The plasmids pcDNA3.1(+) and pcDNA3.1-eGFP were sourced from GenScript Biotech Corporation (Nanjing, China). Chinese hamster ovary cells (CHO-K1) were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). These cells were cultured in a DMEM/F12 medium enriched with 10% fetal bovine serum to ensure optimal growth conditions. Male New Zealand Large White rabbits were acquired from Ningxia Medical University (Yinchuan, China), and bovine positive serum samples for BCoV, bovine diarrheal virus (BVDV), bovine rotavirus (BRV), bovine respiratory syncytial virus (BRSV), and bovine herpesvirus (BHV-1) were stored by our laboratory.

The western and northern regions of China are deemed essential for the cattle business. Consequently, we downloaded the amino acid sequences of the N protein from seven strains of Bovine Coronavirus (BCoV) originating from Northwestern China using the NCBI database. Homology analysis revealed that the sequences were highly conserved, with homology greater than 97.9%. The transmembrane region and signal peptide sequence of the N protein were predicted using DeepTMHMM. N-glycosylation and O-glycosylation sites were forecasted with NetOGlyc and DictyOGlyc, respectively, while potential B cell epitopes were identified via the BepiPred website.

A signal peptide (MKWVTFISLLFLFASAY) was added to the N-terminus and a 6 × His tag to the C-terminus, accompanied by a Kozak motif (GCCACCATGG) preceding the start codon to enhance eukaryotic translation efficiency (10, 11). Restriction sites EcoRI and KpnI were inserted at both ends of the gene, which was subsequently cloned into the pcDNA3.1(+) vector, named pcDNA3.1-N. The sequence was optimized and synthesized by GenScript using CHO cells as the host. The recombinant plasmid was transformed into DH5α cells and cultured overnight in LB liquid medium with 100 μg/mL ampicillin. The plasmid was isolated with an endotoxin-free plasmid mini-prep kit and validated through restriction digestion and DNA sequencing.

A total of 1.2 × 10⁶ CHO-K1 cells were seeded onto a 6-well plate and incubated overnight. Once the cells reached approximately 90% confluency, the medium was replaced with fresh complete culture medium. Lipofectamine 3,000 transfection reagent was used to transiently transfect the cells with the recombinant expression plasmids pcDNA3.1-N and pcDNA3.1-eGFP. Following the established protocol (12), the culture medium was discarded after 24 h, and the cells were subsequently cultured in complete medium supplemented with 800 μg/mL G418. The growth of the CHO-K1 cells was monitored daily, and the medium was refreshed every 48 h to remove dead cells. Incubate for 3 days after the last media change, the cell supernatant was collected and purified using a gravity column filled with Ni-NTA agarose (IMAC) resin (Huiyan biotechnology, Wuhan, China). The purified N protein was then identified by Western blotting (WB) using a sheep anti-rabbit Anti-6 × His tag monoclonal antibody (Abcam, Shanghai, China).

Endotoxin-free recombinant N protein was mixed in a 1:1 ratio with Freund’s Complete Adjuvant (Sigma, Shanghai, China) and administered via multiple subcutaneous injections in the dorsal area of rabbits, with a negative control group established concurrently. Immunizations were carried out biweekly for a total of four times. Only the initial immunization utilized the complete adjuvant, while subsequent ones employed Freund’s Incomplete Adjuvant. After 50 days of immunization, the rabbits were deeply anesthetized, and cardiac puncture was performed to obtain antiserum. The titer of the post-immunization antiserum was subsequently assessed by indirect ELISA, employing the pure N protein as the coating antigen.

In this experiment, rabbit antiserum was purified utilizing gravity columns filled with Protein A Focurose HR resin, with all operations performed at 4°C to reduce protein degradation (13). Briefly, the antiserum was mixed with an equilibration buffer at a 1:2 ratio, filtered through a 0.22 μm filter, and then loaded onto the pre-equilibrated purification column. The column was allowed to stand for 1 h, followed by elution twice with 3 mL of Glycine buffer (pH 3.0), and each tube was neutralized with 80 μL of Tris–HCl buffer (pH 8.8). SDS-PAGE samples were prepared, and the results of the purification were analyzed using Coomassie Brilliant Blue staining. The purified pAb were used as primary antibodies to determine if they could specifically recognize the recombinant N protein via Western blot analysis.

The indirect immunofluorescence assay (IFA) was employed to detect the interaction between rabbit pAb and BCoV (14). MDBK cells in good growth condition were seeded at 1 × 10⁶ cells/well in a 6-well plate. Upon reaching approximately 80% confluency the next day, BCoV was added per well for incubation. After 48 h, cytopathic effects observed under a microscope indicated readiness for IFA. The purified rabbit pAb were diluted 1:2000 as the primary antibody, and TRITC–conjugated Goat Anti-Rabbit IgG (H + L) (Proteintech, Wuhan, China) was diluted 1:1000 as the secondary antibody for incubation. Fluorescence was then examined under a fluorescence microscope. Furthermore, an Anti-6 × His tag monoclonal antibody was employed as the primary antibody in an IFA to detect the expression of N protein in pcDNA3.1-N transfected CHO-K1 cells.

A checkerboard titration method was employed to ascertain the optimal coating concentration of the antigen and the dilution of the antibody. The recombinant N antigen was diluted in carbonate coating buffer to an initial concentration of 10 μg/mL, followed by serial two-fold dilutions. Three replicates were prepared for each dilution, with 100 μL added to each well of an ELISA plate, which was then incubated overnight at 4°C. The following day, the liquid was removed, and the plate was dried. Each well was treated with 150 μL of blocking agent, comprising 3% BSA, 5% BSA, and 5% skim milk powder (SMP), and incubated at 37°C for 2 h. Following three washes of the plate with PBST, high-titer positive serum and blank negative serum were diluted in PBS to four different concentrations (1:100, 1:200, 1:400, 1:800) and incubated at 37°C for 1 h. Subsequently, 100 μL of HRP-conjugated Affinipure Goat Anti-Rabbit IgG was added to each well at four different concentrations (1:5000, 1:8000, 1:10000, and 1:12000) and incubated for four time periods (0.5 h, 1 h, 1.5 h, and 2 h). The plate was subsequently dried, washed three times with wash buffer, and 100 μL of single-component TMB was added to each well. The plate was incubated at 37°C in the dark for 15 min, after which 50 μL of stop solution (2 M H₂SO₄) was added. The absorbance at 450 nm was measured using a microplate reader.

Under the optimal conditions established above, iELISA was performed on 30 BCoV positive bovine serum samples and 30 BCoV negative bovine serum samples, with each set of data replicated three times. The ideal cut-off value, as well as the diagnostic sensitivity and specificity, were established utilizing the Receiver Operating Characteristic (ROC) curve and Youden’s index (15).

The sensitivity of the developed iELISA was evaluated by a two-fold serial dilution of high-titer rabbit pAb, spanning from 2¹³ to 2¹⁹, the maximum dilution factor at which iELISA successfully identified positive rabbit pAb was utilized to assess the assay’s sensitivity. A sample was deemed positive when the ratio of OD450 between positive (P) and negative (N) samples exceeded 2.1. To evaluate the specificity of the method, iELISA was employed to analyze positive serum samples for BCoV, BVDV, BRV, BRSV, and BHV-1, each serum sample was repeated three times.

The intra-assay and inter-assay reproducibility of iELISA was evaluated under the specified conditions applying three bovine positive serum samples and one bovine negative serum samples. Each sample underwent triplicate testing, and the mean, standard deviation, and coefficient of variation (CV) were computed for each sample to evaluate the stability and accuracy of the iELISA.

A total of 58 bovine serum samples were collected from cattle farms in Gansu Province and tested using the developed iELISA and a commercial antibody detection iELISA kit (Keshun Biotechnology, Shanghai, China). The results were compared to evaluate the relative sensitivity and specificity, and the concordance rate between the two methods was calculated. The concordance rate between the developed iELISA and the commercial kit results was determined by counting the number of consistent results between the two tests and dividing it by the total number of samples (16).

Data processing and analysis were performed with GraphPad Prism 8 software. Each experiment had three independent replicates to ensure the reliability and statistical significance of the results.

The amino acid sequence of the N protein was predicted using an online bioinformatics tool. The analysis revealed the presence of three O-glycosylation sites and three N-glycosylation sites (Figures 1A,B), as well as the absence of a signal peptide and no structural domain of transmembrane region (Figure 1C). The N protein was found to be enriched with B-cell epitopes that exhibit strong immunogenicity, which is essential for antibody production (Figure 1D). The N sequence was codon optimized for CHO host cells, leading to an increase in GC content from 46.38 to 58.14%. It was subsequently cloned into the pcDNA3.1(+) vector, named as pcDNA3.1-N (Figure 1E). We identified the size of the N gene as 1,428 bp through double digestion, which was consistent with the theoretical value (Figure 1F). This further confirms the successful construction of the recombinant expression plasmid via DNA sequencing.

We characterized the transfection efficiency of the target gene by observing the expression of green fluorescent protein under fluorescence microscope, we collected CHO-K1 cells at three different time periods of 24, 48 and 36 h after transfection with eGFP, the results showed (Figure 2A) that eGFP protein levels increased with longer transfection times, with a significant rise at 48 h, peaking at 36 h. The results indicated that eGFP protein expression increased with prolonged transfection time, peaking at 36 h. we collected the supernatant of cell culture medium 3 days after transfection and purified it by affinity chromatography using Ni-NTA gravity column. WB and SDS-PAGE results indicated that the purified N protein exhibited a single band at 65 kDa, which exceeds the anticipated size of 53 kDa (Figures 2B,C). This discrepancy may be attributed to post-translational modifications of the recombinant N protein in eukaryotic cells. Subsequent IFA experiments were conducted to further validate the expression of the N protein in CHO-K1 cells (Figure 2D).

We used recombinant N protein as the coated antigen to determine the potency of rabbit pAb serum as 8.192 × 10⁵ (Table 1).

Following the purification of the obtained rabbit serum by affinity chromatography, the structures of reduced and non-reduced rabbit pAb were characterized (Figure 3A). The molecular weights of non-reduced antibodies were approximately 180 KDa, consistent with expectations and maintaining their natural structure, while the molecular weights of reduced antibodies were approximately 55 KDa and 25 KDa, respectively. The WB results demonstrated that the purified rabbit pAb exhibited strong immunoreactivity with the recombinant N protein (Figure 3B). We employed BCoV to infect MDBK cells, and the IFA results showed that the rabbit poly antibody specifically recognized BCoV. These results demonstrated that the rabbit pAb exhibited strong immunoreactivity and specificity toward both recombinant N protein and BCoV, and that it was promising to be developed as a pathogen detection reagent (Figure 3C).

The conditions of iELISA were optimized based on the purified N protein, the results of the checkerboard titration showed that the optimal antigen concentration was 1.25 μg/mL, and the optimal dilution of the antiserum is 200-fold, yielding a P/N value of 29.04 (Figure 4).

Subsequently, other important conditions were also improved, and the experimental results showed that the optimal results were obtained when the blocking agent was 3% BSA, and the dilution of the secondary antibody was 1:8,000 and the incubation time of secondary antibody was 1 h, the best reaction result was achieved (Figure 5).

We employed iELISA to analyze 30 BCoV positive serum samples and 30 negative serum samples for data quantification (Figure 6B), and determined the critical value of the method by producing a subject operating characteristic curve (ROC). The ROC analysis indicated that the diagnostic sensitivity and specificity of iELISA were 89.7 and 100%, respectively, with an optimal cutoff value of 0.986, and a Youden index of 0.897 (Figure 6A). The proximity of the area under the curve (AUC) of the ROC to 1.0 is widely regarded as indicative of assay authenticity. The findings of the iELISA demonstrated an AUC value of 0.989 (95% CI = 0.971 ~ 1.000), signifying that the method exhibits high precision. A sample was classified as “positive” if the measured OD450 value was >0.986, whereas a sample with an OD450 value <0.986 was classified as “negative.”

To ascertain the detection limit of the iELISA based on N protein, the rabbit pAb was diluted 2-fold and analyzed by the established iELISA method. The results showed a P/N value of 1.986 at a dilution of 2¹⁹, thus the maximum dilution at 2¹⁸ (Figure 7).

To evaluate the specificity of iELISA, cross-reactivity tests were performed using bovine positive serum samples of common bovine viruses, including BHV-1, BVDV, BRV, BRSV, and BCoV. The results showed that, except for the BCoV serum samples, the OD450 values of other serum samples were lower than 0.986 with significant differences in values, while the positive serum containing BCoV showed strong reactivity with an OD450 value of 1.125 (Figure 8). The results indicated that the iELISA exhibited no cross-reactivity with other prevalent bovine virus sera and had strong specificity for BCoV sera.

Repeatability was tested using three bovine positive serum samples and one bovine negative serum sample. The results in the table showed that the CV ranged from 1.661 to 3.773% within batches and from 0.291 to 3.381% between batches, which were not more than 5%, leading to good repeatability (Table 2).

Fifty-eight bovine serum samples collected from Gansu Province, China, were tested using a commercial kit and compared with the developed EILSA method. Five BCV-positive samples were detected by the commercial kit, whereas eight positive samples were detected by the developed iELISA method. The diagnostic sensitivity and specificity of the developed ELISA method were 100 and 94.64%, respectively, and the concordance rate between the two methods was 94.83% (Table 3), indicating that the method is suitable for the rapid diagnosis of clinical samples.