Sandwich ELISA Offers Rapid Antigen Test for Feline Coronavirus Detection

Feline coronavirus (FCoV) remains one of the most common viral pathogens in domestic cats, and a small fraction of infections progress to feline infectious peritonitis, a nearly uniformly lethal disease. A research team in Fujian, China, has now developed and validated a double-antibody sandwich enzyme-linked immunosorbent assay (DAS-ELISA) that detects the viral spike protein directly in feline faecal samples and ascitic fluid, offering a low-cost, high-throughput alternative to molecular testing. The study, published in Veterinary Medicine and Science, describes the generation of monoclonal and polyclonal antibodies against a truncated Type I FCoV spike protein and demonstrates that the resulting sandwich assay achieves a clinical sensitivity of 81.3 percent and a specificity of 80.0 percent when benchmarked against reverse transcription quantitative polymerase chain reaction (RT-qPCR).

FCoV is a non-segmented, single-stranded positive-sense RNA virus with a genome of approximately 29 kilobases. The disease it causes was first described in 1963, but the viral aetiology was only established in 1966, when healthy animals inoculated with tissues from diseased cats developed infection, and the agent was confirmed as a coronavirus in 1968. Two biologically distinct pathogenic forms circulate in cat populations. Feline enteric coronavirus (FECV) is highly contagious in multi-cat environments and infects nearly all exposed animals, yet it usually produces only mild, transient diarrhoea or no symptoms at all. Feline infectious peritonitis virus (FIPV), by contrast, is not transmitted by the faecal-oral route; it arises from mutations in FECV within a small proportion of infected cats, shifting viral tropism from intestinal epithelial cells to monocytes and macrophages. Once FIPV replicates inside peritoneal macrophages, it disseminates through the body and produces high-titre progeny virus, culminating in the fatal systemic disease known as feline infectious peritonitis.

The mutations that drive this transition remain incompletely mapped, which makes accurate and early diagnosis of FCoV infection a critical prerequisite for antiviral intervention. The spike (S) protein, a Type I transmembrane glycoprotein divided into an N-terminal S1 receptor-binding domain and a C-terminal S2 domain containing the fusion peptide, sits at the centre of both pathogenesis and immunity. It mediates fusion of the viral envelope with the host cell membrane, determines antigenic classification, and carries abundant epitopes that elicit neutralizing antibodies in infected hosts. FCoV strains are divided into Type I and Type II serotypes on the basis of their S proteins, with Type I being the most primitive and most prevalent serotype worldwide. Because early coronavirus detection strategies have repeatedly benefited from targeting the S protein, the researchers reasoned that S-specific monoclonal antibodies could anchor a robust antigen-detection platform, even though few such assays had previously been reported for FCoV.

Existing diagnostic options each carry drawbacks. RT-qPCR is sensitive and experimentally straightforward, but false negatives can arise with novel variants, false positives from sample contamination, and the technique requires fully equipped laboratories, limiting its use in large-scale clinical screening. Serological assays detect immunoglobulin G antibodies, which only appear days after infection and vary between individuals, so they are poorly suited to early diagnosis. Virus isolation, the classical reference method, is hampered by the difficulty of culturing FCoV in vitro, long turnaround times and demanding operating conditions. Against this backdrop, the team set out to build a sandwich ELISA in which a rabbit polyclonal capture antibody and a mouse monoclonal detection antibody bracket the spike antigen, combining broad binding capacity with high specificity.

To generate the immunogen, the researchers expressed and purified a recombinant fragment of the Type I FCoV spike protein, designated FCoV-SP. Six-week-old BALB/c mice received weekly subcutaneous injections of 50 micrograms of the protein emulsified in Freund’s complete adjuvant for three weeks, while three adult rabbits were immunized with 200 micrograms each following the same schedule. Antibody titres were monitored by indirect ELISA using tail vein blood from mice and auricular artery blood from rabbits. After the final immunization, immunoglobulin G was purified from the rabbit serum with the highest titre on a protein A Sepharose column and stored at minus 80 degrees Celsius for use as the capture antibody.

Monoclonal antibodies were produced by fusing myeloma cells with splenocytes from the immunized mice at a 1:8 cell ratio using 50 percent PEG3350. Hybridomas were cultured in 96-well plates under standard HAT and HT selection regimens, and culture supernatants were screened by an indirect ELISA optimized through checkerboard titration. Positive clones were subcloned by limiting dilution until 100 percent positivity was achieved, and ascites fluid was raised in pristane-primed mice injected with hybridoma cells. A subclone designated E5, whose isotype was identified as IgG2b, emerged as the working monoclonal. Western blotting confirmed that both the E5 monoclonal antibody and the rabbit polyclonal serum specifically recognized the recombinant FCoV-SP at its predicted molecular weight of approximately 32 kilodaltons, validating the reagents at the heart of the assay.

Checkerboard titration established the optimal working conditions: microplates were coated overnight at 4 degrees Celsius with the polyclonal capture antibody at 0.15 micrograms per millilitre, blocked, then incubated with samples or controls for 45 minutes at room temperature. The E5 monoclonal was added at 1.0 micrograms per millilitre for 30 minutes, followed by horseradish peroxidase-conjugated goat anti-mouse IgG diluted 1:10,000 for one hour. The colour reaction was developed with tetramethylbenzidine, stopped with 2 M sulphuric acid, and read at 450 nanometres. Serial twofold dilutions of purified antigen from 3.25 down to 0.0127 micrograms per millilitre generated a standard curve with a coefficient of determination of 0.991, and the analytical limit of detection was calculated as approximately 0.15 micrograms per millilitre, defined as the concentration corresponding to the mean optical density of 20 negative faecal samples plus three standard deviations.

The cut-off value was set at 0.85, derived from the mean OD450 of 0.33 plus three standard deviations of 0.17 among 20 RT-PCR-confirmed negative faecal samples, which were pretreated with 1 percent Triton X-100 and 0.3 percent tri-n-butyl phosphate to inactivate any virus. Repeatability testing across five samples analysed in duplicate over three days yielded intra-assay coefficients of variation between 3.90 and 14.10 percent and inter-assay coefficients between 2.55 and 11.09 percent. One high-signal positive sample exceeded the conventional 10 percent intra-assay threshold at 14.07 percent, likely because small pipetting or washing inconsistencies are proportionally amplified at elevated absorbance, but all values remained below the 20 percent limit generally accepted for exploratory immunoassays. Specificity testing on 48 faecal samples, 12 each from cats infected with feline panleukopenia virus, feline calicivirus, feline herpesvirus and FCoV, produced positive signals only for FCoV-infected samples and the purified spike protein, with no cross-reactions against the other three feline viruses.

For clinical validation, 62 faecal samples were collected from cats in Longyan, Xiamen and Fuzhou with informed owner consent and tested in parallel by DAS-ELISA and RT-qPCR. Of 32 PCR-positive samples, 26 were correctly identified by the ELISA, a positive coincidence rate of 81.3 percent, while 25 of 30 PCR-negative samples were correctly classified, or 83.3 percent. Receiver operating characteristic analysis yielded a sensitivity of 81.3 percent, a specificity of 80.0 percent and an area under the curve of 0.861, indicating clinically useful discriminatory power. The six discordant samples that were PCR-positive but ELISA-negative likely reflect low viral antigen loads early in infection or during recovery, when faecal antigen concentrations fall below the assay threshold even though viral RNA remains detectable. Conversely, five samples were ELISA-positive but PCR-negative; their OD values clustered just above the cut-off, suggesting low-level matrix interference, possibly from endogenous peroxidase contaminants in faeces, rather than genuine antigen binding. The authors note that adding a peroxidase-blocking step or inhibitors to the sample diluent could reduce such false positives, and that equivocal results should be confirmed by RT-PCR.

The study has acknowledged limitations. Because no commercial FCoV antigen-detection ELISA exists, the assay could only be compared against published PCR results rather than a reference kit, and the specificity panel of 12 samples per competing virus is too small to compute robust confidence intervals or definitively exclude cross-reactivity; cohorts of 30 or more per pathogen are recommended for reliable specificity estimates. Cross-reactivity with Type II FCoV strains and canine coronavirus was not systematically tested, and the analytical limit of detection was established with purified antigen in buffer rather than in the faecal matrix, so the true clinical detection limit may be higher. Even so, the combination of acceptable repeatability, strong analytical linearity, no cross-reactivity with the major feline enteric viruses, and agreement with RT-qPCR in clinical samples positions this DAS-ELISA as a promising, affordable front-line tool for FCoV surveillance in multi-cat households, catteries and veterinary clinics, pending larger validation studies and matrix-adjusted sensitivity benchmarks.

Subject of Research: Development and validation of a double-antibody sandwich ELISA for detecting feline coronavirus spike protein antigen

Article Title: Double‐Antibody Sandwich Enzyme‐Linked Immunosorbent Assay for Detecting Feline Coronavirus Infection

Article References: Dong, B., Zou, W., Hu, W., Wu, H., Jing, Y., Xie, F., Zhang, S., & Lin, W. (2026). Double‐Antibody Sandwich Enzyme‐Linked Immunosorbent Assay for Detecting Feline Coronavirus Infection. Veterinary Medicine and Science, 12(6), Article e71289. https://doi.org/10.1002/vms3.71289

Image Credits: AI Generated

DOI: 10.1002/vms3.71289

Keywords: feline coronavirus, DAS-ELISA, spike protein, monoclonal antibody, polyclonal antibody, feline infectious peritonitis, RT-qPCR, veterinary diagnostics, faecal samples, antigen detection, cross-reactivity, diagnostic sensitivity