Development and application of a competitive ELISA for the detection of antibodies against Salmonella Abortusequi in equids

ABSTRACT The high abortion rate associated with Salmonella Abortusequi (S. Abortusequi) infection in equids has re-emerged over the past 10 years and has caused serious economic losses to China. Our previous studies showed that the flagellin FljB gene could distinguish S. Abortusequi from most Salmonella serotypes. In this study, the flagellin antigen was used to develop a competitive enzyme-linked immunosorbent assay (cELISA) that could be used to detect both horse and donkey serum samples using a monoclonal antibody (MAb) that was found to bind to FljB. A cELISA was established using the purified MAb coating of the plate and incubation of the mixture of horseradish peroxidase (HRP)-conjugated FljB antigen with the undiluted serum sample. The performance of the cELISA and the tube agglutination test (TAT) assay was compared with respect to sensitivity and specificity, by testing a panel containing 660 S. Abortusequi-positive and 515 S. Abortusequi-negative serum samples, all of which had been characterized by Western blotting. Receiver operator characteristic (ROC) analyses were performed to determine the cutoff value and estimate the detection specificity (Sp) and sensitivity (Se). ROC analysis showed that the area under the ROC curve (AUC) values of cELISA [AUC = 0.9941; 95% confidence interval (CI), 0.9898–0.9984] were higher than those of TAT (AUC = 0.7705; 95% Cl, 0.7437–0.7972). A cutoff value of 39.5% was selected with Sp and Se values of 100 (95% Cl, 99.26–100.00) and 97.58 (95% Cl, 96.10–98.50), respectively. The cELISA has excellent futures compared with TAT, such as shortened detection time, no need for pre-treatment of sera, and easy interpretation of the results, and is more suitable for disease surveillance.

in equids has re-emerged over the last decade, and currently, the abortion rate in infected mares is between 30% and 100% (24).The significant economic losses in equine husbandry in China are mainly due to the growing number of abortions caused by S. Abortusequi infection (25).Moreover, etiological data based on bacteria isolation have provided evidence of the widespread circulation of S. Abortusequi among horse and donkey farms in Inner Mongolia, Xinjiang, Shandong, Heilongjiang, and Hebei provinces in China (24)(25)(26).
The gold standard for detecting carriers of S. Abortusequi is the detection of the live agent, but this has the disadvantage of being both time-consuming and laborious (27).In addition, vaginal swab samples, fetal tissues, or fecal detection can result in false negatives even in farms at which an outbreak of S. Abortusequi infection has occurred (28).In this case, serological assay detection is considered to be a good complementary method to detect S. Abortusequi infections.The serological assay has excellent futures compared with bacteria isolation, such as shortened detection time, and can be used for monitoring asymptomatic animals.Furthermore, it can be used as a tool for large-scale serological surveys for the detection of antibody against S. Abortusequi.Currently, the most common method for measuring serum antibody against S. Abortusequi is the tube agglutination test (TAT) (15,23).However, the World Organization for Animal Health (WOAH) reports that the TAT is not highly specific and may cross-react with antibodies from other Enterobacterales (29).
Another option for the detection antibodies against S. Abortusequi is the enzymelinked immunosorbent assay (ELISA) assay and includes both indirect (iELISA) and competitive (cELISA) ELISAs.The main advantage of cELISAs over iELISAs is that the cELISAs for the detection of antibodies enable the testing of samples from different animal species.To the best of our knowledge, no specific cELISA serological method capable of the specific identification of S. Abortusequi has been published to date.Previous studies have identified flagellin as a potential candidate for the serological detection of Salmonella, as the protein is abundantly expressed, located on the surface of the bacteria, and has good antigenicity (1,(30)(31)(32)(33).Previous research has resulted in an iELISA based on recombinant flagellin for the detection of antibodies against S. Enteritidis and has demonstrated that the assay has good specificity and did not cross-react with other bacteria (1).Our previous studies showed that the flagellin FljB in nucleotide or amino acid could distinguish S. Abortusequi from most Salmonella serotypes (24,34).In this study, a cELISA based on a monoclonal antibody (MAb) against flagellin FljB, for the detection of S. Abortusequi antibodies in the sera of different animal species was developed and evaluated.

Serum samples
A total of 1,175 sera including 660 S. Abortusequi positive and 515 S. Abortusequi negative serum samples were used in this study (Table 1).One standard positive serum was collected from a horse naturally infected with S. Abortusequi in Heilongjiang for the development of the cELISA.One positive serum from a rabbit experimentally immunized with S. Abortusequi was purchased from the China Institute of Veterinary Drugs Control.The other positive serum samples were collected from farms where S. Abortusequi infection is known to have occurred in Inner Mongolia, Xinjiang, Shandong, and Hebei provinces in China, during 2018-2022.Most of the samples were collected from horses and donkeys in the farms with a clinical abortion caused by S. Abortuse qui infection, which were confirmed with bacteria isolation and Western blotting.Ten antisera positive against equine herpes virus (EHV), equine influenza virus (EIV), equine arteritis virus (EAV), equine infectious anemia virus (EIAV), Theileria equi, Babesia caballi, Streptococcus equi, S. Typhimurium, S. Dublin, and S. Enteritidis were selected for the analytical specificity test.Antisera against EHV, EIV, or EAV were purchased from National Veterinary Services Laboratories Inc in the USA.Antisera against EIAV, S. equi, T. equi, B. caballi, S. Typhimurium, S. Dublin, or S. Enteritidis were prepared using healthy horses and stored in our laboratory, which were tested specific to the mentioned pathogens using commercial antibody detection kits.Negative serum samples were collected from healthy horses or donkeys in Xinjiang, Chongqing, Gansu, Henan, Jiangsu, and Hubei provinces.

Bacterial strains
To better characterize the MAbs and to determine both the binding capability of the MAbs versus S. Abortusequi from different farms and the extent of any cross-reactions against other Salmonella serotypes, four Salmonella serotypes were used in this study.
The five different S. Abortusequi strains used in this study were isolated by our labora tory from the tissues of clinically aborted foals.The other three Salmonella serotypes, including S. Typhimurium, S. Enteritidis, and S. Dublin, were kind gifts from Prof. Guoqiang Zhu at the College of Veterinary Medicine, Yangzhou University (Yangzhou, Jiangsu, China).

MAb production and characterization
S. Abortusequi recombinant FljB protein was produced as previously described (1).MAbs against FljB were produced through the cell fusion method, with methods that were performed as previously described (35).One Balb/c mouse was primed subcutaneously with purified FljB in Freund complete adjuvant and was boosted intraperitoneally with the same antigen in phosphate-buffered saline (PBS) 3 weeks later.Hybridoma cells were screened using iELISA.The specificity of the MAbs was further evaluated using iELISA, to compare the pattern of reactivity of the generated MAbs with the S. Abortusequi and a spectrum of Salmonella of different serotypes.cELISAs were used to analyze the reciprocal competi tion between the selected MAbs and the natural S. Abortusequi infections in horse sera.Direct ELISAs were set up to evaluate the capability of coated MAbs to efficiently bind the HRP-conjugated FljB antigen.

ELISA for MAb characterization
MAbs were characterized through iELISA, direct ELISA, and cELISA performed as described previously (36).For the iELISAs, five S. Abortusequi strains and three other Salmonella strains, including S. Typhimurium, S. Enteritidis, and S. Dublin, were adsorbed onto microplates by overnight incubation at 4°C; the plates were then sequentially incubated with hybridoma supernatants and HRP-conjugated goat anti-mouse IgG antibodies for 45 min at 37°C.All washing steps were performed three times with washing buffer (PBS containing 0.1% Tween 20, PBST).After incubation, the plate was washed again and incubated with freshly prepared 3,3′-5,5′-tetramethyl benzidine (TMB) peroxidase substrate (GalaxyBio, Beijing, China) for 10 min at 37°C.The reaction was stopped by adding 2-M H 2 SO 4 , and the optical density at the 450-nm wavelength (OD 450 ) was measured using the VersaMax Microplate Reader (BioTek, Winooski, VT, USA).
For the competitive binding assays, cELISAs were designed to analyze the capability of sera naturally infected with S. Abortusequi to inhibit the binding of MAbs to S. Abortusequi FljB protein.One hundred-microliter samples of S. Abortusequi sera were incubated for 45 min at 37°C in FljB antigen-coated plates, and, after washing, 50 µL of hybridoma supernatant was added to each sample.Binding of the MAb was assessed with HRP-conjugated anti-mouse IgG, and the colorimetric reaction was performed as previously described.The competitively reactive MAbs were further evaluated.Briefly, microplates were coated overnight at 4°C with different MAbs at 2-µg/mL concentration, followed by the addition of a serially diluted HRP-conjugated FljB (1:3,000 to 1:9,000).

cELISA for the detection of S. Abortusequi antibody
One hundred microliters of the mixture of the diluted HRP-conjugated FljB antigen with the serially diluted serum sample was incubated for 30 min at 37°C in MAb-coated microplates.After washing three times with PBST, 100 µL of TMB peroxidase substrate (GalaxyBio, Beijing, China) was added to each well, and the plates were incubated for 10 min at 37°C.The absorbance values obtained were expressed as percent inhibition (PI) of the negative control computed as follows: (ODNC -ODSample)/(ODNC -ODPC) × 100%.
In order to standardize the cELISA, the factors that could affect the performance of the assay were assessed by testing a series of sera.The series consisted of one high titer (TAT titer, 32) and one low titer (TAT titer, 2) S. Abortusequi-positive serum samples and three other serum samples positive for other Salmonella serotypes (S.Typhimurium-, S. Enteritidis-, and S. Dublin-positive sera).Two factors were assessed: (i) reaction strategies of cELISAs and (ii) screening for serum dilution.Two response strategies were assessed.In the first strategy, the mixture of HRP-conjugated FljB antigen with the serially diluted serum sample was first incubated for 30 min at 37°C, and after incubation, the mixture was transferred into the coated antibodies plate and incubated for another 30 min at 37°C.In the second response strategy, the mixture of HRP-conjugated FljB antigen and serially diluted serum sample was incubated directly with the coated antibodies for 30 min at 37°C.

TAT
To further evaluate and compare the sensitivities of cELISA and TAT, the same serum samples were tested using both cELISA and TAT.TAT was performed as described previously (37).The mixture of the S. Abortusequi antigen with the serially diluted serum sample was incubated for 20 hr at 37°C in tubes.

Statistical analysis
Receiver operator characteristic (ROC) curve analysis was constructed in GraphPad Prism8 (GraphPad Software, USA), which is a statistical tool for evaluating the discrimina tory power of a diagnostic test.The area under the ROC curve (AUC) was employed to assess the accuracy of the tests distinguished between non-informative (AUC = 0.5), less accurate (0.5 < AUC ≤ 0.7), moderately accurate (0.7 < AUC ≤ 0.9), highly accurate (0.9 < AUC < 1), and perfect tests (AUC = 1) (38).The Wilson-Brown test was used to determine the 95% confidence intervals (CIs).

Selection and characterization of MAbs
In order to evaluate the specificity of the MAbs, hybridomas were screened using parallel iELISA against S. Abortusequi as well as Salmonella of different types.Only MAbs recognizing the S. Abortusequi strain were preserved.The specificity of anti-FljB MAbs was further confirmed using the results of competitive binding assays, and the binding of the various anti-FljB MAbs was inhibited only by S. Abortusequi positive sera.This reactivity means that the selected MAbs were suitable for the development of a cELISA.
Based on the results of the iELISAs, eight MAbs specific to S. Abortusequi were also tested using cELISAs.Three competitively reactive MAbs (1A10, 3B7, and 3B6) were selected with the cELISAs.
Moreover, the capability of the anti-FljB MAbs to efficiently bind the HRP-conjugated FljB antigen when used as coating antibodies in a direct ELISA was analyzed.Only one coated MAb (1A10) gave the strongest signal and was selected (Fig. 1).

Setup and standardization of cELISA for the detection of S. Abortusequi antibodies
A cELISA with purified 1A10 MAb adsorbed to the plate was then conducted.The one-step response strategy showed the best performance and speed and was selected for further evaluation (Fig. 2).Then, to confirm the optimal dilution of the test serum, serial dilutions of the five sera, ranging from 1 to 128 times dilutions, were tested with cELISA.A clear-cut separation between S. Abortusequi and other three Salmonellapositive sera was seen in the cELISA even if the serum was not diluted (Fig. 3A).The maximum dilution that still resulted in a positive value in the cELISA was a dilution of 64 and 4 for high-titer (TAT titer, 32) and low titer (TAT titer, 2) S. Abortusequi-positive sera, respectively.The specificity of the detection by the cELISA was investigated using a panel of antisera for other viruses or bacteria.The results showed that for all sera except S. Abortusequi-positive serum, no positive results were detected (Fig. 3B).The above results were suggestive of a satisfactory analytical specificity and sensitivity of the cELISA.

Clinical detection performances
The assay performances of the cELISA were validated by assessing 1,175 S. Abortuse qui-positive and S. Abortusequi-negative serum samples originating from horses and donkeys (Table 1).Samples were taken from the field and reference sera, were classi fied as S. Abortusequi positive or negative according to their known origin, and were confirmed with Western blotting.The ROC curves were constructed based on the previous classification of the sera into positive and negative.When compared with the Western blotting results, the AUC values showed that the TAT is moderately accurate (AUC = 0.7705; 95% Cl, 0.7437-0.7972)but has very low sensitivity (Se: 54.05; 95% CI, 50.28-57.86)(Fig. 4A; Table 1).Unexpectedly, the AUC values showed that the cELISA (cutoff value: 38.2% of inhibition) is highly accurate (AUC = 0.9941; 95% Cl 0.9898-0.9984)and has very high sensitivity (Se: 98.03; 95% CI, 96.66-98.85)and specificity (Sp: 99.81; 95% CI, 98.91-99.99)(Fig. 4B).Considering the excellent clinical test performance and to avoid false-positive results in routine activity, a cELISA cutoff value of 39.5% was selected with Se and Sp values of 97.58 (95% Cl, 96.10-98.50)and 100 (95% Cl, 99.26-100.00).When the cutoff value of cELISA was determined to be 39.5%, 883 of 1,175 samples tested in the TAT had the same results as the cELISA but 292 had divergent results between the two tests (Fig. 5A), whereas 1,162 of 1,175 samples yielded the same test results in both the Western blotting and cELISA with only 13 samples yielding differing results using the two methods (Fig. 5B).

DISCUSSION
The re-emergence of S. Abortusequi infection in equids has caused serious economic losses to China.Furthermore, horses and donkeys have been found infected with the bacterium in more and more provinces in China (24,25).S. Abortusequi infection could therefore have wider economic impacts than currently predicted.
The major goal of this study was to develop a rapid and simple cELISA method based on a MAb to detect antibodies against S. Abortusequi in different animal species.The cELISA described here has a direct advantage over iELISA because secondary antibodies specific to the immunoglobulins of the species being tested are not required.It was developed by using a MAb specific to S. Abortusequi and exhibited no cross-reactions to other Salmonella strains.We tested two different reaction strategies to select our cELISA and chose that characterized by optimum performance combined with a reduction in assay time compared with the current best available TAT test (Fig. 2).This assay consisted of purified MAb adsorbed to the plate with the addition of the mixtures of HRP-conjuga ted FljB antigen together with the serum sample in a single step.Interestingly, even when the serum being tested was undiluted, the assay was still able to discriminate well between S. Abortusequi-positive and S. Abortusequi-negative samples (Fig. 3A).
The developed cELISA was further validated with a large number of serum samples from different species obtained from the field and with other sera against EHV, EIV, EAV, EIAV, S. equi, T. equi, and B. caballi and three Salmonella serovars including S. Dublin, S. Typhimurium, and S. Enteritidis (Table 1).Although Salmonella is categorized into more than 2,600 Salmonella serotypes, there are few serotypes of Salmonella causing equine infections that have been reported.S. Abortusequi is the most common serotype of Salmonella that infects equines in China.According to our data, all Salmonella isolates in samples from equine abortions were identified as S. Abortusequi serotype.However, S. Typhimurium, S. Enteritidis, and S. Dublin serovars may also cause infection in equines in rare cases (11).Therefore, S. Typhimurium, S. Enteritidis, and S. Dublin serovars were selected to evaluate the specificity of the cELISA.Sera originating from horses strongly positive against other viruses or bacteria were classified as a negative population, better confirming the specificity of the cELISA (Fig. 3B).
The total positive detection rates of the cELISA and TAT in positive serum samples were 98.03% (647/660) and 53.79% (355/660), respectively.The 355 samples that tested positive in the TAT all also tested positive in the cELISA (Fig. 5A).These results showed that the cELISA has excellent detection performance and discriminating power with a high Se and Sp.Moreover, the performance of the cELISA showed a very good agree ment (98.89%, 1,162/1,175) with Western blotting (Fig. 5B) compared to the TAT assays (74.21%, 872/1,175)).
In general, the cELISA was in very good agreement with the Western blotting assays, drawing attention to some outliers in this study (Fig. 5B).Thirteen samples originating from horse or donkey farms tested positive by Western blotting but resulted negative in the cELISA.Further analysis showed that the 13 sera were only weakly positive in Western blotting.Considering that these 13 samples originated from areas where S. Abortusequi is known to be present, we think that these were not false positives, and we suppose that the lack of correspondence between the two assays could be due to a higher sensitivity of the Western blotting method than cELISA.
Furthermore, a series of other advantageous properties of the cELISA, such as the feasibility of using the sera without dilution, and the one-step response strategy make the cELISA ideal for screening and confirming of S. Abortusequi infection in equids.Since positive antibody presumably takes days to develop following infection, this limitation should be noted when the cELISA is used to detect serum samples.

FIG 1
FIG 1 ELISA block titration with the indicated different MAbs and dilutions of HRP-conjugated FljB.

FIG 2 FIG 3
FIG 2 Schematic diagram of solid phase competition ELISA.Anti-S.Abortusequi serum and 1A10 MAb, which precoated in the plate, competitively react specifically with FljB-HRP.The conjugation is washed away during the cELISA procedure.Little or no color is visible in the positive samples.Conversely, negative samples have a darker color.

FIG 4 FIG 5
FIG4 ROC curve analysis.ROC curves for TAT (A) and cELISA (B) using the panel of 1,175 sera described above.

TABLE 1
Description of serum samples used in this study

Serum classification No. of sera Species Western blot results cELISA a TAT b TAT (titers) Anamnestic data
f S. equi, S. Dublin, S. Typhimurium, S. Enteritidis, T. equi, and B. caballi were prepared using healthy horses.a cELISA, competitive enzyme-linked immunosorbent assay.b TAT, agglutination test.c EHV, equine herpes virus.d EIV, equine influenza virus.e EAV, equine arteritis virus.f EIAV, equine infectious anemia virus.