Easily Prepared Nanobody-horseradish Peroxidase Fusion Protein-based Immunoassay to Detect Antibodies Against Genotype 2 Porcine Reproductive and Respiratory Syndrome Viruses

Background: Nanobodies are the third generation of genetically-engineered antibodies, possessing advantages of small molecular weight, simple genetic engineering, and low production cost for promising diagnostic application. In this study, a competitive ELISA (cELISA) using nanobody-horseradish peroxidase (HRP) fusion protein was utilized as an ultrasensitive probe for the rst time to detect antibodies against genotype 2 porcine reproductive and respiratory syndrome virus (PRRSV). In addition, a platform for easily producing the nanobody-HRP fusion protein against PRRSV was designed and developed. Results: In the present study, three nanobodies against the PRRSV-N protein were screened by Camel immunization, library construction, and phage display. Subsequently, a recombinant HEK293S cell line stably secreting nanobody-HRP fusion protein against PRRSV-N protein was successfully designed and constructed using the lentivirus transduction assay. Using the cell lines, the fusion protein was easily produced. Then, a novel cELISA was developed using the nanobody-HRP fusion protein for detecting the antibodies against genotype 2 PRRSV in pig sera, exhibiting a cut ‐ off value of 23.19% and good sensitivity (100%), specicity, and reproducibility. The compliance rate of cELISA with a commercial IDEXX ELISA kit was 96.4%. By testing the sequential sera from the challenged pigs, the results showed that the sensitivity of cELISA was higher than the commercial IDEXX ELISA kit. In addition, the commercial IDEXX ELISA kit can be combined with the developed cELISA for the differential detection of antibodies against genotype 1 and 2 PRRSV in pig sera. Conclusions: By screening the three nanobodies against the genotype 2 PRRSV-N protein, a recombinant HEK293S cell line stably secreting nanobody-HRP fusion PRRSV-N a cELISA was for detecting antibodies against genotype 2 PRRSV fusion high sensitivity, specicity, and platform is simple and low cost. The PRRSV-N protein specic nanobodies were screened by three rounds of panning using phage display technology, as previously described, with the following modications Briey, the VHH phage library was rescued via M13K07 helper phage. The 96-well plates (Nunc, Denmark) were coated with the recombinant PRRSV-N protein (4 µg/well) overnight at 4°C for the three rounds of panning. The coated plates were blocked with 200 µL of 2.5% skim milk at 37°C for 1 h and washed with 0.05% PBS’T (1 L PBS with 0.5 mL Tween-20). Then, the above rescued recombinant phage (5 × 10 11 pfu/mL) were added to the plates and incubated at room temperature (RT) for 1 h. After the plates were washed again, the binding phages were eluted using 100 µL 0.1 M trimethylamine (Sigma, St. Louis, MO, USA) and neutralized with same volume of 1 M Tris-HCl (pH 7.4). Subsequently, the growth log phase of E. coli TG1 was infected with the eluted phages and amplied for further rounds of selection. The enrichment of specic phage particles was analyzed using anti-M13/HRP conjugate ELISA combined with phage titration after three rounds of panning. Finally, the 96 colonies were picked randomly and induced with IPTG (1 mM) to express soluble VHHs with an HA-Tag. These recombinant VHHs-HA-Tag proteins were extracted and tested for their capacity to recognize the PRRSV-N protein using iELISA with an anti-HA-Tag antibody as the rst antibody (GenScript, Biotech Corp, China). Finally, the positive clones were sequenced, and the nanobodies were grouped according to the hypervariable complementary-determining region 3 (CDR3) sequence. cell HEK293S cells were transduced using above and with After 48 h, the transduced cells were observed under a uorescence microscope. Then, HEK293S cells with uorescence were sorted by High-speed sorting ow cytometer (BD, US). The biological activity (titres and stability) of nanobody-HRP fusion proteins produced by the transient transfection and stable expressing system were compared. Firstly, the titers of the nanobody-HRP fusion proteins from the two systems were tested with direct ELISA using a checkerboard titration. Different amounts of the PRRSV-N protein, i.e. 100, 200, 400, and 800 ng/well, were coated into the 96-well plates, and the different dilution ratios of the nanobody-HRP fusions (1:10, 1:100, 1:500, and 1:1000) were tested. The titer was assessed when the OD 450nm value of the direct ELISA was 1.0, and the stability of the fusion proteins produced by the two systems was also evaluated. The two methods were independently repeated ve times for producing the fusion proteins. Then, the fusion proteins from the different production batches were detected using direct ELISA with a 1:100 dilution. Subsequently, the stability of fusion proteins from the stable expression system was tested from the second, forth, sixth, eighth, and tenth generations and the supernatants from these generations were detected using direct ELSIA with dilutions of 1:10, 1:100, and 1:1000. In addition, the operations of the two systems were compared based on the procedures for producing the fusion proteins. the immunized camel. Positive rate analysis by colony PCR revealed that 96% of these colonies contained a correct insert corresponding to the size of VHH genes. Then, 50 randomly clones were selected, sequenced, and analyzed. The results show that each clone was manifested to contain a distinct VHH sequence (data not shown), suggesting the good diversity and high quality of the library.


Introduction
Nowadays, conventional polyclonal and monoclonal antibodies are widely used as indispensable reagents for the development of disease diagnostic kits [1]. Nevertheless, the traditional antibodies have some shortcomings that limit their application in related elds. For example, polyclonal antibodies suffer from batch-to-batch variability, while monoclonal antibodies have high costs and di cult genetic manipulation for production. Thus, there is an urgent need to develop strategies aimed at the production of alternative scaffolds [2]. In recent years, single-domain antibodies (sdAbs), also known as nanobodies, are derived from the heavy chain antibody variable region (VHH) in camelids and are universally preferred over traditional antibodies [3]. Compared with traditional antibodies, nanobodies exhibit more attractive features for diagnostic application, such as small volume (15 kDa), easy genetic manipulation, and high stability [4][5][6][7][8][9]. Recently, nanobodies have been fused with horseradish peroxidase (HRP) for the The 217 negative sera for anti-PRRSV antibodies were obtained from the healthy pigs and were veri ed to be negative via a commercial ELISA kit (IDEXX, Westbrook, ME, USA). Serum samples from the pigs challenged with PRRSV HuN4, SD16, and CH-1R strains were used to evaluate the cELISA assay [25][26][27].
To determine the coincidence rate of cELISA, 381 challenged sera samples were collected from our previous animal experiments [28,29]. Meanwhile, 450 clinical sera samples were collected from various farms in Shandong and detected using both the commercial ELISA kit and cELISA.

Animal experiments
Nine four-week-old piglets were obtained from a PRRS-free farm and tested for negative PRRSV via the detection of real-time RT-PCR and anti-PRRSV antibodies [30,31]. The piglets were randomly divided into three groups (3 pigs per group), which were separately raised in different isolation rooms. Group 1 piglets were intranasally administered with 1 × 10 6.5 TCID 50 of PRRSV NADC30-like strain, group 2 with 2 × 10 5 TCID 50 of PRRSV GZ11-G1 strain, and group 3 with 2 mL MARC-145 cell culture supernatant as the negative control. Serum samples were collected at 0, 1, 3, 5, 7, 10, 14, 21 and 28 days post-inoculation (dpi) then used for the detection of antibodies against PRRSV-N protein by cELISA and a commercial IDEXX ELISA kit.

Bactrian camel immunization and library construction
A four-year-old male Bactrian camel was immunized subcutaneously with the killed PRRSV (CH1a strain) with six times [32]. Every two weeks, the camel was immunized once and after six immunizations, the serum samples from the Bactrian camel was collected and detected for anti-PRRSV antibodies using iELISA. Four days after the last immunization, peripheral blood mononuclear cells (PBMCs) were isolated from 200 mL blood sample by Leucosep® tubes (Greiner Bio-One, Frickenhausen, Germany). Total RNA was extracted from the 1 × 10 7 PBMCs and reverse transcribed into cDNA. Then, the Camellidae heavy chain-only antibodies (VHH) genes were ampli ed using the cDNA as a template by nested PCR, as described previously [33]. The rst PCR products (~ 700 bp) ampli ed with the CALL001 and CALL002 primers (Table 1) were puri ed according to the instructions of the EasyPure Quick Gel extraction kit (TransGen Biotech, Beijing, China). The second PCR with primers VHH-FOR and VHH-REV (Table 1) was ampli ed using the rst puri ed PCR products as the template. The nal puri ed PCR products (~ 400 bp)   were ligated into the phagemid vector pMECS with Not I and Pst I enzymes sites by T4 DNA ligase (NEB,  Ipswich, MA, USA). Then, the ligation products were electro-transformed into competent E. coli TG1 cells. Cells were cultured on LB agar plates with 2% glucose and 100 µg/mL ampicillin at 37°C overnight.
Subsequently, the bacterial colonies were scraped from the plates, re-suspended in phosphate buffer saline (0.01 M PBS, pH 7.2) to prepare the VHH phage library against PRRSV, and stored at -80°C. The ORF7 gene encoding the PRRSV-N protein was ampli ed using an infectious PRRSV cDNA clone pBAC-SD16 as the template [34]. The PCR products were puri ed and cloned into the pET-28a prokaryotic expression vector (Novagen, Darmstadt, Germany). After sequencing, the recombinant positive plasmid was named pET28a-N. The primers of PCR ampli cation are listed in Table 1. After the pET28a-N plasmids were transformed into E. coli BL21 (DE3) (TransGen Biotech, Beijing, China), the recombinant PRRSV-N protein was expressed and puri ed based on the previous descriptions [35]. Brie y, the positive bacteria were induced with 0.1 mM isopropyl-β-thiogalactopyranoside (IPTG) for 6 h at 37°C. The bacteria were collected and re-suspended in Buffer A (20 mM Tris, 300 mM NaCl, pH 7.2-7.4). The supernatant of bacterial solution was collected after the induced bacteria were ultrasonicated and centrifuged. Subsequently, the supernatant containing recombinant PRRSV-N protein was puri ed using a Ni-NTA column (Roche, Mannheim, Germany) and eluted with Buffer B (20 mM Tris, 300 mM NaCl, and 250 mM imidazole, pH 7.2-7.4). Finally, the expression, puri cation, and antigenicity of the recombinant PRRSV-N protein were analyzed by SDS-PAGE and Western blot with the positive pig sera for PRRSV.
Screening and identi cation of PRRSV-N speci c nanobodies The PRRSV-N protein speci c nanobodies were screened by three rounds of panning using phage display technology, as previously described, with the following modi cations [32]. Brie y, the VHH phage library was rescued via M13K07 helper phage. The 96-well plates (Nunc, Denmark) were coated with the recombinant PRRSV-N protein (4 µg/well) overnight at 4°C for the three rounds of panning. The coated plates were blocked with 200 µL of 2.5% skim milk at 37°C for 1 h and washed with 0.05% PBS'T (1 L PBS with 0.5 mL Tween-20). Then, the above rescued recombinant phage (5 × 10 11 pfu/mL) were added to the plates and incubated at room temperature (RT) for 1 h. After the plates were washed again, the binding phages were eluted using 100 µL 0.1 M trimethylamine (Sigma, St. Louis, MO, USA) and neutralized with same volume of 1 M Tris-HCl (pH 7.4). Subsequently, the growth log phase of E. coli TG1 was infected with the eluted phages and ampli ed for further rounds of selection. The enrichment of speci c phage particles was analyzed using anti-M13/HRP conjugate ELISA combined with phage titration after three rounds of panning. Finally, the 96 colonies were picked randomly and induced with IPTG (1 mM) to express soluble VHHs with an HA-Tag. These recombinant VHHs-HA-Tag proteins were extracted and tested for their capacity to recognize the PRRSV-N protein using iELISA with an anti-HA-Tag antibody as the rst antibody (GenScript, Biotech Corp, China). Finally, the positive clones were sequenced, and the nanobodies were grouped according to the hypervariable complementary-determining region 3 (CDR3) sequence.

Establishment of HEK293S cell lines stable expression of nanobody-HRP fusion protein
To select the best nanobody to construct the stably expressed cells for producing the nanobody-HRP fusion protein, the different fusion proteins were rst expressed with transient transfection. The recombinant plasmids were constructed based on the previous descriptions [10,36]. The VHH gene was ampli ed using primers Nb-F and Nb-R (Table 1) with pMECS-VHH plasmid as the template. Then, the PCR products and pCMV-N1-HRP vector were both digested via the Pst I and Not I enzymes and ligated with T4 ligation enzymes to create the recombinant pCMV-N1-Nbs-HRP plasmids. Next, the HEK293T cells were transfected with the recombinant plasmids to produce the nanobody-HRP fusion proteins using polyetherimide (PEI, Warrington, USA) agents. The cell supernatant containing nanobody-HRP fusion proteins were collected after transfection for 60-72 h. The cELISA procedure was used to select the best nanobody, which is described below. The highest percent competitive inhibition (PI) values of the nanobody-HRP fusion protein were select for constructing the stably expressing cells.
The platform of HEK293S cell line stably expressing the nanobody-HRP fusion proteins was designed as following: the secreting signal sequence, an HA tag, VHH, HRP and His tag sequence were obtained from the pCMV-N1-Nbs-HRP with the digestion of enzymes EcoR I and BamH I. Then, the gene was ligated into pLVX-IRES-ZsGreen lentivirus vector digested with the same two enzymes. To produce lentivirus particles, the HEK293T cells were co-transfected with pLVX-IRES-ZsGreen-Nb-HRP, psPAX2 and pMD2.0G plasmids using X-tremeGENE HP DNA Transfection Reagent (Roche, Basel, Switzerland) according to the manufacturer's instructions. After transfection for 60 h, the packaging lentivirus was observed under a uorescence microscope, and the cell culture supernatant was collected. HEK293S cells were transduced using the above recombinant lentiviruses and supplemented with 10 µg/ml of PolyBrene (Sigma, St. Louis, MO, USA). After 48 h, the transduced cells were observed under a uorescence microscope. Then, HEK293S cells with green uorescence were sorted by High-speed sorting ow cytometer (BD, US).
Biological activity analysis of nanobody-HRP fusion proteins produced by the two systems The biological activity (titres and stability) of nanobody-HRP fusion proteins produced by the transient transfection and stable expressing system were compared. Firstly, the titers of the nanobody-HRP fusion proteins from the two systems were tested with direct ELISA using a checkerboard titration. Different amounts of the PRRSV-N protein, i.e. 100, 200, 400, and 800 ng/well, were coated into the 96-well plates, and the different dilution ratios of the nanobody-HRP fusions (1:10, 1:100, 1:500, and 1:1000) were tested. The titer was assessed when the OD 450nm value of the direct ELISA was 1.0, and the stability of the fusion proteins produced by the two systems was also evaluated. The two methods were independently repeated ve times for producing the fusion proteins. Then, the fusion proteins from the different production batches were detected using direct ELISA with a 1:100 dilution. Subsequently, the stability of fusion proteins from the stable expression system was tested from the second, forth, sixth, eighth, and tenth generations and the supernatants from these generations were detected using direct ELSIA with dilutions of 1:10, 1:100, and 1:1000. In addition, the operations of the two systems were compared based on the procedures for producing the fusion proteins.

Development of competitive ELISA using nanobodyHRP fusion proteins for detecting anti-PRRSV antibodies
The cELISA was developed using the nanobody-HRP fusion proteins as a probe according to a reported procedure [10]. Firstly, the optimal amount of antigen and dilution of fusion protein were determined using a checkerboard titration test with direct ELISA. Different amounts of the PRRSV-N protein (100, 200, 400, and 800 ng/well) were coated into the 96-well plates, then the dilution ratios of fusion proteins of 1:10, 1:100, 1:500, 1:1000 were tested. Finally, the optimal amount of antigen and fusions were selected when the OD 450nm value of the direct ELISA was 1.0 and the amount of coated antigen was the lowest.
Secondly, the optimal dilution ratio of pig sera was determined. Five separate positive and negative pig sera were diluted at 1:10, 1:20, 1:40, 1:80, and 1:160 and tested with the cELISA. The optimal serum dilution was determined according to the smallest ratio of OD 450nm values between the positive and negative sera (P/N). Finally, the times of incubation and color reaction after the addition of tetramethylbenzidine (TMB) were separately optimized. The incubation times of the mixtures containing the nanobody-HRP fusions and the positive or negative sera with coated PRRSV-N protein were tested at 20, 30, and 40 min. After incubation, TMB was added to color and tested after 10, 15, and 20 min. The smallest ratio of OD 450nm values between the positive and negative sera was selected as the optimal incubation and colorimetric reaction times.
After optimizing the above conditions, cELISA was performed as follows. (1) The 96-well ELISA plate was coated with the optimal amount of PRRSV-N recombinant protein and incubated overnight at 4°C. (2) The plate was blocked with 200 µL 2.5% (w/v) non-fat dry milk in PBS'T at room temperature (RT) for 1 h after washed three times with PBS'T. (3) After washed with PBS'T again, each well was added into 100 µL of testing mixtures containing the optimal dilutions of serum sample and nanobody-HRP fusions in 2.5% (w/v) non-fat dry milk, then incubated for optimal times at RT. (4) After the plate was washed again in the same way, TMB (100 µL/well) was added and incubated for optimal times at RT. (5) Finally, 3 M H 2 SO 4 (50 µL/well) was used to stop the colorimetric reaction, and the OD 450nm values were read using an automated ELISA plate reader (Bio-Rad, USA).
Determination of cut-off value, sensitivity, speci city and repeatability of the cELISA The PI values were calculated with following formula: PI (%) = [1-(OD 450nm value of testing serum sample/OD 450nm value of negative sample)] × 100%. The 217 negative pig serum samples for anti-PRRSV antibodies were used to determine the cut-off value. After these sera were detected using the developed cELISA, the cut-off value was calculated by the mean PI of 217 negative serum samples plus 3 standard deviations to ensure 99% con dence for the negative sera samples within this range [37].
The sensitivity of cELISA was evaluated by testing sera from the different dpi of the three challenged NADC30-like PRRSV pigs as well as the 164 PRRSV-clinical positive sera con rmed by the commercial ELISA kit. In addition, double dilutions (from 1:10 to 1:5120) of ve positive pig sera for anti-PRRSV antibodies were tested using cELISA to determine the lowest detection dilution.
The speci city of the cross-competing assay was assessed between the nanobody-HRP fusions and antibodies against other swine viruses, including porcine parvovirus (PPV), porcine circovirus type 2 (PCV2), porcine pseudorabies virus (PRV), transmissible gastroenteritis virus (TGEV), porcine epidemic diarrhea virus (PEDV), and swine in uenza virus (SIV). Standard positive sera for anti-the other swine viruses antibodies were con rmed by the commercial ELISA kit. Total 164 PRRSV-clinical negative sera were also tested with the cELISA. To further con rm whether cELISA can detect anti-genotype 1 PRRSV antibodies, the sera from the pigs pre-and post-challenged with GZ11-G1 strain (genotype 1) were tested. Meanwhile, the sera samples from the pigs challenged with HuN4, SD16, and CH-1R strains of genotype 2 PRRSV strains were evaluated to determine whether cELISA can detect antibodies against different genotype 2 PRRSV isolates.
To determine the reproducibility of cELISA, eight separate positive and negative clinical pig serum samples were tested and used to perform the intra-assay and inter-assay variabilities. The inter-assay variation (between plates) and intra-assay variation (within a plate) were evaluated by the coe cient of variation (CV). Each sample was tested using three different plates to determine the inter-assay CV, while three replicates within each plate were used to calculate the intra-assay CV [38].

Comparisons of competitive ELISA with commercial ELISA kit
To evaluate the coincidence of cELISA with the commercial ELISA kit, 381 serum samples from challenged pigs with PRRSV and 450 clinical pig serum samples were tested with each method and analyzed via SPSS software. Among these sera, the results reveal inconsistencies between the two detection methods for these, thus IFA veri cation was subsequently performed.

Statistical analysis
Statistical analysis was performed using GraphPad Prism version 5.0 (GraphPad Software, San Diego, CA, USA). Student's t-test and Kappa index values were calculated to estimate the platform for HEK293S cell lines stably expressing nanobody-HRP fusion protein, as well as the coincidence between cELISA and the commercial ELISA kit. Repeatability was assessed using CV (CV = SD/Mean), where a CV value less than 15% for the intra-plate assay was considered an acceptable repeatability level for the assay. These calculations were performed using SPSS software (Version 20, http://www.spss.com.cn).

Results
Expression and puri cation of the recombinant PRRSVN protein.
SDS-PAGE analysis showed that the recombinant PRRSV-N protein was successfully expressed in soluble form with the expected size of 17 kDa (Fig. 1a). In addition, the high purity of the target protein was obtained with the Ni-resin puri cation (Fig. 1a). To further identify the expression and antigenicity of the target protein, the results of Western blot indicate that the recombinant PRRSV-N protein can react with the positive pig sera for anti-PRRSV antibodies (Fig. 1b). The puri ed recombinant PRRSV-N protein was used as the coating antigens to screen speci c nanobodies and develop the cELISA.
Screening and identi cation of anti-PRRSV-N protein speci c nanobodies A phage display VHH library consisting of approximately 3.2 × 10 8 individual colonies was constructed from PBMCs of the immunized camel. Positive rate analysis by colony PCR revealed that 96% of these colonies contained a correct insert corresponding to the size of VHH genes. Then, 50 randomly clones were selected, sequenced, and analyzed. The results show that each clone was manifested to contain a distinct VHH sequence (data not shown), suggesting the good diversity and high quality of the library.
After three rounds of biopanning, the speci c VHHs phage particles against PRRSV-N protein were enriched ( Table 2). 96 mono-clones were randomly selected and expressed periplasmic extracts from the third round of screening for further iELISA detection. The results reveal that 36 individual colonies were identi ed for speci c binding to the PRRSV-N protein (Fig. 2a). Subsequently, the above 36 colonies were sequenced, and three different PRRSV-N speci c nanobodies were screened based on the amino acid sequence classi cation of the CDR3 hypervariable region (named PRRSV-N-Nb1, -Nb2, and -Nb3). The deduced amino acid sequences of the three nanobodies were aligned with the human VH sequence, for which the numbering and CDRs follow the method described by Kabat et al. [39]. Alignment results suggest that PRRSV-N-Nb1 and -Nb2 have typical hydrophilic amino acid substitutions in the framework-2 regions Val37, Gly44, Leu45, and Trp47 (located on the VH-VL interface region of VHs) (Fig. 2b). In addition, iELISA results showed that the three nanobodies could speci cally bind to recombinant PRRSV-N protein but could not cross-react with the NDV-NP recombinant protein (Fig. 2c). Rather, the recombinant NDV-NP protein was expressed using the same vector pET-28a, which also had a 6×His-Tag, excluding the possibility that these nanobodies might recognize the 6×His region. Moreover, the titres of the periplasmic extracts of PRRSV-N-Nb1 and -Nb2 were higher than that of PRRSV-N-Nb3 (Fig. 2d). The results of Western blot and direct ELISA reveal that the PRRSV-Nb1-HRP, -Nb2-HRP, and -Nb3-HRP fusion proteins were successfully expressed in the form of secretion ( Fig. 3a and b). The three fusions can speci cally bind to the PRRSV-N protein (Fig. 3c), indicating that the fusions do not change the antigenic reaction of the nanobody. However, the titres of the supernatant containing the PRRSV-N-Nb1-HRP and -Nb2-HRP fusion proteins were signi cantly higher than that of PRRSV-N-Nb3-HRP (Fig. 3b).
Therefore, the PRRSV-N-Nb1-HRP and -Nb2-HRP fusion proteins were chosen to further evaluate cELISA. Compared with the two fusions blocked by the positive pig sera for anti-PRRSV antibodies to bind the antigen, cELISA results indicate that the blocking rate of PRRSV-N-Nb1-HRP was higher than that of PRRSV-N-Nb2-HRP (Fig. 3d). Therefore, PRRSV-N-Nb1 was selected to construct the stably expressed cells for producing the PRRSV-N-Nb1-HRP fusion protein and for further development of cELISA.
In order to conveniently and quickly produce fusion proteins, the HEK293S cell lines stably expressing PRRSV-N-Nb1HRP fusion proteins were successfully established. The positive recombinant HEK293S cells were observed under a uorescence microscope (Fig. 3e), then the cell supernatant was collected and analyzed for the antigenic activity of PRRSV-N-Nb1-HRP fusion proteins using cELISA. As shown in Fig. 3f, the stably expressed fusion proteins from the HEK293S cells can be still blocked to bind the antigens by the positive sera, which is consistent with the expression by transient transfection.

Comparisons of the two platforms of transient transfection and stable expression
Compared with the transient transfection system, titers of the PRRSV-N-Nb1-HRP fusion protein from the recombinant HEK293S cell lines were higher (Fig. 4a). In addition, with 100 ng/well PRRSV-N protein, the OD 450nm value was approximately 1.0 with the dilution of 1:100 for the fusion from stably expressed cells (Fig. 4a). However, the OD 450nm value reached 1.0 at 200 ng/well of coated antigens with the same dilution for the transient transfection (Fig. 4a). The two systems were independently repeated ve times. Although signi cant differences were observed for the three independent experiments of the transient transfection system, no signi cant difference was noted for stably expressed cells (Fig. 4b). This suggests a greater stability of the expressed recombinant HEK293S cell lines than the transient system.
Moreover, the recombinant HEK293S cell lines for stable expression of PRRSV-N-Nb1-HRP fusion protein could be passed continuously for 8 generations with no difference in titers (Fig. 4c).
In the procedure of the stable expression system producing PRRSV-N-Nb1-HRP fusion proteins, cells were cultured for 48 h then the supernatant was collected for direct used (Fig. 4e). However, the transient transfection system required plating, while plasmids were extracted for each time then transfected into cells. After 48-72 h of transfection, the supernatant was collected for direct use. This production cycle took approximately 132 h (Fig. 4d), and the system required extra costs for the plasmid extraction kit and transfection reagent. The operation procedures of the two platforms indicate that the stably expressed platform is less complex and less costly than the transient transfection system.
Competitive ELISA using the PRRSV-N-Nb1HRP fusion proteins as reagents The optimal concentration of coated PRRSV-N proteins was determined to be 100 ng/well, and the optimal dilution of PRRSV-N-Nb1HRP fusion proteins was identi ed as 1:100 using a checker board titration assay ( Table 3). The optimal dilution of the tested pig serum sample was determined as 1:20 based on the different dilutions of 5 positive and negative sera producing the lowest P/N ( Table 4). The optimized incubation time of the sera and PRRSV-N-Nb1-HRP fusion protein mixtures was found to be 30 min, and the optimal colorimetric reaction time was 15 min (Table 5).  Note: Five positive and negative sera were tested using cELISA. The best dilution was selected when the OD 450nm values of positive to negative (P/N) sera was smallest. Note: The best competition time and colorimetric reaction time was also selected when the OD 450nm values of positive to negative (P/N) sera was smallest.
To determine the cut-off value of cELISA, the results reveal that the average PI (X) value of 217 negative sera was 2.49% with an SD of 6.9%. The cut-off value of cELISA was determined to be 23.19% (2.49% + 3SD). Therefore, the PI of pig serum sample ≥ 23.19% is considered positive, while PI < 23.19 % is negative.
Sensitivity, speci city and reproducibility of the competitive ELISA The sera from the pre-and post-challenged pigs with NADC30-like PRRSV and 164 positive clinical serum samples were tested to assess the sensitivity of cELISA. Seropositivity was rst observed at 5 dpi in one of the three pigs, and all sera were still positive for anti-PRRSV antibodies until 28 dpi (Fig. 5a). Comparatively, the seropositivity was rst observed at 7 dpi with the commercial IDEXX ELISA kit (Fig. 5b). These results suggest a higher sensitivity of cELISA compared to the commercial IDEXX ELISA kit. For the 164 positive clinical serum samples, the PI values of 109 samples were greater than 80%, and only 8 samples had PI values from 23.19-30% (Fig. 5c). These results indicate that the sensitivity of cELISA for testing clinical pig sera was 100%. For the different dilutions of the 5 positive pig sera using cELISA, sera at a dilution of 1:1280 were negative, and those at 1:320 were positive (Fig. 5d).
To determine the speci city of cELISA, antisera against other swine viruses, including PPV, PCV2, PRV, TGEV, PEDV and SIV, were tested using cELISA, using 6 PRRSV positive sera samples as the positive control. According to the results, the PI values of 6 positive serum samples were 79%-91%, while the PI values of antisera against other swine viruses were 1%-19% (Fig. 6a). Furthermore, the 164 negative sera were detected using the cELISA with PI values ranging 1%-20% (Fig. 6b). To further evaluate whether cELISA can test anti-genotype 1 PRRSV antibodies, sera from the pre-and post-challenged pigs with GZ11-G1 strain (genotype 1) were examined. The results reveal that all sera, until 28 dpi, were positive via detection with the commercial IDEXX ELISA kit, but all were negative using the developed cELISA (Fig. 6c). Meanwhile, the sera were also tested from the pre-and post-challenged pigs with HuN4, SD16, and CH-1R strains of genotype 2 PRRSV. Accordingly, seropositivity was rst observed at 7 dpi, and all sera remained positive for anti-PRRSV HuN4, SD16 and CH-1R strains of antibodies until 28 or 21 dpi ( Fig. 7d-f). Taken together, these results con rm that the developed cELISA can speci cally detect antigenotype 2 PRRSV antibodies.
To analyze the reproducibility of cELISA, eight separate positive and negative clinical serum samples were tested and used to evaluate the intra-assay and inter-assay variabilities. The intra-assay CV of the PI were analyzed in the range of 0.55%-4.64% with a median value of 2.6%, while the range for the interassay CV was 1.57%-9.53% with a median value of 5.55% (Table 6). These data indicate that the cELISA method exhibits good reproducibility. Inter assay precision (CV%) 1.57-9.53 5.55 Note: Intra assay precision: Determined from three repetitions (well-to-well) of 8 serum samples in the same method. Inter assay precision: Determined from three repetitions (plate-to-plate) at different time.
Agreements of competitive ELISA and commercial IDEXX ELISA kit To assess the consistency of cELISA, 381 pig sera from the challenged pigs at different dpi (0-28 dpi) were tested with both cELISA and commercial IDEXX ELISA kit. The results of both methods coincided in 378 (205+/173-) of the 381 serum samples with an agreement rate of 99.2% (Kappa = 0.98). In addition, for the 450 clinical pig sera collected from various farms in Shandong, an agreement rate of 96.4% (Kappa = 0.82) was determined for the two detection methods (Table 7). Statistical analysis further indicates that cELISA had a high level of consistency with the commercial IDEXX ELISA kit, and no signi cant differences were observed between cELISA and the commercial IDEXX ELISA kit (Kappa values > 0.4) ( Table 7). Next, the sera with inconsistent results were veri ed by IFA. The results showed that the three sera with inconsistent results among 381 challenged sera were positive for IFA, positive for cELISA, and negative for the commercial IDEXX ELISA kit (Fig. 7a). As shown in Fig. 7b and c, 16 of 450 clinical sera had inconsistent results. Among them, four serum samples were positive using cELISA, negative using commercial IDEXX ELISA kit, and positive for IFA (Fig. 7b). The remaining twelve sera were negative for cELISA and positive for the commercial IDEXX ELISA kit, while IFA revealed seven positive sera and ve negative sera (Fig. 7c). It can be suggested that the seven serum samples were likely to be genotype 1 PRRSV. While cELISA has a high consistency with the commercial IDEXX ELISA kit, and it exhibits a higher sensitivity, which is promising for clinical testing.  The kappa value > 0.4 was regarded as signi cant difference

Discussion
The sensitivity and speci city of different ELISA methods can be determined by using antibodies are critical reagents [40]. Despite the use of traditional antibodies, including polyclonal and monoclonal antibodies, for developing ELISA [41], these antibodies have high production costs and require enzyme labelling [42]. In the latest research, nanobodies have acquired increased attention as the smallest known antigen binding antibody with simple genetic manipulation and, thus, as a promising new generation antibody for diagnostic applications [43,44]. Therefore, nanobody-HRP fusion proteins have been designed and used to develop ELISA for detecting antibodies against different virus. Compared to traditional antibodies for the commercial ELISA kit, the nanobody-HRP fusion protein is simple and inexpensive to produce and does not require puri cation or enzyme-labelling. In this study, a nanobody-HRP fusion protein against anti-PRRSV-N protein was produced and used for the rst time as a probe to develop a cELISA for detecting anti-PRRSV antibodies in pig sera. The procedures were performed according to a previous study but with some modi cations [36].
In a previous work, a platform was constructed to produce the nanobody-HRP fusion proteins via transient transfection of HEK293T cells. However, the production of the platform was time-consuming, laborious, costly, and not suitable for mass production [10]. To solve these problems, another platform using HEK293S cell lines for stably expressing the nanobody-HRP fusion proteins was constructed in the present study, particularly avoiding the trouble of each transfection. In addition, the titers of nanobody-HRP fusion proteins in the supernatant from the recombinant HEK293S cell lines were found to be than those of the compared transient transfection system. Moreover, produced nanobody-HRP fusion proteins from different batches were stable using the stable expression system. In conclusion, the procedures of the stable expression system are simple and easy for mass production PRRS is one of the most common and economically-important infectious diseases of swine globally. Clinical signs of PRRS are not characteristic, and sometimes, the course of PRRSV infection is subclinical, thus laboratory detection methods are necessary for diagnosis. At present, ELISA is the most popular method that is also used to monitor the antibody level on a population. Among the available commercial ELISA kits for detecting anti-PRRSV antibodies, the IDEXX PRRS X3 Ab Test is the most widely used and generally recognized as the de facto gold standard [45]. However, this commercial iELISA kit requires the use of a second antibody and, as such, is expensive for mass clinical application.
Comparatively, the cELISA based on the nanobody-HRP fusion protein developed in the present study demonstrated simple and low-cost production. In addition, the sensitivity of cELISA is higher than the commercial ELISA kit. More importantly, the developed cELISA has a high agreement with the IDEXX PRRS X3 Ab Test. These advantages suggest that the developed cELISA has a good prospect of market application and promotion.
In the present study, the developed cELISA can detect the antibodies against different genotype 2 PRRSV isolates but not genotype 1 PRRSV. This suggests that the epitope recognized by PRRSV-N-Nb1 may be only located in the N protein of genotype 2 PRRSV. The homology of the amino acid sequence of the genotype 2 PRRSV N protein ranges between 96%-100% [46], indicating that the assay may detect antibodies against most of the PRRSV genotype 2 strains. In the future, more experiments will be needed to determine the key amino acids of the epitope and to further analyze the epitope of amino acid from different PRRSV isolates. Such study will reveal whether the antibodies against all the genotype 2 PRRSV isolates can be detected by the developed cELISA. According to our knowledge, the commercial IDEXX ELISA kit can detect antibodies against both genotype 1 and 2 PRRSV. Therefore, utilizing this kit following the developed cELISA may be advantageous for the differential diagnoses of genotype 1 and 2 PRRSV.
Although the developed cELISA has simple operation, low production cost, and good sensitivity and speci city, the construction of this platform is highly complicated. Speci cally, it requires the immunization of camels, screening of functional nanobodies, and establishing a cell line stably expressing nanobody-HRP fusion proteins. This series of tasks is complicated and demands very skilled experimenters to operate, despite cELISA test only taking 45 min to run. In short, the establishment of this platform is of complicated. Yet, it can be noted that the platform is established successfully, and it becomes very convenient in production and clinical application.

Conclusions
In the present study, three speci c nanobodies against the PRRSV-N protein were screened and identi ed.
Based on these nanobodies, a simple and fast platform for the production of nanobodies-HRP fusion protein was designed. Then, the fusion proteins were utilized as a reagent to develop the cELISA method for detecting anti-genotype 2 PRRSV antibodies in pig sera. The simple operation and low-cost production of nanobody-HRP fusion proteins as an ultrasensitive probe will be a promising tool for developing diagnostic kits for various diseases.       Detection of the pig serum samples with inconsistent results between the developed cELISA and commercial IDEXX ELISA kit by IFA. a Three inconsistent results from 381 challenged sera were tested by IFA, which were positive for cELISA and negative for the commercial IDEXX ELISA kit. b Four inconsistent results from the clinical pig sera were tested by IFA, which were positive for cELISA and negative for the commercial IDEXX ELISA kit. c The remaining twelve sera tested negative using cELISA, but positive using the commercial IDEXX ELISA kit.

Figure 8
Scheme 1 Schematic representation of developing the cELISA to detect PRRSV antibodies using the nanobody-HRP fusion proteins as a reagent. a The platform for stably expressing nanobody-HRP fusion proteins using HEK293S cells. b Competitive ELISA for using the fusion protein as a reagent.