/PD-L1 binding, and vice versa (Fig. 3b). Cell-to-cell conju­­­­­­­­gation assays also demonstrated increased conjugate numbers in APP-overexpressed and CNTN4-overexpressed HEK293FT cells (Fig. 3c). To map interaction domains, we created a DNA construct containing APP isoforms (770, 751, and 695) and CNTN4 C-terminal deletion constructs (#4, #6, #8, and #10) (Fig. 3d), and used a Co-IP assay in HEK293FT cells expressing these constructs. Results showed that CNTN4 bound APP770 and APP751 containing a KPI domain, while APP bound CNTN4 #8 and #10 containing FN (Fibronectin) 1–2 (Fig. 3e). The highest number of conjugates was observed with APP751 and CNTN4 #8 (Fig. 3f). Binding of CNTN4 recombinant protein to cells expressing APP751, but not APP695, and recombinant APP protein to cells expressing CNTN4 #8, but not CNTN4 #6, confirmed that the FN 1–2 domains of CNTN4 physically interact with the KPI domain of APP (Fig. 3g). Inhibition of APP enhances T - cell activation mechanisms We observed increased APP expression in CD4 + and CD8 + T-cells following anti-CD3 antibody stimulation (Fig. 3h and Extended Data Fig. 5a). We then investigated the influence of APP expression on T-cell function by generating App KO mice using CRISPR-Cas9 (Extended Data Fig. 5b). App KO T-cells exhibited increased adhesion to CT26 cells and heightened cytotoxicity compared to App WT T-cells (Fig. 3i). T-cell activation-related processes were induced by anti-CD3 antibody stimulation in both App WT and KO T-cells, but CNTN4 inhibited these processes only in App WT T-cells, not in App KO T-cells (Fig. 3j and Extended Data Fig. 5c-f). CNTN4 inhibited luciferase activity in Jurkat-Lucia TM NFAT cells by approximately 35% when stimulated by anti-CD3 antibody-treated T-cells, but inhibition increased to 82-84% in cells overexpressing full-length APP or intracellular domain-deleted APP (APPΔICD) (Fig. 3k and Extended Data Fig. 5g). These findings indicate that T-cell inhibition by CNTN4 occurs due to decreased adhesion between cancer cells and T-cells caused by the extracellular region of APP rather than changes caused by the intracellular region of APP. We produced an anti-APP antibody (5A7) through immunization and hybridoma generation technologies, which bound APP with an EC 50 of 1.864 nM (Fig. 3l). The CNTN4/APP binding was specifically inhibited by the anti-APP antibody (5A7) (IC 50 = 478 nM) and GENA-104A16 (IC 50 = 0.212 nM) (Fig. 3m). We also evaluated the in vivo effectiveness of the anti-APP antibody (5A7). CT26 cells were subcutaneously injected into both flanks of BALB/c or BALB/c nude mice, and intraperitoneally injected with mIgG1 or anti-APP antibody (5A7) three times at 3-day intervals. The anti-APP antibody (5A7) reduced CT26 tumor volume in the syngeneic mouse model but not in the immunodeficient nude mouse model lacking T-cells (Fig. 3n). This result demonstrates the potent inhibitory effect of the anti-APP antibody (5A7) on APP binding to CNTN4. Clinical implications of CNTN4 expression Generally, elevated PD-L1 levels are associated with a higher likelihood of responsiveness to ICIs; however, some patients with high PD-L1 levels still exhibit a non-responsive phenotype. Consequently, we posited that the refractory response to ICIs observed in PD-L1 positive patients might be linked to CNTN4 expression. To investigate this hypothesis, we assessed the differences in CNTN4 levels between responders and non-responders within a cohort of patients with similarly high PD-L1 expression (Fig. 4a). Among PD-L1 high patients in the immunotherapy-treated gastric cancer cohort, non-responders displayed significantly higher CNTN4 levels compared to responders (p = 0.0148) (Fig. 4b). A similar trend was observed in the immunotherapy-treated lung cancer cohort, with weak statistical significance (p = 0.167) (Fig. 4c). We also confirmed the relationship between various cytotoxic immune-related markers and CNTN4 levels within the PD-L1 high group. Cntn4 was negatively correlated with immune-related cytotoxic markers such as Gzmb, Ifng, Prf1, Cxcl9, and Cxcl10 (Fig. 4d). Further GSEA analysis using hallmark gene sets revealed that biopsy samples from the low CNTN4 expression group exhibited enriched signatures for TNF-α signaling via NF-κB and IFN-γ response (Fig. 4e), consistent with results from mouse tumor tissues depending on CNTN4 expression (Fig. 1e and Extended Data Fig. 3c). Higher CNTN4 levels correlated with poor prognosis in terms of progression-free survival (PFS) and overall survival (OS) (Fig. 4f). These results show that high CNTN4 expression might contribute to the refractory response to ICIs even in high PD-L1 patients, impacting immune-related pathways and survival probability. In the immunotherapy-treated gastric cancer cohort, 18 out of 19 high CNTN4 group patients were non-responders, while 8 out of 18 low CNTN4 group patients were responders. The composition of responders and non-responders based on CNTN4 and PD-L1 levels showed that the responding group primarily consisted of CNTN4-low and PD-L1-high patients (66.7% in responders), while the non-responding group mainly included CNTN4-high patients regardless of PD-L1 levels (64.3% in non-responders) (Fig. 4g). This result suggests that a high PD-L1 level generally implies effective ICI treatment; however, resistance may occur due to high CNTN4 levels even with high PD-L1 levels. Therefore, CNTN4 antibodies should be used first in patients with high CNTN4 levels, regardless of PD-L1 levels. Lastly, we found that CNTN4 was not expressed on any immune cells, unlike APP, which is a CNTN4 binding partner on T-cells, according to the Database of Immune Cell Expression (DICE) (Fig. 4h and Extended Data Fig. 6). In addition, we analyzed PD-L1 and CNTN4 expression using 17 TMA slides and summarized the expression scores for each patient and cancer tissue as a heatmap. We observed CNTN4 in more cancer tissues than PD-L1, with its expression levels exceeding those of PD-L1 (Fig. 4i). These results support that CNTN4 is a tumor-specific target, and that targeting CNTN4 may be and effective treatment strategy for various types of cancer. Discussion CNTN4 as a Novel Immune Checkpoint Protein Our study discovered that CNTN4, expressed in various human tumor tissues, inhibits T-cell activity, indicating its potential role as an inhibitory ICP. We identified APP on T-cells as a binding partner for CNTN4, and upon TCR stimulation, APP expression increased, similar to other ICPs. While the interaction between CNTN4 and APP has been previously reported, our study is the first to establish the CNTN4/APP axis in the TME. This interaction weakens T-cell activity and TCR signaling cascades by reducing conjugation between cancer cells and T-cells. We demonstrated that targeting the CNTN4/APP axis in a syngeneic tumor mouse model reduced tumor growth by enhancing anti-tumor immunity. Clinical analysis revealed an increase in CNTN4 levels among non-responders compared to responders and a negative correlation with various cytotoxic immune-related markers in PD-L1 high patients. High CNTN4 levels were also associated with poor prognosis. In addition, we observed CNTN4 in more cancer tissues than PD-L1, with its expression levels exceeding those of PD-L1. Thus, we propose that blocking the CNTN4/APP interaction could suppress tumor growth by effectively activating T-cells, and the development of drugs targeting CNTN4 could be particularly useful for non-responders to PD-1/PD-L1 blockade and for treating numerous cancer patients. Combining GENA-104A16 with Existing ICI Therapies Immune checkpoint inhibitors (ICIs) have revolutionized oncology by improving survival rates for many cancer patients. However, there are limitations, such as low response and resistance. Hence, new ICPs and target molecules are being studied to expand the use and efficacy of existing ICI therapies. Our study found that the anti-CNTN4 monoclonal antibody GENA-104A16 neutralized T-cell inhibition by CNTN4, demonstrating its anti-tumor and immune-boosting effects. We also confirmed that anti-PD-L1 treatment induced CNTN4 expression in a CT26 syngeneic tumor mouse model and that CNTN4 expression was upregulated in anti-PD-1 non-responders according to clinical cohort analysis. These results suggest that GENA-104A16 could serve as a novel ICI treatment for patients and could be combined or administered sequentially with PD-1/PD-L1 blockade to overcome acquired resistance. Safety Considerations for CNTN4/APP Blockade CNTN4 has been characterized as a key cell adhesion molecule for axon guidance and neural connectivity in neurodevelopment, and previous studies have shown that CNTN4 is a risk gene associated with several neuropsychiatric disorders. APP belongs to the type I transmembrane protein family and has also been extensively studied for its role in the brain. Considering the potential side effects of CNTN4/APP blockade on the nervous system, we confirmed that CNTN4 interacts with APP751 and APP770 isoforms, which are expressed in most tissues, while APP695 isoform is mainly expressed in neurons. This indicates that CNTN4/APP interaction occurs less frequently in the brain and nervous system, and blocking this interaction may have fewer adverse effects on the nervous system. In addition, the expression of CNTN4 in some normal tissues was partially confirmed by IHC, but the expression was negative overall (Extended Data Fig. 7). Thus, it is suggested that a strategy to block the interaction with APP by targeting CNTN4 can reduce side effects on the nervous system and act specifically in cancer. In conclusion, CNTN4 expressed in cancer cells may act as a novel ICP that attenuates TCR signaling by interacting with APP in T-cells. In addition, GENA-104A16 and anti-APP antibody (5A7) can be therapeutic agents that remove tumors by increasing immune activity by suppressing the CNTN4/APP axis. Therefore, the development of new immune checkpoint inhibitory antibodies capable of inhibiting the interaction between CNTN4 and APP may lead to the discovery of therapeutic agents for patients who cannot respond to immunotherapy. Materials & Methods Reagents and antibodies To produce recombinant antibodies, the plasmid vectors were transiently expressed in ExpiCHO-S cells with the ExpiFectamine CHO Kit (ThermoFisher Scientific, Rockford, IL), followed by culturing in ExpiCHO Expression medium at 32–37°C for 7–12 days in shaking flasks and a CO 2 incubator equipped with an orbital shaker. Specifically, 100 µg plasmid vectors (heavy and light chains were cloned into pcDNA3.4 expression vector, Addgene, Watertown, MA) and 800 µL ExpiFectamin CHO Reagent were mixed with OptiPRO SFM (Gibco, UK) medium (19.2 mL final volume) and allowed to stand for 5 min at 27°C. Subsequently, the mixed solution was added to pre-adjusted 6 × 10 6 cells and incubated at 37°C with the condition of 8% CO 2 in humidified air to transfect DNA into cells. At 18 to 22 h post-transfection, 1.5 mL Enhancer 1- and 40-mL Feed, components of the ExpiFectamine CHO Kit, were added to each culture flask. On day 3 post-transfection, 40 mL of feed was added to each flask. Each antibody was purified for several chromatography steps from the cell culture supernatant using an AKTA Avant system (Cytiva, Sweden). First, a protein A column containing Hitrap MabSelect SuRe (Cytiva) was used to capture the antibodies, and Capto S ImpAct and/or Capto adhere chromatography (Cytiva) was used to polish and improve their purities. Purified proteins were concentrated in the formulation buffer by ultrafiltration using an Amicon Ultra 15 (50 kDa cutoff size, Merck, Germany). To confirm the purity of the prepared antibodies, samples were analyzed by SDS-PAGE and size exclusion HPLC (Agilent, Palo Alto, CA) with a TSKgel G3000SWXL column (Tosoh, Japan). The biotinylated antibody was prepared according to the procedure recommended by the manufacturer of the biotinylation kit (Pierce, Rockford, IL). Antibodies for biotinylation were washed five times with 1× PBS, pH 7.4 (Corning Costar, Corning, NY) using Amicon Ultra 0.5 (Merck) to remove sources of primary buffer, followed by a concentration above 2.0 mg/mL. Then, 157 µL biotin reagent (dissolved in 1× PBS, pH 7.4) per 1 mg antibody was added to the above sample and incubated for 1 h with gentle rotation. Non-reacted biotin was removed by ultrafiltration using an Amicon Ultra 4 (Merck). Anti-human CD3, anti-human CD28, anti-mouse CD3, and anti-mouse CD28 antibodies were acquired from R&D Systems (Minneapolis, MN). anti-His, anti-ZAP70, anti-p-ZAP70, anti-LCK, anti-p-LCK, anti-ERK1/2, anti-p-ERK1/2, anti-JNK, anti-p-JNK, anti-p38, anti-p-p38, anti-β-actin, anti-p65, anti-NFATc1, anti-GAPDH, anti-Histone H3, HRP-conjugated anti-mouse or rabbit IgG, anti-rabbit or mouse IgG antibodies were obtained from Cell Signaling Technology (Danvers, MA). IL-2 recombinant protein (rIL-2) was acquired from Roche (Switzerland). CellTracker CMFDA-green, CMRA-orange, BMQC-violet, CFSE-green, and Lipofectamine 2000 were obtained from Invitrogen (Carlsbad, CA). Human and mouse PD-L1 and CNTN4 (His or Fc tag) recombinant proteins were acquired from Sino Biological, Inc. (China). Human APP (His or Fc tag) recombinant proteins were acquired from BioLegend (San Diego, CA). The plasmid DNA purification kit and WEST-ZOL western Blot Detection kit were purchased from iNtRON Biotechnology (Korea). PrimeSTAR HS DNA polymerase was purchased from TaKaRa Bio, Inc. (Japan). Restriction enzymes and ligases were purchased from New England Biolabs Inc. (Beverly, MA). Unless otherwise stated, all chemical reagents were purchased from Sigma-Aldrich. Cells Jurkat-Lucia™ NFAT (TIB-152, ATCC, Manassas, VA), HEK293FT (CVCL_6911, ThermoFisher Scientific), CT26 (CRL-2638, ATCC), B16F10 (CRL-6475, ATCC), and U-2 OS (HTB-96, ATCC) cells were maintained in RPMI-1640, McCoy’s 5A, or Dulbecco’s modified Eagle’s medium (Gibco) supplemented with 10% (v/v) FBS (Gibco). CT26/Cntn4 cells were purchased from GenScript (China). To establish stable CNTN4 KD U-2 OS cell lines, CNTN4 three sgRNAs (Thermo Fisher Scientific) and TrueCut™ Cas9 Protein v2 (Thermo Fisher Scientific) were electroporated with Neon™ Transfection system (Invitrogen) into U-2 OS cells. Human PBMCs were isolated using the Ficoll-Paque Plus (GE Healthcare Life Sciences, Piscataway, NJ). Mouse bone marrow was harvested by flushing using syringe-containing media after cutting off the ends of the femurs and tibiae. Naïve CD3 + , CD4 + , or CD8 + T-cells were purified from the mouse spleen or human PBMCs by negative selection using a T-cell isolation kit (Miltenyi Biotec, Germany). The purity of each population was confirmed to be more than 95% using flow cytometry. To generate clonal expansion T-cells, naïve T-cells were incubated in 2 µg/mL anti-CD3/28 coated culture plates with 50 U/mL rIL-2 for 48 h and cultured for an additional five days with 50 U/mL rIL-2. Animals C57BL/6 and BALB/c mice were purchased from Orient Bio Inc. (Korea). All the mice were housed under specific pathogen-free conditions. All experimental methods and protocols were approved by the Institutional Animal Care and Use Committee of the GenNBio Animal Research Center and were carried out in accordance with the approved guidelines. cDNA constructs To generate the CNTN4 and APP constructs, CNTN4 and APP clones coding for the full-length open reading frame were purchased from OriGene (Rockville, MD). CNTN4 #4, #6, #8, APP751, APP695, and APPΔICD genes were generated by standard or overlapping PCR and subcloned into the pCMV6-Entry vector (Origene). Immunohistochemistry assay (IHC) Various tissue slides (provided by SuperBioChips) were incubated with anti-CNTN4 (Abcam, Cambridge, MA) or anti-PD-L1 (Ventana, Tucson, AZ) antibody, followed by diaminobenzidine staining (Ventana), according to the manufacturer’s instructions. The stained slides were photographed using an Aperio AT2 Scanner (Leica, Germany) and recorded by a pathologist. Flow cytometric analysis Cells were suspended in FACS buffer (2% FBS and 1 mM EDTA in PBS). For CNTN4 expression, GENA-104A16 or hIgG isotype antibody was added, and APC-conjugated hIgG secondary antibody (BioLegend) was sequentially added for 30 min at 4°C. For APP expression, anti-2D5-Biotin or mIgG-Biotin antibodies were added, and PE-conjugated Streptavidin secondary antibody (BioLegend) was sequentially treated for 30 min at 4°C. For protein binding, His-tagged recombinant protein was added and FITC-conjugated anti-His secondary antibody (BioLegend) was sequentially added for 30 min at 4°C. The cells were then assessed on a FACSCanto II (BD Biosciences, San Jose, CA), and the data were analyzed using Flowjo software (Treestar, San Carlos, CA). Characterization (proliferation, migration, and invasion) of cancer cells with different CNTN4 expression Cancer cells were plated, divided, and grown in 6-well plates for 0–4 days. Cell proliferation was measured by cell counting using a Luna cell counter (Logos Biosystems, Korea). For migration studies, a standard assay was used to determine the number of cells that traversed a porous polycarbonate membrane (8 µm pore size) in response to a chemo-attractant (higher serum concentration) using the Cytoselect 24-well cell migration assay (Cell Biolabs, San Diego, CA). Invasion was measured using the BD BioCoat Matrigel invasion chambers (BD Biosciences). In both assays, the cancer cells were suspended in serum-free media. The cells (3 × 10 5 /chamber) were seeded in an upper chamber in serum-free media for migration and invasion assays. The lower chamber was filled with media containing 10% FBS. After incubation for 18 h (invasion) or 40 h (migration) at 37°C in a 5% CO 2 incubator, cells passing through polycarbonate membrane were stained for 10 min using the provided staining solution, washed, and dried. Stained membranes were incubated for 10 min with extraction buffer, and the lysates were measured using a microplate reader (SpectraMax M2; Molecular Devices, Sunnyvale, CA) with an optical density (OD) of 560 nm. SoftMax Pro 7 software was used for the analysis. All reactions were performed in triplicates. Tumorgraft For the tumor growth experiments, BALB/c or C57BL/6 mice were subcutaneously injected with 1 × 10 6 CT26, CT26/Cntn4, or B16F10 cells per mouse. Tumor size was measured thrice a week until the endpoint, and tumor volume was calculated as length × width 2 × 0.5. When the average tumor size reached a volume of approximately 100 mm 3 , the experimental treatments were initiated. Tumor-bearing mice were intraperitoneally injected with mIgG1, anti-mPD-L1, anti-mPD-1 (BioXCell, West Lebanon, NH), GENA-104A16. mIgG1 or anti-APP antibody (5A7) on days 0, 3, 6, 9, and 12. Whole transcriptomic analysis of tumor tissues The tumor tissues were gently removed using forceps, and the surgical scissors were transferred to PBS. RNA was extracted using an RNA extraction kit (Qiagen, Germany). Next, 151-bp paired-end libraries were constructed from 1 µg of RNA using the TruSeq RNA Sample Prep Kit v2 (Illumina). Whole transcriptome sequencing (WTS) was performed using an Illumina HiSeq instrument. RNA-seq reads from each WTS experiment were aligned to the mouse reference genome (GRCm38) using STAR 18 aligner. Gene expression was quantified by RSEM 19 . TPM values were used to identify gene sets enriched in the tumors. GSEA (Gene set enrichment analysis) was performed using the Java GSEA desktop application (GSEA v2.1.0) 20 . Gene sets that were upregulated or downregulated with an FDR < 0.25 were considered significant. In vitro plate-bound T cell proliferation or neutralization assay Anti-CD3 antibody (human: 4 µg/mL, mouse: 3 µg/mL) and CNTN4 or PD-L1 recombinant protein (human: 50–200 nM, mouse: 37.5 ~ 150 nM) were diluted in PBS and incubated overnight for coating at 37°C on 96-well non-treated cell culture plates. For the neutralization assay, the plate was washed with PBS and incubated at 37°C for 1 h after the antibody treatment. T-cells were stained with CFSE at 37°C for 10 min, washed with the culture medium, and then added to the culture at 2 × 10 5 T-cells/well on a cell culture plate. After 72 h, dividing cells were evaluated on a FACSCanto II, and data were analyzed using FlowJo software. Measure of cytokine secretion The amount of human or mouse IFN-γ or TNF-α in the T cell culture supernatants was determined by ELISA using a Duo Set ELISA kit (R&D System). The manufacturer's recommendations were followed. 96-well plates were coated with 100 µL/well of capture antibody overnight and blocked with 100 µL/well of ELISA diluent in D.W. 100 µL of the samples were applied to the wells in duplicates. For the standard curve, a recombinant standard was used at concentrations between 2,000 pg/mL and 62.5 ng/mL. Streptavidin-HRP conjugate was diluted 1:40 in the substrate solution. The reaction was stopped by adding 50 µL/well of 2 N H 2 SO 4 and the plates were analyzed at 450 nm using a SpectraMax M2 microplate reader. SoftMax Pro 7 software was used for the analysis. All reactions were performed in triplicates. Affinity of GENA-104A16 against recombinant proteins or overexpressing HEK293FT cells by indirect ELISA or flow cytometry, respectively To measure the affinity of antibodies against recombinant proteins, 96-well ELISA plates were coated with 5 nM of human or mouse recombinant protein at 4°C overnight. The coated ELISA plate was washed three times and blocked with PBST (PBS containing 0.05% Tween 20). A tenfold serial dilution of antibodies (seven or eleven points) in PBS was added, incubated for 1 h at 37°C, and then washed three times with PBST. HRP-conjugated anti-human or anti-mouse antibodies were added (1:3,000 dilution in PBST), incubated for 1 h at 37°C, and washed three times with PBST. ABTS substrate solution (2,2ʹ-azinobis [3-ethylbenzothiazoline-6-sulfonic acid]-diammonium salt, ThermoFisher Scientific) was added and incubated for 5 min at 27°C. Absorbance at 405 nm was determined using a microplate reader (BioTek Synergy 4, Winooski, VT). The half-maximal binding concentration (EC 50 ) value was determined with a nonlinear four-parameter logistic curve using GraphPad PRISM 9.3.1 (GraphPad Software Inc., San Diego, CA). To measure the affinity of GENA-104A16 for HEK293FT cells, CNTN4 overexpression HEK293FT cells (3 × 10 5 ) were washed three times with PBS and resuspended in 500 µL of FACS buffer. The threefold serial dilution of GENA-104A16 (12.54 pM ~ 66.67 nM, 14 points) in FACS buffer was added, incubated for 1 h at 4℃, and then washed three times with FACS buffer. APC-conjugated anti-human IgG was added (1:400 dilution in FACS buffer), incubated for 30 min at 4°C, and washed thrice with FACS buffer. Cells were assessed on a FACSCanto II, and the data were analyzed using FlowJo software. MFI ( M ean F luorescent I ntensity) values were obtained from the histograms and were used to plot the binding curves. Dose-response curves were generated by plotting the resulting MFI against the log concentration of GENA-104A16, and the EC 50 value was analyzed by nonlinear regression curve fitting using GraphPad PRISM. Biolayer interferometry Binding kinetic measurements were performed using the Octet R8 system (Sartorius, Germany) at 30℃ at 1,000 rpm agitation (Sartorius). GENA-104A16 was loaded onto anti-human Fc biosensor 2nd generation biosensor (AHC2) at 1.0 µg/mL in running buffer for 0.02% PBST for 5 min. Subsequently, the tips were transferred to running buffer for 180 s for sensor rinsing. For kinetic analyses, the association with CNTN4 (varying concentrations ranging from 51.9 ~ 415.4 ng/mL in running buffer) was measured for 600 s, followed by dissociation for 600 s (in running buffer). In each experiment, one negative control (reference well or blank) was measured, where the capture antibody was incubated with running buffer instead of the antigen. Data fitting and analysis were performed with the Octet®BLI Analysis 12.2 program (Sartorius) using a 1:1 binding model after Savizky-Golay filtering. Immune cell profiling For immune profiling, the tumors were harvested on day 4. Tumors were chopped (Dorco, Korea) and transferred in RPMI-1640 media supplemented with 0.25 mg/mL hyaluronidase Type IV-S, 50 µg/mL DNase type 1, 2.5 mg/mL collagenase type 1, 1.5 mg/mL collagenase type 2, and 1 mg/mL collagenase type 4. The samples were then incubated at 37°C for 50 min and filtered using a 70 µm cell strainer (BD Biosciences). Tumor cells (1 × 10 6 cells/well) were treated with anti-mouse CD16/32 (BD Biosciences) at 4°C for 10 min to block the Fc receptor. Surface staining was conducted and BD Pharmingen™ Mouse Foxp3 Buffer Set (BioLegend) solution was added. Then, intracellular staining was performed. Stained cells were acquired using FACSCanto II, and the data were analyzed using FlowJo software. The antibody information was as follows: APC/Cy7 anti-mouse CD45 antibody (BioLegend, 103116), Pacific blue anti-mouse CD3 antibody (Biolegend, 100334), PE/Cy7 anti-mouse CD4 antibody (BioLegend, 100422), PE anti-mouse CD8 antibody (BioLegend, 100708), APC anti-mouse CD25 antibody (BioLegend, 101910), PE anti-mouse FoxP3 antibody (Biolegend, 126404), GENA-104A16, APC/Cy7 Rat IgG2b, κ isotype control antibody (Biolegend, 400624), Pacific blue Rat IgG2b, κ isotype control antibody (Biolegend, 400925), PE/Cy7 Rat IgG2b, κ isotype control antibody (Biolegend, 400618), PE Rat IgG2b, κ isotype control antibody (Biolegend, 400608), APC Rat IgG2b, κ isotype control antibody (Biolegend, 400612), hIgG4 (Sino Biological Inc., HG4K), Goat anti-Human IgG (H + L) Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 647 (Invitrogen, A21445) scRNA analysis for population of tumor-infiltrating immune cells Library preparation and pre-processing scRNA-seq The tumor tissues were gently removed using forceps, and the surgical scissors were transferred to PBS. Tumor tissues were placed in a petri dish and finely chopped with a razor blade. The chopped tissues were transferred to a 15 mL conical tube, filled with 5 mL of 1⋅ tumor dissociation buffer (Miltenyi Biotec), diluted in RPMI, and incubated at 37°C in a shaking incubator for 30 min. The cells were suspended in 3 mL RBC lysis buffer (BioLegend) and incubated for 5 min at 27°C. Cells were washed with PBS and resuspended in CELLBANKER (AMSBIO, Cambridge, MA). Library preparation and sequencing were performed by Macrogen (Korea). The library was prepared using Chromium Next Gem Single cell kit v3 and Single Cell 3ʹ v3 Gel Beads (10⋅ Genomics), and the library was sequenced using the Macrogen sequencing system. The raw data were processed using CellRanger v3.1 with default settings. Reads with the same cell barcode, UMIs ( U nique M olecular I dentifier), and genes were grouped to calculate the number of UMIs per gene per cell using the “count” command. The UMI count tables for each cellular barcode were used for further analyses. The matrix data from CellRanger were processed using the Seurat package (version 4.3.0) in R software (version 4.2.1) for each individual sample. To prepare the Seurat object, we first processed gene expression data from each sample using the Read 10⋅ function, and then ran the CreateSeuratObject function with metadata. Whole Seurat objects were integrated into one object to reduce the batch effect and perform post-analyses. Low-quality cells were discarded from the whole single-cell data for post-analysis. Cells with less than 200 expressed genes were discarded. Cells with expression levels lower than the 5th percentile or higher than the 95th percentile were discarded. Cells a mitochondrial gene expression percentage was higher than 15% were discarded. Data integration, clustering, and annotation The merged Seurat object, including quality-checked cells, was split individually by treatment type and then normalized using the SCTransform function with the top 2,000 highly variable genes. To obtain highly variable genes, the SelectIntegrationFeatures function was used. Cell clustering and UMAP visualization were performed using FindClusters and RunUMAP (Uniform manifold approximation and projection), respectively. The integrated and clustered cells were then annotated manually. First, lymphocytes were identified based on the expression of Ptprc (CD45). If the cell’s expression of Ptprc was ≥ 1, it was annotated as a lymphocyte and the other cells were annotated as non-lymphocytes. The lymphocytes were sub-annotated using the following gene expression profiles: CD8 T ( CD3g , CD8a ), CD4 T ( Cd3g , Cd4 ), natural killer ( Ncr1 ), macrophages ( Adgre1 ), dendritic cells ( Siglech ), monocytes ( Cd14 , Csf1r ), and neutrophils ( Cd14 , Cxcr2 ). If there are some cells could not be annotated by those markers, they were annotated as ‘Others’ or ‘Un-annotated’ The expression of target genes such as cytotoxic markers and immune checkpoints was assessed. Cytotoxic markers ( Ifng, Gzma, Gzmb, Tnf, Prf1 ) were assessed in CD8 + T-cells and immune checkpoints ( Pdcd1, Havcr2, Tigit, Ctla4, Lag3, Icos ) were assessed in CD4 + T-cells. Conjugation assay T-cell parts (PD-1, APP, APP751, and APP695 overexpressing HEK293FT cells, mouse CD3 + T-cells, and human CD3 + T-cells, 3 × 10 5 ) and target part cells (PD-L1, CNTN4, CNTN4 #6, or CNTN4 #8 overexpressing HEK293FT cells, CT26 cells, and U-2 OS or CNTN4 KD U-2 OS cells, 3 × 10 5 ) were stained with CMFDA and CMRA or Violet BMQC, respectively, according to the manufacturer’s protocol. For conjugation, equal volumes of T -cell parts and target part cells were mixed and incubated with or without GENA-104A16 for indicated time points in a humidified 5% CO 2 incubator at 37℃. The relative proportions of green, orange, violet, and green-orange events or green-violet events in each tube were determined by two-color flow cytometry using a FACSCantoII and analyzed with FlowJo software. The number of gated events counted per sample was at least 10,000. The percentage of conjugated T-cells was determined as the number of dual-labeled (CMFDA- and CMRA-positive or CMFDA- and BMQC-positive) events divided by the number of CMFDA-positive T-cells. All reactions were performed in duplicates. In vitro cytotoxicity assay Human PBMCs or mouse splenocyte and bone marrow (1 × 10 8 ) were incubated with U-2 OS or CT26 (1 × 10 7 ) lysates for 24 h, respectively, in a humidified 5% CO 2 incubator at 37℃. Human or mouse CD3 + T-cells were purified from the mixture by positive selection using a T-cell isolation kit (EasySep, StemCell Technologies, Seattle, WA) and then subjected to clonal expansion. For flow cytometric cytotoxicity assays, cancer cells were washed in PBS and stained with CMFDA, according to the manufacturer’s protocol. T-cells were mixed CMFDA-labeled cancer cells (E:T = 10:1 or 5:1) and incubated for indicated time points in a humidified 5% CO 2 incubator at 37℃. After incubation, 10 µL of a 5 µg/mL solution of 7-AAD (BioLegend) was added to the cell suspension and incubated for 10 min on ice. Cells were evaluated on a FACSCanto II, and data were analyzed using FlowJo software. Western blotting Cells were lysed in ice-cold lysis buffer (50 mM Tris-HCl, pH 7.4, containing 150 nM NaCl, 1% Triton X-100, and one tablet of complete protease inhibitors) for 15 min on ice. Cell lysates were centrifuged at 16,000 × g for 30 min at 4°C, and the supernatants were eluted with SDS sample buffer (100 mM Tris-HCl, pH 6.8, 4% SDS, and 20% glycerol with bromophenol blue) and heated for 5 min. The proteins were separated by SDS-PAGE on 10 ~ 15% gels and were transferred to PVDF (Bio-Rad, Richmond, CA) membranes using a Mini Trans-Blot Cell (Bio-Rad). The membrane was blocked in 5% BSA (1 h), rinsed, and incubated with the appropriate antibodies in TBST (Tris-buffered saline containing 0.1% Tween 20) and 0.5% BSA overnight. Excess primary antibody was removed by washing the membrane thrice with TBST. The membranes were then incubated with 0.1 µg/mL peroxidase-conjugated secondary antibodies (anti-rabbit or anti-mouse) for 30 min. After three washes with TBST, bands were visualized using western blotting detection reagents and exposed to LOURMAT CHEMI-DOC (Eppendorf, Germany). Subcellular fractionation Cells were washed with PBS, harvested, and resuspended in hypotonic buffer (20 mM Tris-HCl, pH 7.4, 10 mM KCl, 2 mM MgCl 2 , 1 mM EGTA, 0.5 mM DTT, 0.5 mM PMSF, and one tablet of complete protease inhibitors), incubated for 5 min followed by the addition of NP-40 to a final concentration of 0.1%. After 3 min of incubation, the cytoplasm and nucleus were separated by centrifugation at 800 × g for 8 min. Subsequently, to ensure the removal of nuclear remnants, the cytoplasmic fractions were centrifuged at 1,500 × g for 5 min, and the supernatant was collected as the final cytoplasmic fraction. The nucleus was purified by 10 min incubation in isotonic lysis buffer (20 mM Tris-HCl, pH 7.4, 150 mM KCl, 2 mM MgCl 2 , 1 mM EGTA, 0.3% NP-40, 0.5 mM DTT, 0.5 mM PMSF, and one tablet of complete protease inhibitors), and centrifuged at 700 × g for 7 min. Chromatin-IP assay (ChIP) Cells were fixed with formaldehyde (final 1%) for 30 min, harvested by scraper, and centrifuged at 13,000 rpm for 10 min at 4℃. Pellet was resuspended in 800 µL of SDS lysis buffer (1% SDS, 10 mM EDTA, and 50 mM Tris-HCl, pH 8.0). To shear chromatin, the lysate was sonicated on ice for 3 min with sonicator (Bandelin, Germany) tip of 3 mm in diameter at 30% amplitude and 0.5 cycle. Samples were centrifuged at 13,000 rpm for 10 min at 4℃, and 200 µL of supernatant was divided into aliquots for subsequent 10-fold dilution in ChIP dilution buffer (0.01% SDS, 1% Triton X-100, 1.2 mM EDTA, pH 8.0, 167 mM NaCl, and 16.7 mM Tris-HCl, pH 8.0). To test the amount of input DNA for each sample, a 20 µL diluted aliquot was saved for further processing in parallel with all other samples at the reversal of the cross-linking step. Every 2 mL sample of chromatin was precleared with 60 µL of 50% slurry of protein A/G agarose beads (v/v) containing 200 µg/mL herring sperm DNA for 1 h at 4℃ on rotating wheel, after which the beads were pelleted and the supernatant was transferred to a new tube. IgG, anti-p65, or anti-NFATc1 antibody (10 µg) was added to the precleared chromatin sample and incubated overnight at 4°C on a rotating wheel. Immune complexes were collected using 60 µL of 50% slurry of protein A/G agarose beads, 200 µg/mL herring sperm DNA, and rotating for 3 h at 4℃, followed by centrifugation at 1,000 rpm for 1 min at 4℃. The beads were washed for 5 min in low salt wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 500 mM NaCl, and 20 mM Tris-HCl, pH 8.0), LiCl wash buffer (0.25 M LiCl, 1% NP-40, 1% sodium deoxycholate, 1 mM EDTA, and 10 mM Tris-HCl, pH 8.0), and twice with TE buffer. Chromatin complexes were eluted from the beads in 30 min with 400 µL of elution buffer (1% SDS and 0.1 M NaHCO 3 ) at 27°C. To reverse cross-linking, 200 mM NaCl was added, and the samples were incubated at 65°C for 4 h. To digest proteins, samples were incubated at 45°C for 90 min after the addition of the following: 10 mM EDTA, 50 µg/mL proteinase K, and 40 mM Tris, pH 6.5. Samples were extracted twice with phenol/chloroform/isoamyl alcohol (25:24:1) and DNA was precipitated with 20 µg of glycogen and two volumes of 100% ethanol. Pellets were collected by centrifugation at 13,000 rpm for 10 min at 4°C. The samples were resuspended in 100 µL of deionized water and stored at -80°C. Real-time quantitative RT-PCR (qRT-PCR) Real-time qPCR was conducted in a CFX384 Touch Real-Time PCR Detection System (Bio-Rad, Hercules, CA) using a TOPreal™ SYBR Green RT-qPCR High-ROX Kit (Enzynomics, Korea) and gene-specific primers (forward and reverse pairs, respectively) as follows: p65 binding site on Ifn-γ gene promoter, 5ʹ-AGTAGGTATTTTACTAATCAC-3ʹ and 5 ʹ -AGGAAACTCTTGGGCTTCTCA-3 ʹ; Nfatc1 binding site on Ifn-γ gene promoter, 5 ʹ -TCATGGTTTGAGAAGCCCAAG-3 ʹ and 5 ʹ -TATGGTTTTGTGGCATGTTAG-3 ʹ; negative control region on Ifn-γ gene promoter, 5 ʹ -GTAAGTATGAATTCTTAATAA-3 ʹ and 5 ʹ -ACTCCGTAGTAAGTTGGACAG-3 ʹ, distal region on Tnf-α gene promoter, 5ʹ -AGGAGTGGGAGGGTGGGGGAG-3ʹ and 5ʹ -GGGAGACATGATATTGAGGAG-3ʹ, proximal region on Tnf-α gene promoter, 5ʹ -TGTCCCATTTAGAAATCAAAA-3ʹ; and 5ʹ -AAAAGCTCTCATTCAACCCTC-3ʹ, negative control region on Tnf-α gene promoter, 5ʹ -GTGAGTGTCTGGGCAACCCTT-3ʹ and 5ʹ -CGTTCATTCATCTCTCTGTGC-3ʹ, p65 binding site on IFN-γ gene promoter, 5ʹ -CTAGGCTGGTCTCAAACTCCT-3ʹ and 5ʹ -ATCAATATACTACATTGTTAA-3ʹ, NFATc1 binding site on IFN-γ gene promoter, 5ʹ -TTGTTCCCAACCACAAGCAAA-3ʹ and 5ʹ -GATGAGACAGACCCATTATGC-3ʹ, negative control region on IFN-γ gene promoter, 5ʹ -GTAAGTATGACTTTTTAATAG-3ʹ and 5ʹ -CTCTATACAAACTGAGCAGAA-3ʹ, p65 binding site #1 on TNF-α gene promoter, 5ʹ -CGGGGCTGTCCCAGGCTTGTC-3ʹ and 5ʹ -GATAGAACTAGAACTGGGAGG-3ʹ, p65 binding site #2 on TNF-α gene promoter, 5ʹ -TTTTCCTGCATCCTGTCTGGA-3ʹ and 5ʹ -CATCAAGGATACCCCTCACA-3ʹ, NFATc1 binding site on TNF-α gene promoter, 5ʹ -TTGTGTGTCCCCAACTTTCCA-3ʹ and 5ʹ -GGCTGGGTGTGCCAACAACTG-3ʹ, negative control region on TNF-α gene promoter, 5ʹ -GTGAGTGCCTGGCCAGCCTTC-3ʹ and 5ʹ -ATCCTGTCTCTCCATCTTTCT-3ʹ. IgG was used as a negative control. The immunoprecipitated DNA was normalized using the following formula: immunoprecipitated DNA = 2 −(ΔCt of the immuno−precipitated DNA − ΔCt of input) , where Ct is the threshold cycle value. In each sample, immunoprecipitated DNA was normalized to the input. Immunoprecipitation (IP) Cell lysates were precleared, and supernatants were incubated overnight with antibodies at 4°C, followed by incubation with protein A/G agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA). The beads were collected, washed with PBS, and resuspended in equal volumes of 5× SDS loading buffer. Immunoprecipitated proteins were separated by SDS-PAGE on 12% gels and analyzed by western blotting as described above. Competitive binding assay using the GENA-104A16 The 96-well ELISA plate was coated with 10 nM of the target recombinant protein at 4°C overnight. The coated ELISA plate was washed thrice and blocked with PBST. A five-fold serial dilution of GENA-104A16 or anti-APP antibody (5A7) (0.0064–500 nM, eight points) and 300 nM His-tagged binding recombinant protein in PBS was added, incubated for 1 h at 4°C, and then washed three times with PBST. Anti-His antibody was added, incubated at RT, and washed three times with PBST. Peroxidase-conjugated anti-mouse IgG antibody was added (1:1,000 dilution in PBST), incubated for 1 h at 4°C, and washed three times with PBST. TMB substrate solution (3,3ʹ,5,5ʹ -tetramethylbenzidine, ThermoFisher Scientific) was added and incubated for 10 min at RT. The reaction was stopped with H 2 SO 4 and absorbance was measured at 405 nm using a microplate reader. Binding (%) of CNTN4 and APP = (O.D. value of APP and CNTN4 binding in the presence of GENA-104A16) / (O.D. value of APP and CNTN4 binding in the absence of GENA-104A16) × 100 The IC 50 value was determined with normalized nonlinear logistic curve using the GraphPad PRISM. Statistical analyses Statistical analyses were performed using Prism 9.3.1 (GraphPad) or R software (version 4.2.1). Differences between two variables and multiple variables were assessed using the Student’s t -test and ANOVA with Tukey's multiple comparison test, respectively. Associations between two continuous variables were estimated using Pearson’s correlation test. Associations between two discrete variables were estimated using the chi-square test. Differences were considered significant if the p -value was less than 0.05. All statistical methods and significance thresholds are described in the corresponding figure legends. Declarations Disclosure of Potential Conflict of Interest No potential conflicts of interest were disclosed. Author Contributions B.-N.J designed, performed, analyzed the experiments, and prepared the manuscript; S.K and Y.K wrote the manuscript and analyzed experimental results and clinical data; M.C, K.A.P, C.L, and H.P designed the experiments and supervised the manuscript; H.Y, H.K, Y.H, and Y.Y.K, performed the experiments; C.P, and G.K analyzed the bioinformatic data. All the authors have revised the manuscript accordingly. Acknowledgements This research was supported by the Korea Drug Development Fund funded by the Ministry of Science and ICT, the Ministry of Trade, Industry, and Energy, and the Ministry of Health and Welfare (RS-2022-00166157, RS-2023-00217717, Republic of Korea). References Ceeraz, S., Nowak, E. C., Burns, C. M. & Noelle, R. J. Immune checkpoint receptors in regulating immune reactivity in rheumatic disease. Arthritis Res Ther 16 , 1–12 (2014). Pardoll, D. M. The blockade of immune checkpoints in cancer immunotherapy. Nature Reviews Cancer 2012 12:4 12 , 252–264 (2012). Sharpe, A. H., Wherry, E. J., Ahmed, R. & Freeman, G. J. The function of programmed cell death 1 and its ligands in regulating autoimmunity and infection. Nature Immunology vol. 8 239–245 Preprint at https://doi.org/10.1038/ni1443 (2007). Dong, H. et al. Tumor-associated B7-H1 promotes T-cell apoptosis: A potential mechanism of immune evasion. Nature Medicine 2002 8:8 8 , 793–800 (2002). Syn, N. L., Teng, M. W. L., Mok, T. S. K. & Soo, R. A. De-novo and acquired resistance to immune checkpoint targeting. Lancet Oncol 18 , e731–e741 (2017). Johnson, D. B., Chandra, S. & Sosman, J. A. Immune Checkpoint Inhibitor Toxicity in 2018. JAMA 320 , 1702–1703 (2018). Martins, F. et al. New therapeutic perspectives to manage refractory immune checkpoint-related toxicities. Lancet Oncol 20 , e54–e64 (2019). Wilkinson, R. W. & Leishman, A. J. Further advances in cancer immunotherapy: Going beyond checkpoint blockade. Front Immunol 9 , 1082 (2018). Tang, T. et al. Advantages of targeting the tumor immune microenvironment over blocking immune checkpoint in cancer immunotherapy. Signal Transduction and Targeted Therapy 2021 6:1 6 , 1–13 (2021). Liu, D. et al. Integrative molecular and clinical modeling of clinical outcomes to PD1 blockade in patients with metastatic melanoma. Nature Medicine 2019 25:12 25 , 1916–1927 (2019). Ren, D. et al. Predictive biomarkers and mechanisms underlying resistance to PD1/PD-L1 blockade cancer immunotherapy. Mol Cancer 19 , 1–19 (2020). Wang, Q. & Wu, X. Primary and acquired resistance to PD-1/PD-L1 blockade in cancer treatment. Int Immunopharmacol 46 , 210–219 (2017). Jenkins, R. W., Barbie, D. A. & Flaherty, K. T. Mechanisms of resistance to immune checkpoint inhibitors. British Journal of Cancer 2018 118:1 118 , 9–16 (2018). Hansford, L. M. et al. Cloning and characterization of the human neural cell adhesion molecule, CNTN4 (alias BIG-2). Cytogenet Genome Res 101 , 17–23 (2003). Mercati, O. et al. Contactin 4, -5 and -6 differentially regulate neuritogenesis while they display identical PTPRG binding sites. Biol Open 2 , 324–334 (2013). Oguro-Ando, A. et al. Cntn4, a risk gene for neuropsychiatric disorders, modulates hippocampal synaptic plasticity and behavior. Translational Psychiatry 2021 11:1 11 , 1–15 (2021). Barbari, C. et al. Immunotherapies and Combination Strategies for Immuno-Oncology. International Journal of Molecular Sciences 2020, Vol. 21, Page 5009 21 , 5009 (2020). Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29 , 15–21 (2013). Li, B. & Dewey, C. N. RSEM: Accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics 12 , 1–16 (2011). Subramanian, A. et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 102 , 15545–15550 (2005). Additional Declarations There is NO Competing Interest. Supplementary Files ExtendedDataFigures.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-2979573","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":205126212,"identity":"3447c006-df3d-48f8-b327-470312ee5e48","order_by":0,"name":"Bu-Nam Jeon","email":"","orcid":"","institution":"Genome and Company","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Bu-Nam","middleName":"","lastName":"Jeon","suffix":""},{"id":205126213,"identity":"43a386ac-ba2d-4bec-8e51-3ff19afe0ce5","order_by":1,"name":"Sujeong Kim","email":"","orcid":"","institution":"Gwangju Institute of Science and Technology (GIST)","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Sujeong","middleName":"","lastName":"Kim","suffix":""},{"id":205126214,"identity":"0ed60149-5a8e-4bc3-ab6d-b92eaff58f71","order_by":2,"name":"Yunjae Kim","email":"","orcid":"","institution":"Gwangju Institute of Science and Technology (GIST)","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Yunjae","middleName":"","lastName":"Kim","suffix":""},{"id":205126215,"identity":"1dd1870b-0620-4309-8693-d5fff8dcf062","order_by":3,"name":"Hyunkyung Yu","email":"","orcid":"","institution":"Genome\u0026Company","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Hyunkyung","middleName":"","lastName":"Yu","suffix":""},{"id":205126216,"identity":"10f672ec-f921-4ae0-a65e-4d7ed0341a90","order_by":4,"name":"Hyunuk Kim","email":"","orcid":"","institution":"Genome\u0026Company","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Hyunuk","middleName":"","lastName":"Kim","suffix":""},{"id":205126217,"identity":"51233c9d-cef5-483e-b5e2-7fa6eb38a54d","order_by":5,"name":"Youngeun Ha","email":"","orcid":"","institution":"Genome\u0026Company","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Youngeun","middleName":"","lastName":"Ha","suffix":""},{"id":205126218,"identity":"64588b1e-c9ce-4a6b-98ed-30694ed3dc17","order_by":6,"name":"Yun Yeon Kim","email":"","orcid":"","institution":"Genome and Company","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Yun","middleName":"Yeon","lastName":"Kim","suffix":""},{"id":205126219,"identity":"bbf2ca3d-2c66-45ff-9299-4b35051da613","order_by":7,"name":"Changho Park","email":"","orcid":"","institution":"Genome\u0026Company","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Changho","middleName":"","lastName":"Park","suffix":""},{"id":205126220,"identity":"eb6a13d9-af5d-46a1-8f00-8a98134f2abe","order_by":8,"name":"Gihyeon Kim","email":"","orcid":"","institution":"Genome\u0026Company","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Gihyeon","middleName":"","lastName":"Kim","suffix":""},{"id":205126221,"identity":"d18241dd-1692-4991-ae77-c5e756b38ae9","order_by":9,"name":"Miyoung Cha","email":"","orcid":"","institution":"Genome\u0026Company","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Miyoung","middleName":"","lastName":"Cha","suffix":""},{"id":205126222,"identity":"c3eb6c1e-ef84-4970-8fc2-860e55417bd8","order_by":10,"name":"Karolina Palucka","email":"","orcid":"","institution":"The Jackson Laboratory for Genomic Medicine","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Karolina","middleName":"","lastName":"Palucka","suffix":""},{"id":205126223,"identity":"52928409-ee6e-413d-8d50-d53d00862c83","order_by":11,"name":"Charles Lee","email":"","orcid":"","institution":"The Jackson Laboratory for Genomic Medicine","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Charles","middleName":"","lastName":"Lee","suffix":""},{"id":205126224,"identity":"9bf0b0cc-348e-421e-b84b-da251550a719","order_by":12,"name":"Hansoo Park","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6UlEQVRIiWNgGAWjYBACAzB5wKa+n4GxASLEQ5yWNMaZDUAtB0jQcphxA0g5UVrM+c8e/FxxhpnZ+Pzi5s8fGOzkGXjOPsCrxXJGXrLkmRtsbGY3HrZJHGBINmzgbTfA77AbPAaSDR94eMxuHGwDOow5gYGfjYBfzp8x/tnwQULCeMbB5g8HGOqJ0HIgx0yy4YaBgQF/YwPQYYcTGHjbCGi5kZdm2XAmIUHiBmObxBmD44ZtPMcIOezs4ZsNx/4n8Pcff/yhoqJanp8nDb8WRDRIJDCAo4mAT5C18B8grHYUjIJRMApGJgAAxEtKBd0QvRcAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0003-1027-8850","institution":"Gwangju Institute of Science and Technology (GIST)","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Hansoo","middleName":"","lastName":"Park","suffix":""}],"badges":[],"createdAt":"2023-05-25 07:15:44","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-2979573/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-2979573/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":38307126,"identity":"688d574f-caf2-48ec-8e11-ae78dd30f813","added_by":"auto","created_at":"2023-06-09 18:18:58","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":7651791,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCNTN4 Overexpression in Tumors and Impact on Immune Response.\u003c/strong\u003e a, Identification of CNTN4 as a novel immune checkpoint target and therapeutic antibody development process targeting CNTN4. b, Examination of CNTN4 mRNA expression across various normal human tissues, with the majority showing low CNTN4 expression (z-scores \u0026lt; 5) as per BioGPS data derived from U133plus2 Affimetrix microarray. c, Validation of CNTN4 IHC staining in diverse human cancer tissues using a verified IHC antibody, examined across 17 TMA slides encompassing approximately 900 human tumors. d, Percentage of cancer samples presenting a CNTN4 IHC score of +1 or above across various human cancer types. e, Comprehensive transcriptomic analysis of CT26 and CT26/Cntn4 tumors; GSEA for Hallmark databases depicted in the left panel and a heatmap displaying the expression patterns of genes linked to each hallmark pathway in the right panel. f, Investigation of CNTN4 effects on human T-cells: human blood-derived CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T-cells were activated with anti-CD3/28 antibodies (2 µg/mL, respectively) and treated with rIL-2. The binding of CNTN4 protein (His tagged) was determined by flow cytometry, with mean fluorescence intensity (MFI) presented in bar graph format. Additionally, the impact of CNTN4 and PD-L1 recombinant proteins (50 ~ 200 nM) on T-cell proliferation and cytokine release (IFN-γ and TNF-α) was studied over 72 hours, with results analyzed by FACS and ELISA respectively. Statistical analyses were conducted using GraphPad PRISM and involved one-way ANOVA, with Tukey's multiple tests further ascertaining significance. The significance is denoted *, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05; **, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01; ***, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.005; ****, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001 for comparisons to anti-CD3 only, and as #, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05; ##, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01; ###, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.005; ####, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001 for comparisons to PD-L1 at the same concentration.\u003c/p\u003e","description":"","filename":"Fig.1.png","url":"https://assets-eu.researchsquare.com/files/rs-2979573/v1/e0550788b88df016f8f41b9e.png"},{"id":38307129,"identity":"679b8a8c-ae80-4442-bf0a-3e90892b7b30","added_by":"auto","created_at":"2023-06-09 18:18:58","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":3889215,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGENA-104A16 Enhances T-Cell Mediated Cytotoxicity of CNTN4-Expressing Tumors in vivo.\u003c/strong\u003e a, GENA-104A16's binding affinity to human or mouse CNTN4 measured using sandwich ELISA. EC50 values were obtained using a nonlinear four-parameter logistic curve via GraphPad PRISM. b, BioLayer Interferometry (BLI) sensorgrams demonstrating kinetic analysis of CNTN4's binding to GENA-104A16, with measurements taken during 600 seconds of association and dissociation. c, Evaluating the impact of GENA-104A16 (0.75 ~ 3.0 µM) on T-cell proliferation and cytokine secretion (IFN-γ and TNF-α) in the presence of CNTN4 on an anti-CD3 antibody pre-coated plate. d, Analysis of conjugation formation and cytotoxicity. The percentage of conjugates and apoptotic U-2 OS cells were determined using flow cytometry in the presence or absence of GENA-104A16. e, Confirmation of T-cell activation signaling pathway-related proteins' phospho-form expression through Western blotting after treatment with GENA-104A16. f, Validation of p65 or NFATc1 binding to the IFN-γ gene promoter via ChIP and qRT-PCR after an incubation period of 12 hours. g, h, i, Assessment of tumor growth kinetics in mice injected with MC38 (g), CT26 (h), or CT26/Cntn4 (i) cells, and treated with intraperitoneal injections of GENA-104A16. j, k, Analysis of tumor-infiltrating cell populations and marker gene expressions in CD8\u003csup\u003e+\u003c/sup\u003e T-cells from CT26/Cntn4 tumor tissues treated with GENA-104A16, using single-cell RNA sequencing. l, Tumor growth kinetics in mice injected with B16F10 cells and treated with ICIs. All statistical analyses were conducted using GraphPad PRISM and involved one-way or two-way ANOVA, with Tukey's multiple tests further determining significance. *, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05; **, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01; ***, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.005; ****, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001 and compared to the isotype antibody treatment #, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05; ##, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01; ###, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.005; ####, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Fig.2.png","url":"https://assets-eu.researchsquare.com/files/rs-2979573/v1/ae6df9d97544c097ff6d26c7.png"},{"id":38307723,"identity":"84793c8e-50fc-4434-a1f7-5e31d0912dc2","added_by":"auto","created_at":"2023-06-09 18:26:58","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":7634299,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eModulation of T-Cell Activities and Enhancement of Anti-Tumor Effects through APP Targeting, CNTN4's Binding Partner.\u003c/strong\u003e a, Screening of human CNTN4 for binding across 5,647 human plasma membrane proteins. b, Evaluation of protein binding in HEK293FT cells transfected with PD-1, APP, PD-L1, or CNTN4 expression plasmids, assessed via flow cytometry after sequential treatments. c, Analysis of conjugation formation in transfected HEK293FT cells, using flow cytometry to determine the percentage of conjugates. d, Illustration of the structures of CNTN4 deletion constructs and APP isoforms. e, Co-immunoprecipitation analysis of APP isoforms and CNTN4 deletion constructs derived from transfected HEK293FT cells. f, Evaluation of conjugation formation in transfected HEK293FT cells, expressing different isoforms of APP and CNTN4 deletion constructs. g, Assessment of protein binding in transfected HEK293FT cells overexpressing various isoforms of APP and CNTN4 constructs. h, APP expression analysis in mouse clonal expansion CD4\u003csup\u003e+\u003c/sup\u003e and CD8a\u003csup\u003e+\u003c/sup\u003e T-cells stimulated with anti-CD3 antibody, assessed via FACS. i, Conjugation formation and cytotoxicity analysis in CT26 cells incubated with App WT or KO CD3\u003csup\u003e+\u003c/sup\u003e T-cells. j, Assessment of T-cell proliferation and cytokine secretion in CD4\u003csup\u003e+\u003c/sup\u003e T-cells isolated from App WT or KO mouse spleen and incubated with PD-L1 or CNTN4 recombinant protein. k, Evaluation of Jurkat-LuciaTM NFAT cell activity after transfection with APP or APPΔICD expression plasmid and subsequent incubation with CNTN4 recombinant protein. l, Measurement of the binding affinity of anti-APP antibody (5A7) to APP using sandwich ELISA. m, Evaluation of anti-APP antibody (5A7) or GENA-104A16's inhibitory effect on CNTN4 and APP binding through competitive ELISA. n, Examination of tumor growth kinetics in mice injected with CT26 cells and treated with intraperitoneal injections of anti-APP antibody (5A7). All statistical analyses were conducted using GraphPad PRISM, employing one-way ANOVA, followed by Tukey's multiple tests for significance determination.\u003cstrong\u003e \u003c/strong\u003eCompared to anti-CD3 only *, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05; **, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01; ***, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.005; ****, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001; compared to the \u003cem\u003eApp\u003c/em\u003e WT T-cells #, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05; ##, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01; ###, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.005; ####, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001; n.s., not significant\u003c/p\u003e","description":"","filename":"Fig.3.png","url":"https://assets-eu.researchsquare.com/files/rs-2979573/v1/4dea3fda90db858ab6f63d83.png"},{"id":38308408,"identity":"a828de4f-8688-456b-a497-569581cd7709","added_by":"auto","created_at":"2023-06-09 18:34:58","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":4091025,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eClinical Significance of CNTN4 Expression in Immunotherapy-Treated Cancer Cohorts. \u003c/strong\u003ea, Schematic outlining the classification of immunotherapy-treated cancer cohorts into responders and non-responders based on response evaluation criteria. Patients were further stratified into PD-L1-high and -low groups, and CNTN4 levels were compared between responders and non-responders in the PD-L1-high groups. b, c, Comparison of relative CNTN4 expression levels between responders and non-responders in the PD-L1-high groups in gastric cancer cohort (b) and lung cancer cohort (c). d, Demonstration of the negative correlation between CNTN4 and various immune-related markers in the PD-L1-high groups of the gastric cancer cohort. e, Gene Set Enrichment Analysis (GSEA) conducted between the CNTN4-high and -low groups, stratified by the median CNTN4 value in the gastric cancer cohort. f, Comparison of progression-free survival and overall survival between the CNTN4-high and -low groups in the gastric cancer cohort, again divided by the median CNTN4 value. g, Presentation of differences in the composition of responders and non-responders between CNTN4-high and -low groups, including schematic representations of CNTN4 and PD-L1 expression within each responder and non-responder group. h, Examination of CNTN4 expression across several immune cell types utilizing the Database of Immune Cell Expression (DICE). i, A heatmap displaying CNTN4 and PD-L1 expression scores through Immunohistochemical (IHC) analysis for each cancer patient, with corresponding CNTN4 and PD-L1 IHC staining of human cancer tissues. Statistical analyses involved the use of Student’s t-test and chi-squared test (b, c), Pearson's correlation test (d), and long-rank test (g).\u003c/p\u003e","description":"","filename":"Fig.4.png","url":"https://assets-eu.researchsquare.com/files/rs-2979573/v1/e9ed5952d1bb7540bad9e273.png"},{"id":39278978,"identity":"ad0394cf-c60c-4be7-8833-a9226d40ef22","added_by":"auto","created_at":"2023-06-29 07:32:01","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2402630,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-2979573/v1/5653466f-9a99-4223-80ab-7cc9efa0bae6.pdf"},{"id":38307130,"identity":"6c09cb94-16b7-4cb5-835e-25ffa53ba776","added_by":"auto","created_at":"2023-06-09 18:18:58","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":17795501,"visible":true,"origin":"","legend":"","description":"","filename":"ExtendedDataFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-2979573/v1/0478de4324396bf0e5d1a291.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"CNTN4/APP axis of cancer cells and T-cells","fulltext":[{"header":"Main text","content":"\u003cp\u003eThe tumor microenvironment (TME) allows cancer cells to evade the immune system through interactions between T lymphocytes and immune checkpoints (ICPs)\u003csup\u003e1,2\u003c/sup\u003e. Immune checkpoint inhibitors (ICIs) have emerged as innovative cancer treatments, with the most notable targets being PD-1 (Programmed cell death-protein 1) and PD-L1 (Programmed cell death-ligand 1)\u003csup\u003e3,4\u003c/sup\u003e. These ICIs hold promise for activating therapeutic antitumor immunity and overcoming immune evasion. However, recent studies reveal that the response rate to ICIs remains low, and some patients exhibit resistance\u003csup\u003e5\u0026ndash;9\u003c/sup\u003e. Numerous attempts have been made to address this resistance, but challenges persist\u003csup\u003e10\u0026ndash;13\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eCNTN4, a contactin family immunoglobulin member, functions as an axon-associated synaptic cell adhesion molecule in the brain as a GPI (Glycosylphosphatidylinositol)-anchored membrane protein\u003csup\u003e14\u003c/sup\u003e. Research has explored the relationship between CNTN4 and brain and nervous system development, finding Cntn4 expression in the olfactory bulb, thalamus, hippocampus, and cerebral cortex, suggesting that it plays an important role in neuropsychiatric phenotypes, similar to other contactins\u003csup\u003e15,16\u003c/sup\u003e.\u003csup\u003e\u0026nbsp;\u003c/sup\u003eHowever, the immunomodulatory roles of CNTN4 within the TME remain unknown.\u003c/p\u003e\n\u003cp\u003eIn this study, we demonstrate that CNTN4 is highly expressed in various cancer tissues and suppresses T cell activation by binding to APP on T cells. Targeting the CNTN4/APP axis could enhance T-cell activation-related processes and increase anti-tumor effects in a syngeneic tumor mouse model. Furthermore, we assess the clinical implications of CNTN4 in immunotherapy-treated cancer cohorts (Fig. 1a).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCNTN4 is overexpressed in tumor tissues\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe examined the expression of CNTN4 in normal human tissues and found that most normal tissues did not express CNTN4 (Fig. 1b; a z-score \u0026gt; 5 suggests gene expression in that tissue). In contrast, IHC analysis confirmed CNTN4 expression in 21 types of human cancer tissues. Notably, 13 types of human cancer tissues (lung, breast, thyroid, sarcoma, colon, bladder, prostate, skin, liver, endometrium, stomach, pancreas, and gallbladder) exhibited a high positivity (+1 or higher) rate of over 50% (Fig. 1c, d; Extended Data Fig. 1a). These data indicate that CNTN4 is expressed in various cancer tissues but not in normal tissues, potentially affecting tumor formation or growth.\u003c/p\u003e\n\u003cp\u003eTo investigate the characteristics of cancer cells in relation to CNTN4 expression, we utilized \u003cem\u003eCNTN4\u003c/em\u003e KD (Knockdown) U-2 OS cells and Cntn4 overexpression in CT26 cells (referred to as CT26/Cntn4). Both \u003cem\u003eCNTN4\u003c/em\u003e KD U-2 OS cells and CT26/Cntn4 cells displayed no significant differences in \u003cem\u003ein vitro\u003c/em\u003e cellular proliferation, migration, and invasiveness according to CNTN4 expression (Extended Data Fig. 1b, c). Although CNTN4 expression did not alter cancerous characteristics \u003cem\u003ein vitro\u003c/em\u003e, whole transcriptomic analysis of \u003cem\u003ein vivo\u003c/em\u003e CT26 and CT26/Cntn4 tumors revealed 6,098 differentially expressed genes (FDR; False discovery rate \u0026lt; 0.1 \u0026amp; Fold change \u0026gt;2). Downregulated hallmark pathways in CT26/Cntn4 tumors compared to CT26 tumors included TNF-\u0026alpha; signaling via NF-kB (FDR = 0.034), apoptosis (FDR = 0.117), and inflammatory responses (FDR = 0.145) (Fig. 1e). These results suggest that CNTN4 in tumor tissues may reduce apoptosis and immune response sensitivity.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSuppression of T cell activity by CNTN4\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe next\u0026nbsp;investigated whether CNTN4 could\u0026nbsp;influence immune cell\u0026nbsp;activity, given\u0026nbsp;the\u0026nbsp;reduced\u0026nbsp;sensitivity of CNTN4-overexpressed tumors to immune response. While\u0026nbsp;CNTN4 did not affect the differentiation\u0026nbsp;or\u0026nbsp;activity of B cells, NK (Natural killer) cells, DCs (Dendritic cells),\u0026nbsp;or\u0026nbsp;macrophages, it\u0026nbsp;bound to human or mouse T-cells in a dose-dependent manner and\u0026nbsp;constantly\u0026nbsp;inhibited\u0026nbsp;their\u0026nbsp;proliferation and cytokine secretion (IFN-g\u0026nbsp;and TNF-a)\u0026nbsp;using plate-bound anti-CD3 antibody (Fig. 1f, Extended Data Fig. 1d). The inhibition of T-cells\u0026nbsp;by CNTN4\u0026nbsp;was higher than\u0026nbsp;that by\u0026nbsp;PD-L1 at high concentrations, indicating\u0026nbsp;that CNTN4 could be a\u0026nbsp;novel\u0026nbsp;immune checkpoint candidate that negatively regulates T-cell activity\u003cstrong\u003eure 2: GENA-104A16 Enhances T-Cell Mediated Cytotoxicity of CNTN4-Expressing Tumors in vivo.\u003c/strong\u003e a, GENA-104A16\u0026apos;s binding affinity to human or mouse CNTN4 measured using sandwich ELISA. EC50 values were obtained using a nonlinear four-parameter logistic curve via GraphPad PRISM. b, BioLayer Interferometry (BLI) sensorgrams demonstrating kinetic analysis of CNTN4\u0026apos;s binding to GENA-104A16, with measurements taken during 600 seconds of association and dissociation. c, Evaluating the impact of GENA-104A16 (0.75 ~ 3.0 \u0026micro;M) on T-cell proliferation and cytokine secretion (IFN-\u0026gamma; and TNF-\u0026alpha;) in the presence of CNTN4 on an anti-CD3 antibody pre-coated plate. d, Analysis of conjugation formation and cytotoxicity. The percentage of conjugates and apoptotic U-2 OS cells were determined using flow cytometry in the presence or absence of GENA-104A16. e, Confirmation of T-cell activation signaling pathway-related proteins\u0026apos; phospho-form expression through Western blotting after treatment with GENA-104A16. f, Validation of p65 or NFATc1 binding to the IFN-\u0026gamma; gene promoter via ChIP and qRT-PCR after an incubation period of 12 hours. g, h, i, Assessment of tumor growth kinetics in mice injected with MC38 (g), CT26 (h), or CT26/Cntn4 (i) cells, and treated with intraperitoneal injections of GENA-104A16. j, k, Analysis of tumor-infiltrating cell populations and marker gene expressions in CD8\u003csup\u003e+\u003c/sup\u003e T-cells from CT26/Cntn4 tumor tissues treated with GENA-104A16, using single-cell RNA sequencing. l, Tumor growth kinetics in mice injected with B16F10 cells and treated with ICIs. All statistical analyses were conducted using GraphPad PRISM and involved one-way or two-way ANOVA, with Tukey\u0026apos;s multiple tests further determining significance.\u0026nbsp;*, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05; **, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01; ***, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.005; ****, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001 and compared to the isotype antibody treatment #, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05; ##, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01; ###, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.005; ####, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGENA-104 neutralizes CNTN4-induced suppression of T-cell\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo examine whether CNTN4 blockade could reverse its T-cell suppression, we generated a monoclonal antibody against CNTN4 (named GENA-104A16) and determined its binding affinity to human or mouse CNTN4. GENA-104A16 bound to both human and mouse CNTN4 (with over 95% homology between them), exhibiting EC\u003csub\u003e50\u003c/sub\u003e values of 0.166 nM and 0.167 nM, respectively (Fig. 2a). The bio-layer interferometry (BLI)-derived KD of 90 pM for the recombinant CNTN4 protein and GENA-104A16 indicated high-affinity antigen-binding capability (Fig. 2b). Testing GENA-104A16 binding to CNTN4-overexpressed HEK293FT cells yielded an EC\u003csub\u003e50\u003c/sub\u003e value of 0.316 nM (Extended Data Fig. 2a), and GENA-104A16 neutralized CNTN4-mediated T-cell inhibition in a dose-dependent manner (Fig. 2c). Additionally, GENA-104A16 treatment dose-dependently increased CD3\u003csup\u003e+\u003c/sup\u003e T/U-2 OS conjugates and CD3\u003csup\u003e+\u003c/sup\u003e T-cell cytotoxicity against U-2 OS (Fig. 2d). In the case of \u003cem\u003eCNTN4\u003c/em\u003e KD U-2 OS cells, conjugates and CD3\u003csup\u003e+\u003c/sup\u003e T-cell cytotoxicity were higher than those of U-2 OS cells, regardless of GENA-104A16 treatment (Fig. 2d). These results demonstrate that GENA-104A16 is a potent inhibitor of CNTN4 binding to T-cells, with sub-nanomolar inhibitory activity.\u003c/p\u003e\n\u003cp\u003eTo elucidate the T-cell regulation mechanisms of CNTN4 and GENA-104A16, we assessed their effects on human TCR signaling. CNTN4 inhibited the phosphorylation of TCR signaling-related proteins (ZAP70, PI3K, LCK, ERK1/2, JNK, and p38), and GENA-104A16 neutralized this phenomenon (Fig. 2e and Extended Data Fig. 2b). CNTN4 also impeded the translocation of transcription factors (p65 and NFATc1) from the cytoplasm to the nucleus upon T-cell activation, which GENA-104A16 treatment counteracted (Extended Data Fig. 2c). ChIP assays revealed that CNTN4 inhibited the binding of transcription factors to target gene promoters (IFN-g and TNF-a) in human T-cells stimulated by anti-CD3 antibody, and GENA-104A16 neutralized the decrease in transcription factor binding to these sites (Fig. 2f, Extended Data Fig. 2d-g). These findings demonstrate that GENA-104A16 promotes T-cell activation and TCR signaling cascades which are inhibited by CNTN4.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGENA-104A16 promotes killing of CNTN4-expressing tumors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe subsequently examined the \u003cem\u003ein vivo\u003c/em\u003e effectiveness of GENA-104A16. Mice received subcutaneous injections of MC38, CT26, or CT26/Cntn4 cells, and intraperitoneal injections of mIgG1 or GENA-104A16.mIgG1 (1 mg/kg or 3 mg/kg) five times at three-day intervals. In MC38 with no CNTN4 expression, GENA-104A16 had no effect on tumor reduction (Fig. 2g). The treatment of GENA-104A16 has led to a significant reduction in CT26 tumor size, a decrease that becomes even more pronounced when CNTN4 is overexpressed, as demonstrated in CT26/Cntn4 tumors (Fig. 2h, i). These results reveal that the efficacy of GENA-104A16 is dependent on CNTN4 expression. In light of Cntn4 expression modulating T-cell activation, we analyzed the immunological status associated with Cntn4 expression through FACS analysis and whole transcriptomic analysis in tumor tissues. GENA-104A16 treatment increased the population of CD4\u003csup\u003e+\u003c/sup\u003eCD3\u003csup\u003e+\u003c/sup\u003e/CD45\u003csup\u003e+\u003c/sup\u003e T-cells and decreased the population of Tregs (Foxp3\u003csup\u003e+\u003c/sup\u003eCD25\u003csup\u003e+\u003c/sup\u003e/CD4\u003csup\u003e+\u003c/sup\u003e) in CT26 cancer tissues compared to mIgG1 treatment, with more pronounced changes in CT26/Cntn4 cancer tissues (Extended Data Fig. 3a, b). Intriguingly, CT26/Cntn4 tumors exhibited a significantly higher tumor-infiltrating Treg population than CT26, indicating that elevated CNTN4 expression promotes immune evasion (Extended Data Fig. 3a, b). Consistent with FACS analysis, whole transcriptomic analysis revealed that TNF-a signaling and IFN-g response pathway gene sets tended to increase in GENA-104A16-treated CT26/Cntn4 compared to control IgG1-treated CT26/Cntn4, while CT26/Cntn4 tumors reduced sensitivity to immune response compared to CT26 tumors (Extended Data Fig. 3c).\u003c/p\u003e\n\u003cp\u003eWe performed single-cell RNA (scRNA) sequencing and analysis to further assess the characteristics of TILs in CT26/Cntn4 tumors from mice treated with GENA-104A16 (Extended Data Fig. 4a\u0026ndash;d). We obtained 17,563 cells and calculated gene expression counts for each single cell, then distinguished them into non-lymphocytes and lymphocytes based on Ptprc (Cd45) gene expression, a lymphocyte marker. The 15,082 cells with low Ptprc expression (count \u0026lt; 1) were classified as non-lymphocytes and divided into 12 clusters, while the remaining 6,098 cells expressing Ptprc (count \u0026ge; 1) were classified as lymphocytes and divided into 11 clusters. Immune cells were classified into CD4\u003csup\u003e+\u003c/sup\u003e T-cells, CD8\u003csup\u003e+\u003c/sup\u003e T-cells, monocytes, macrophages, NK cells, pDC, and neutrophils by profiling the expression of variable genes in each cluster. Interestingly, a higher proportion of TILs was observed in GENA-104A16 treated tumors (18.63%) compared to mIgG1 treated tumors (9.57%) (Fig. 2j). Both CD4\u003csup\u003e+\u003c/sup\u003e T-cells (0.34% vs. 1.24%) and CD8\u003csup\u003e+\u003c/sup\u003e T-cells (2.81% vs. 4.18%) were more abundant in GENA-104A16-treated tumors (Fig. 2j), with most of these CD8\u003csup\u003e+\u003c/sup\u003e cells expressing cytotoxic markers (Fig. 2k). These results suggest that GENA-104A16 can neutralize CNTN4-mediated T-cell suppression, enhance T-cell activity, and sequentially activate the cancer-immunity cycle, thereby increasing the number of TILs.\u003c/p\u003e\n\u003cp\u003eSince the effectiveness of many immunotherapeutic agents as monotherapy options has been a recent major concern, we sought to investigate the potential for combining GENA-104A16 with traditional therapies or other immunotherapeutic agents to achieve maximum therapeutic benefit for a wide range of patients\u003csup\u003e17\u003c/sup\u003e. We examined the antitumor effect of GENA-104A16 in combination with anti-PD-L1 or anti-PD-1 antibodies. Since CT26 tumors exhibit a significant single effect of GENA-104A16, determining whether the combination with other ICIs is synergistic or additive is challenging. Therefore, B16F10 cells, which express minimal CNTN4, were subcutaneously injected into both flanks of C57BL/6 mice, and intraperitoneally injected with antibodies (mIgG1, GENA-104A16.mIgG1, anti-PD-L1 antibody, or a combination of GENA-104A16.mIgG1 and anti-mPD-L1 antibodies) four times at three-day intervals. While GENA-104A16 administration alone did not reduce B16F10 tumor volume, the combination of anti-mPD-L1 antibody and GENA-104A16 further increased the reduction in tumor volume (Fig. 2l, left). Similarly, the combination of anti-mPD-1 antibody and GENA-104A16 resulted in a significant decrease in tumor volume, whereas monotherapy did not have a significant effect (Fig. 2l, right). These findings suggest that GENA-104A16 may enhance the antitumor effects when combined with other immune checkpoint inhibitors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCNTN4\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003einteraction\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;with APP on T-cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe identified a receptor on T-cells that binds to CNTN4 through receptor screening, which revealed interactions with two APP isoforms (APP770 and APP751) (Fig. 3a). The CNTN4 protein specifically bound to APP-overexpressed HEK293FT cells, akin to PD-1/PD-L1 binding, and vice versa (Fig. 3b). Cell-to-cell conju\u0026shy;\u0026shy;\u0026shy;\u0026shy;\u0026shy;\u0026shy;\u0026shy;\u0026shy;gation assays also demonstrated increased conjugate numbers in APP-overexpressed and CNTN4-overexpressed HEK293FT cells (Fig. 3c). To map interaction domains, we created a DNA construct containing APP isoforms (770, 751, and 695) and CNTN4 C-terminal deletion constructs (#4, #6, #8, and #10) (Fig. 3d), and used a Co-IP assay in HEK293FT cells expressing these constructs. Results showed that CNTN4 bound APP770 and APP751 containing a KPI domain, while APP bound CNTN4 #8 and #10 containing FN (Fibronectin) 1\u0026ndash;2 (Fig. 3e). The highest number of conjugates was observed with APP751 and CNTN4 #8 (Fig. 3f). Binding of CNTN4 recombinant protein to cells expressing APP751, but not APP695, and recombinant APP protein to cells expressing CNTN4 #8, but not CNTN4 #6, confirmed that the FN 1\u0026ndash;2 domains of CNTN4 physically interact with the KPI domain of APP (Fig. 3g).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInhibition of APP enhances T\u003c/strong\u003e\u003cstrong\u003e-\u003c/strong\u003e\u003cstrong\u003ecell activation mechanisms\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe observed increased APP expression in CD4\u003csup\u003e+\u003c/sup\u003e and CD8\u003csup\u003e+\u003c/sup\u003e T-cells following anti-CD3 antibody stimulation (Fig. 3h and Extended Data Fig. 5a). We then investigated the influence of APP expression on T-cell function by generating \u003cem\u003eApp\u003c/em\u003e KO mice using CRISPR-Cas9 (Extended Data Fig. 5b). \u003cem\u003eApp\u003c/em\u003e KO T-cells exhibited increased adhesion to CT26 cells and heightened cytotoxicity compared to \u003cem\u003eApp\u003c/em\u003e WT T-cells (Fig. 3i). T-cell activation-related processes were induced by anti-CD3 antibody stimulation in both \u003cem\u003eApp\u0026nbsp;\u003c/em\u003eWT and KO T-cells, but CNTN4 inhibited these processes only in \u003cem\u003eApp\u0026nbsp;\u003c/em\u003eWT T-cells, not in \u003cem\u003eApp\u0026nbsp;\u003c/em\u003eKO T-cells (Fig. 3j and Extended Data Fig. 5c-f). CNTN4 inhibited luciferase activity in Jurkat-Lucia\u003csup\u003eTM\u003c/sup\u003e NFAT cells by approximately 35% when stimulated by anti-CD3 antibody-treated T-cells, but inhibition increased to 82-84% in cells overexpressing full-length APP or intracellular domain-deleted APP (APP\u0026Delta;ICD) (Fig. 3k and Extended Data Fig. 5g). These findings indicate that T-cell inhibition by CNTN4 occurs due to decreased adhesion between cancer cells and T-cells caused by the extracellular region of APP rather than changes caused by the intracellular region of APP.\u003c/p\u003e\n\u003cp\u003eWe produced an anti-APP antibody (5A7) through immunization and hybridoma generation technologies, which bound APP with an EC\u003csub\u003e50\u003c/sub\u003e of 1.864 nM (Fig. 3l). The CNTN4/APP binding was specifically inhibited by the anti-APP antibody (5A7) (IC\u003csub\u003e50\u003c/sub\u003e = 478 nM) and GENA-104A16 (IC\u003csub\u003e50\u003c/sub\u003e = 0.212 nM) (Fig. 3m). We also evaluated the \u003cem\u003ein vivo\u003c/em\u003e effectiveness of the anti-APP antibody (5A7). CT26 cells were subcutaneously injected into both flanks of BALB/c or BALB/c nude mice, and intraperitoneally injected with mIgG1 or anti-APP antibody (5A7) three times at 3-day intervals. The anti-APP antibody (5A7) reduced CT26 tumor volume in the syngeneic mouse model but not in the immunodeficient nude mouse model lacking T-cells (Fig. 3n). This result demonstrates the potent inhibitory effect of the anti-APP antibody (5A7) on APP binding to CNTN4.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical implications of CNTN4 expression\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGenerally, elevated PD-L1 levels are associated with a higher likelihood of responsiveness to ICIs; however, some patients with high PD-L1 levels still exhibit a non-responsive phenotype. Consequently, we posited that the refractory response to ICIs observed in PD-L1 positive patients might be linked to CNTN4 expression. To investigate this hypothesis, we assessed the differences in CNTN4 levels between responders and non-responders within a cohort of patients with similarly high PD-L1 expression (Fig. 4a). Among PD-L1 high patients in the immunotherapy-treated gastric cancer cohort, non-responders displayed significantly higher CNTN4 levels compared to responders (p = 0.0148) (Fig. 4b). A similar trend was observed in the immunotherapy-treated lung cancer cohort, with weak statistical significance (p = 0.167) (Fig. 4c). We also confirmed the relationship between various cytotoxic immune-related markers and CNTN4 levels within the PD-L1 high group. Cntn4 was negatively correlated with immune-related cytotoxic markers such as Gzmb, Ifng, Prf1, Cxcl9, and Cxcl10 (Fig. 4d). Further GSEA analysis using hallmark gene sets revealed that biopsy samples from the low CNTN4 expression group exhibited enriched signatures for TNF-\u0026alpha; signaling via NF-\u0026kappa;B and IFN-\u0026gamma; response (Fig. 4e), consistent with results from mouse tumor tissues depending on CNTN4 expression (Fig. 1e and Extended Data Fig. 3c). Higher CNTN4 levels correlated with poor prognosis in terms of progression-free survival (PFS) and overall survival (OS) (Fig. 4f). These results show that high CNTN4 expression might contribute to the refractory response to ICIs even in high PD-L1 patients, impacting immune-related pathways and survival probability.\u003c/p\u003e\n\u003cp\u003eIn the immunotherapy-treated gastric cancer cohort, 18 out of 19 high CNTN4 group patients were non-responders, while 8 out of 18 low CNTN4 group patients were responders. The composition of responders and non-responders based on CNTN4 and PD-L1 levels showed that the responding group primarily consisted of CNTN4-low and PD-L1-high patients (66.7% in responders), while the non-responding group mainly included CNTN4-high patients regardless of PD-L1 levels (64.3% in non-responders) (Fig. 4g). This result suggests that a high PD-L1 level generally implies effective ICI treatment; however, resistance may occur due to high CNTN4 levels even with high PD-L1 levels. Therefore, CNTN4 antibodies should be used first in patients with high CNTN4 levels, regardless of PD-L1 levels.\u003c/p\u003e\n\u003cp\u003eLastly, we found that CNTN4 was not expressed on any immune cells, unlike APP, which is a CNTN4 binding partner on T-cells, according to the Database of Immune Cell Expression (DICE) (Fig. 4h and Extended Data Fig. 6). In addition, we analyzed PD-L1 and CNTN4 expression using 17 TMA slides and summarized the expression scores for each patient and cancer tissue as a heatmap. We observed CNTN4 in more cancer tissues than PD-L1, with its expression levels exceeding those of PD-L1 (Fig. 4i). These results support that CNTN4 is a tumor-specific target, and that targeting CNTN4 may be and effective treatment strategy for various types of cancer.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003e \u003cb\u003eCNTN4 as a Novel Immune Checkpoint Protein\u003c/b\u003e \u003c/p\u003e \u003cp\u003eOur study discovered that CNTN4, expressed in various human tumor tissues, inhibits T-cell activity, indicating its potential role as an inhibitory ICP. We identified APP on T-cells as a binding partner for CNTN4, and upon TCR stimulation, APP expression increased, similar to other ICPs. While the interaction between CNTN4 and APP has been previously reported, our study is the first to establish the CNTN4/APP axis in the TME. This interaction weakens T-cell activity and TCR signaling cascades by reducing conjugation between cancer cells and T-cells. We demonstrated that targeting the CNTN4/APP axis in a syngeneic tumor mouse model reduced tumor growth by enhancing anti-tumor immunity. Clinical analysis revealed an increase in CNTN4 levels among non-responders compared to responders and a negative correlation with various cytotoxic immune-related markers in PD-L1 high patients. High CNTN4 levels were also associated with poor prognosis. In addition, we observed CNTN4 in more cancer tissues than PD-L1, with its expression levels exceeding those of PD-L1. Thus, we propose that blocking the CNTN4/APP interaction could suppress tumor growth by effectively activating T-cells, and the development of drugs targeting CNTN4 could be particularly useful for non-responders to PD-1/PD-L1 blockade and for treating numerous cancer patients.\u003c/p\u003e\n\u003ch3\u003eCombining GENA-104A16 with Existing ICI Therapies\u003c/h3\u003e\n\u003cp\u003eImmune checkpoint inhibitors (ICIs) have revolutionized oncology by improving survival rates for many cancer patients. However, there are limitations, such as low response and resistance. Hence, new ICPs and target molecules are being studied to expand the use and efficacy of existing ICI therapies. Our study found that the anti-CNTN4 monoclonal antibody GENA-104A16 neutralized T-cell inhibition by CNTN4, demonstrating its anti-tumor and immune-boosting effects. We also confirmed that anti-PD-L1 treatment induced CNTN4 expression in a CT26 syngeneic tumor mouse model and that CNTN4 expression was upregulated in anti-PD-1 non-responders according to clinical cohort analysis. These results suggest that GENA-104A16 could serve as a novel ICI treatment for patients and could be combined or administered sequentially with PD-1/PD-L1 blockade to overcome acquired resistance.\u003c/p\u003e\n\u003ch3\u003eSafety Considerations for CNTN4/APP Blockade\u003c/h3\u003e\n\u003cp\u003eCNTN4 has been characterized as a key cell adhesion molecule for axon guidance and neural connectivity in neurodevelopment, and previous studies have shown that CNTN4 is a risk gene associated with several neuropsychiatric disorders. APP belongs to the type I transmembrane protein family and has also been extensively studied for its role in the brain. Considering the potential side effects of CNTN4/APP blockade on the nervous system, we confirmed that CNTN4 interacts with APP751 and APP770 isoforms, which are expressed in most tissues, while APP695 isoform is mainly expressed in neurons. This indicates that CNTN4/APP interaction occurs less frequently in the brain and nervous system, and blocking this interaction may have fewer adverse effects on the nervous system. In addition, the expression of CNTN4 in some normal tissues was partially confirmed by IHC, but the expression was negative overall (Extended Data Fig.\u0026nbsp;7). Thus, it is suggested that a strategy to block the interaction with APP by targeting CNTN4 can reduce side effects on the nervous system and act specifically in cancer.\u003c/p\u003e \u003cp\u003eIn conclusion, CNTN4 expressed in cancer cells may act as a novel ICP that attenuates TCR signaling by interacting with APP in T-cells. In addition, GENA-104A16 and anti-APP antibody (5A7) can be therapeutic agents that remove tumors by increasing immune activity by suppressing the CNTN4/APP axis. Therefore, the development of new immune checkpoint inhibitory antibodies capable of inhibiting the interaction between CNTN4 and APP may lead to the discovery of therapeutic agents for patients who cannot respond to immunotherapy.\u003c/p\u003e"},{"header":"Materials \u0026 Methods","content":"\u003cp\u003e \u003cb\u003eReagents and antibodies\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo produce recombinant antibodies, the plasmid vectors were transiently expressed in ExpiCHO-S cells with the ExpiFectamine CHO Kit (ThermoFisher Scientific, Rockford, IL), followed by culturing in ExpiCHO Expression medium at 32\u0026ndash;37\u0026deg;C for 7\u0026ndash;12 days in shaking flasks and a CO\u003csub\u003e2\u003c/sub\u003e incubator equipped with an orbital shaker. Specifically, 100 \u0026micro;g plasmid vectors (heavy and light chains were cloned into pcDNA3.4 expression vector, Addgene, Watertown, MA) and 800 \u0026micro;L ExpiFectamin CHO Reagent were mixed with OptiPRO SFM (Gibco, UK) medium (19.2 mL final volume) and allowed to stand for 5 min at 27\u0026deg;C. Subsequently, the mixed solution was added to pre-adjusted 6 \u0026times; 10\u003csup\u003e6\u003c/sup\u003e cells and incubated at 37\u0026deg;C with the condition of 8% CO\u003csub\u003e2\u003c/sub\u003e in humidified air to transfect DNA into cells. At 18 to 22 h post-transfection, 1.5 mL Enhancer 1- and 40-mL Feed, components of the ExpiFectamine CHO Kit, were added to each culture flask. On day 3 post-transfection, 40 mL of feed was added to each flask. Each antibody was purified for several chromatography steps from the cell culture supernatant using an AKTA Avant system (Cytiva, Sweden). First, a protein A column containing Hitrap MabSelect SuRe (Cytiva) was used to capture the antibodies, and Capto S ImpAct and/or Capto adhere chromatography (Cytiva) was used to polish and improve their purities. Purified proteins were concentrated in the formulation buffer by ultrafiltration using an Amicon Ultra 15 (50 kDa cutoff size, Merck, Germany). To confirm the purity of the prepared antibodies, samples were analyzed by SDS-PAGE and size exclusion HPLC (Agilent, Palo Alto, CA) with a TSKgel G3000SWXL column (Tosoh, Japan). The biotinylated antibody was prepared according to the procedure recommended by the manufacturer of the biotinylation kit (Pierce, Rockford, IL). Antibodies for biotinylation were washed five times with 1\u0026times; PBS, pH 7.4 (Corning Costar, Corning, NY) using Amicon Ultra 0.5 (Merck) to remove sources of primary buffer, followed by a concentration above 2.0 mg/mL. Then, 157 \u0026micro;L biotin reagent (dissolved in 1\u0026times; PBS, pH 7.4) per 1 mg antibody was added to the above sample and incubated for 1 h with gentle rotation. Non-reacted biotin was removed by ultrafiltration using an Amicon Ultra 4 (Merck).\u003c/p\u003e \u003cp\u003eAnti-human CD3, anti-human CD28, anti-mouse CD3, and anti-mouse CD28 antibodies were acquired from R\u0026amp;D Systems (Minneapolis, MN). anti-His, anti-ZAP70, anti-p-ZAP70, anti-LCK, anti-p-LCK, anti-ERK1/2, anti-p-ERK1/2, anti-JNK, anti-p-JNK, anti-p38, anti-p-p38, anti-β-actin, anti-p65, anti-NFATc1, anti-GAPDH, anti-Histone H3, HRP-conjugated anti-mouse or rabbit IgG, anti-rabbit or mouse IgG antibodies were obtained from Cell Signaling Technology (Danvers, MA).\u003c/p\u003e \u003cp\u003eIL-2 recombinant protein (rIL-2) was acquired from Roche (Switzerland). CellTracker CMFDA-green, CMRA-orange, BMQC-violet, CFSE-green, and Lipofectamine 2000 were obtained from Invitrogen (Carlsbad, CA). Human and mouse PD-L1 and CNTN4 (His or Fc tag) recombinant proteins were acquired from Sino Biological, Inc. (China). Human APP (His or Fc tag) recombinant proteins were acquired from BioLegend (San Diego, CA).\u003c/p\u003e \u003cp\u003eThe plasmid DNA purification kit and WEST-ZOL western Blot Detection kit were purchased from iNtRON Biotechnology (Korea). PrimeSTAR HS DNA polymerase was purchased from TaKaRa Bio, Inc. (Japan). Restriction enzymes and ligases were purchased from New England Biolabs Inc. (Beverly, MA). Unless otherwise stated, all chemical reagents were purchased from Sigma-Aldrich.\u003c/p\u003e \u003cp\u003e \u003cb\u003eCells\u003c/b\u003e \u003c/p\u003e \u003cp\u003eJurkat-Lucia\u0026trade; NFAT (TIB-152, ATCC, Manassas, VA), HEK293FT (CVCL_6911, ThermoFisher Scientific), CT26 (CRL-2638, ATCC), B16F10 (CRL-6475, ATCC), and U-2 OS (HTB-96, ATCC) cells were maintained in RPMI-1640, McCoy\u0026rsquo;s 5A, or Dulbecco\u0026rsquo;s modified Eagle\u0026rsquo;s medium (Gibco) supplemented with 10% (v/v) FBS (Gibco). CT26/Cntn4 cells were purchased from GenScript (China). To establish stable \u003cem\u003eCNTN4\u003c/em\u003e KD U-2 OS cell lines, CNTN4 three sgRNAs (Thermo Fisher Scientific) and TrueCut\u0026trade; Cas9 Protein v2 (Thermo Fisher Scientific) were electroporated with Neon\u0026trade; Transfection system (Invitrogen) into U-2 OS cells. Human PBMCs were isolated using the Ficoll-Paque Plus (GE Healthcare Life Sciences, Piscataway, NJ). Mouse bone marrow was harvested by flushing using syringe-containing media after cutting off the ends of the femurs and tibiae. Na\u0026iuml;ve CD3\u003csup\u003e+\u003c/sup\u003e, CD4\u003csup\u003e+\u003c/sup\u003e, or CD8\u003csup\u003e+\u003c/sup\u003e T-cells were purified from the mouse spleen or human PBMCs by negative selection using a T-cell isolation kit (Miltenyi Biotec, Germany). The purity of each population was confirmed to be more than 95% using flow cytometry. To generate clonal expansion T-cells, na\u0026iuml;ve T-cells were incubated in 2 \u0026micro;g/mL anti-CD3/28 coated culture plates with 50 U/mL rIL-2 for 48 h and cultured for an additional five days with 50 U/mL rIL-2.\u003c/p\u003e \u003cp\u003e \u003cb\u003eAnimals\u003c/b\u003e \u003c/p\u003e \u003cp\u003eC57BL/6 and BALB/c mice were purchased from Orient Bio Inc. (Korea). All the mice were housed under specific pathogen-free conditions. All experimental methods and protocols were approved by the Institutional Animal Care and Use Committee of the GenNBio Animal Research Center and were carried out in accordance with the approved guidelines.\u003c/p\u003e \u003cp\u003e \u003cb\u003ecDNA constructs\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo generate the CNTN4 and APP constructs, CNTN4 and APP clones coding for the full-length open reading frame were purchased from OriGene (Rockville, MD). CNTN4 #4, #6, #8, APP751, APP695, and APPΔICD genes were generated by standard or overlapping PCR and subcloned into the pCMV6-Entry vector (Origene).\u003c/p\u003e \u003cp\u003e \u003cb\u003eImmunohistochemistry assay (IHC)\u003c/b\u003e \u003c/p\u003e \u003cp\u003eVarious tissue slides (provided by SuperBioChips) were incubated with anti-CNTN4 (Abcam, Cambridge, MA) or anti-PD-L1 (Ventana, Tucson, AZ) antibody, followed by diaminobenzidine staining (Ventana), according to the manufacturer\u0026rsquo;s instructions. The stained slides were photographed using an Aperio AT2 Scanner (Leica, Germany) and recorded by a pathologist.\u003c/p\u003e \u003cp\u003e \u003cb\u003eFlow cytometric analysis\u003c/b\u003e \u003c/p\u003e \u003cp\u003eCells were suspended in FACS buffer (2% FBS and 1 mM EDTA in PBS). For CNTN4 expression, GENA-104A16 or hIgG isotype antibody was added, and APC-conjugated hIgG secondary antibody (BioLegend) was sequentially added for 30 min at 4\u0026deg;C. For APP expression, anti-2D5-Biotin or mIgG-Biotin antibodies were added, and PE-conjugated Streptavidin secondary antibody (BioLegend) was sequentially treated for 30 min at 4\u0026deg;C. For protein binding, His-tagged recombinant protein was added and FITC-conjugated anti-His secondary antibody (BioLegend) was sequentially added for 30 min at 4\u0026deg;C. The cells were then assessed on a FACSCanto II (BD Biosciences, San Jose, CA), and the data were analyzed using Flowjo software (Treestar, San Carlos, CA).\u003c/p\u003e \u003cp\u003e \u003cb\u003eCharacterization (proliferation, migration, and invasion) of cancer cells with different CNTN4 expression\u003c/b\u003e \u003c/p\u003e \u003cp\u003eCancer cells were plated, divided, and grown in 6-well plates for 0\u0026ndash;4 days. Cell proliferation was measured by cell counting using a Luna cell counter (Logos Biosystems, Korea). For migration studies, a standard assay was used to determine the number of cells that traversed a porous polycarbonate membrane (8 \u0026micro;m pore size) in response to a chemo-attractant (higher serum concentration) using the Cytoselect 24-well cell migration assay (Cell Biolabs, San Diego, CA). Invasion was measured using the BD BioCoat Matrigel invasion chambers (BD Biosciences). In both assays, the cancer cells were suspended in serum-free media. The cells (3 \u0026times; 10\u003csup\u003e5\u003c/sup\u003e/chamber) were seeded in an upper chamber in serum-free media for migration and invasion assays. The lower chamber was filled with media containing 10% FBS. After incubation for 18 h (invasion) or 40 h (migration) at 37\u0026deg;C in a 5% CO\u003csub\u003e2\u003c/sub\u003e incubator, cells passing through polycarbonate membrane were stained for 10 min using the provided staining solution, washed, and dried. Stained membranes were incubated for 10 min with extraction buffer, and the lysates were measured using a microplate reader (SpectraMax M2; Molecular Devices, Sunnyvale, CA) with an optical density (OD) of 560 nm. SoftMax Pro 7 software was used for the analysis. All reactions were performed in triplicates.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTumorgraft\u003c/b\u003e \u003c/p\u003e \u003cp\u003eFor the tumor growth experiments, BALB/c or C57BL/6 mice were subcutaneously injected with 1 \u0026times; 10\u003csup\u003e6\u003c/sup\u003e CT26, CT26/Cntn4, or B16F10 cells per mouse. Tumor size was measured thrice a week until the endpoint, and tumor volume was calculated as length \u0026times; width\u003csup\u003e2\u003c/sup\u003e \u0026times; 0.5. When the average tumor size reached a volume of approximately 100 mm\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e, the experimental treatments were initiated. Tumor-bearing mice were intraperitoneally injected with mIgG1, anti-mPD-L1, anti-mPD-1 (BioXCell, West Lebanon, NH), GENA-104A16. mIgG1 or anti-APP antibody (5A7) on days 0, 3, 6, 9, and 12.\u003c/p\u003e \u003cp\u003e \u003cb\u003eWhole transcriptomic analysis of tumor tissues\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe tumor tissues were gently removed using forceps, and the surgical scissors were transferred to PBS. RNA was extracted using an RNA extraction kit (Qiagen, Germany). Next, 151-bp paired-end libraries were constructed from 1 \u0026micro;g of RNA using the TruSeq RNA Sample Prep Kit v2 (Illumina). Whole transcriptome sequencing (WTS) was performed using an Illumina HiSeq instrument. RNA-seq reads from each WTS experiment were aligned to the mouse reference genome (GRCm38) using STAR\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e aligner. Gene expression was quantified by RSEM\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. TPM values were used to identify gene sets enriched in the tumors. GSEA (Gene set enrichment analysis) was performed using the Java GSEA desktop application (GSEA v2.1.0)\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Gene sets that were upregulated or downregulated with an FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.25 were considered significant.\u003c/p\u003e \u003cp\u003e \u003cb\u003eIn vitro plate-bound T cell proliferation or neutralization assay\u003c/b\u003e \u003c/p\u003e \u003cp\u003eAnti-CD3 antibody (human: 4 \u0026micro;g/mL, mouse: 3 \u0026micro;g/mL) and CNTN4 or PD-L1 recombinant protein (human: 50\u0026ndash;200 nM, mouse: 37.5\u0026thinsp;~\u0026thinsp;150 nM) were diluted in PBS and incubated overnight for coating at 37\u0026deg;C on 96-well non-treated cell culture plates. For the neutralization assay, the plate was washed with PBS and incubated at 37\u0026deg;C for 1 h after the antibody treatment. T-cells were stained with CFSE at 37\u0026deg;C for 10 min, washed with the culture medium, and then added to the culture at 2 \u0026times; 10\u003csup\u003e5\u003c/sup\u003e T-cells/well on a cell culture plate. After 72 h, dividing cells were evaluated on a FACSCanto II, and data were analyzed using FlowJo software.\u003c/p\u003e \u003cp\u003e \u003cb\u003eMeasure of cytokine secretion\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe amount of human or mouse IFN-γ or TNF-α in the T cell culture supernatants was determined by ELISA using a Duo Set ELISA kit (R\u0026amp;D System). The manufacturer's recommendations were followed. 96-well plates were coated with 100 \u0026micro;L/well of capture antibody overnight and blocked with 100 \u0026micro;L/well of ELISA diluent in D.W. 100 \u0026micro;L of the samples were applied to the wells in duplicates. For the standard curve, a recombinant standard was used at concentrations between 2,000 pg/mL and 62.5 ng/mL. Streptavidin-HRP conjugate was diluted 1:40 in the substrate solution. The reaction was stopped by adding 50 \u0026micro;L/well of 2 N H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e and the plates were analyzed at 450 nm using a SpectraMax M2 microplate reader. SoftMax Pro 7 software was used for the analysis. All reactions were performed in triplicates.\u003c/p\u003e \u003cp\u003e \u003cb\u003eAffinity of GENA-104A16 against recombinant proteins or overexpressing HEK293FT cells by indirect ELISA or flow cytometry, respectively\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo measure the affinity of antibodies against recombinant proteins, 96-well ELISA plates were coated with 5 nM of human or mouse recombinant protein at 4\u0026deg;C overnight. The coated ELISA plate was washed three times and blocked with PBST (PBS containing 0.05% Tween 20). A tenfold serial dilution of antibodies (seven or eleven points) in PBS was added, incubated for 1 h at 37\u0026deg;C, and then washed three times with PBST. HRP-conjugated anti-human or anti-mouse antibodies were added (1:3,000 dilution in PBST), incubated for 1 h at 37\u0026deg;C, and washed three times with PBST. ABTS substrate solution (2,2ʹ-azinobis [3-ethylbenzothiazoline-6-sulfonic acid]-diammonium salt, ThermoFisher Scientific) was added and incubated for 5 min at 27\u0026deg;C. Absorbance at 405 nm was determined using a microplate reader (BioTek Synergy 4, Winooski, VT). The half-maximal binding concentration (EC\u003csub\u003e50\u003c/sub\u003e) value was determined with a nonlinear four-parameter logistic curve using GraphPad PRISM 9.3.1 (GraphPad Software Inc., San Diego, CA).\u003c/p\u003e \u003cp\u003eTo measure the affinity of GENA-104A16 for HEK293FT cells, CNTN4 overexpression HEK293FT cells (3 \u0026times; 10\u003csup\u003e5\u003c/sup\u003e) were washed three times with PBS and resuspended in 500 \u0026micro;L of FACS buffer. The threefold serial dilution of GENA-104A16 (12.54 pM\u0026thinsp;~\u0026thinsp;66.67 nM, 14 points) in FACS buffer was added, incubated for 1 h at 4℃, and then washed three times with FACS buffer. APC-conjugated anti-human IgG was added (1:400 dilution in FACS buffer), incubated for 30 min at 4\u0026deg;C, and washed thrice with FACS buffer. Cells were assessed on a FACSCanto II, and the data were analyzed using FlowJo software. MFI (\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eM\u003c/span\u003eean \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eF\u003c/span\u003eluorescent \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eI\u003c/span\u003entensity) values were obtained from the histograms and were used to plot the binding curves. Dose-response curves were generated by plotting the resulting MFI against the log concentration of GENA-104A16, and the EC\u003csub\u003e50\u003c/sub\u003e value was analyzed by nonlinear regression curve fitting using GraphPad PRISM.\u003c/p\u003e \u003cp\u003e \u003cb\u003eBiolayer interferometry\u003c/b\u003e \u003c/p\u003e \u003cp\u003eBinding kinetic measurements were performed using the Octet R8 system (Sartorius, Germany) at 30℃ at 1,000 rpm agitation (Sartorius). GENA-104A16 was loaded onto anti-human Fc biosensor 2nd generation biosensor (AHC2) at 1.0 \u0026micro;g/mL in running buffer for 0.02% PBST for 5 min. Subsequently, the tips were transferred to running buffer for 180 s for sensor rinsing. For kinetic analyses, the association with CNTN4 (varying concentrations ranging from 51.9\u0026thinsp;~\u0026thinsp;415.4 ng/mL in running buffer) was measured for 600 s, followed by dissociation for 600 s (in running buffer). In each experiment, one negative control (reference well or blank) was measured, where the capture antibody was incubated with running buffer instead of the antigen. Data fitting and analysis were performed with the Octet\u0026reg;BLI Analysis 12.2 program (Sartorius) using a 1:1 binding model after Savizky-Golay filtering.\u003c/p\u003e \u003cp\u003e \u003cb\u003eImmune cell profiling\u003c/b\u003e \u003c/p\u003e \u003cp\u003eFor immune profiling, the tumors were harvested on day 4. Tumors were chopped (Dorco, Korea) and transferred in RPMI-1640 media supplemented with 0.25 mg/mL hyaluronidase Type IV-S, 50 \u0026micro;g/mL DNase type 1, 2.5 mg/mL collagenase type 1, 1.5 mg/mL collagenase type 2, and 1 mg/mL collagenase type 4. The samples were then incubated at 37\u0026deg;C for 50 min and filtered using a 70 \u0026micro;m cell strainer (BD Biosciences). Tumor cells (1 \u0026times; 10\u003csup\u003e6\u003c/sup\u003e cells/well) were treated with anti-mouse CD16/32 (BD Biosciences) at 4\u0026deg;C for 10 min to block the Fc receptor. Surface staining was conducted and BD Pharmingen\u0026trade; Mouse Foxp3 Buffer Set (BioLegend) solution was added. Then, intracellular staining was performed. Stained cells were acquired using FACSCanto II, and the data were analyzed using FlowJo software. The antibody information was as follows: APC/Cy7 anti-mouse CD45 antibody (BioLegend, 103116), Pacific blue anti-mouse CD3 antibody (Biolegend, 100334), PE/Cy7 anti-mouse CD4 antibody (BioLegend, 100422), PE anti-mouse CD8 antibody (BioLegend, 100708), APC anti-mouse CD25 antibody (BioLegend, 101910), PE anti-mouse FoxP3 antibody (Biolegend, 126404), GENA-104A16, APC/Cy7 Rat IgG2b, κ isotype control antibody (Biolegend, 400624), Pacific blue Rat IgG2b, κ isotype control antibody (Biolegend, 400925), PE/Cy7 Rat IgG2b, κ isotype control antibody (Biolegend, 400618), PE Rat IgG2b, κ isotype control antibody (Biolegend, 400608), APC Rat IgG2b, κ isotype control antibody (Biolegend, 400612), hIgG4 (Sino Biological Inc., HG4K), Goat anti-Human IgG (H\u0026thinsp;+\u0026thinsp;L) Cross-Adsorbed Secondary Antibody, Alexa Fluor\u0026trade; 647 (Invitrogen, A21445)\u003c/p\u003e \u003cp\u003e \u003cb\u003escRNA analysis for population of tumor-infiltrating immune cells\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eLibrary preparation and pre-processing scRNA-seq\u003c/em\u003e \u003c/p\u003e \u003cp\u003eThe tumor tissues were gently removed using forceps, and the surgical scissors were transferred to PBS. Tumor tissues were placed in a petri dish and finely chopped with a razor blade. The chopped tissues were transferred to a 15 mL conical tube, filled with 5 mL of 1\u0026sdot; tumor dissociation buffer (Miltenyi Biotec), diluted in RPMI, and incubated at 37\u0026deg;C in a shaking incubator for 30 min. The cells were suspended in 3 mL RBC lysis buffer (BioLegend) and incubated for 5 min at 27\u0026deg;C. Cells were washed with PBS and resuspended in CELLBANKER (AMSBIO, Cambridge, MA). Library preparation and sequencing were performed by Macrogen (Korea). The library was prepared using Chromium Next Gem Single cell kit v3 and Single Cell 3ʹ v3 Gel Beads (10\u0026sdot; Genomics), and the library was sequenced using the Macrogen sequencing system. The raw data were processed using CellRanger v3.1 with default settings. Reads with the same cell barcode, UMIs (\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eU\u003c/span\u003enique \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eM\u003c/span\u003eolecular \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eI\u003c/span\u003edentifier), and genes were grouped to calculate the number of UMIs per gene per cell using the \u0026ldquo;count\u0026rdquo; command. The UMI count tables for each cellular barcode were used for further analyses. The matrix data from CellRanger were processed using the Seurat package (version 4.3.0) in R software (version 4.2.1) for each individual sample. To prepare the Seurat object, we first processed gene expression data from each sample using the Read 10\u0026sdot; function, and then ran the CreateSeuratObject function with metadata. Whole Seurat objects were integrated into one object to reduce the batch effect and perform post-analyses. Low-quality cells were discarded from the whole single-cell data for post-analysis. Cells with less than 200 expressed genes were discarded. Cells with expression levels lower than the 5th percentile or higher than the 95th percentile were discarded. Cells a mitochondrial gene expression percentage was higher than 15% were discarded.\u003c/p\u003e \u003cp\u003e \u003cem\u003eData integration, clustering, and annotation\u003c/em\u003e \u003c/p\u003e \u003cp\u003eThe merged Seurat object, including quality-checked cells, was split individually by treatment type and then normalized using the SCTransform function with the top 2,000 highly variable genes. To obtain highly variable genes, the SelectIntegrationFeatures function was used. Cell clustering and UMAP visualization were performed using FindClusters and RunUMAP (Uniform manifold approximation and projection), respectively. The integrated and clustered cells were then annotated manually. First, lymphocytes were identified based on the expression of \u003cem\u003ePtprc\u003c/em\u003e (CD45). If the cell\u0026rsquo;s expression of \u003cem\u003ePtprc\u003c/em\u003e was \u0026ge;\u0026thinsp;1, it was annotated as a lymphocyte and the other cells were annotated as non-lymphocytes. The lymphocytes were sub-annotated using the following gene expression profiles: CD8 T (\u003cem\u003eCD3g\u003c/em\u003e, \u003cem\u003eCD8a\u003c/em\u003e), CD4 T (\u003cem\u003eCd3g\u003c/em\u003e, \u003cem\u003eCd4\u003c/em\u003e), natural killer (\u003cem\u003eNcr1\u003c/em\u003e), macrophages (\u003cem\u003eAdgre1\u003c/em\u003e), dendritic cells (\u003cem\u003eSiglech\u003c/em\u003e), monocytes (\u003cem\u003eCd14\u003c/em\u003e, \u003cem\u003eCsf1r\u003c/em\u003e), and neutrophils (\u003cem\u003eCd14\u003c/em\u003e, \u003cem\u003eCxcr2\u003c/em\u003e). If there are some cells could not be annotated by those markers, they were annotated as \u0026lsquo;Others\u0026rsquo; or \u0026lsquo;Un-annotated\u0026rsquo; The expression of target genes such as cytotoxic markers and immune checkpoints was assessed. Cytotoxic markers (\u003cem\u003eIfng, Gzma, Gzmb, Tnf, Prf1\u003c/em\u003e) were assessed in CD8\u003csup\u003e+\u003c/sup\u003e T-cells and immune checkpoints (\u003cem\u003ePdcd1, Havcr2, Tigit, Ctla4, Lag3, Icos\u003c/em\u003e) were assessed in CD4\u003csup\u003e+\u003c/sup\u003e T-cells.\u003c/p\u003e \u003cp\u003e \u003cb\u003eConjugation assay\u003c/b\u003e \u003c/p\u003e \u003cp\u003eT-cell parts (PD-1, APP, APP751, and APP695 overexpressing HEK293FT cells, mouse CD3\u003csup\u003e+\u003c/sup\u003e T-cells, and human CD3\u003csup\u003e+\u003c/sup\u003e T-cells, 3 \u0026times; 10\u003csup\u003e5\u003c/sup\u003e) and target part cells (PD-L1, CNTN4, CNTN4 #6, or CNTN4 #8 overexpressing HEK293FT cells, CT26 cells, and U-2 OS or \u003cem\u003eCNTN4\u003c/em\u003e KD U-2 OS cells, 3 \u0026times; 10\u003csup\u003e5\u003c/sup\u003e) were stained with CMFDA and CMRA or Violet BMQC, respectively, according to the manufacturer\u0026rsquo;s protocol. For conjugation, equal volumes of T -cell parts and target part cells were mixed and incubated with or without GENA-104A16 for indicated time points in a humidified 5% CO\u003csub\u003e2\u003c/sub\u003e incubator at 37℃. The relative proportions of green, orange, violet, and green-orange events or green-violet events in each tube were determined by two-color flow cytometry using a FACSCantoII and analyzed with FlowJo software. The number of gated events counted per sample was at least 10,000. The percentage of conjugated T-cells was determined as the number of dual-labeled (CMFDA- and CMRA-positive or CMFDA- and BMQC-positive) events divided by the number of CMFDA-positive T-cells. All reactions were performed in duplicates.\u003c/p\u003e \u003cp\u003e \u003cb\u003eIn vitro cytotoxicity assay\u003c/b\u003e \u003c/p\u003e \u003cp\u003eHuman PBMCs or mouse splenocyte and bone marrow (1 \u0026times; 10\u003csup\u003e8\u003c/sup\u003e) were incubated with U-2 OS or CT26 (1 \u0026times; 10\u003csup\u003e7\u003c/sup\u003e) lysates for 24 h, respectively, in a humidified 5% CO\u003csub\u003e2\u003c/sub\u003e incubator at 37℃. Human or mouse CD3\u003csup\u003e+\u003c/sup\u003e T-cells were purified from the mixture by positive selection using a T-cell isolation kit (EasySep, StemCell Technologies, Seattle, WA) and then subjected to clonal expansion. For flow cytometric cytotoxicity assays, cancer cells were washed in PBS and stained with CMFDA, according to the manufacturer\u0026rsquo;s protocol. T-cells were mixed CMFDA-labeled cancer cells (E:T\u0026thinsp;=\u0026thinsp;10:1 or 5:1) and incubated for indicated time points in a humidified 5% CO\u003csub\u003e2\u003c/sub\u003e incubator at 37℃. After incubation, 10 \u0026micro;L of a 5 \u0026micro;g/mL solution of 7-AAD (BioLegend) was added to the cell suspension and incubated for 10 min on ice. Cells were evaluated on a FACSCanto II, and data were analyzed using FlowJo software.\u003c/p\u003e \u003cp\u003e \u003cb\u003eWestern blotting\u003c/b\u003e \u003c/p\u003e \u003cp\u003eCells were lysed in ice-cold lysis buffer (50 mM Tris-HCl, pH 7.4, containing 150 nM NaCl, 1% Triton X-100, and one tablet of complete protease inhibitors) for 15 min on ice. Cell lysates were centrifuged at 16,000 \u0026times; \u003cem\u003eg\u003c/em\u003e for 30 min at 4\u0026deg;C, and the supernatants were eluted with SDS sample buffer (100 mM Tris-HCl, pH 6.8, 4% SDS, and 20% glycerol with bromophenol blue) and heated for 5 min. The proteins were separated by SDS-PAGE on 10\u0026thinsp;~\u0026thinsp;15% gels and were transferred to PVDF (Bio-Rad, Richmond, CA) membranes using a Mini Trans-Blot Cell (Bio-Rad). The membrane was blocked in 5% BSA (1 h), rinsed, and incubated with the appropriate antibodies in TBST (Tris-buffered saline containing 0.1% Tween 20) and 0.5% BSA overnight. Excess primary antibody was removed by washing the membrane thrice with TBST. The membranes were then incubated with 0.1 \u0026micro;g/mL peroxidase-conjugated secondary antibodies (anti-rabbit or anti-mouse) for 30 min. After three washes with TBST, bands were visualized using western blotting detection reagents and exposed to LOURMAT CHEMI-DOC (Eppendorf, Germany).\u003c/p\u003e \u003cp\u003e \u003cb\u003eSubcellular fractionation\u003c/b\u003e \u003c/p\u003e \u003cp\u003eCells were washed with PBS, harvested, and resuspended in hypotonic buffer (20 mM Tris-HCl, pH 7.4, 10 mM KCl, 2 mM MgCl\u003csub\u003e2\u003c/sub\u003e, 1 mM EGTA, 0.5 mM DTT, 0.5 mM PMSF, and one tablet of complete protease inhibitors), incubated for 5 min followed by the addition of NP-40 to a final concentration of 0.1%. After 3 min of incubation, the cytoplasm and nucleus were separated by centrifugation at 800 \u0026times; \u003cem\u003eg\u003c/em\u003e for 8 min. Subsequently, to ensure the removal of nuclear remnants, the cytoplasmic fractions were centrifuged at 1,500 \u0026times; \u003cem\u003eg\u003c/em\u003e for 5 min, and the supernatant was collected as the final cytoplasmic fraction. The nucleus was purified by 10 min incubation in isotonic lysis buffer (20 mM Tris-HCl, pH 7.4, 150 mM KCl, 2 mM MgCl\u003csub\u003e2\u003c/sub\u003e, 1 mM EGTA, 0.3% NP-40, 0.5 mM DTT, 0.5 mM PMSF, and one tablet of complete protease inhibitors), and centrifuged at 700 \u0026times; \u003cem\u003eg\u003c/em\u003e for 7 min.\u003c/p\u003e \u003cp\u003e \u003cb\u003eChromatin-IP assay (ChIP)\u003c/b\u003e \u003c/p\u003e \u003cp\u003eCells were fixed with formaldehyde (final 1%) for 30 min, harvested by scraper, and centrifuged at 13,000 rpm for 10 min at 4℃. Pellet was resuspended in 800 \u0026micro;L of SDS lysis buffer (1% SDS, 10 mM EDTA, and 50 mM Tris-HCl, pH 8.0). To shear chromatin, the lysate was sonicated on ice for 3 min with sonicator (Bandelin, Germany) tip of 3 mm in diameter at 30% amplitude and 0.5 cycle. Samples were centrifuged at 13,000 rpm for 10 min at 4℃, and 200 \u0026micro;L of supernatant was divided into aliquots for subsequent 10-fold dilution in ChIP dilution buffer (0.01% SDS, 1% Triton X-100, 1.2 mM EDTA, pH 8.0, 167 mM NaCl, and 16.7 mM Tris-HCl, pH 8.0). To test the amount of input DNA for each sample, a 20 \u0026micro;L diluted aliquot was saved for further processing in parallel with all other samples at the reversal of the cross-linking step. Every 2 mL sample of chromatin was precleared with 60 \u0026micro;L of 50% slurry of protein A/G agarose beads (v/v) containing 200 \u0026micro;g/mL herring sperm DNA for 1 h at 4℃ on rotating wheel, after which the beads were pelleted and the supernatant was transferred to a new tube. IgG, anti-p65, or anti-NFATc1 antibody (10 \u0026micro;g) was added to the precleared chromatin sample and incubated overnight at 4\u0026deg;C on a rotating wheel. Immune complexes were collected using 60 \u0026micro;L of 50% slurry of protein A/G agarose beads, 200 \u0026micro;g/mL herring sperm DNA, and rotating for 3 h at 4℃, followed by centrifugation at 1,000 rpm for 1 min at 4℃. The beads were washed for 5 min in low salt wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 500 mM NaCl, and 20 mM Tris-HCl, pH 8.0), LiCl wash buffer (0.25 M LiCl, 1% NP-40, 1% sodium deoxycholate, 1 mM EDTA, and 10 mM Tris-HCl, pH 8.0), and twice with TE buffer. Chromatin complexes were eluted from the beads in 30 min with 400 \u0026micro;L of elution buffer (1% SDS and 0.1 M NaHCO\u003csub\u003e3\u003c/sub\u003e) at 27\u0026deg;C. To reverse cross-linking, 200 mM NaCl was added, and the samples were incubated at 65\u0026deg;C for 4 h. To digest proteins, samples were incubated at 45\u0026deg;C for 90 min after the addition of the following: 10 mM EDTA, 50 \u0026micro;g/mL proteinase K, and 40 mM Tris, pH 6.5. Samples were extracted twice with phenol/chloroform/isoamyl alcohol (25:24:1) and DNA was precipitated with 20 \u0026micro;g of glycogen and two volumes of 100% ethanol. Pellets were collected by centrifugation at 13,000 rpm for 10 min at 4\u0026deg;C. The samples were resuspended in 100 \u0026micro;L of deionized water and stored at -80\u0026deg;C.\u003c/p\u003e \u003cp\u003e \u003cb\u003eReal-time quantitative RT-PCR (qRT-PCR)\u003c/b\u003e \u003c/p\u003e \u003cp\u003eReal-time qPCR was conducted in a CFX384 Touch Real-Time PCR Detection System (Bio-Rad, Hercules, CA) using a TOPreal\u0026trade; SYBR Green RT-qPCR High-ROX Kit (Enzynomics, Korea) and gene-specific primers (forward and reverse pairs, respectively) as follows: p65 binding site on \u003cem\u003eIfn-γ\u003c/em\u003e gene promoter, 5ʹ-AGTAGGTATTTTACTAATCAC-3ʹ and 5 ʹ -AGGAAACTCTTGGGCTTCTCA-3 ʹ; Nfatc1 binding site on \u003cem\u003eIfn-γ\u003c/em\u003e gene promoter, 5 ʹ -TCATGGTTTGAGAAGCCCAAG-3 ʹ and 5 ʹ -TATGGTTTTGTGGCATGTTAG-3 ʹ; negative control region on \u003cem\u003eIfn-γ\u003c/em\u003e gene promoter, 5 ʹ -GTAAGTATGAATTCTTAATAA-3 ʹ and 5 ʹ -ACTCCGTAGTAAGTTGGACAG-3 ʹ, distal region on \u003cem\u003eTnf-α\u003c/em\u003e gene promoter, 5ʹ -AGGAGTGGGAGGGTGGGGGAG-3ʹ and 5ʹ -GGGAGACATGATATTGAGGAG-3ʹ, proximal region on \u003cem\u003eTnf-α\u003c/em\u003e gene promoter, 5ʹ -TGTCCCATTTAGAAATCAAAA-3ʹ; and 5ʹ -AAAAGCTCTCATTCAACCCTC-3ʹ, negative control region on \u003cem\u003eTnf-α\u003c/em\u003e gene promoter, 5ʹ -GTGAGTGTCTGGGCAACCCTT-3ʹ and 5ʹ -CGTTCATTCATCTCTCTGTGC-3ʹ, p65 binding site on \u003cem\u003eIFN-γ\u003c/em\u003e gene promoter, 5ʹ -CTAGGCTGGTCTCAAACTCCT-3ʹ and 5ʹ -ATCAATATACTACATTGTTAA-3ʹ, NFATc1 binding site on \u003cem\u003eIFN-γ\u003c/em\u003e gene promoter, 5ʹ -TTGTTCCCAACCACAAGCAAA-3ʹ and 5ʹ -GATGAGACAGACCCATTATGC-3ʹ, negative control region on \u003cem\u003eIFN-γ\u003c/em\u003e gene promoter, 5ʹ -GTAAGTATGACTTTTTAATAG-3ʹ and 5ʹ -CTCTATACAAACTGAGCAGAA-3ʹ, p65 binding site #1 on \u003cem\u003eTNF-α\u003c/em\u003e gene promoter, 5ʹ -CGGGGCTGTCCCAGGCTTGTC-3ʹ and 5ʹ -GATAGAACTAGAACTGGGAGG-3ʹ, p65 binding site #2 on \u003cem\u003eTNF-α\u003c/em\u003e gene promoter, 5ʹ -TTTTCCTGCATCCTGTCTGGA-3ʹ and 5ʹ -CATCAAGGATACCCCTCACA-3ʹ, NFATc1 binding site on \u003cem\u003eTNF-α\u003c/em\u003e gene promoter, 5ʹ -TTGTGTGTCCCCAACTTTCCA-3ʹ and 5ʹ -GGCTGGGTGTGCCAACAACTG-3ʹ, negative control region on \u003cem\u003eTNF-α\u003c/em\u003e gene promoter, 5ʹ -GTGAGTGCCTGGCCAGCCTTC-3ʹ and 5ʹ -ATCCTGTCTCTCCATCTTTCT-3ʹ. IgG was used as a negative control. The immunoprecipitated DNA was normalized using the following formula: immunoprecipitated DNA\u0026thinsp;=\u0026thinsp;2\u003csup\u003e\u0026minus;(ΔCt of the immuno\u0026minus;precipitated DNA \u0026minus; ΔCt of input)\u003c/sup\u003e, where Ct is the threshold cycle value. In each sample, immunoprecipitated DNA was normalized to the input.\u003c/p\u003e \u003cp\u003e \u003cb\u003eImmunoprecipitation (IP)\u003c/b\u003e \u003c/p\u003e \u003cp\u003eCell lysates were precleared, and supernatants were incubated overnight with antibodies at 4\u0026deg;C, followed by incubation with protein A/G agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA). The beads were collected, washed with PBS, and resuspended in equal volumes of 5\u0026times; SDS loading buffer. Immunoprecipitated proteins were separated by SDS-PAGE on 12% gels and analyzed by western blotting as described above.\u003c/p\u003e \u003cp\u003e \u003cb\u003eCompetitive binding assay using the GENA-104A16\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe 96-well ELISA plate was coated with 10 nM of the target recombinant protein at 4\u0026deg;C overnight. The coated ELISA plate was washed thrice and blocked with PBST. A five-fold serial dilution of GENA-104A16 or anti-APP antibody (5A7) (0.0064\u0026ndash;500 nM, eight points) and 300 nM His-tagged binding recombinant protein in PBS was added, incubated for 1 h at 4\u0026deg;C, and then washed three times with PBST. Anti-His antibody was added, incubated at RT, and washed three times with PBST. Peroxidase-conjugated anti-mouse IgG antibody was added (1:1,000 dilution in PBST), incubated for 1 h at 4\u0026deg;C, and washed three times with PBST. TMB substrate solution (3,3ʹ,5,5ʹ -tetramethylbenzidine, ThermoFisher Scientific) was added and incubated for 10 min at RT. The reaction was stopped with H\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e and absorbance was measured at 405 nm using a microplate reader.\u003c/p\u003e \u003cp\u003eBinding (%) of CNTN4 and APP = (O.D. value of APP and CNTN4 binding in the presence of GENA-104A16) / (O.D. value of APP and CNTN4 binding in the absence of GENA-104A16) \u0026times; 100\u003c/p\u003e \u003cp\u003eThe IC\u003csub\u003e50\u003c/sub\u003e value was determined with normalized nonlinear logistic curve using the GraphPad PRISM.\u003c/p\u003e \u003cp\u003e \u003cb\u003eStatistical analyses\u003c/b\u003e \u003c/p\u003e \u003cp\u003eStatistical analyses were performed using Prism 9.3.1 (GraphPad) or R software (version 4.2.1). Differences between two variables and multiple variables were assessed using the Student\u0026rsquo;s \u003cem\u003et\u003c/em\u003e-test and ANOVA with Tukey's multiple comparison test, respectively. Associations between two continuous variables were estimated using Pearson\u0026rsquo;s correlation test. Associations between two discrete variables were estimated using the chi-square test. Differences were considered significant if the \u003cem\u003ep\u003c/em\u003e-value was less than 0.05. All statistical methods and significance thresholds are described in the corresponding figure legends.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eDisclosure of Potential Conflict of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo potential conflicts of interest were disclosed.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eB.-N.J designed, performed, analyzed the experiments, and prepared the manuscript; S.K and Y.K wrote the manuscript and analyzed experimental results and clinical data; M.C, K.A.P, C.L, and H.P designed the experiments and supervised the manuscript; H.Y, H.K, Y.H, and Y.Y.K, performed the experiments; C.P, and G.K analyzed the bioinformatic data. All the authors have revised the manuscript accordingly.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research was supported by the Korea Drug Development Fund funded by the Ministry of Science and ICT, the Ministry of Trade, Industry, and Energy, and the Ministry of Health and Welfare (RS-2022-00166157, RS-2023-00217717, Republic of Korea).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eCeeraz, S., Nowak, E. C., Burns, C. M. \u0026amp; Noelle, R. J. Immune checkpoint receptors in regulating immune reactivity in rheumatic disease. \u003cem\u003eArthritis Res Ther\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 1\u0026ndash;12 (2014).\u003c/li\u003e\n\u003cli\u003ePardoll, D. M. The blockade of immune checkpoints in cancer immunotherapy. \u003cem\u003eNature Reviews Cancer 2012 12:4\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 252\u0026ndash;264 (2012).\u003c/li\u003e\n\u003cli\u003eSharpe, A. H., Wherry, E. J., Ahmed, R. \u0026amp; Freeman, G. J. The function of programmed cell death 1 and its ligands in regulating autoimmunity and infection. \u003cem\u003eNature Immunology\u003c/em\u003e vol. 8 239\u0026ndash;245 Preprint at https://doi.org/10.1038/ni1443 (2007).\u003c/li\u003e\n\u003cli\u003eDong, H. \u003cem\u003eet al.\u003c/em\u003e Tumor-associated B7-H1 promotes T-cell apoptosis: A potential mechanism of immune evasion. \u003cem\u003eNature Medicine 2002 8:8\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 793\u0026ndash;800 (2002).\u003c/li\u003e\n\u003cli\u003eSyn, N. L., Teng, M. W. L., Mok, T. S. K. \u0026amp; Soo, R. A. De-novo and acquired resistance to immune checkpoint targeting. \u003cem\u003eLancet Oncol\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, e731\u0026ndash;e741 (2017).\u003c/li\u003e\n\u003cli\u003eJohnson, D. B., Chandra, S. \u0026amp; Sosman, J. A. Immune Checkpoint Inhibitor Toxicity in 2018. \u003cem\u003eJAMA\u003c/em\u003e \u003cstrong\u003e320\u003c/strong\u003e, 1702\u0026ndash;1703 (2018).\u003c/li\u003e\n\u003cli\u003eMartins, F. \u003cem\u003eet al.\u003c/em\u003e New therapeutic perspectives to manage refractory immune checkpoint-related toxicities. \u003cem\u003eLancet Oncol\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, e54\u0026ndash;e64 (2019).\u003c/li\u003e\n\u003cli\u003eWilkinson, R. W. \u0026amp; Leishman, A. J. Further advances in cancer immunotherapy: Going beyond checkpoint blockade. \u003cem\u003eFront Immunol\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 1082 (2018).\u003c/li\u003e\n\u003cli\u003eTang, T. \u003cem\u003eet al.\u003c/em\u003e Advantages of targeting the tumor immune microenvironment over blocking immune checkpoint in cancer immunotherapy. \u003cem\u003eSignal Transduction and Targeted Therapy 2021 6:1\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 1\u0026ndash;13 (2021).\u003c/li\u003e\n\u003cli\u003eLiu, D. \u003cem\u003eet al.\u003c/em\u003e Integrative molecular and clinical modeling of clinical outcomes to PD1 blockade in patients with metastatic melanoma. \u003cem\u003eNature Medicine 2019 25:12\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 1916\u0026ndash;1927 (2019).\u003c/li\u003e\n\u003cli\u003eRen, D. \u003cem\u003eet al.\u003c/em\u003e Predictive biomarkers and mechanisms underlying resistance to PD1/PD-L1 blockade cancer immunotherapy. \u003cem\u003eMol Cancer\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 1\u0026ndash;19 (2020).\u003c/li\u003e\n\u003cli\u003eWang, Q. \u0026amp; Wu, X. Primary and acquired resistance to PD-1/PD-L1 blockade in cancer treatment. \u003cem\u003eInt Immunopharmacol\u003c/em\u003e \u003cstrong\u003e46\u003c/strong\u003e, 210\u0026ndash;219 (2017).\u003c/li\u003e\n\u003cli\u003eJenkins, R. W., Barbie, D. A. \u0026amp; Flaherty, K. T. Mechanisms of resistance to immune checkpoint inhibitors. \u003cem\u003eBritish Journal of Cancer 2018 118:1\u003c/em\u003e \u003cstrong\u003e118\u003c/strong\u003e, 9\u0026ndash;16 (2018).\u003c/li\u003e\n\u003cli\u003eHansford, L. M. \u003cem\u003eet al.\u003c/em\u003e Cloning and characterization of the human neural cell adhesion molecule, CNTN4 (alias BIG-2). \u003cem\u003eCytogenet Genome Res\u003c/em\u003e \u003cstrong\u003e101\u003c/strong\u003e, 17\u0026ndash;23 (2003).\u003c/li\u003e\n\u003cli\u003eMercati, O. \u003cem\u003eet al.\u003c/em\u003e Contactin 4, -5 and -6 differentially regulate neuritogenesis while they display identical PTPRG binding sites. \u003cem\u003eBiol Open\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 324\u0026ndash;334 (2013).\u003c/li\u003e\n\u003cli\u003eOguro-Ando, A. \u003cem\u003eet al.\u003c/em\u003e Cntn4, a risk gene for neuropsychiatric disorders, modulates hippocampal synaptic plasticity and behavior. \u003cem\u003eTranslational Psychiatry 2021 11:1\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 1\u0026ndash;15 (2021).\u003c/li\u003e\n\u003cli\u003eBarbari, C. \u003cem\u003eet al.\u003c/em\u003e Immunotherapies and Combination Strategies for Immuno-Oncology. \u003cem\u003eInternational Journal of Molecular Sciences 2020, Vol. 21, Page 5009\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 5009 (2020).\u003c/li\u003e\n\u003cli\u003eDobin, A. \u003cem\u003eet al.\u003c/em\u003e STAR: ultrafast universal RNA-seq aligner. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 15\u0026ndash;21 (2013).\u003c/li\u003e\n\u003cli\u003eLi, B. \u0026amp; Dewey, C. N. RSEM: Accurate transcript quantification from RNA-Seq data with or without a reference genome. \u003cem\u003eBMC Bioinformatics\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 1\u0026ndash;16 (2011).\u003c/li\u003e\n\u003cli\u003eSubramanian, A. \u003cem\u003eet al.\u003c/em\u003e Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e \u003cstrong\u003e102\u003c/strong\u003e, 15545\u0026ndash;15550 (2005).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-2979573/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-2979573/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eImmune checkpoint inhibitors have significantly advanced tumor treatment, but their limited benefits and strong responses in only a subset of patients persist as challenges. CNTN4, a neuronal membrane protein involved in cell adhesion and synapse signaling, has unclear immunomodulatory functions. In this study, we reveal the immune checkpoint role of CNTN4 in T-cell proliferation and activation both \u003cem\u003ein vitro\u003c/em\u003e and \u003cem\u003ein vivo\u003c/em\u003e. We found that CNTN4, highly expressed in numerous tumor tissues, impedes T-cell proliferation, cytotoxicity, and the secretion of cytotoxic cytokines \u003cem\u003ein vitro\u003c/em\u003e. On T cells, CNTN4 binds to two APP isoforms, APP770 and APP751, which results in attenuated TCR signaling and diminished cell adhesion capacity. To target this interaction, we developed GENA-104A16 against CNTN4 and an anti-APP antibody (5A7) that blocks the binding between CNTN4 and APP. Administering these two antibodies demonstrated anti-tumor effects in a syngeneic tumor mouse model and increased tumor-infiltrating lymphocytes within tumor tissues \u003cem\u003ein vivo\u003c/em\u003e. Furthermore, elevated CNTN4 levels are associated with poor prognosis and negatively correlated with various cytotoxic immune-related markers. In conclusion, CNTN4 serves as a bona fide immune checkpoint protein and represents a promising therapeutic target for developing immunotherapeutic drugs.\u003c/p\u003e","manuscriptTitle":"CNTN4/APP axis of cancer cells and T-cells","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2023-06-09 18:18:53","doi":"10.21203/rs.3.rs-2979573/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a29f284b-bd7d-4b6d-86c4-9a41b55577cc","owner":[],"postedDate":"June 9th, 2023","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":21936990,"name":"Biological sciences/Immunology/Immunotherapy/Immunosuppression"},{"id":21936991,"name":"Biological sciences/Cancer/Tumour immunology"},{"id":21936992,"name":"Biological sciences/Cancer/Cancer microenvironment"}],"tags":[],"updatedAt":"2023-06-29T07:31:47+00:00","versionOfRecord":[],"versionCreatedAt":"2023-06-09 18:18:53","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-2979573","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-2979573","identity":"rs-2979573","version":["v1"]},"buildId":"WrCJVZZCHTDjtuVLN7oU0","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}