New Sandwich ELISA Reveals Fc Fusion Extends Feline GM-CSF Half-Life Eightfold in Cats

Feline granulocyte-macrophage colony-stimulating factor, better known as GM-CSF, sits at the center of myeloid immune regulation in cats. The cytokine drives the proliferation and differentiation of neutrophils, macrophages, and dendritic cells, and it underpins anti-infectious immunity and hematopoietic homeostasis. Because feline panleukopenia virus and feline immunodeficiency virus frequently trigger severe leukopenia, GM-CSF has long been viewed as a promising therapeutic target for restoring white blood cell counts in infected cats. Clinical studies have shown that recombinant human GM-CSF can raise leukocyte numbers in some FIV-infected animals, but the native molecule is cleared from the body so rapidly that frequent administration would be required, a constraint that has severely limited clinical translation.

To overcome this pharmacokinetic bottleneck, researchers at Huazhong Agricultural University and Hubei Jiangxia Laboratory in Wuhan constructed a recombinant GM-CSF-Fc protein using Fc fusion technology, a strategy that exploits the neonatal Fc receptor recycling pathway to extend the in vivo retention of cytokines. Evaluating such a construct, however, demanded an analytical tool capable of quantifying differently engineered forms of recombinant feline GM-CSF simultaneously. No commercial feline-specific GM-CSF detection kit exists, and because GM-CSF sequences diverge substantially among species, heterologous cross-species methods often show poor recognition efficiency and potential immunological cross-reactivity. Traditional bioactivity assays, meanwhile, are labor-intensive, insufficiently sensitive, and poorly reproducible in complex biological matrices such as serum.

Reporting in Frontiers in Veterinary Science, the team described how they expressed recombinant feline GM-CSF in a Chinese hamster ovary, or CHO, eukaryotic expression system and used the purified protein to immunize rabbits, generating high-affinity polyclonal antibodies. Two recombinant fusion proteins were designed for the study: GM-CSF-His, carrying a Twin-Strep-tag and a 10×His tag at the C-terminus, and GM-CSF-Fc, carrying a Twin-Strep-tag and a feline IgG Fc fragment. After affinity purification, SDS-PAGE analysis showed molecular weights of approximately 25 kDa for GM-CSF-His and 55 kDa for GM-CSF-Fc, consistent with theoretical values, and Western blot confirmed that both proteins were specifically recognized by anti-feline GM-CSF antibodies. Grayscale quantification revealed purities of 84.23 percent for GM-CSF-His and 93.40 percent for GM-CSF-Fc.

The immunization protocol induced robust and stable specific antibody responses in New Zealand White rabbits. Indirect ELISA testing of serially diluted immune sera showed that the highest-titer antiserum still maintained a signal-to-noise ratio above the positive threshold at a dilution of 1:8,000. After purification by saturated ammonium sulfate precipitation and Protein A affinity chromatography, the qualified polyclonal antibodies provided reliable raw materials for constructing the sandwich ELISA detection system, which paired a commercial capture antibody with a biotinylated rabbit polyclonal detection antibody raised against untagged recombinant feline GM-CSF.

Establishing the double-antibody sandwich ELISA required systematic optimization of key reaction parameters through checkerboard titration. The optimal conditions proved to be a capture antibody concentration of 6 μg/mL paired with a detection antibody at 2 μg/mL, blocking with 5 percent bovine serum albumin for 2 hours, and a 1:10,000 dilution of streptavidin-HRP incubated for 1 hour. These settings yielded the highest signal-to-noise ratios at every optimization step, effectively maintaining strong positive detection signals while reducing background interference from non-specific binding.

Methodological validation demonstrated excellent analytical performance. The assay exhibited strong linearity within 7.8 to 2,000 pg/mL, with a coefficient of determination of 0.993, and the final validated quantitative range was 50 to 2,000 pg/mL, spanning the expected concentration fluctuations of recombinant GM-CSF under physiological and immune-activated conditions in cats. The limit of detection was calculated at 8.12 pg/mL and the calculated limit of quantification at 27.44 pg/mL, while the experimentally validated lower limit of quantification was 50 pg/mL. Quality control samples at six concentrations showed coefficients of variation below 10 percent, relative bias within 15 percent, and total error below 20 percent, all complying with accepted guidelines for bioanalytical method validation. Critically, the assay showed no cross-reactivity with the related feline cytokines IL-3, IL-5, M-CSF, and G-CSF at any tested concentration.

The core innovation lies in the assay’s ability to quantify differently engineered forms of recombinant feline GM-CSF with comparable recovery. Concentration-response curves generated for GM-CSF-His and GM-CSF-Fc under identical conditions produced typical sigmoidal shapes, with EC50 values of 918.0 pg/mL and 1,064 pg/mL respectively, a ratio of 1.16, and statistical comparison using the extra sum-of-squares F test showed no significant difference in EC50 or Hill slope. Serum spike recovery experiments in blank feline serum yielded recoveries of 98.05 to 110.83 percent for GM-CSF-His and 84.86 to 90.35 percent for GM-CSF-Fc across validated dilution factors of 1:5 to 1:80, with coefficients of variation at or below 2.76 percent. These results confirm that both protein formats can be accurately quantified in feline serum using the same calibration curve, overcoming the limitation of conventional assays optimized for a single protein format.

With the validated assay in hand, the researchers turned to the pharmacokinetic question at the heart of the study. Six healthy adult cats were randomly assigned to two groups and subcutaneously injected with 10 μg/kg of either GM-CSF-His or GM-CSF-Fc. Peripheral blood samples were collected from the cephalic vein at sixteen time points spanning 0 to 216 hours post-administration, and serum concentrations were quantified using the new ELISA with non-compartmental analysis performed in DAS 2.0 software. The results were striking: Fc fusion substantially prolonged the apparent terminal elimination half-life of immunoreactive GM-CSF from 5.05 ± 0.91 hours for GM-CSF-His to 40.89 ± 3.41 hours for GM-CSF-Fc, an approximately 8.1-fold extension. Individual half-life values showed no overlap between groups, and the mean difference of 35.84 hours was statistically significant.

The 8.1-fold extension sits at the higher end of the range reported for long-acting veterinary cytokines. For comparison, an Fc-fused feline interferon-ω2 extended the plasma half-life approximately 5.9-fold, reaching 7.0-fold only after introducing YTE mutations, while the long-acting feline G-CSF fusion protein APB-F1 achieved roughly a 5-fold extension. The authors suggest the difference may be attributed to distinct receptor-mediated clearance rates, the impact of the Fc fusion site on receptor binding, and differences in glycosylation patterns. Beyond half-life, the Fc-fused construct also showed markedly higher total drug exposure, with the area under the concentration-time curve increasing more than sixfold and apparent clearance falling from 51.95 to 8.52 mL/h/kg.

The study acknowledges several limitations, including the small sample size of three cats per group, the absence of comprehensive stability and matrix-effect evaluations, and the fact that only immunoreactive GM-CSF was measured rather than biologically active protein, meaning the observed prolongation of serum persistence should not be equated with prolonged biological activity. Nevertheless, the work provides a versatile and sensitive analytical platform for preclinical evaluation of Fc-fused feline GM-CSF formulations and serves as an important methodological reference for quantitative assessment of Fc-fused veterinary biologicals. Because the assay recognizes the GM-CSF moiety irrespective of C-terminal modification, it minimizes methodological discrepancies and enables accurate characterization of pharmacokinetic differences between recombinant variants, a capability the authors expect to support novel biopharmaceutical development of Fc-fusion veterinary cytokines well beyond this single target.

Subject of Research: Development and validation of a sandwich ELISA for recombinant feline GM-CSF and its application to determining the serum half-life of Fc-fused GM-CSF in cats

Article Title: Development of a sandwich ELISA for recombinant feline GM-CSF and determination of its serum half-life

Article References: Yu, C., Wang, M., Chen, L., Wang, T., Zhao, Y., Song, H., & Jin, M. (2026). Development of a sandwich ELISA for recombinant feline GM-CSF and determination of its serum half-life. Frontiers in Veterinary Science, 13, Article 1958789. https://doi.org/10.3389/fvets.2026.1958789

Image Credits: AI Generated

DOI: 10.3389/fvets.2026.1958789

Keywords: feline GM-CSF, sandwich ELISA, Fc fusion, pharmacokinetics, half-life, cytokine, veterinary biopharmaceuticals, CHO expression, polyclonal antibodies, leukopenia, recombinant protein, cats