Dual-Target Antibody Strikes Two Inflammatory Pathways to Ease Colitis in Mice

Inflammatory bowel disease affects millions of people worldwide, and for a large share of patients the biologic drugs now on the market simply stop working. A team of researchers reporting in the journal iScience has now built a single antibody molecule capable of simultaneously blocking two of the most important inflammatory drivers of Crohn’s disease and ulcerative colitis, and the molecule has shown striking results in a humanized mouse model of colitis. The work, led by Xia Li, Hao Peng, Feng Hao, Yi Gu, Jinying Ning, and Changyuan Yu, describes a bispecific antibody that targets both TL1A, a member of the tumor necrosis factor superfamily, and the shared p40 subunit of the interleukin-12 and interleukin-23 cytokines. By shutting down both signaling routes at once, the antibody suppressed the pathogenic Th1 and Th17 immune responses that fuel chronic intestinal inflammation and significantly eased disease in mice.

The rationale for the dual approach rests on decades of immunological evidence. TL1A, the only known ligand for the death receptor 3 signaling receptor, is strongly upregulated in the inflamed intestinal tissue of patients with inflammatory bowel disease. When it engages DR3, it activates NF-κB and MAPK pathways inside T cells, promoting co-stimulation, inflammatory cytokine secretion, and even the intestinal fibrosis that complicates long-standing disease. Meanwhile, IL-12 and IL-23 are heterodimeric cytokines that share the p40 subunit and steer T cell differentiation down two destructive paths: IL-12 drives Th1 cells and interferon-gamma production through STAT4, while IL-23 stabilizes Th17 cells and their signature IL-17 output through STAT3. Crucially, these pathways do not operate in isolation. TL1A enhances IL-12-mediated Th1 polarization and IL-23-driven Th17 differentiation, forming a feedforward loop that amplifies mucosal damage. Existing single-target biologics, including ustekinumab and risankizumab, leave 30 to 50 percent of patients without durable remission, precisely because redundant and interconnected pathways allow inflammation to escape.

To solve the engineering challenge of hitting two targets with one molecule, the team used a common light chain architecture combined with knobs-into-holes Fc engineering, which forces the two different heavy chains to pair correctly while eliminating light chain mispairing, a notorious problem in bispecific antibody manufacturing. The resulting antibody, named KA-4316, was produced in ExpiCHO cells at high expression levels of 478.29 milligrams per liter and purified to exceptional monomer purity of 97.35 percent. Size-exclusion chromatography, SDS-polyacrylamide gel electrophoresis, and differential scanning fluorimetry confirmed robust assembly at the expected 144 kilodalton molecular weight and a favorable melting temperature of 68.62 degrees Celsius. These biophysical benchmarks matter enormously for translational development, because a bispecific antibody that aggregates or assembles inconsistently cannot be manufactured reliably at clinical scale.

Functional characterization showed that KA-4316 binds its two targets with sub-nanomolar affinity: 8.72 × 10⁻¹¹ molar for human TL1A and 2.50 × 10⁻¹⁰ molar for IL-12/23p40. Reciprocal sequential biolayer interferometry and sandwich ELISA assays confirmed that the antibody can engage both antigens at the same time without steric interference between its two arms. The antibody recognized both soluble and membrane-associated TL1A, and it bound the cynomolgus monkey orthologs of both targets, a critical property for non-human primate pharmacology, while showing no detectable binding to the mouse versions. In neutralization assays, KA-4316 blocked the interaction of IL-12 and IL-23 with the IL-12Rβ1 receptor subunit with IC50 values of 12.88 and 16.81 nanomolar respectively, and it prevented TL1A from engaging DR3 with an IC50 of 2.76 nanomolar, with weaker activity against the decoy receptor DcR3.

In cell-based assays, the bispecific antibody functioned as a potent dual inhibitor of inflammatory signaling. It suppressed TL1A-driven NF-κB activation in DR3 reporter cells with an IC50 of 0.49 nanomolar, neutralized TL1A-induced apoptosis in TF-1 cells, blocked IL-12-induced STAT4 activation and interferon-gamma secretion from NK cells, and dampened IL-23-mediated STAT3 activation. Most tellingly, when human peripheral blood mononuclear cells were stimulated with the synergistic combination of TL1A, IL-12, and IL-18, KA-4316 suppressed interferon-gamma release with an IC50 of 0.47 nanomolar, outperforming either parental monospecific antibody alone and matching their combined administration. Chou-Talalay combination index analysis of the parental antibody mixture produced values below one across multiple effect levels, quantitatively confirming pharmacological synergy between the two pathways. The authors note candidly that the bispecific format itself offered only a modest, statistically insignificant potency improvement over simply mixing the two parental antibodies, meaning the value of the molecule lies in its convenience, pharmacokinetics, and unified manufacturing rather than intrinsic functional superiority.

To make the antibody clinically practical, the team engineered a half-life-extended variant called KA-5817 by introducing the YTE mutations, M252Y, S254T, and T256E, into the Fc region. These mutations enhance binding to the neonatal Fc receptor at acidic pH, which recycles IgG molecules and prolongs their persistence in circulation. In FcRn-humanized mice receiving a single 10 milligram per kilogram intravenous dose, KA-5817 exhibited a serum half-life of 21.53 days, compared with 8.32 days for the original KA-4316, a nearly threefold extension that could translate into less frequent dosing and better patient compliance. A pilot pharmacokinetic and safety study in cynomolgus monkeys at 3 and 10 milligrams per kilogram showed dose-dependent elimination kinetics complicated by anti-drug antibody formation at the higher dose, but comprehensive hematology and serum biochemistry revealed no treatment-related hematologic, hepatic, or renal toxicity at either dose level.

The decisive test came in a clinically translatable colitis model using C57BL/6 mice triple-humanized for TL1A, IL23A, and IL12B, meaning the animals carry human versions of all three relevant genes. Acute colitis was induced with dextran sulfate sodium in drinking water for six days, and mice received intravenous antibody treatments on days 0, 3, and 7 at a total dose of 20 milligrams per kilogram. Compared with an isotype control, single monospecific antibodies against TL1A or IL-12/23p40 produced only partial relief. KA-5817, however, dramatically reduced weight loss, disease activity scores combining stool consistency and rectal bleeding, and colon shortening, performing on par with an equimolar physical mixture of the two monospecific antibodies. Histological examination of colon tissue confirmed the benefit: where isotype-treated mice showed extensive epithelial erosion, crypt loss, and dense inflammatory infiltration, KA-5817-treated animals retained intact crypt architecture with substantially reduced immune cell invasion of the mucosa.

Mechanistic analysis of immune cells isolated from the colonic lamina propria revealed how the antibody works. Flow cytometry showed that KA-5817 significantly reduced the frequency of pathogenic Th1 cells, the CD4-positive interferon-gamma producers, while Th17 cells showed a trend toward reduction, and regulatory T cell proportions were unchanged. In colon tissue homogenates, the antibody substantially lowered concentrations of the pro-inflammatory cytokines TNF-α and IL-6 and diminished phosphorylated STAT3, the downstream signaling node through which IL-23 sustains Th17-driven autoimmunity. Together these data sketch a coherent mechanism: dual blockade breaks the synergistic feedforward loop between TL1A and the IL-12/23 axis, restrains the differentiation and activity of the two T cell subsets that dominate intestinal pathology, and calms cytokine production in the inflamed tissue itself.

The findings arrive at a moment of intense commercial and clinical interest in TL1A as a drug target. Pfizer’s anti-TL1A monoclonal antibody PF-06480605 has already shown promise in patients, and PF-07261271, a TL1A × IL-12/23p40 bispecific developed by Pfizer and Roche, is in phase 2b trials for ulcerative colitis using CrossMab technology, which the authors argue suffers from structural asymmetry and intellectual property constraints. Their common light chain design, they contend, simplifies manufacturing and analytical characterization while maintaining potency. The team is also forthright about limitations: the DSS colitis model is driven largely by acute epithelial injury and innate immunity rather than the chronic, T cell-mediated immunopathology of human disease, and the monkey pharmacokinetic study used only one animal per dose. Future work in adoptive T cell transfer models, which generate persistent Th1 and Th17-dominated inflammation, will be needed to confirm translational potential. Still, with strong preclinical efficacy, favorable safety signals, and a manufacturable molecular design, this dual-targeting strategy represents a compelling next-generation candidate for patients whose inflammation currently slips past single-pathway drugs.

Subject of Research: Development of a TL1A and IL-12/23p40 bispecific antibody for treating inflammatory bowel disease

Article Title: Bispecific antibodies targeting TL1A and IL-12/23p40 suppress pathogenic Th1/Th17 responses and ameliorate colitis in mice

Article References: Bispecific antibodies targeting TL1A and IL-12/23p40 suppress pathogenic Th1/Th17 responses and ameliorate colitis in mice. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: bispecific antibody, TL1A, IL-12/23p40, inflammatory bowel disease, colitis, Th1 cells, Th17 cells, DR3, STAT3, FcRn, half-life extension, preclinical study