FTL008.16, a 5T4-Conditional 4-1BB bispecific antibody, potently enhances antitumor immunity via tumor-directed t-cell activation Authors: Shuang Liu, Ziqing Ren, Yongyan Tu, Pengyu Chen, Yinglu Zhang, Shengyan Zhao, Xiaofeng Chen, Han Deng, Shuyun Yang, Feng Qu, Ming Zhang, Lantu Gou, Qing Li, Ying Huang, Fanxin Ma, Jinliang Yang Published: 2026 ABSTRACT AbstractThe clinical development of 4-1BB agonists for cancer immunotherapy has been constrained by dose-limiting toxicities, particularly hepatotoxicity. To overcome this challenge, we developed FTL008.16, a novel bispecific antibody targeting 4-1BB (CD137) on T cells and the tumor-associated antigen 5T4. FTL008.16 is a human IgG1 antibody bearing an effector-deficient Fc region that specifically binds 4-1BB at an epitope overlapping with that of its natural ligand (4-1BBL), potentially mimicking physiological 4-1BBL-mediated co-stimulation. As demonstrated through in vitro reporter assays and primary cell-based systems, FTL008.16 elicits dose-dependent 4-1BB activation strictly contingent on 5T4 engagement, enhancing T-cell proliferation, cytokine secretion, and tumor cell lysis, thereby demonstrating potent antitumor efficacy in preclinical models. In vivo, FTL008.16 exhibited superior efficacy over a conventional anti-4-1BB monoclonal antibody, suppressing established tumors in both CT26 and MC38 models, with complete regression observed in CT26 tumors. In cynomolgus monkeys, FTL008.16 demonstrated an excellent safety profile, with no elevations in alanine aminotransferase (ALT) or aspartate aminotransferase (AST) observed even at 50 mg/kg. Importantly, this 5T4-restricted activation mechanism confines immune stimulation to the tumor microenvironment, thereby circumventing the systemic toxicity associated with conventional 4-1BB agonists. Collectively, these findings establish FTL008.16 as a conditionally active immunotherapeutic agent that successfully achieves both robust antitumor efficacy and a favorable safety profile.Keywords: 4-1BB (CD137), 5T4, bispecific antibody, cancer immunotherapy, T-cell activation Introduction In recent years, immunotherapy has emerged as a transformative treatment modality in oncology. Specifically, immune checkpoint inhibitors (ICIs) targeting cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), programmed cell death protein 1 (PD-1), and its ligand PD-L1 have demonstrated durable responses across various malignancies. However, their clinical utility remains limited by heterogeneous response rates and immune-related adverse events. 1-3 To overcome these limitations, drug development has shifted toward next-generation immunotherapies designed for tumor-localized activation, notably those targeting the T-cell co-stimulatory receptor 4-1BB (CD137). 4 Tumor antigen-dependent 4-1BB agonism selectively activates tumor-infiltrating lymphocytes (TILs) and natural killer (NK) cells within the tumor microenvironment (TME), enhancing their effector function and survival while avoiding systemic toxicity. 5 The rationale for targeting 4-1BB is underpinned by its potent biology and restricted activation profile. As a member of the tumor necrosis factor receptor superfamily (TNFRSF), 4-1BB is expressed on activated immune cells, with its expression markedly upregulated on T cells and NK cells following T-cell receptor (TCR) engagement. 6-10 Ligation of 4-1BB with its natural ligand (4-1BBL/TNFSF9) triggers potent downstream signaling that delivers a critical co-stimulatory signal for cytotoxic T cells. 11-14 Crucially, this signal can drive persistent functional modifications in tumor-specific T cells, and its activity is naturally confined to microenvironments where both antigen and ligand are present—such as tumors and draining lymph nodes. 15 , 16 This combination of potent efficacy and inherent spatial restriction makes 4-1BB an ideal candidate for safe and effective immunotherapy. 17 The therapeutic potential of 4-1BB co-stimulation in oncology has been extensively validated in preclinical studies. However, the clinical application of first-generation anti-4-1BB monoclonal antibodies (mAbs), such as urelumab and utomilumab, was limited by systemic toxicity. 18 , 19 Specifically, urelumab caused dose-limiting hepatotoxicity, while utomilumab showed limited efficacy as a monotherapy. 20 , 21 To overcome these challenges, next-generation bispecific antibodies (BsAbs) have been engineered to achieve tumor-restricted 4-1BB activation through co-engagement of tumor-associated antigens (TAAs). Preclinical evaluation of various TAA × 4-1BB bispecific constructs—including HER2 × 4-1BB, 22 CD30 × 4-1BB, 23 and B7H3 × 4-1BB—has confirmed enhanced tumor specificity with reduced off-target toxicity. 24 This strategy has advanced to Phase I/II clinical trials with Fc-silenced, conditionally active BsAbs targeting PD-L1 × 4-1BB and CLDN18.2 × 4-1BB, 25 , 26 where early results indicate promising safety profiles and antitumor activity in patients with solid tumors. The success of this approach critically depends on the selection of an optimal TAA. The oncofetal antigen 5T4 (trophoblast glycoprotein) represents a prime candidate, as it exhibits minimal expression in normal adult tissues but is frequently overexpressed in diverse carcinomas—including colorectal, gastric, lung, ovarian, and renal cell carcinomas. 27-29 High 5T4 expression is closely associated with tumor invasion, metastasis, and poor patient prognosis, further supporting its relevance as a therapeutic target. 30 , 31 Importantly, 5T4 is stably expressed on the tumor cell surface with limited heterogeneity in primary lesions, a key feature that ensures effective and consistent co-engagement for tumor-restricted 4-1BB activation. 32 This rationale is further strengthened by ongoing clinical programs against 5T4, including antibody-drug conjugates, CAR-T cell therapies, and bispecific antibodies, which have demonstrated favorable safety profiles and preliminary antitumor activity in early-phase trials. 33-35 In this study, we describe a novel BsAb designated FTL008.16, which consists of a tumor-targeting human IgG1 (hIgG1) arm and two 4-1BB-agonistic single-chain variable fragments (scFvs). 5T4 was selected as the targeting moiety based on its tumor-restricted overexpression profile. 36 , 37 Compared with the parental anti-4-1BB mAb FTL001 (developed in-house), FTL008.16 demonstrates enhanced T-cell co-stimulatory activity in vitro and improved tumor localization. Mechanistically, structural analyzes reveal that FTL008.16 binds human 4-1BB at an epitope overlapping with that of 4-1BBL, competitively blocking natural ligand engagement. Fc-silencing mutations were introduced to minimize FcγR interactions, ensuring strict 5T4-dependent crosslinking for conditional 4-1BB activation at sites of dual antigen co-expression. FTL008.16 potently induces 4-1BB signaling and amplifies T-cell responses in a 5T4-dependent manner both in vitro and in vivo, while mitigating systemic immune toxicity and hepatotoxicity. Collectively, our findings establish 5T4-directed 4-1BB activation as a promising therapeutic strategy for solid tumors that successfully decouples potent antitumor efficacy from systemic toxicity. Material and methods Cell lines and primary cells All cell lines were maintained following the supplier's recommended protocols. The HEK-Blue™ 4-1BB reporter cell line (InvivoGen), stably expressing human 4-1BB and an NF-κB-inducible secreted embryonic alkaline phosphatase (SEAP) reporter gene, was used to evaluate 4-1BB agonist activity. Stable CHO cell lines expressing human 5T4 were established via lentiviral transduction, with empty vector-transduced CHO cells serving as controls. For 4-1BB expression, 293T cells were transfected with human 4-1BB, followed by fluorescence-activated cell sorting (FACS) to isolate single-cell clones (designated 293Th4-1BB). Similarly, CT26 wild-type (WT) cells (CT26.WT, ATCC) and MC38-WT (MC38.WT, ATCC) cells were genetically engineered to express human 5T4, with single-cell clones (CT26h5T4 and MC38h5T4, respectively) isolated by FACS. Human tumor cell lines endogenously expressing 5T4 (DLD-1 and MDA-MB-231) were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). All human peripheral blood mononuclear cells (PBMCs) were purchased from Milecell Biotechnologies. These cells were isolated by the supplier from the peripheral blood of healthy donors using Ficoll-Paque (Cytiva, #17144002) according to the manufacturer's instructions, in compliance with relevant ethical standards and guidelines. Human T cells were isolated through immunomagnetic sorting (Miltenyi, #130-097-043) using CD3 antibody-conjugated magnetic beads. CD8+ T cells were purified from PBMCs by negative selection using a commercial CD8+ T Cell Isolation Kit (Miltenyi, #130-096-495). NK cells were similarly isolated through negative selection (NK Cell Isolation Kit, Miltenyi, #130-092-657) and then pre-stimulated with 10 ng/mL of human IL-2 (Acro, #GMP-L02H14) overnight at 37 °C. Expression and purification of recombinant antibodies Synthetic genes encoding the anti-h4-1BB monoclonal antibody (FTL001), anti-h5T4 monoclonal antibody (HB5-2-6), or anti-4-1BB × 5T4 bispecific antibody (BsAb, FTL008.16) were cloned into expression vectors. The constructed plasmids were linearized by restriction digestion and transfected into CHO cells (Horizon, #HD-BioP3) for stable cell line development. The stably transfected cells were cultured in Dynamis™ AGT™ medium (Gibco, #A26175-04) at 37 °C for 7 to 15 days in a CO₂ shaker incubator. Glucose (Sigma, #G7021-5KG) was added periodically during culture. When cell viability dropped below 70%, the cultures were terminated, and the recombinant antibodies were harvested. The antibodies were purified from culture supernatants using Protein A affinity chromatography (NMab Pro, Suzhou Nanomicro Technology). The column was equilibrated with 20 mM Tris-HCl containing 150 mM NaCl (pH 7.4), loaded with sample, and washed sequentially with 3 column volumes (CV) each of 20 mM citrate/1 M NaCl (pH 6.0) and 20 mM citrate (pH 5.0). Elution was performed with 20 mM citrate (pH 3.5), neutralized to pH 6.5 with 1 M Tris-HCl, followed by dialysis against PBS. Protein concentration was determined by spectrophotometry (Nano-400A, Allsheng), while antibody purity and molecular weight were verified by 4−12% Bis-Tris SDS-PAGE. Surface plasmon resonance (SPR) SPR was performed on a Biacore 8 K instrument at 25 °C using HBS-EP + as running buffer (Cytiva, #BR100669). His-tagged 5T4 and 4-1BB proteins were immobilized on an NTA sensor chip according to the manufacturer's instructions. Antibodies were serially diluted twofold and injected at 30 μL/min with 120 seconds association and 600-1200-second dissociation phases. The sensor chip surface was regenerated between cycles using 350 mM EDTA containing 0.5% SDS at a flow rate of 30 μL/min for 60 seconds. Kinetic data were analyzed using a 1:1 binding model with the Biacore evaluation software. Binding ELISA High-binding 96-well plates were coated overnight at 4 °C with 1 µg/mL recombinant human 4-1BB or 5T4 protein (100 µL/well) in carbonate coating buffer. After PBST washing, nonspecific binding sites were blocked with 1% BSA (Solarbio, #A8020) in PBS. Serially diluted antibodies (initial concentrations: 5 μg/mL for h4-1BB, 2 μg/mL for h5T4; fivefold dilutions) were incubated for 1 hour at 37 °C. Antigen-antibody interactions were detected using HRP-conjugated anti-human IgG Fc antibody (1:5,000 dilution, Jackson, #109-035-098) followed by TMB substrate development. The enzymatic reaction was terminated with 0.5 M H₂SO₄, and absorbance was quantified at 450 nm using a microplate reader (BioTek Synergy LX). Blocking ELISA Human 4-1BBL was immobilized onto 96-well plates as described above. Antibodies were serially diluted, mixed with a fixed concentration of h4-1BB, and then added to the wells, followed by incubation at 37 °C for 1 hour. After washing, an HRP-conjugated goat anti-His antibody was added, and TMB peroxidase substrate was used for signal detection. The percentage of antibody remaining bound to h4-1BB was plotted, and the half-maximal inhibitory concentration (IC50) values were calculated using GraphPad Prism software. Antibody-dependent cellular cytotoxicity (ADCC) effector function assay ADCC reporter activity was assessed using Jurkat-Lucia™ NFAT-CD16 cells (InvivoGen, #jktl-nfat-cd16) as effector cells and 293Th4-1BB and CT26h5T4 cells as target cells at an effector-to-target (E:T) ratio of 2:1. After co-culture for 6 hours at 37 °C, luciferase activity was measured using Quanti-LUC (InvivoGen, #rep-qlc2). Rituximab (Rituxan®, Genentech) and human IgG1 (hIgG1) isotype control were included as positive and negative controls, respectively. Flow cytometric binding assay Flow cytometry was used to analyze the binding ability of the HEK-Blue™ 4-1BB reporter cell line and 5T4-expressing cell lines (MDA-MB-231, DLD-1, CT26h5T4, MC38h5T4). Diluted antibodies were added to the cell suspension at a density of 1 × 10⁵ cells/well and incubated for 40 minutes at 4 °C. Subsequently, the cells were washed twice with PBS to remove unbound antibodies, followed by incubation with FITC-conjugated anti-human IgG (H + L) secondary antibody (ZSGB-BIO, #ZF-0308) at 4 °C for 40 minutes. Finally, fluorescence was measured on a flow cytometer (Beckman CytoFLEX), and mean fluorescence intensity (MFI) was analyzed using the manufacturer's software. Reporter cell line assay 4-1BB pathway activation was assessed using the HEK-Blue™ NF-κB reporter cell line, which overexpresses human 4-1BB and an NF-κB-inducible secreted embryonic alkaline phosphatase (SEAP) gene. Reporter cells were co-cultured with human 5T4-expressing tumor cells in the presence of serially diluted antibodies for 48 hours at 37 °C. NF-κB-mediated SEAP production was quantified by measuring the absorbance after adding QUANTI-Blue™ Solution (InvivoGen, #QBLA-42-14), using a BioTek Synergy LX microplate reader. Western blotting Human 293Th4-1BB cells were stimulated with FTL008.16, FTL001, or control IgG1 at 37 °C for 30 minutes and lysed in NP-40 buffer containing protease and phosphatase inhibitors. Proteins were resolved on 4–12% Bis-Tris gels, transferred to PVDF membranes, and probed overnight at 4 °C with antibodies against phospho-TRAF2 (Selleck, #F1348; 1:1,000), TRAF2 (Selleck, #F2419; 1:1,000), and GAPDH (Proteintech, #60004-1-Ig; 1:5,000). After incubation with HRP-conjugated goat anti-mouse (Jackson, #115-035-003) and goat anti-rabbit (Jackson, #111-035-003) secondary antibodies (1:5,000), blots were visualized using ECL Plus substrate and imaged on an Azure imaging system (Azure Biosystems). Band intensity was quantified using Quantity One software (Bio-Rad). 4-1BB expression profiling and immune activation in vitro 4-1BB Expression Analysis. Primary human PBMCs were co-cultured for 72 hours with CT26h5T4 or CT26-WT cells at a 3:1 ratio in RPMI 1640/10% FBS containing soluble anti-CD3 antibody (BioLegend, #317326) at concentrations of 0.01 or 0.1 µg/mL. 4-1BB expression on viable CD4+ T cells, CD8+ T cells, and CD56+ NK cells was analyzed by flow cytometry using the following antibodies: anti-CD4-PE/Cy7 (BioLegend, #300512), anti-CD8-BV421 (BD, #562428), anti-CD56-PerCP (BioLegend, #362526), and anti-4-1BB-APC (BD, #550890). To assess binding, healthy donor PBMCs were activated for 72 hours with plate-bound anti-CD3 (BioLegend, #317326) and anti-CD28 (BioLegend, #302934) antibodies (2 μg/mL each). Activated cells were incubated with FTL008.16 (0.25–2,500 nM) and stained with anti-CD3-FITC (BD, #555329), anti-CD4-PerCP (Invitrogen, #45-0049-42), anti-CD8-APC (BD, #555369), and a PE-conjugated anti-human IgG antibody (Absin, #bs-0297M-PE) to detect FTL008.16 binding. The binding level was quantified based on the MFI in CD3+CD4+ and CD3+CD8+ T-cell populations. For proliferation assays, CHOh5T4 cells were treated with 50 µg/mL mitomycin C (MCE, #HY-13316) for 1 hour at 37 °C to prevent overgrowth, then seeded at 3 × 10⁴ cells/well. Human PBMCs were labeled with CFSE (MCE, #HY-D0938) according to the manufacturer’s instructions. Labeled PBMCs (1.2 × 10⁵ cells/well) were stimulated with 10 ng/mL soluble anti-CD3 antibody in the presence of serial dilutions of FTL008.16 or hIgG1 isotype control. After 96 hours of incubation at 37 °C, cells were harvested, blocked with Fc receptor blocking solution (BioLegend, #422302), and stained with 7-AAD (BD, #559925) for viability, as well as anti-CD3-PE (BD, #555340) and anti-CD8-APC (BD, #555369) for surface markers. CD8+ T-cell proliferation was analyzed and quantified by flow cytometry. T-cell activation and IFN-γ secretion. For the functional assay, CHOh5T4 cells were pre-treated with mitomycin C for 1 hour at 37 °C one day prior to co-culture. Pre-activated human T cells were co-cultured with CHOh5T4 cells at a 5:1 E:T ratio in the presence of serially diluted FTL008.16 or control antibodies (starting at 5 µg/mL). After 72 hours, supernatant concentrations of IFN-γ were measured using a commercial ELISA kit (Dakewe Biotech, #1110003). Cytokine release assay PBMCs isolated from healthy donors were seeded in 96-well flat-bottom plates at 2 × 10⁵ cells/well in RPMI 1640/10% FBS. For solid-phase stimulation, plates were pre-coated with protein and washed before PBMC addition. For liquid-phase stimulation, antibodies were directly added to wells at 0.1–10 µg/mL. Cells were incubated for 24 or 48 hours, with staphylococcal enterotoxin B (SEB; 100 ng/mL) as a positive control. Cytokine levels (IFN-γ, TNF-α, IL-2, IL-4, IL-6, IL-10, IL-17A) were measured using a commercial Cytometric Bead Array kit (BD, #560484) on a CytoFlex system according to the manufacturer's instructions. PBMC-mediated target cell lysis assay For the PBMC-mediated cytotoxicity assay, anti-human CD3 antibody was immobilized at 5 µg/mL on assay plates overnight at 4 °C. After removing unbound antibody, human PBMCs (2.5 × 10⁴ cells/well) were co-cultured with 5T4-positive DLD-1 target cells or 5T4-negative HEK 293 T target cells (5 × 10³ cells/well) for 72 hours at 37 °C in a 5% CO₂ humidified incubator. Cell viability was determined using the CellTiter-Glo Luminescent Cell Viability Assay (Promega, #G7570) according to the manufacturer’s instructions. Luminescence was measured using a BioTek Synergy LX microplate reader, and cell viability was normalized to target-cell-only control wells. Mice and tumor models C57BL/6-hCD137 mice were obtained from Biocytogen, and BALB/c-hCD137 mice were from GemPharmatech. All experiments used mice aged 8–12 weeks. C57BL/6-hCD137 mice were injected subcutaneously (SC) in the upper flank with 5 × 10⁵ MC38h5T4 cells in 100 μL Dulbecco’s phosphate-buffered saline (DPBS; Gibco, #C14190500BT). When the tumor volume reached 100 ± 20 mm³, 40 mice were stratified into five groups (n = 8/group, female mice) based on body weight and tumor volume: vehicle control (G1), FTL008.16 dose groups (0.68, 1.36 mg/kg for G2 and G3, respectively), FTL001 (1 mg/kg, G4), and HB5-2-6 (anti-5T4 mAb, 1 mg/kg, G5). Treatments were administered via tail vein injection twice weekly for 8 doses (days 0–25). BALB/c-hCD137 mice were injected SC in the upper flank with 5 × 10⁶ CT26h5T4 cells in 200 μL DPBS. When the tumor volume reached 100 ± 20 mm³, 50 mice were stratified into five groups (n = 10/group; 5 males + 5 females) based on body weight and tumor volume. Five groups were evaluated: vehicle (G1), FTL008.16 (0.68, 1.36, and 2.72 mg/kg, G2–G4) and FTL001 (anti-4-1BB mAb, 1 mg/kg, G5). Treatments were administered via tail vein injection twice weekly for 8 doses (days 0–24). At the experimental endpoint, mice were euthanized by 95% CO₂ inhalation. Tumors were subsequently harvested, and their volume was determined using digital calipers with the formula: volume (mm³) = ½ × length × width². The tumors were then weighed to record the final mass. Body weight was monitored regularly throughout the study to assess systemic health and treatment-related toxicity. The predefined humane endpoints for survival analysis were a tumor volume exceeding 2,500 mm³ or body weight loss exceeding 20%. Data are presented as mean ± SEM. Analyzes were performed using SPSS 24.0, and graphs were generated with GraphPad Prism 8.0. In vivo imaging of FTL008.16 biodistribution To assess in vivo drug distribution, fluorescent dye IVIS 800 was conjugated to FTL008.16, hIgG1 isotype control, and FTL001 via amine coupling according to the manufacturer’s instructions. Unconjugated dye was removed by purification. These labeled proteins were injected via the tail vein into BALB/c-hCD137 mice bearing CT26h5T4 tumors of approximately 200 mm³ at a volume of 100 µL and a dose of 1.5 mg/kg. Fluorescent signals were detected at 4, 8, 22, and 48 hours post-injection using an IVIS Spectrum in vivo imaging system (PerkinElmer) with excitation and emission wavelengths of 740 nm and 845 nm, respectively. Toxicity study in cynomolgus monkeys Cynomolgus monkeys (five males and five females per group; n = 10/group) received weekly intravenous injections of FTL008.16 at 0 (vehicle control), 5, 15, or 50 mg/kg for five weeks (days 0, 8, 15, 22, and 29). Interim necropsy was performed on day 36 (three males and three females per group), and terminal necropsy followed a 6-week recovery period on day 78 (remaining animals). A comprehensive set of safety parameters was evaluated. This included mortality and clinical signs (via detailed twice-daily observations and daily cage-side monitoring); body weight (measured weekly); food consumption (quantified daily); hematology (including red blood cell count, white blood cell count and differential, platelet count, hemoglobin, and hematocrit); serum biochemistry (including ALT, AST, ALP, total protein, albumin, creatinine, and BUN); and histopathology (comprehensive microscopic examination of tissues and organs with severity grading). The study was conducted in compliance with Good Laboratory Practice (GLP) guidelines at Xishan Zhongke (Suzhou, China). Terminal necropsy was conducted under deep anesthesia induced by Tiletamine-Zolazepam (Zoletil® 50, 5 mg/kg, IM), followed by euthanasia through femoral artery exsanguination. The entire process adhered to standard veterinary practices and animal welfare guidelines. Statistical analyzes were performed using Provantis 10.5.0.3. Continuous data (e.g., body weight, clinical pathology, organ weights) were analyzed separately by sex and are presented as mean ± SD. Statistical analysis All data are presented as mean ± SEM or mean ± SD as indicated in the figure legends. For comparisons among multiple treatment groups and a single control group, one-way ANOVA with Dunnett's multiple comparison test was used. For comparisons between two specific groups, unpaired two-tailed Student's t-test was applied. p < 0.05 was considered statistically significant. All statistical analyzes were performed using GraphPad Prism 8.0 (GraphPad Software, San Diego, CA, USA). All experiments were repeated at least three times with independent biological replicates, and representative results are shown. ARRIVE guidelines This study was conducted and reported in accordance with the ARRIVE guidelines 2.0. Results Development of a novel tumor-targeting 4-1BB × 5T4 bispecific antibody (FTL008.16) We first generated and characterized FTL001, a high-affinity agonistic monoclonal antibody (mAb) targeting human 4-1BB. Structural modeling revealed that FTL001 recognizes a distinct epitope on 4-1BB compared with urelumab, utomilumab, and 4-1BBL (Figure 1a). In a competitive ELISA against human 4-1BBL (Figure 1b), FTL001 showed strong blocking activity with an IC50 of 1.015 μg/mL, superior to utomilumab (IC50 = 6.160 μg/mL) and comparable to FTL008.16 (IC50 = 1.249 μg/mL). Urelumab exhibited only minimal blocking, and the hIgG1 isotype control showed none, confirming that FTL001 efficiently competes with 4-1BBL for 4-1BB binding. A novel tumor-targeting anti-4-1BB × 5T4 bispecific antibody (FTL008.16) exhibits unique dual-target binding properties. (a) Structural interactions of 4-1BB with its ligand 4-1BBL (bottom left) or agonistic antibodies (urelumab, top left; utomilumab, top right; FTL001, bottom right) are illustrated. The four cysteine-rich domains (CRDs) of 4-1BB are color-coded: CRD1 (tan), CRD2 (cornflower blue), CRD3 (medium slate blue), and CRD4 (peru). (b) Competitive binding of FTL001, FTL008.16, utomilumab, urelumab, and hIgG1 isotype control to h4-1BB pre-bound with 4-1BBL was assessed by ELISA. IC50 values are indicated; data are mean ± SD (n = 3). (c) Schematic design of FTL008.16: an anti-5T4 F(ab’)₂ fragment for tumor targeting is fused to a mutated hIgG1 Fc region linked to an anti-4-1BB scFv, enabling conditional 4-1BB engagement. (d) Antibody-dependent cellular cytotoxicity (ADCC) induced by FTL008.16 was assessed using an ADCC reporter bioassay. Rituximab-treated Raji cells served as the positive control, and 293Th4-1BB cells incubated with hIgG1 served as the negative control. RLU, relative light units. Data are mean ± SEM (n = 3). (e) Surface plasmon resonance confirmed the binding of FTL008.16 to the targets 4-1BB (left) and 5T4 (right) at the indicated concentrations. (f) ELISA demonstrated that FTL008.16 simultaneously bound 5T4 (left) and 4-1BB (right) at the indicated antibody concentrations. We next constructed the 4-1BB × 5T4 BsAb FTL008.16 by fusing the anti-4-1BB scFv derived from FTL001 to the C-terminus of a high-affinity anti-5T4 IgG1 (HB5-2-6 clone; KD = 18.71 pM) 38 and introducing L234A, L235A, and P329G (LALA-PG) mutation to abolish FcγR binding (Figure 1c). Since this substitution silences FcγR-mediated effector functions, we assessed its ADCC-inducing potential. Consistent with the silenced Fc function, FTL008.16 did not induce ADCC against 5T4-expressing CT26h5T4 cells or 293Th4-1BB cells, whereas the positive control rituximab triggered robust cytotoxicity (Figure 1d). SPR analysis confirmed that FTL008.16 bound human 4-1BB with high affinity (KD = 3.90 nM) and human 5T4 with even higher affinity (KD = 24.2 pM) (Figure 1e). A dual-binding ELISA further validated its bispecific nature by demonstrating simultaneous binding to both h5T4 and h4-1BB (Figure 1f). FTL008.16 binds to 5T4 and 4-1BB and induces potent, antigen-dependent 4-1BB agonistic activity To characterize the binding profile of FTL008.16, we confirmed its specific binding to both 5T4-positive tumor cells and 4-1BB-expressing cells (Supplementary Figure S1). We next assessed FTL008.16-mediated 4-1BB signaling using an NF-κB reporter system co-cultured with 5T4-positive tumor lines, including engineered high-5T4 models (CT26h5T4, MC38h5T4) and endogenous 5T4-expressing lines (MDA-MB-231, DLD-1). FTL008.16 exhibited potent, 5T4-dependent agonistic activity, with EC50 values of 6.65–8.46 ng/mL in high-5T4 cells, 12.86 ng/mL in MDA-MB-231 cells, and 23.10 ng/mL in DLD-1 cells, correlating with cellular 5T4 expression levels. In contrast, FTL001 showed weak, 5T4-independent activity across all lines (Figure 2a). Using isogenic MC38 cells, FTL008.16 exhibited strict 5T4 dependence, with approximately 50-fold lower potency in 5T4-negative versus 5T4-positive cells (EC50 = 12,565 vs. 236.1 ng/mL), whereas FTL001 remained weakly active irrespective of 5T4 expression (Figure 2b). Consistent with a tumor antigen-dependent crosslinking mechanism, FTL008.16-driven signaling was restricted to 5T4-positive tumor tissues and absent in 5T4-negative normal tissues (Figure 2c). FTL008.16 exhibits tumor-restricted 4-1BB agonistic signaling in human cells invitro. (a) Concentration-dependent 4-1BB co-stimulatory activity of FTL008.16 was assessed in HEK-Blue h4-1BB reporter cells co-cultured with 5T4-positive tumor cell lines (CT26h5T4, MC38h5T4, MDA-MB-231, or DLD-1), demonstrating 5T4-restricted agonism. (b) Concentration-dependent 4-1BB co-stimulatory activity of FTL008.16 and FTL001 was assessed in HEK-Blue h4-1BB reporter cells co-cultured with 5T4-positive MC38h5T4 or 5T4-negative MC38-WT cells. FTL008.16 exhibited 5T4-restricted agonism, whereas FTL001 showed comparably weak activity across both cell lines. (c) Schematic working model of tumor-restricted agonism by FTL008.16. In normal tissues (5T4-), binding solely to 4-1BB results in minimal receptor clustering. In contrast, within the tumor microenvironment (5T4+), simultaneous engagement of 4-1BB on immune cells and 5T4 on tumor cells drives robust 4-1BB clustering and agonism. (d) Western blot analysis of TRAF2 phosphorylation in 293Th4-1BB cells treated with FTL008.16, FTL001, or hIgG1 isotype control for 30 minutes. pTRAF2 and total TRAF2 levels were normalized to GAPDH. *p < 0.05, **p < 0.01 vs. hIgG1 control. Members of the TNFRSF, including 4-1BB, recruit TRAF adapter proteins to initiate intracellular signaling. 39 TRAF2, a key E3 ubiquitin ligase, links 4-1BB to the NF-κB pathway via K63-linked ubiquitination. To determine whether FTL008.16 activates this signaling axis, we assessed TRAF2 phosphorylation at Ser11, which is required for its E3 ligase activity. In 293Th4-1BB cells, FTL008.16 induced a dose-dependent increase in phospho-TRAF2 (pTRAF2) levels without affecting total TRAF2 or GAPDH (Figure 2d). At 50 nM and 5 nM, FTL008.16 significantly enhanced pTRAF2 levels compared with hIgG1 control, an effect comparable to that of FTL001. These findings demonstrate that FTL008.16 specifically engages 4-1BB to activate downstream TRAF2 signaling. FTL008.16 enhances 5T4-dependent T-cell responses and cytotoxicity in vitro Upon stimulation and co-culture with tumor cells, 4-1BB expression was most prominently upregulated on CD8+ T cells (Figure 3a), identifying them as the primary target population for 4-1BB-directed therapy. Reflecting this expression pattern, FTL008.16 showed concentration-dependent binding to activated T cells, with the highest level of binding observed on CD8+ T cells (Figure 3b). FTL008.16 induced a dose-dependent expansion of CD8+ T cells in CD3-stimulated PBMCs co-cultured with 5T4-expressing CHO cells, as measured by absolute cell counts at 96 hours, whereas the isotype control hIgG1 had no effect (Figure 3c). Functionally, when co-cultured with 5T4-expressing CHO cells, FTL008.16 induced potent, dose-dependent IFN-γ production from both PBMCs and purified T cells (Figure 3d). Critically, this IFN-γ response was strictly dependent on 5T4 expression, as robust cytokine production was elicited by CT26h5T4 cells but not by their 5T4-negative counterparts (Figure 3e). FTL008.16 exhibits 5T4-dependent immune cell activation and tumor cell killing in PBMC co-cultures. (a) PBMCs were stimulated with anti-CD3 (0, 10, or 100 ng/mL), and 4-1BB expression on CD4+ T, CD8+ T, and NK cells was analyzed in co-culture with CT26h5T4 or CT26-WT cells. (b) Concentration-dependent binding of FTL008.16 to anti-CD3/CD28-activated T-cell subsets (CD3+, CD4+, CD8+) was assessed by flow cytometry. (c) CFSE-labeled PBMCs were stimulated with anti-CD3 and serial dilutions of FTL008.16 or hIgG1 isotype control in the presence of CHOh5T4 cells. CD8+ T-cell counts were assessed by flow cytometry at 96 hours. (d) IFN-γ release from PBMCs and purified T cells was measured after co-culture with 5T4-expressing tumor cells and treatment with FTL008.16 or FTL001 control. (e) Purified T cells and mitomycin C-treated CT26-WT or CT26h5T4 tumor cells were incubated with serial dilutions of FTL008.16. IFN-γ secretion in supernatants was measured by ELISA. (f) FTL008.16 does not induce cytokine release from PBMCs of healthy donors (n = 3), as assessed by cytometric bead array (CBA). (g) PBMC-mediated cytotoxicity of FTL001 and FTL008.16 against 5T4-positive (DLD-1) and 5T4-negative (HEK 293T) cells. **p < 0.01, ***p < 0.001 vs. hIgG1 control. To assess nonspecific inflammatory cytokine release, human PBMCs from healthy donors were treated with FTL008.16. Analysis of supernatants collected at 24 and 48 hours revealed that FTL008.16 did not induce any nonspecific production of pro-inflammatory cytokines at any concentration tested (Figure 3f). In contrast, both FTL001 and FTL008.16 significantly enhanced PBMC-mediated cytotoxicity against 5T4-positive DLD-1 cells in a dose-dependent manner compared with the hIgG1 control (FTL001: p < 0.01 at both 10 μg/mL and 0.1 μg/mL; FTL008.16: p < 0.001 at 10 μg/mL, p < 0.01 at 0.1 μg/mL). Importantly, neither antibody induced significant cytotoxicity in 5T4-negative HEK 293 T cells (Figure 3g). Together, these findings underscore the tumor-restricted activity of FTL008.16, demonstrating its ability to elicit potent antitumor immunity without triggering systemic inflammatory responses. Antitumor efficacy of FTL008.16 in humanized mouse models To evaluate the antitumor efficacy of FTL008.16, MC38h5T4 tumors were established in C57BL/6-hCD137 knock-in (KI) mice (Figure 4a). Treatment with FTL008.16 at 0.68 mg/kg (G2) and 1.36 mg/kg (G3) significantly suppressed tumor growth compared with the vehicle control (G1; p < 0.01 and p < 0.0001, respectively; Figure 4c). The reference anti-4-1BB mAb FTL001 (1 mg/kg, G4) also showed significant efficacy (p < 0.01), whereas the anti-5T4 mAb HB5-2-6 (1 mg/kg, G5) did not. Notably, at equimolar doses, FTL008.16 (1.36 mg/kg) demonstrated statistically superior antitumor efficacy compared with FTL001 (1 mg/kg) at the experimental endpoint (p < 0.05), and also exhibited significantly enhanced antitumor activity relative to HB5-2-6 (1 mg/kg; p < 0.01). Consistent with tumor volume data, tumor weight was significantly reduced in G2–G4 (p < 0.01; Figure 4b). All treatments were well tolerated with no significant body weight loss observed (Figure 4c). FTL008.16 exhibits antitumor activity in hCD137-humanized mice. (a) C57BL/6-hCD137 knock-in (KI) mice (n = 8 per group) were subcutaneously inoculated with MC38h5T4 tumor cells. Seven days post-inoculation, when the mean tumor volume reached approximately 100 mm³, tumor-bearing mice were randomly allocated to groups and dosed with the indicated antibodies twice weekly (8 doses total). (b) Tumor weights of MC38h5T4 tumors in C57BL/6-hCD137 KI mice at the end of the study. (c) Tumor growth curves (left) and body weight changes (right) of MC38h5T4 tumor-bearing mice described in (a). Data are presented as mean ± SEM. (d) BALB/c-hCD137 KI mice (n = 10 per group) were subcutaneously inoculated with CT26h5T4 tumor cells. Seven days post-inoculation, when the mean tumor volume reached approximately 100 mm³, tumor-bearing mice were randomly allocated to groups and dosed with the indicated antibodies twice weekly (8 doses total). (e) Tumor weights of CT26h5T4 tumors in BALB/c-hCD137 KI mice at the end of the study. (f) Tumor growth curves (left) and body weight changes (right) of CT26h5T4 tumor-bearing mice described in (d). Data are presented as mean ± SEM. Statistical significance was assessed by one-way ANOVA with Dunnett's multiple comparison test (b, c, e, f). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (g) In vivo biodistribution of FTL008.16 and control antibodies in BALB/c-hCD137 KI mice bearing CT26h5T4 tumors, assessed by fluorescence imaging. To further evaluate the antitumor activity of FTL008.16, CT26h5T4 tumors were established in BALB/c-hCD137 KI mice (Figure 4d). FTL008.16 treatment at 0.68, 1.36, and 2.72 mg/kg (G2–G4) dose-dependently inhibited tumor growth (p < 0.05, p < 0.001 vs. G1; Figure 4f), with the 2.72 mg/kg dose (G4) exhibiting the strongest effect. The anti-4-1BB mAb FTL001 (1 mg/kg, G5) also significantly suppressed tumor growth (p < 0.05). These findings were corroborated by significant reductions in tumor weight (G2–G5, p < 0.05, p < 0.01; Figure 4e). All treatments were well tolerated, with no significant body weight loss (Figure 4f). We next assessed the biodistribution of FTL008.16 in tumor-bearing BALB/c-hCD137 KI mice by in vivo imaging (Figure 4g). FTL008.16 exhibited markedly enhanced tumor-specific accumulation compared with the hIgG control and FTL001, with peak signal intensity observed at 48 hours. The intense and localized fluorescence of FTL008.16 at tumor sites contrasted with the broader systemic distribution of the control antibodies, demonstrating superior tumor targeting. This distinct biodistribution profile confirms that the superior antitumor efficacy of FTL008.16 is driven by its 5T4-dependent tumor localization. FTL008.16 is well tolerated and does not induce systemic cytokine release in cynomolgus monkeys The safety profile of FTL008.16 was evaluated in cynomolgus monkeys following repeated administration of three dose levels (5, 15, and 50 mg/kg; Figure 5a). The molecule demonstrated cross-reactivity with cynomolgus 4-1BB and 5T4 with affinities comparable to the human orthologs (Figure 5b, Supplementary Figure S2). FTL008.16 was well tolerated across all dose groups, with no treatment-related adverse effects observed in mortality, clinical observations, or physiological parameters. Critically, even at the highest dose of 50 mg/kg, there were no indications of hepatotoxicity, as evidenced by the absence of elevations in ALT/AST levels (Figure 5c) and no histopathological signs of immune-mediated liver inflammation (Figure 5d). Furthermore, no biologically relevant changes were noted in hematological parameters (e.g., peripheral blood lymphocyte counts; Supplementary Figure S3). Notably, serum cytokine levels (including IL-2, IL-6, IL-8, IL-10, IFN-γ, and TNF-α) were undetectable or remained below the lower limit of quantification (LLOQ) in all animals, corroborating the absence of systemic cytokine release. Safety and target engagement profile of FTL008.16 in cynomolgus monkeys. (a) Cynomolgus monkeys (n = 10 per group) received five weekly intravenous injections of FTL008.16 at 0 (vehicle control), 5, 15, or 50 mg/kg. Initial sacrifice was performed on day 36, and terminal sacrifice on day 78 after a recovery period. (b) Binding of FTL008.16 to human (green) and cynomolgus monkey (red) 4-1BB (left) and 5T4 (right), measured by ELISA. EC50 values are indicated. (c) Serum clinical chemistry parameters, including ALT and AST, were measured throughout the study. Data are presented as mean ± SEM. (d) Representative histopathological images of major organs (liver, spleen, kidney, colon, and mesenteric lymph node) stained with hematoxylin and eosin (H&E). Scale bar, 50 µm. Discussion The clinical development of agonistic 4-1BB antibodies has been hampered by dose-limiting hepatotoxicity or suboptimal efficacy. 40 To overcome this challenge, we developed FTL008.16, a 4-1BB × 5T4 bispecific antibody designed to confine 4-1BB agonism to 5T4-positive tumors. FTL008.16 was engineered based on the principle that safe and potent TNFRSF agonism depends on precise control of antibody properties, including affinity, Fc function, and crosslinking. 41 , 42 Accordingly, we incorporated a 161-fold affinity bias for 5T4 over 4-1BB (KD = 24.2 pM vs. 3.90 nM)—a property known to enhance efficacy and limit immune escape—with an Fc-silencing LALA-PG mutation. 43 , 44 As a result, productive 4-1BB clustering is induced exclusively upon dual antigen engagement within 5T4-positive tumor microenvironments, thereby averting off-target toxicity. The distinct 4-1BB epitope targeted by FTL008.16, which partially overlaps with the 4-1BBL binding site, fundamentally differentiates it from clinical-stage predecessors. Unlike urelumab, a ligand non-blocker that binds a distal epitope, or utomilumab, which sterically hinders ligand binding, 45-47 FTL008.16 confers dual attributes by virtue of its unique binding mode: ligand-blocking capacity and a markedly improved safety profile. The translational value of this safety profile was evidenced by the complete absence of hepatotoxicity in cynomolgus monkeys even at 50 mg/kg (Figure 5c), demonstrating a wide therapeutic window. Furthermore, the tumor-selective mechanism was corroborated by the lack of nonspecific pro-inflammatory cytokine release from healthy human cells (Figure 3f). Taken together, these findings position FTL008.16 as a promising clinical candidate that effectively dissociates potent antitumor T-cell activation from systemic, on-target off-tumor toxicity, thereby addressing a key limitation of conventional 4-1BB agonizts. The functional consequence of this tumor-selective design is reflected in the strict antigen-dependent activity of FTL008.16. A key finding of this study is the correlation between FTL008.16 potency and target antigen density. Across cell lines with varying 5T4 expression levels—including engineered high-5T4 models (CT26h5T4, MC38h5T4) and tumor lines with endogenous moderate expression (MDA-MB-231, DLD-1)—FTL008.16 demonstrated potent agonistic activity that scaled with available 5T4 levels (Figure 2a). Notably, even in DLD-1 cells, which represent a clinically relevant model of moderate antigen expression, FTL008.16 exhibited a meaningful functional advantage over FTL001, with an approximately 3.2-fold improvement in EC50 (23.10 ng/mL vs. 74.59 ng/mL) and enhanced PBMC-mediated cytotoxicity at pharmacological concentrations (Figure 3g). This expression-guided profile is particularly relevant for heterogeneous solid tumors, suggesting that FTL008.16 may retain efficacy across 5T4-positive tumors while sparing antigen-negative tissues, a strategy known to mitigate 4-1BB-mediated toxicity. 48 The strict 5T4 dependence of FTL008.16 was further corroborated by the approximately 50-fold reduction in potency observed in 5T4-negative isogenic cells compared with their 5T4-positive counterparts (EC50 = 12,565 vs. 236.1 ng/mL; Figure 2b). This antigen-dependent selectivity establishes a broad safety margin that distinguishes FTL008.16 from conventional non-targeted 4-1BB agonizts. Functionally, FTL008.16 preferentially engages CD8+ T cells and potently induces IFN-γ upon tumor recognition. This mechanism aligns with recent findings demonstrating that tumor-targeted 4-1BB agonism can effectively enhance tumor-infiltrating T-cell responses. 49 Collectively, this tumor-selective activation with antigen density-guided potency positions FTL008.16 as a promising next-generation 4-1BB agonist for 5T4-positive malignancies. Mechanistically, our studies revealed that FTL008.16 activates the canonical 4-1BB pathway via TRAF2 phosphorylation at Ser11 (Figure 2d), initiating the K63-linked ubiquitination cascade essential for NF-κB signaling. 50 , 51 This integrated mechanism—combining precise tumor targeting with amplified intracellular signaling—distinctly positions FTL008.16. It avoids the constitutive agonism and toxicity of non-targeted antibodies such as urelumab, 52 while surpassing checkpoint-modulating BsAbs (e.g., PD-L1 × 4-1BB), which primarily rely on localization without a comparable enhancement in intrinsic signaling potency. 53 , 54 In summary, these findings highlight the potential of FTL008.16 as a tumor-directed 4-1BB agonist with improved activity and safety for the treatment of 5T4-positive cancers. To validate whether this tumor-selective mechanism translates into an improved in vivo safety profile, we evaluated FTL008.16 in a humanized 4-1BB mouse model. Consistent with its design, FTL008.16 administration across low, medium, and high doses induced no body weight loss or other overt signs of toxicity (Figure 4b and 4e). This favorable tolerability stands in stark contrast to the dose-limiting toxicities associated with the clinical-stage agonist urelumab, which include body weight loss, splenomegaly, and CD8+ T-cell-mediated liver inflammation. 18 , 55 We next assessed whether this enhanced safety correlated with precise tumor targeting. Fluorescence imaging revealed efficient and specific accumulation of FTL008.16 in 5T4-positive tumors (Figure 4g), consistent with the restricted expression pattern of 5T4. 56 Collectively, these data demonstrate that the excellent in vivo safety of FTL008.16 is a direct consequence of its ability to localize 4-1BB agonism to the tumor microenvironment, thereby circumventing systemic immune activation. These findings robustly support the concept that conditional, tumor antigen-dependent activation is a viable strategy to unlock the therapeutic potential of 4-1BB agonism. In summary, we have developed FTL008.16, a novel 4-1BB × 5T4 bispecific antibody incorporating LALA-PG Fc-silencing mutation that abolishes FcγR binding, which achieves conditional 4-1BB agonism specifically on 5T4-positive tumor cells. This tumor-selective mechanism drives potent antitumor efficacy in preclinical models while successfully avoiding the systemic immune activation and hepatotoxicity that have plagued previous 4-1BB-targeted therapies. The compelling preclinical efficacy and safety data provide a strong rationale for clinical translation. Accordingly, a Phase I clinical trial (CTR20251519) has been initiated to evaluate the safety, pharmacokinetics, pharmacodynamics, and antitumor activity of FTL008.16 in patients with advanced solid tumors. This trial will critically assess whether this targeted approach can deliver a superior therapeutic window for cancer patients. To further advance the translational potential of FTL008.16 and strengthen the clinical relevance of our findings, future studies should focus on validating the key signaling events identified in this work. Specifically, humanized mouse models should be used to directly confirm TRAF2 phosphorylation at Ser11, K63-linked ubiquitination, and NF-κB activation in tumor tissues, which would provide robust in vivo support for our well-characterized in vitro observations. Furthermore, investigations using patient-derived models that recapitulate the heterogeneous expression of 5T4 are warranted for evaluating efficacy under clinically relevant conditions, as such models better reflect the complex tumor microenvironment. Collectively, these future endeavors will not only deepen our mechanistic understanding of FTL008.16 but also provide a stronger foundation for its ongoing clinical development and eventual translation into cancer immunotherapy. Supplementary Material Supplementary Material E10_Efficacy_study_hCD137_mice__2_.pdf Supplementary Material Supplementary Materials.docx Acknowledgments The authors extend their sincere gratitude to Qing Li, Fanxin Ma, and Ying Huang for their foundational contributions to the conceptualization of the 4-1BB × 5T4 bispecific antibody format. The authors also gratefully acknowledge Sound Biopharmaceuticals Co., Ltd. for providing essential technical expertise and support in conducting the animal studies and in vitro experiments. Funding Statement This work was supported by the CAMS Innovation Fund for Medical Sciences (CIFMS) (2021-I2M-5-075) and the National Major Scientific and Technological Special Project for Significant New Drugs Development (2018ZX09201018, 2017ZX09302010). Disclosure of potential conflicts of interest No potential conflict of interest was reported by the author(s). Data availability statement The datasets presented in the study are included in the article/supplementary file. Further inquiries can be directed to the corresponding authors. Ethical approval All animal procedures for the MC38h5T4 and CT26h5T4 model experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Biocytogen and GemPharmatech under protocols PS-01-2304129 and GPTAP20221114-3, respectively. The experimental period for these models spanned from 14 December 2022 to 9 June 2023, and all procedures were performed in accordance with the National Research Council's Guide for the Care and Use of Laboratory Animals. The cynomolgus monkey toxicity study was performed from 21 December 2022 to 20 August 2023 at Xishan Zhongke (Suzhou, China) in compliance with Good Laboratory Practice (GLP) guidelines and was approved by the IACUC of Xishan Zhongke under protocol IP22030213. All efforts were made to minimize animal suffering and to use the minimum number of animals necessary for data reliability. Supplementary material Supplemental data for this article can be accessed at https://doi.org/10.1080/2162402X.2026.2684136. References Associated Data Supplementary Materials Supplementary Material E10_Efficacy_study_hCD137_mice__2_.pdf Supplementary Material Supplementary Materials.docx Data Availability Statement The datasets presented in the study are included in the article/supplementary file. Further inquiries can be directed to the corresponding authors.