A bispecific antibody targeting PD-L1/TNFR2 increases tumor targeting and enhances antitumor efficacy in colorectal cancer

Abstract

Background

Immune checkpoint inhibitors (ICIs) have shown limited efficacy in colorectal cancer (CRC), largely due to immunosuppressive tumor microenvironment (TME) including regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs). Additionally, the off-target effects of ICIs can reduce drug accumulation in tumor tissues and lead to immune-related adverse events, further compromising their clinical utility.

Methods

Using knob-into-hole technology, we developed ATAPL1, a bispecific antibody targeting both programmed death-ligand 1 (PD-L1) and tumor necrosis factor receptor 2 (TNFR2). In multiple subcutaneous and orthotopic CRC mouse models, ATAPL1 was evaluated for its tumor targeting, antitumor efficacy, immunomodulatory effects within the TME, and potential combinations with standard chemotherapies.

Results

Compared with αPD-L1 or αTNFR2 monotherapy, ATAPL1 showed enhanced tumor accumulation and significantly improved antitumor efficacy, with no significant toxicity. It effectively modulated the TME by decreasing immunosuppressive Tregs and monocytic MDSCs (M-MDSCs), while promoting CD8⁺ T-cell activation and macrophage function. ATAPL1 also primed long-term immune surveillance, suggesting durable antitumor activity. Furthermore, ATAPL1 overcame chemotherapy resistance and exhibited superior therapeutic effects when combined with FOLFOX or FOLFIRI, compared with either treatment alone.

Conclusion

These findings support ATAPL1 as a promising dual-targeted therapeutic strategy for CRC. It offers the potential to overcome immune and chemotherapy resistance, reduce off-target effects, and improve treatment outcomes through enhanced immune activation and tumor targeting.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Immune checkpoint inhibitors, such as α programmed death-ligand 1 (PD-L1), exhibit limited clinical efficacy in colorectal cancer, largely attributable to a profoundly immunosuppressive tumor microenvironment, such as regulatory T cells (Tregs) and myeloid-derived suppressor cells. Furthermore, their therapeutic potential is compromised by significant off-target effects, which contribute to immune-related adverse events and suboptimal drug accumulation within tumor tissues.

WHAT THIS STUDY ADDS

• Our study presents ATAPL1, a novel bispecific antibody that targets both PD-L1 and tumor necrosis factor receptor 2 (TNFR2). This antibody enhances tumor-specific targeting to avoid off-target effects and remodels the immunosuppressive tumor microenvironment by depleting Tregs and monocytic myeloid-derived suppressor cells (M-MDSCs) while activating CD8⁺ T cells. Furthermore, our findings demonstrate that PDL1/TNFR2 blockade overcomes chemotherapy resistance. ATAPL1 also induces long-term antitumor immunity, yielding superior synergistic effects when combined with standard care regimens.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• The adoption of a dual-targeted therapeutic strategy—with complementary targets like PD-L1 and TNFR2—offers the potential to surmount immune and chemotherapy resistance, minimize off-target effects, and lay a foundation for future clinical trials focused on colorectal cancer.

Background

Colorectal cancer (CRC) ranks among the leading causes of cancer-related deaths worldwide, and patients who develop resistance to conventional treatments face limited therapeutic alternatives.^{1 2} While immune checkpoint inhibitors (ICIs), such as programmed death-ligand 1 (PD-L1) and programmed cell death protein 1 (PD-1) blockade, have shown efficacy in several types of cancer, their impact on CRC has been restricted due to both inherent and acquired resistance mechanisms.^{3 4} Immune-suppressive cells, such as regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs), play a crucial role in the development of resistance to immunotherapy.5,8 In addition, off-target effects of ICIs may reduce drug accumulation within the tumor microenvironment (TME) and lead to immune-related adverse events (irAEs), further limiting the clinical benefit.

A promising strategy to overcome the limitations of therapeutic resistance and off-target effects is the development of bispecific antibodies. These antibodies, which simultaneously target two different antigens, offer several advantages over traditional monoclonal antibodies.^{9 10} Bispecific antibodies can synergistically activate antitumor immunity through multiple mechanisms, providing a more robust immune response. The dual-target approach not only enhances therapeutic efficacy but also addresses the complexity and heterogeneity of the TME, which often limits the effectiveness of monotherapies.11,13 Furthermore, bispecific antibodies have the potential to reduce off-target effects, improving the precision of therapy by selectively targeting the specific cells and pathways involved in immune evasion and tumor progression.^{14 15}

Recent research has highlighted tumor necrosis factor receptor 2 (TNFR2) as a promising therapeutic target for overcoming resistance to immunotherapy.^{16 17} TNFR2 is an inducible receptor primarily expressed on immune cells, with its expression significantly upregulated in inflammatory microenvironments, particularly in Tregs and monocytic MDSCs (M-MDSCs). Targeting TNFR2 may help disrupt the immunosuppressive effects of Tregs and MDSCs, thereby enhancing the antitumor immune response.18,21 The combination of TNFR2 targeting with ICIs has been explored in several preclinical studies, and antibodies targeting TNFR2 are already being evaluated in clinical trials.22,25

In this study, we propose the development of ATAPL1, a bispecific antibody targeting both PD-L1 and TNFR2, aiming to overcome therapeutic resistance and reduce off-target effects associated with ICIs in CRC. We evaluated the preclinical efficacy of ATAPL1 in CRC mouse models, focusing on its ability to remodel the TME and elicit a more robust and sustained antitumor immune response. Additionally, we determined the tumor-targeting capability of ATAPL1 and its safety profile. Finally, the therapeutic efficacy of chemotherapy combined with ATAPL1 was evaluated in CRC models. This approach provides a novel strategy to enhance therapeutic efficacy and broaden its potential application in clinical treatment of CRC.

Materials and methods

Cell lines and cells culture

Cell lines MC38, CT26, NIH/3T3, HCT116 and CHO-K1 were purchased from ATCC. MC38^OVA, MC38^Luc, MC38^OVA/Luc cell line was derived from MC38 cells transfected with lentivirus encoding OVA or Luciferase. CHO^mTNFR2, CHO^mPDL1, CHO^hTNFR2 and CHO^hPDL1 cell lines were derived from CHO-K1 cells transfected with lentivirus encoding mouse TNFR2, mouse PD-L1, human TNFR2 and human PD-L1, respectively. CHO-K1 cells were cultured in Dulbecco’s modified eagle medium (DMEM)/F12 medium (Gibco), and other cells were cultured in DMEM medium (Gibco). All mediums were supplemented with 10% fetal bovine serum (TransGen, China) and 1% penicillin-streptomycin solution (Yeasen, China). Cells were cultured in a humidified incubator with an atmosphere containing 5% CO₂ at 37°C.

Recombinant antibodies construction

The αTNFR2 construct comprises the sequence of the anti-mouse TNFR2 antibody, as derived from patent US20220073634A1, with the Fc region of mouse IgG2a. The αPD-L1 is composed of the variable heavy (VH) and variable light (VL) chains sequence of atezolizumab, linked to the Fc region of mouse IgG2a. ATAPL1 is assembled by combining αTNFR2 and αPD-L1 through the “Knob into hole” technology. The αhTNFR2 is composed of anti-human TNFR2 (ViroTher Biopharm, China) linked to the Fc region of human IgG1. The αhPD-L1 is composed of the VH and VL sequence of atezolizumab, linked to the Fc region of human IgG1. hATAPL1 is assembled by combining αhTNFR2 and αhPD-L1 through the “Knob into hole” technology. The recombinant antibodies were expressed using ExpiCHO (Thermo Fisher Scientific) and purified through protein A affinity chromatography.

ELISA

To quantify the binding ability of ATAPL1 in vitro, a polystyrene microplate was pre-coated with 300 ng/well mouse PD-L1-his (ViroTher Biopharm, China) or mouse TNFR2-his (Sino Biological). Then, 100 μL of purified protein (gradient dilution) was added and incubated for 2 hours at 37°C. After washing, anti-IgG Fc tag antibody (Horseradish Peroxidase (HRP)) (Immuno Way, RS0001) was added for 1 hour incubation at 37°C. 3,3’,5,5’-tetramethylbenzidine (TMB) (Beyotime, P0210) was then added as the substrate, and the absorbance was read at 450 nm.

To quantify the binding ability of hATAPL1 in vitro, a polystyrene microplate was pre-coated with 100 ng/well human PD-L1-his (ViroTher Biopharm, China) or human TNFR2-his (Sino Biological). Then, 100 μL of purified protein (gradient dilution) was added and incubated for 2 hours at 37°C. After washing, anti-human IgG Fc tag antibody (HRP) (GenScript, A01854-200) was added for 1 hour incubation at 37°C. TMB (Beyotime, P0210) was then added as the substrate, and the absorbance was read at 450 nm.

Mouse interleukin (IL)-6 (BioLegend, 431304), tumor necrosis factor (TNF)-α (BioLegend, 430904) and interferon (IFN)-γ (BioLegend, 438804) were quantified using ELISA kits.

CD8⁺ T cells activation

MC38^OVA tumor cells and OT-1-derived CD8⁺ T cells, isolated using anti-CD8 microbeads (Miltenyi, 130-045-201), are co-cultured at a ratio of 1:5 (tumor cells to T cells) in the presence of 10 μg/mL of antibody (Isotype, αTNFR2, αPD-L1, ATAPL1). After 48 hours of incubation, flow cytometry is used to analyze CD69 expression on CD8⁺ T cells as an activation marker. Additionally, the supernatant is collected and analyzed by ELISA to measure IFN-γ secretion.

T cells proliferation

OT-1-derived T cells isolated from the spleen were labeled with carboxyfluorescein succinimidyl ester (CFSE) and co-cultured with MC38^OVA tumor cells at a 1:5 ratio (tumor cells to T cells) in the presence of 10 μg/mL of antibody (Isotype, αTNFR2, αPD-L1, ATAPL1). After 48 hours of incubation, T-cell proliferation was assessed by flow cytometry. The CFSE dilution was used as a marker of T-cell proliferation, with increased dilution indicating cell division.

Cytotoxicity of CD8⁺ T cells

OT-1-derived CD8⁺ T cells from the spleen were co-cultured with MC38^OVA tumor cells at various ratios 1:5 (tumor cells to T cells) in the presence of 10 μg/mL antibody (Isotype, αTNFR2, αPD-L1, ATAPL1) to evaluate the impact of antibody treatment on T cell-mediated cytotoxicity. After 24 hours, T cell-mediated killing of tumor cells was assessed by the Cell Counting Kit-8 (CCK-8) assay to evaluate tumor cell viability and T-cell cytotoxicity.

Phagocytosis

Macrophages are generated by differentiating bone marrow-derived macrophages (BMDM) from mouse bone marrow cells in the presence of macrophage colony-stimulating factor (M-CSF) for 6 days. After differentiation, the macrophages are polarized by adding 30 ng/mL IL-4 and IL-10 for 48 hours. These macrophages are then co-cultured with PD-L1-overexpressing CT26 tumor cells at a 1:1 ratio in a 12-well plate. 10 μg/mL of antibody (Isotype, αTNFR2, αPD-L1, ATAPL1) is added to the co-culture, and after 4 hours, phagocytosis is assessed by flow cytometry. The tumor cells are labeled with CFSE (Beyotime, C0051), and macrophages are identified using CD45 markers.

H&E staining assay

To determine the potential toxicity of ATAPL1, organs including the kidney, liver, spleen, lung, brain, heart, and colon were obtained from mice after treatment, fixed with 4% paraformaldehyde for 30 hours, and embedded in paraffin (Servicebio, G1101). 4 µm sections were cut and stained with H&E.

Flow cytometry

For extracellular staining, single-cell suspensions were incubated with fluorescent antibodies at room temperature for 15 min. Afterward, cells were fixed using a 4% (w/v) paraformaldehyde (PFA) solution and analyzed with either the Attune NxT Cytometer (Thermo Fisher) or CytoFLEX Flow Cytometer (Beckman). For intracellular staining, cell membranes were permeabilized using the Fixation/Permeabilization Solution Kit (BD Cytofix/Cytoperm, 554714), followed by a 30 min incubation with primary antibodies in the dark. After washing with phosphate-buffered saline, cells were resuspended in 4% (w/v) PFA solution and analyzed by flow cytometry. Data analysis was performed using FlowJo software. The following antibodies were used for flow cytometry staining: CD45-APC-Cy7 (BD), CD11b-FITC (BD), F4/80-BV421 (BD), Ly6G-PE (BD), CD206-Alexa Fluor 647 (BD), CD3-FITC (BD), CD4-Alexa Fluor 700 (BD), CD4-BV421 (BD), CD8-PE-Cy7 (BD), PD-L1-BV605 (BD), TNFR2-BV421 (BioLegend), CD69-BB700 (BD), CD25-PE (BD), GzmB-PE (BD), CD107a-APC AF700 (BD), Lag3-PE (BD), Tim3-APC (BD), IFN-γ-BV421 (BD), FOXP3-Alexa Fluor 647 (BD), CD86-PE-Cy7 (BD), Ly6C-BV605 (BD) and FVS510 (BD, 564406) for live/dead cells staining.

Animal experiments

6–8 weeks old Balb/c mice, C57BL/6 mice and NOD-Prkdc^scid Il2rg^null (NCG) mice were purchased from the Model Animal Research Center of Nanjing University (Nanjing, China).

To establish subcutaneous mouse tumor models, 2×10⁶ MC38 tumor cells were implanted subcutaneously into the right flank of C57BL/6 mice, while CT26 tumor cells were implanted in the same manner into Balb/c mice. When the tumors reached approximately 100 mm³, the mice were randomly divided into groups and injected with antibodies. Tumor diameters were measured using a vernier caliper, and tumor volume was calculated using the formula 0.5×length×width².

In the humanized mouse model, NCG mice were subcutaneously inoculated with 5×10⁶ human HCT116 tumor cells. When the tumors reached approximately 100 mm³, the mice were randomly grouped and human peripheral blood mononuclear cells (PBMCs) (Milestone Biotechnologies, China) were intravenously injected (5×10⁶ PBMCs). On days 1, 4, and 7 post-PBMC injection, antibodies were administered and tumor volume was monitored.

For the depletion of CD8⁺ T cells, CD4⁺ T cells, or macrophages, mice were injected intraperitoneally (i.p.) with 100 µg/dose of anti-CD8α antibody (Bio X Cell, BP0117), anti-CD4 antibody (Bio X Cell, BE0003-1), or 200 µL/dose clodronate liposomes (CL, Yeasen 40338ES10) on the day before treatment and the day after the second dose of antibodies.

For the combination therapy of ATAPL1 with FOLFOX, CT26 tumor-bearing mice were i.p. injected with 50 mg/kg 5-Fluorouracil (5-FU) (Heowns, F-80675) and 10 mg/kg leucovorin (Macklin, F809591) on days 0, 3, 6, and 9, following the initial treatment. Additionally, 5 mg/kg oxaliplatin (Hengrui, Oxaliplatin for Injection) and 100 µg of antibodies were administered on days 0 and 6.

For the combination therapy of ATAPL1 with FOLFIRI, MC38 tumor-bearing mice were i.p. injected with 50 mg/kg 5-FU and 10 mg/kg leucovorin on days 0, 3, 6, and 9, following the initial treatment. Additionally, 50 mg/kg irinotecan (MCE, 100286-90-6) and 100 µg of antibodies were administered on days 0 and 6.

To create an MC38 orthotopic tumor model, first harvest the subcutaneous MC38^Luc tumor from a donor mouse. After euthanasia, carefully excise the tumor in a sterile environment and cut it into 1–2 mm³ pieces. Next, anesthetize a recipient C57BL/6 mouse and make a small incision in the abdominal area to expose the colon. Form a small pocket in the serosal layer of the colon and implant the tumor fragment. Secure the pocket with surgical glue and close the abdominal incision with sterile sutures. After surgery, monitor the mouse for recovery and tumor growth over the course of 7 days. The mice were then randomly assigned to groups and i.p. injected with antibodies. Tumor volume was tracked via Bioluminescence Imaging using D-luciferin.

All animals were euthanized when the tumor volume reached 2,000 mm³.

Statistical analyses

All statistical analyses were performed using Prism V.9.4.0 (GraphPad Software, California, USA). Data are presented as the means±SD or means±SEM. The significance was analyzed using Student’s t-test or analysis of variance. Survival curves were plotted according to the Kaplan-Meier method and compared using a log-rank (Mantel-Cox) test. P values<0.05 were considered significant and are indicated as *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Results

TNFR2⁺ Tregs and M-MDSCs contribute to αPD-L1 resistance in mouse colorectal tumor model

We first evaluated the therapeutic efficacy of PD-L1 antibody (αPD-L1) in the CT26 tumor-bearing mouse model and found that αPD-L1 treatment did not significantly inhibit tumor growth, nor did it notably extend the survival of the mice (figure 1A and B). These results indicated the presence of primary resistance to αPD-L1 in the CT26 mouse model. Our previous research has shown that the persistent presence of Treg and M-MDSC cells in the TME is associated with immune therapy resistance, and that most of these immunosuppressive cells express TNFR2.²⁶ Then, we analyzed the changes in Treg and M-MDSC cells in the CT26 TME after αPD-L1 treatment. We found that αPD-L1 treatment did not significantly alter the populations of Treg and M-MDSC cells in TME (figure 1C), and further analysis showed over 60% of these cells were TNFR2 positive (figure 1D). Similarly, in the MC38 tumor-bearing mouse model, we also observed over 60% of Treg and M-MDSC cells in TME expressed TNFR2 following αPD-L1 treatment (online supplemental figure S1A and B). Based on these findings, we hypothesized that combining αPD-L1 and TNFR2 antibody (αTNFR2) could overcome the resistance to αPD-L1 therapy in colorectal tumor model. On combination treatment of αPD-L1 and αTNFR2 (Combo), we observed significant control of tumor growth and a notable extension of survival in the tumor-bearing mice (figure 1E–1H). Taken together, these results suggested that TNFR2⁺ Treg and M-MDSC cells were involved in resistance to PD-L1 antibody therapy, and that targeting both PD-L1 and TNFR2 exerted a synergistic antitumor effect in the mouse CRC model.

Figure 1. TNFR2⁺ Tregs and M-MDSCs contribute to αPD-L1 resistance in mouse colorectal tumor model. (A) 6–8 weeks old Balb/c mice bearing subcutaneous CT26 colorectal tumor received i.p. three doses of 200 µg Isotype, or αPD-L1 on days 0, 3 and 6. (B) Tumor growth and Kaplan-Meier survival curves were monitored. Six mice each group. (C and D) 6–8 weeks old Balb/c mice bearing subcutaneous CT26 colorectal tumor received i.p. two doses of 200 µg Isotype, or αPD-L1 on days 0 and 3, 48 hours after the second injection, tumor tissues were harvested and single-cell suspensions were analyzed by flow cytometry, eight mice each group. (C) The percentages of tumor infiltrating Tregs and M-MDSCs of bulk tumor cells. (D) The TNFR2 expression on the Tregs and M-MDSCs after αPD-L1 treatment. (E) Experimental scheme of CT26 colorectal tumor model. 6–8 weeks old Balb/c mice bearing subcutaneous colorectal carcinoma received i.p. three doses of 200 µg Isotype, αPD-L1, αTNFR2 or Combo treatment on days as indicated. Seven mice each group. (F and G) Tumor growth and (H) Kaplan-Meier survival curves were monitored. Tumor growth is shown as means±SEM. The remaining data are shown as means±SD. ns, not significant, *p<0.05, **p<0.01, ****p<0.0001. i.p., intraperitoneally; MDSC, myeloid-derived suppressor cell; PD-L1, programmed death-ligand 1; TNFR2, tumor necrosis factor receptor 2; Treg, regulatory T cell; M-MDSC, monocytic MDSC.

ATAPL1 exhibits significantly increased accumulation in TME, and activities on immune cells

Next, we analyzed the expression patterns of TNFR2 and PD-L1 in colon adenocarcinoma (COAD) patients (GEPIA2 database), and found that TNFR2 was upregulated, whereas the expression of PD-L1 remained nearly unchanged in tumor tissue compared with adjacent normal tissue (figure 2A). Notably, the expression level of TNFR2 in COAD tumors was substantially higher than that of PD-L1 (figure 2A). Due to TNFR2 was primarily inducibly expressed and predominantly found on immune cells, we compared TNFR2 expression in tumor tissues with that in peripheral blood leukocytes in the CT26 tumor-bearing mouse model. We found that TNFR2 expression was significantly upregulated in tumor tissues, particularly on CD11b⁺ myeloid-derived cells and CD4⁺ T cells (figure 2B and C). We also compared TNFR2 expression in tumor tissues with that in immune cells from the spleen and draining lymph nodes. The results also showed that TNFR2 expression was significantly elevated in tumor-associated immune cells (figure 2D). These results suggest that TNFR2 is a good target for CRC treatment.

Figure 2. ATAPL1 exhibits significantly increased accumulation in TME, and activities on immune cells. (A) The expression levels of CD274 and TNFRSF1B in COAD were analyzed using the GEPIA2 database (http://gepia2.cancer-pku.cn/). (B) The TNFR2 expression on WBC and tumor tissue cells from naïve or CT26 tumor-bearing mice was analyzed by flow cytometry. (C) The TNFR2 expression on CD45⁺CD11b⁺ and CD45⁺CD4⁺ cells in WBC and tumor tissue cells from naïve or CT26 tumor-bearing mice was analyzed by flow cytometry (n=5). (D) The TNFR2 expression on CD45⁺ cells in draining lymph node, spleen cells and tumor tissue cells (n=5). (E) Schematic diagram of the construction of ATAPL1. (F) The ability of ATAPL1 to bind mouse PD-L1 and TNFR2 was calculated using ELISA. (G and H) The binding of ATAPL1 to CHO^mPDL1 cells or CHO^mTNFR2 was analyzed by flow cytometry (n=3). (I) 6–8 weeks old Balb/c mice bearing subcutaneous CT26 colorectal tumor received i.p. 200 µg ATAPL1 or αPD-L1, the distribution of antibodies in vivo was calculated using ELISA (n=3). (J) CD8⁺ T cells isolated from the spleen of OT-1 mice were pre-stained by CFSE, and then co-cultured with MC38^OVA tumor cells in the presence of 10 µg/mL antibodies as indicated, 48 hours later, T-cell proliferation was analyzed by flow cytometry (n=4). (K–M) MC38^OVA tumor cells were co-cultured with CD8⁺ T cells isolated from the spleen of OT-1 mice in the presence of 10 µg/mL antibodies as indicated, 48 hours after incubation, (K) the CD69 expression on CD8⁺ T cells was analyzed by flow cytometry (n=3). (L) The expression of IFN-γ in the co-culture supernatant was measured by ELISA (n=6). (M) The cytotoxic effect of CD8⁺ T cells on MC38^OVA tumor cells was assessed by the CCK-8 assay (n=6). (N) Macrophages induced from BMDMs were co-cultured with PD-L1-overexpressing CT26 tumor cells labeled with CFSE in the presence of 10 µg/mL antibodies as indicated, 4 hours later, phagocytosis was determined by flow cytometry (n=4). The data are shown as means±SD. ns, not significant, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. BMDM, bone marrow-derived macrophages; COAD, colon adenocarcinoma; IFN, interferon; i.p., intraperitoneally; KIH, knob-into-hole; PD-L1, programmed death-ligand 1; TME, tumor microenvironment; TNFR2, tumor necrosis factor receptor 2; WBC, white blood cell; CCK8, Cell Counting Kit-8; CFSE, carboxyfluorescein succinimidyl ester; MFI, mean fluorescence intensity; scFv, single-chain variable fragment.

Based on these observations, we developed a bispecific antibody named ATAPL1, consisting of αPD-L1 and αTNFR2 (single-chain variable fragment (scFv)), using knob-into-hole technology (figure 2E). We confirmed that ATAPL1 effectively binds to both PD-L1 and TNFR2 targets, with half maximal effective concentration (EC₅₀) values of 17.7 ng/mL and 3.7 ng/mL (figure 2F), respectively. Additionally, ATAPL1 effectively bound to both TNFR2 and PD-L1 expressed on the surface of cells (figure 2G and H). Importantly, ATAPL1 demonstrated more than 2.8–4.1-fold higher tumor accumulation compared with αPD-L1 within 48 hours in tumor tissues following intraperitoneal injection of αPD-L1 or ATAPL1 (figure 2I). These results suggest that systemic delivery of ATAPL1 increases tumor targeting efficacy.

Next, we examined the immunomodulatory activity of ATAPL1 on CD8⁺ T cells. In vitro, both ATAPL1 and αTNFR2 significantly promoted CD8⁺ T-cell proliferation (figure 2J and S2A), indicating that TNFR2 blockade enhanced CD8⁺ T-cell expansion. Moreover, ATAPL1 significantly enhanced CD8⁺ T-cell activation (figure 2K), promoting the secretion of IFN-γ (figure 2L), and boosting the cytotoxic activity of T cells against tumor cells (figure 2M).

In addition to CD8⁺ T cells, we also evaluated the impact of ATAPL1 on macrophages. Surprisingly, ATAPL1 massively enhanced macrophage phagocytosis of tumor cells up to 2.5-fold compared with either isotype, αPD-L1 or αTNFR2 (figure 2N and S2B).

Previous studies have established that cancer-associated fibroblasts (CAFs) express both TNFR2 and PD-L1,^{27 28} providing a strong molecular rationale for our hypothesis that ATAPL1 may directly target CAFs. To investigate it, we pretreated 3T3-derived CAFs with ATAPL1 and subsequently evaluated its effect on the cytotoxic activity of T cells. Our findings revealed that ATAPL1 pretreatment significantly mitigated CAF-mediated suppression of T-cell cytotoxicity, demonstrating its ability to effectively disrupt CAF-derived immunosuppressive signals (online supplemental figure S2C).

Taken together, these results indicated that the bispecific ATAPL1 could bind to both PD-L1 and TNFR2, increasing antibody accumulation in TME, activating CD8⁺ T-cell proliferation and function, and enhancing macrophage phagocytosis.

Therapeutic efficacy of ATAPL1 in colorectal cancer mouse models

Subsequently, we investigated the antitumor effects of ATAPL1 in a CRC CT26 mouse model (figure 3A). We found that, compared with monotherapy with either αPD-L1, αTNFR2 or combination of αPD-L1 and αTNFR2, ATAPL1 significantly inhibited tumor growth and extended the survival of the mice (figure 3B and C). Moreover, we monitored the body weight change of treated mice and found that the ATAPL1-treated group exhibited the most stable body weight among all the groups (figure 3D).

Figure 3. Therapeutic efficacy of ATAPL1 in colorectal cancer mouse models. (A) 6–8 weeks old subcutaneous CT26 tumor-bearing Balb/c mice received i.p. three doses of 200 µg Isotype, αTNFR2, αPD-L1, Combo or ATAPL1 treatment on days as indicated. Seven mice each group. (B) Tumor growth, (C) Kaplan-Meier survival curves and (D) the body weight change was monitored. (E) Experimental scheme of MC38^Luc orthotopic tumor model. 6–8 weeks old C57BL/6 mice bearing orthotopic colorectal carcinoma received i.p. five doses of 200 µg Isotype, αTNFR2, αCD47 or ATAPL1 treatment on days as indicated. Six mice each group. (F) Tumor growth and (G) Kaplan-Meier survival curves were monitored. (H) 6–8 weeks old subcutaneous CT26 tumor-bearing Balb/c mice received i.p. three doses of 6 mg/kg Isotype, 6 mg/kg ATAPL1, 3 mg/kg ATAPL1 or 1 mg/kg ATAPL1 treatment on days as indicated. (I) Tumor growth was monitored. Five mice each group. Tumor growth is shown as means±SEM. The remaining data are shown as means±SD. ns, not significant, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. i.p., intraperitoneally; PD-L1, programmed death-ligand 1; TNFR2, tumor necrosis factor receptor 2.

Additionally, we also evaluated the therapeutic effect of ATAPL1 in an orthotopic CRC MC38 mouse model (figure 3E). The results showed that ATAPL1 significantly inhibited the tumor growth and prolonged the survival of the mice compared with monotherapy with either isotype, αPD-L1 or αTNFR2 (figure 3F and G).

To determine whether the antitumor activity of ATAPL1 was dose-dependent, we assessed its therapeutic efficacy at varying doses in the CT26 tumor-bearing mouse model (figure 3H). The results showed a clear dose-dependent response, with significantly enhanced antitumor effects observed at higher doses (figure 3I).

To assess the safety profile of ATAPL1, we first performed histological analyses of major organs—including the heart, liver, spleen, lungs, kidneys, brain, and intestines—in CT26 tumor-bearing mice after treatment. Compared with the isotype control group, ATAPL1 treatment did not induce observed pathological alterations in these tissues (online supplemental figure S3A). Furthermore, we evaluated the systemic inflammatory response by measuring peripheral blood cytokine levels in naive mice. While ATAPL1-treated mice exhibited a modest increase in inflammatory cytokines compared with untreated controls, and this elevation did not reach statistical significance, suggesting minimal systemic immune activation (online supplemental figure S3B). Additionally, liver function markers (alanine aminotransferase (ALT) and aspartate aminotransferase (AST)), kidney function markers (blood urea nitrogen (BUN), urea (UREA), and creatinine (CREA)), and hematological parameters remained within normal ranges after ATAPL1 treatment (online supplemental figures S3C and D), collectively indicating a favorable safety profile.

Taken together, these findings indicated that ATAPL1 represented a safe and effective therapeutic strategy for overcoming PD-L1 antibody resistance in CRC models.

Immune microenvironment remodeling by ATAPL1 in colorectal cancer models

To further explore the impact of ATAPL1 on the TME, we evaluated the changes in the TME following ATAPL1 treatment in the CT26 tumor-bearing mouse model. The results revealed that ATAPL1 significantly reduced Tregs and M-MDSCs in TME (figure 4A). In contrast, ATAPL1 treatment significantly increased CD8⁺ T cells in TME compared with αPD-L1 (figure 4B) and upregulated their expression of granzyme B and CD107a (figure 4C). Interestingly, ATAPL1 also reduced the expression of Tim3 and Lag3 on CD8⁺ T cells (figure 4D). These findings suggest that ATAPL1 treatment effectively activates CD8⁺ T cells while preventing their exhaustion. Similar results were observed in the MC38 tumor-bearing mouse model, with ATAPL1 treatment significantly increasing the proportions of lymphocytes, CD8⁺ T cells, and activated IFN-γ⁺CD8⁺ T cells in the tumor tissue (online supplemental figure S4).

Figure 4. Immune microenvironment remodeling by ATAPL1 in colorectal cancer models. (A–G) Experimental scheme of flow cytometric analysis after treatment. 6–8 weeks old subcutaneous CT26 tumor-bearing Balb/c mice received i.p. two doses of 200 µg Isotype, αTNFR2, αPD-L1, or ATAPL1 treatment. Tumor tissues were harvested for analysis at 48 hours after the second injection. Six mice per group. (A) The percentages of tumor-infiltrating Tregs and M-MDSCs of intratumoral cells. (B) The percentages of tumor-infiltrating CD8⁺ T cells of intratumoral cells. (C) The expression of GzmB and CD107a on the CD8⁺ T cells. (D) The expression of Tim3 and Lag3 on the CD8⁺T cells. (E) The percentages of tumor-infiltrating TAMs of intratumoral cells. (F) The expression of CD86 and MHC-II on the TAMs. (G) The expression of CD206 on the TAMs. (H and I) 6–8 weeks Balb/c mice bearing subcutaneous CT26 tumor received i.p. three doses of 200 µg Isotype or ATAPL1 on days as indicated. 200 µL/dose CL, 100 µg/dose of anti-CD8α or anti-CD4 were injected i.p. 1 day before the above treatment and repeated 1 day after the second dose. Seven mice per group. (I) Tumor growth was monitored. (J) The cured CT26 tumor-bearing mice post-ATAPL1 treatment were rechallenged with 5×10⁶ CT26 cells implanted at the contralateral site. Age-matched naïve mice were used as controls. Tumor growth was monitored. Five mice per group. (K) Spleen cells from ATAPL1-treated, cured CT26 tumor-bearing mice were co-cultured with CT26 cells in vitro (Cured+CT26 group). Spleen cells from cured mice were co-cultured with 4T1 cells (CR+4T1 group) as the control group, and spleen cells from naive mice were co-cultured with CT26 cells (naive+CT26 group) as another control group. IFN-γ expression in CD8⁺ T cells was measured by ELISA at 48 hours (n=4). Tumor growth is shown as means±SEM. The remaining data are shown as means±SD. ns, not significant, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. CL, clodronate liposomes; GzmB, granzyme B; IFN, interferon; i.p., intraperitoneally; MDSC, myeloid-derived suppressor cell; MHC, major histocompatibility complex; PD-L1, programmed death-ligand 1; TAM, tumor-associated macrophage; TNFR2, tumor necrosis factor receptor 2; Treg, regulatory T cell; Lag3, lymphocyte activation gene 3; M-MDSC, monocytic MDSC; MFI, mean fluorescence intensity; Tim3, T cell immunoglobulin and mucin-domain containing-3.

Additionally, we analyzed the changes in tumor-associated macrophages (TAMs) within the TME. Although ATAPL1 treatment did not significantly affect the infiltration of TAMs (figure 4E), it did lead to a marked upregulation of CD86 and major histocompatibility complex (MHC)-II expression on TAMs (figure 4F), while the expression of CD206 was significantly reduced compared with isotype control (figure 4G). These results indicate that ATAPL1 treatment also remodels TAMs towards antitumoral phenotypes.

To identify the key effector cells responsible for the antitumor effect of ATAPL1, we assessed its therapeutic efficacy in mouse models with specific immune cell deletions, including CD4⁺ T cells, CD8⁺ T cells, and TAMs (figure 4H). The results demonstrated that deletion of either CD8⁺ T cells or TAMs significantly inhibited the antitumor effects of ATAPL1, suggesting that the antitumor efficacy of ATAPL1 relied on the functionality of both CD8⁺ T cells and TAMs (figure 4I).

To determine whether ATAPL1 treatment primes long-term immune surveillance, we re-implanted CT26 tumor cells into mice previously cured by ATAPL1 treatment. These mice showed complete inhibition of tumor growth (figure 4J). Furthermore, co-culturing splenocytes from these cured mice with CT26 tumor cells led to specific T-cell activation, further confirming that ATAPL1 treatment induced long-term immune memory (figure 4K).

In conclusion, these results demonstrated that ATAPL1 treatment remodels the tumor immune microenvironment by inhibiting M-MDSC and Treg cells, enhancing the antitumor activity of CD8⁺ T cells and macrophages, and inducing long-term antitumor immune memory.

Evaluation of ATAPL1 in humanized colorectal cancer mouse models

Building on previous experiments conducted in mouse CRC models, we sought to verify whether ATAPL1 exerts antitumor effects in human CRC models. To this end, we constructed a bispecific antibody targeting both human PD-L1 and TNFR2, named hATAPL1. hATAPL1 could effectively bind to human PD-L1 and TNFR2, with corresponding EC₅₀ values of 214 ng/mL and 102 ng/mL, respectively (figure 5A). Additionally, hATAPL1 also bound effectively to human PD-L1 and TNFR2 on the surface of cells (figure 5B and C).

Figure 5. Evaluation of ATAPL1 in humanized colorectal cancer mouse models. (A) The ability of hATAPL1 to bind human PD-L1 or human TNFR2 was calculated through ELISA. (B and C) The binding of hATAPL1 to CHO^hPDL1 cells or CHO^hTNFR2 was analyzed by flow cytometry (n=3). (D) 6–8 weeks old HCT116 tumor-bearing NCG mice were intravenously injected with 5×10⁶ PBMCs 1 day before i.p. three doses of 200 µg Isotype, αTNFR2, αCD47 or ATAPL1 on days as indicated. Six mice per group. (E and F) Tumor growth was monitored in HCT116 tumor-bearing NCG mice. Tumor growth is shown as means±SEM. **p<0.01, ****p<0.0001. i.p., intraperitoneally; i.v., intravenous; PD-L1, programmed death-ligand 1; TNFR2, tumor necrosis factor receptor 2; EC50, half maximal effective concentration; MFI, mean fluorescence intensity; PBMC, peripheral blood mononuclear cells.

We then further evaluated the antitumor efficacy of hATAPL1 in a HCT116 tumor-bearing human PBMC-immune-reconstructed NCG mouse model (figure 5D). Compared with monotherapies with either human PD-L1 or TNFR2 antibodies, hATAPL1 significantly controlled tumor growth (figure 5E and F). These results indicated that hATAPL1 demonstrated significantly enhanced antitumor effects in humanized CRC mouse model and held strong potential for clinical development.

ATAPL1 significantly improves the antitumor effects of either FOLFOX or FOLFIRI chemotherapy in colorectal cancer

Chemotherapy regimens, including FOLFOX and FOLFIRI, are widely used in the treatment of patients with CRC.^{29 30} To explore the potential of combining ATAPL1 with these chemotherapy regimens, we first assessed the expression of TNFR2 and PD-L1 in CT26 tumor tissues after FOLFOX treatment. We found that FOLFOX treatment significantly upregulated the expression of both TNFR2 and PD-L1 in tumor tissues (figure 6A), which provided the rationale for combining ATAPL1 with FOLFOX.

Figure 6. ATAPL1 significantly improves the antitumor effects of either FOLFOX or FOLFIRI chemotherapy in colorectal cancer. (A) 6–8 weeks old Balb/c mice bearing subcutaneous CT26 colorectal tumor received i.p. FOLFOX on every 3 days, 48 hours after the third injection, tumor tissues were harvested and the TNFR2 and PD-L1 expression on the tumor tissue cells were analyzed by flow cytometry, five mice each group. (B) Experimental scheme of ATAPL1 and FOLFOX combination therapy in CT26 tumor-bearing mouse model. 6–8 weeks old subcutaneous CT26 tumor-bearing Balb/c mice received i.p. ATAPL1 or FOLFOX on days as indicated. Seven mice per group. (C and D) Tumor growth and (E) Kaplan-Meier survival curves were monitored. (F) 6–8 weeks old C57BL/6 mice bearing subcutaneous MC38 colorectal tumor received i.p. FOLFIRI on every 3 days, 48 hours after the third injection, tumor tissues were harvested and the TNFR2 and PD-L1 expression on the tumor tissue cells was analyzed by flow cytometry, five mice each group. (G) Experimental scheme of ATAPL1 and FOLFIRI combination therapy in MC38 tumor-bearing mouse model. 6–8 weeks old subcutaneous MC38 tumor-bearing C57BL/6 mice received i.p. ATAPL1 or FOLFIRI on days as indicated. Six mice per group. (H and I) Tumor growth and (J) Kaplan-Meier survival curves were monitored. Tumor growth is shown as means±SEM. The remaining data are shown as means±SD. ns, not significant, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. i.p., intraperitoneally; PD-L1, programmed death-ligand 1; TME, tumor microenvironment; TNFR2, tumor necrosis factor receptor 2; 5-FU, 5-Fluorouracil; MFI, mean fluorescence intensity.

We then evaluated the effects of combining FOLFOX with ATAPL1 in the CT26 tumor-bearing mouse model (figure 6B). Our results showed that the combination therapy significantly controlled tumor growth and extended the survival of the mice compared with either treatment alone, with 1/7 of the mice achieving complete remission (figure 6C–6E). In contrast, the combination of FOLFOX and αPD-L1 did not significantly improve tumor control over FOLFOX monotherapy, suggesting that the TNFR2 signaling pathway might play a crucial role in maintaining resistance to treatment in CRC (figure 6C–6E).

Similarly, we observed that FOLFIRI treatment led to the upregulation of TNFR2 and PD-L1 expression in MC38 tumor tissues (figure 6F). We therefore evaluated the potential of combining FOLFIRI with ATAPL1 in the MC38 tumor-bearing mouse model (figure 6G). The results demonstrated that the combination of FOLFIRI and ATAPL1 significantly inhibited tumor growth compared with either treatment alone (figure 6H and I). Notably, the FOLFIRI and ATAPL1 combination significantly extended the survival of the mice compared with FOLFIRI alone, with 1/6 of the mice achieving complete remission, whereas the FOLFIRI and αPD-L1 combination did not (figure 6J).

In conclusion, our results demonstrated that ATAPL1 enhanced antitumor efficacy of FOLFOX or FOLFIRI chemotherapy, and offered substantial potential for improving treatment outcomes in CRC models.

Discussion

While ICIs targeting PD-L1 have shown effectiveness in various cancers, their utility in CRC has been limited by immune suppression within the TME. Specifically, Tregs and MDSCs, which express TNFR2, contribute significantly to this immune suppression, undermining the efficacy of PD-L1 blockade. The results of our study demonstrate the promising therapeutic potential of ATAPL1, a bispecific antibody targeting both PD-L1 and TNFR2, effectively disrupts these suppressive immune cells, thereby promoting a more robust antitumor immune response in CRC mouse models.

Our data demonstrate that ATAPL1 exhibits superior antitumor efficacy compared with PD-L1 monotherapy in both microsatellite stable (MSS, CT26) and microsatellite instability (MSI, MC38) models,^{31 32} suggesting MSI-independent therapeutic benefits. Clinically, these findings hold particular significance as MSS tumors typically exhibit primary resistance to current checkpoint inhibitors due to their immunologically “cold” phenotype. The consistent responses observed position our PD-L1/TNFR2 dual-targeting approach as a promising strategy to overcome current MSI-based therapeutic limitations. Future studies should validate these conclusions in more additional MSS and MSI models.

Tregs and MDSCs promote tumor progression and enhance resistance to therapy by exerting immunosuppressive effects in the TME, inhibiting the function and activity of effector T cells, and facilitating immune evasion.^{33 34} Our findings demonstrate that the dual blockade of PD-L1 and TNFR2 with ATAPL1 effectively reduces the populations of these immune suppressive cells, and enhances the activation and proliferation of CD8⁺ T cells, a critical component of antitumor immunity. Surprisingly, ATAPL1 markedly promotes macrophage phagocytosis, contributing to tumor clearance and antigen presentation via upregulated MHC-II. These immune alterations suggest that ATAPL1 works by both inhibiting suppressive immune mechanisms and activating effector cells, thereby improving the immune response against the tumor. This dual blockade approach and its underlying mechanism may have broad implications for overcoming immune therapy resistance in various tumor types. It highlights the importance of analyzing immune suppressive cell populations within tumors as predictive biomarkers for immunotherapy efficacy, as well as a potential strategy for reversing immune resistance in patients.

ATAPL1, as a single bispecific antibody engineered with “knob-into-hole” technology, offers several distinct advantages over the combination therapy of αPD-L1 and αTNFR2. Unlike combination therapy, which suffers from mismatched drug clearance rates, bispecific antibody ensures synchronized target engagement with a single optimized pharmacokinetics profile, enabling less frequent dosing and improved patient compliance.^{35 36} As a single biologic entity, it eliminates the need for separate good manufacturing practice (GMP) production processes required for dual-antibody therapies, substantially reducing manufacturing complexity and costs. Importantly, ATAPL1’s targeting of co-expressed PD-L1 and TNFR2 in tumor tissues enhances tumor-selective accumulation through avidity effects, which not only lowers the effective therapeutic dose but also minimizes systemic toxicity by reducing off-target exposure, representing a comprehensive improvement over conventional combination immunotherapy approach.

The off-target effects of systemic PD-L1 delivery can lead to irAEs, primarily due to the immune system attacking non-tumor tissues during treatment.^{37 38} PD-L1 inhibitors enhance T-cell activity, which may trigger immune responses not only against tumor cells but also against normal tissues, such as the liver, gastrointestinal tract, or endocrine system.39,41 These off-target effects complicate the management of irAEs and may lead to treatment interruptions or delays, ultimately affecting the patient’s prognosis and treatment safety.3842,44 Since TNFR2 is primarily expressed on immune cells and is significantly upregulated in tumor tissues compared with peripheral tissues, the dual blockade of PD-L1 and TNFR2 with ATAPL1 offers a unique advantage over PD-L1 monotherapy. ATAPL1 can accumulate more selectively within the TME, enhancing the effectiveness of PD-L1 inhibition specifically in the tumor tissue. This not only facilitates the action of the PD-L1 antibody within the tumor but also helps to reduce the peripheral levels of PD-L1 antibody. ATAPL1 treatment was well-tolerated in vivo, showing no significant organ toxicity or upregulation of inflammatory cytokines. This favorable safety profile, combined with its robust therapeutic efficacy, underscores ATAPL1’s potential as a promising clinical candidate for CRC. Our approach—enhancing drug accumulation within the tumor while lowering peripheral levels—offers distinct advantages over direct combination therapies. In this sense, ATAPL1’s dual mechanism of action serves as a “two birds with one stone” solution, optimizing both efficacy and safety.

TAMs play an important role in tumor progression and treatment resistance. TAMs promote tumor immune escape by secreting immunosuppressive factors and inhibiting the function of effector T cells, thereby reducing antitumor immune responses.45,47 During treatment, the immunosuppressive effect of TAMs leads to tumor resistance to chemotherapy and immunotherapy, significantly affecting treatment outcomes.^{48 49} However, as macrophages, TAMs also possess certain antitumor properties, such as phagocytosing tumor cells and secreting pro-inflammatory factors to exert some antitumor effects.^{50 51} Our results demonstrate that the therapeutic efficacy of ATAPL1 relies on the functional cooperation of both CD8⁺ T cells and TAMs. This not only emphasizes the importance of CD8⁺ T cells in mediating antitumor immunity but also reveals the pivotal role of TAMs in the tumor response. Our findings suggest that the direct depletion of TAMs within the TME does not necessarily enhance the effectiveness of immunotherapy. Given that TAMs in the TME are often functionally impaired, unable to effectively phagocytize tumor cells, present antigens, or activate T cells, we propose that reprogramming the functional activity of TAMs is more advantageous than simply eliminating them. ATAPL1 treatment appears to restore TAM function by promoting a shift from the immunosuppressive M2 phenotype to the pro-inflammatory M1 phenotype and enhancing MHC-II expression. This reprogramming of TAMs contributes to the overall antitumor response, offering valuable insights for therapeutic strategies that target TAMs—not through depletion, but by functional reprogramming.

Chemotherapy drug resistance is a major challenge in cancer treatment. Tumor cells evade the effects of chemotherapy through various mechanisms, including the upregulation of inhibitory molecules that suppress antitumor immune responses, further enhancing resistance.⁵² As chemotherapy is used repeatedly, tumor cells accumulate more resistance characteristics, ultimately leading to treatment failure.^{53 54} Therefore, overcoming chemotherapy drug resistance is key to improving cancer treatment outcomes. In addition to direct antitumor effects, ATAPL1 was able to overcome chemotherapy resistance when combined with conventional regimens such as FOLFOX and FOLFIRI. Chemotherapy treatments lead to the upregulation of immune checkpoint molecules like PD-L1 and TNFR2, which act as “brakes” on the immune system, and hinder the immune response against the tumor. These molecules contribute to immune resistance, and our results indicate that ATAPL1 enhances the efficacy of these chemotherapies by targeting both PD-L1 and TNFR2 pathways, effectively disrupting immune suppression within the TME. One challenge in targeting TME homeostasis is that disrupting it can trigger upregulation of compensatory immune suppressive signals, leading to both innate and acquired resistance. PD-L1 and TNFR2 are key components of this feedback loop, and blocking them is essential for overcoming the immune suppression induced by chemotherapy. Therefore, inhibiting these “brake” signals allows for better antitumor outcomes with FOLFOX and FOLFIRI, maximizing their therapeutic potential. Our findings not only provide a promising solution to chemotherapy resistance but also offer valuable insights for the design of future clinical trials targeting immune checkpoints.

Despite the promising results showing the potential of ATAPL1 in overcoming immune resistance and enhancing antitumor immunity in CRC, there are several limitations in our study that warrant further exploration. First, the use of mouse models, including CT26 and MC38, while valuable, may not fully replicate the complex immune microenvironment of human CRC. Further validation in more clinically relevant models, such as humanized or patient-derived xenografts, is needed. Additionally, the exact mechanisms by which ATAPL1 modulates immune cell populations and CAFs, require more detailed investigation. Although ATAPL1 demonstrated remarkable antitumor effects in mice, its clinical translation poses challenges, including the production, stability, and potential immunogenicity of the bispecific antibody. Furthermore, variability in patient responses could be an issue, and identifying biomarkers for patient selection will be crucial. Lastly, while the combination of ATAPL1 with chemotherapy showed enhanced efficacy, the potential synergistic effects with other immunotherapies have yet to be fully explored. Future studies addressing these limitations will be essential for optimizing ATAPL1’s therapeutic potential in clinical settings.

In conclusion, ATAPL1 represents a novel and effective approach to overcoming immune resistance in CRC. By targeting both PD-L1 and TNFR2, ATAPL1 not only improves tumor targeting and antitumor immunity but also provides a promising strategy for overcoming both immune and chemotherapy resistance. These findings suggest that ATAPL1 has significant potential to improve treatment outcomes for patients with CRC, particularly those who do not respond well to current therapeutic strategies.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.