Instrument-Free and Visual Detection of Salmonella Based on Magnetic Nanoparticles and an Antibody Probe Immunosensor

Salmonella, a common foodborne pathogen, causes many cases of foodborne illness and poses a threat to public health worldwide. Immunological detection systems can be combined with nanoparticles to develop sensitive and portable detection technologies for timely screening of Salmonella infections. Here, we developed an antibody-probe-based immuno-N-hydroxysuccinimide (NHS) bead (AIB) system to detect Salmonella. After adding the antibody probe, Salmonella accumulated in the samples on the surfaces of the immuno-NHS beads (INBs), forming a sandwich structure (INB–Salmonella–probes). We demonstrated the utility of our AIB diagnostic system for detecting Salmonella in water, milk, and eggs, with a sensitivity of 9 CFU mL−1 in less than 50 min. The AIB diagnostic system exhibits highly specific detection and no cross-reaction with other similar microbial strains. With no specialized equipment or technical requirements, the AIB diagnostic method can be used for visual, rapid, and point-of-care detection of Salmonella.


Introduction
Salmonella is a common foodborne pathogen that infects humans and many other animals [1,2]. Cramps, diarrhea, vomiting, and fever are the most frequently reported symptoms of salmonellosis worldwide [3][4][5]. The elderly, the immunocompromised, and infants are the most commonly infected patients, experiencing significant morbidity and mortality [6,7]. It is estimated that approximately 94 million humans are infected with Salmonella globally each year, of which 80.3 million cases are foodborne [8][9][10]. Food is necessary for human survival, but it is often contaminated with Salmonella. The main carriers of Salmonella are poultry products, but other undercooked or raw meats, dairy products, and other industrialized foods are also easily contaminated with Salmonella, which can then infect humans [1]. Incidences of Salmonella infections have been reported in both developed and developing countries, and cases of Salmonella infections have increased in recent decades [11].

Generation of Specific mAbs Against Salmonella
Two stable positive hybridomas were screened through three subcloning cycles from twentythree originally positive wells, designated as 2F1 and 1B12 ( Figure 2A). Reactivity of the two mAbs was determined via enzyme-linked immunosorbant assay (ELISA). The results showed that both mAbs reacted with Salmonella ( Figure 2B). The immunoglobulin isotypes of 2F1 and 1B12 were determined using a mouse monoclonal antibody isotyping kit. Figure 2C shows 2F1 and 1B12 isotyped as IgG3, and the light chains of the two mAbs belong to the kappa chain. The two mAbs were used to produce ascites. The ascites was purified using protein A-sepharose and tested via sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) ( Figure 2D). The titers of mAbs 2F1 and 1B12 were evaluated via ELISA, and the titers of both mAbs reached 1:204800 ( Figure  2E,F). The KD values of mAb 2F1 and 1B12 were measured as described in our previous published study [22], and were calculated as KD = 3.677 ± 0.33 nM for mAb 2F1 and KD = 1.126 ± 0.15 nM for mAb 1B12.

Generation of Specific mAbs Against Salmonella
Two stable positive hybridomas were screened through three subcloning cycles from twenty-three originally positive wells, designated as 2F1 and 1B12 ( Figure 2A). Reactivity of the two mAbs was determined via enzyme-linked immunosorbant assay (ELISA). The results showed that both mAbs reacted with Salmonella ( Figure 2B). The immunoglobulin isotypes of 2F1 and 1B12 were determined using a mouse monoclonal antibody isotyping kit. Figure 2C shows 2F1 and 1B12 isotyped as IgG3, and the light chains of the two mAbs belong to the kappa chain. The two mAbs were used to produce ascites. The ascites was purified using protein A-sepharose and tested via sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) ( Figure 2D). The titers of mAbs 2F1 and 1B12 were evaluated via ELISA, and the titers of both mAbs reached 1:204800 ( Figure 2E,F). The K D values of mAb 2F1 and 1B12 were measured as described in our previous published study [22], and were calculated as K D = 3.677 ± 0.33 nM for mAb 2F1 and K D = 1.126 ± 0.15 nM for mAb 1B12.

Synthesis of the HRP mAb Probes
The purified mAbs 2F1 and 1B12 were dialyzed in phosphate-buffered saline (PBS) to remove the Tris-HCl and glycine. The purified mAbs and HRP were then coupled by an aldehyde-amino bridge under NaIO4 and NaBH4 ( Figure 3A). The ratio of conjugation was calculated as 1 mg HRP/2.5 mg antibody. After conjugation, we acquired about 25 mg mAb probe containing 10 mg HRP. Next, the activities and titers of the two HRP mAb probes were determined by ELISA ( Figure 3B,C).

Synthesis of the HRP mAb Probes
The purified mAbs 2F1 and 1B12 were dialyzed in phosphate-buffered saline (PBS) to remove the Tris-HCl and glycine. The purified mAbs and HRP were then coupled by an aldehyde-amino bridge under NaIO 4 and NaBH 4 ( Figure 3A). The ratio of conjugation was calculated as 1 mg HRP/2.5 mg antibody. After conjugation, we acquired about 25 mg mAb probe containing 10 mg HRP. Next, the activities and titers of the two HRP mAb probes were determined by ELISA ( Figure 3B,C).

Synthesis of the HRP mAb Probes
The purified mAbs 2F1 and 1B12 were dialyzed in phosphate-buffered saline (PBS) to remove the Tris-HCl and glycine. The purified mAbs and HRP were then coupled by an aldehyde-amino bridge under NaIO4 and NaBH4 ( Figure 3A). The ratio of conjugation was calculated as 1 mg HRP/2.5 mg antibody. After conjugation, we acquired about 25 mg mAb probe containing 10 mg HRP. Next, the activities and titers of the two HRP mAb probes were determined by ELISA ( Figure 3B,C).

Characterization of the Paired Antibodies
The reactivity and specificity of mAb 1B12 and 2F1 were evaluated via ELISA and Western blot. Figure 4A shows that both mAb 1B12 and 2F1 specifically recognized Salmonella and did not cross-react with similar microbial strains. The Western blot results showed that both mAbs recognized the different proteins on the surface of Salmonella ( Figure 4B). The optimal mAbs used to establish the AIB system were screened based on double sandwich ELISA (DAS-ELISA). HRP-labeled mAbs (1B12 and 2F1) and unlabeled mAbs (1B12 and 2F1) were constructed in each group for the DAS-ELISA, which showed that the group composed of mAb 2F1 and 1B12 was more effective than the other groups ( Figure 4C). In addition, we evaluated the specificity of the mAb 2F1 and 1B12 combination. Figure 4D shows that the group of composed of 1B12 and 2F1 displayed high specificity and did not recognize the control strains (E. coli, S. aureus, K. pneumoniae, Shigella, A. baumannii, P. aeruginosa, and Streptococcus). Figure 4C,D show that the mAb 2F1 and 1B12 group was the optimal combination for detection of Salmonella.

Characterization of the Paired Antibodies
The reactivity and specificity of mAb 1B12 and 2F1 were evaluated via ELISA and Western blot. Figure 4A shows that both mAb 1B12 and 2F1 specifically recognized Salmonella and did not crossreact with similar microbial strains. The Western blot results showed that both mAbs recognized the different proteins on the surface of Salmonella ( Figure 4B). The optimal mAbs used to establish the AIB system were screened based on double sandwich ELISA (DAS-ELISA). HRP-labeled mAbs (1B12 and 2F1) and unlabeled mAbs (1B12 and 2F1) were constructed in each group for the DAS-ELISA, which showed that the group composed of mAb 2F1 and 1B12 was more effective than the other groups ( Figure 4C). In addition, we evaluated the specificity of the mAb 2F1 and 1B12 combination. Figure 4D shows that the group of composed of 1B12 and 2F1 displayed high specificity and did not recognize the control strains (E. coli, S. aureus, K. pneumoniae, Shigella, A. baumannii, P. aeruginosa, and Streptococcus). Figure 4C,D show that the mAb 2F1 and 1B12 group was the optimal combination for detection of Salmonella.

INB Preparation and Characterization
The INBs were generated by conjugating NHS modified magnetic beads (NHS beads) with mAb 2F1 via covalent coupling ( Figure 5A). The prepared INBs were evaluated via SDS-PAGE and Western blot. The results demonstrated that mAb 2F1 had conjugated on the surface of the NHS beads ( Figure 5B,C).

INB Preparation and Characterization
The INBs were generated by conjugating NHS modified magnetic beads (NHS beads) with mAb 2F1 via covalent coupling ( Figure 5A). The prepared INBs were evaluated via SDS-PAGE and Western blot. The results demonstrated that mAb 2F1 had conjugated on the surface of the NHS beads ( Figure 5B,C).

AIB System Optimization
In the AIB system, we optimized each preparation step to achieve the best detection effect. Using gradient dilutions and plate counts, the average binding efficiency of the INBs was 90%. The optimum INB capture period was confirmed using various incubation times from 10 to 60 min. Figure  6A,B show that the INBs completely captured the Salmonella within 30 min. No significant differences were observed as the incubation time increased. The optimal incubation period, during which the mAb 1B12 probes formed sandwich products, was evaluated over 10 to 60 min ( Figure 6C). The optical density from the formed sandwich products was determined via microplate reader. The sandwich products quickly formed within 20 min, demonstrating the high affinity between the mAb 1B12 probes and Salmonella ( Figure 6D).

AIB System Optimization
In the AIB system, we optimized each preparation step to achieve the best detection effect. Using gradient dilutions and plate counts, the average binding efficiency of the INBs was 90%. The optimum INB capture period was confirmed using various incubation times from 10 to 60 min. Figure 6A,B show that the INBs completely captured the Salmonella within 30 min. No significant differences were observed as the incubation time increased. The optimal incubation period, during which the mAb 1B12 probes formed sandwich products, was evaluated over 10 to 60 min ( Figure 6C). The optical density from the formed sandwich products was determined via microplate reader. The sandwich products quickly formed within 20 min, demonstrating the high affinity between the mAb 1B12 probes and Salmonella ( Figure 6D).

Assessment of Salmonella Detection Using the AIB System
We initially tested the AIB system specificity using seven similar microbial strains and found that the AIB system was highly specific for detecting Salmonella. The blue color ( Figure 7A) and strong optical density ( Figure 7C) were observed only in the presence of Salmonella. The controls showed no significant changes. The AIB system sensitivity was tested with different Salmonella concentrations ranging from 9 × 10 7 to 9 × 10 0 colony-forming units (CFU) mL −1 . Figure 7B,D show the blue color and optical density at different Salmonella concentrations, as recorded by the AIB system. We then set up a plotted linear curve using the different Salmonella concentrations. Figure 7E shows a good linear relationship (R 2 = 0.9945).

Assessment of Salmonella Detection Using the AIB System
We initially tested the AIB system specificity using seven similar microbial strains and found that the AIB system was highly specific for detecting Salmonella. The blue color ( Figure 7A) and strong optical density ( Figure 7C) were observed only in the presence of Salmonella. The controls showed no significant changes. The AIB system sensitivity was tested with different Salmonella concentrations ranging from 9 × 10 7 to 9 × 10 0 colony-forming units (CFU) mL −1 . Figure 7B,D show the blue color and optical density at different Salmonella concentrations, as recorded by the AIB system. We then set up a plotted linear curve using the different Salmonella concentrations. Figure 7E shows a good linear relationship (R 2 = 0.9945).

Figure 7.
Evaluation of the AIB system. Specificity assay of the AIB system for Salmonella detection: (A) images and (C) optical density of Salmonella and non-target strains in the AIB system. Sensitivity assay of the AIB system for Salmonella detection: (B) change in color and (D) optical density due to different concentrations of Salmonella (from 10 7 CFU mL −1 to 10 0 CFU mL −1 ). (E) Plotted linear curve of the AIB system with Salmonella ranging from 10 7 to 10 0 CFU mL −1 .

Salmonella Detection by the AIB System in Artificially Contaminated Samples
To evaluate the AIB system performance, we used milk and egg samples contaminated with different concentrations of Salmonella ranging from 10 5 to 10 0 CFU mL −1 . Figure 8A,B show the changes in blue color and strong optical density in the milk and egg samples contaminated with different Salmonella concentrations. These results confirmed that the new AIB system can rapidly and accurately detect Salmonella, even in complex samples such as milk and eggs. Besides, the prepared milk and egg samples were used for the extraction of the genome and then applied for PCR and LAMP assays using the reported primers [23]. Figure 8C,D show that the sensitivity values of PCR and LAMP were 10 2 CFU and 10 1 CFU, respectively, in both milk and egg samples.

Salmonella Detection by the AIB System in Artificially Contaminated Samples
To evaluate the AIB system performance, we used milk and egg samples contaminated with different concentrations of Salmonella ranging from 10 5 to 10 0 CFU mL −1 . Figure 8A,B show the changes in blue color and strong optical density in the milk and egg samples contaminated with different Salmonella concentrations. These results confirmed that the new AIB system can rapidly and accurately detect Salmonella, even in complex samples such as milk and eggs. Besides, the prepared milk and egg samples were used for the extraction of the genome and then applied for PCR and LAMP assays using the reported primers [23]. Figure 8C,D show that the sensitivity values of PCR and LAMP were 10 2 CFU and 10 1 CFU, respectively, in both milk and egg samples.

Discussion
Salmonella is an important human pathogen worldwide, infecting humans and various other animals. Improper cooking and processing of animal-derived foods (e.g., raw milk, meat, and eggs) are the main mechanisms by which Salmonella infects humans [24]. Salmonellosis, one of the most important zoonoses, which mainly causes severe foodborne gastroenteritis and bacterial diarrhea, is a huge public health problem [25]. Foods such as meat, milk, and eggs are essential for human survival but are often contaminated with Salmonella [24]. Studies have reported that cases of Salmonella poisoning have occurred in powdered infant formula [26], raw milk [27], eggs [28], and meat [13,29]; thus, the food industry must prioritize developing innovative methods for detecting Salmonella.
Conventional culture-based methods are considered the gold standard for detecting Salmonella in various samples. However, these are labor-intensive and time-consuming, usually taking 2 to 3 days [30]. Thus, these methods are unsuitable for rapid detection. Recently, rapid detection methods have been developed to detect Salmonella that target the nucleic acid, including PCR [13], real-time PCR [31], immunocaptured-PCR (IC-PCR) [32], and LAMP [8,33]. The disadvantages of PCR and IC-PCR detection methods are that they require instruments, professional personnel, bacterial enrichment, and genome or plasmid extraction [15,34]. LAMP is a novel amplification approach developed by Notomi et al. [35], which is rapid and highly specific, and has been applied for various pathogens, including parasites [36,37], fungi [38], bacteria [20], and viruses [39,40]. The biggest disadvantage of LAMP is that it produces aerosol during detection, leading to many false positives.

Discussion
Salmonella is an important human pathogen worldwide, infecting humans and various other animals. Improper cooking and processing of animal-derived foods (e.g., raw milk, meat, and eggs) are the main mechanisms by which Salmonella infects humans [24]. Salmonellosis, one of the most important zoonoses, which mainly causes severe foodborne gastroenteritis and bacterial diarrhea, is a huge public health problem [25]. Foods such as meat, milk, and eggs are essential for human survival but are often contaminated with Salmonella [24]. Studies have reported that cases of Salmonella poisoning have occurred in powdered infant formula [26], raw milk [27], eggs [28], and meat [13,29]; thus, the food industry must prioritize developing innovative methods for detecting Salmonella.
Conventional culture-based methods are considered the gold standard for detecting Salmonella in various samples. However, these are labor-intensive and time-consuming, usually taking 2 to 3 days [30]. Thus, these methods are unsuitable for rapid detection. Recently, rapid detection methods have been developed to detect Salmonella that target the nucleic acid, including PCR [13], real-time PCR [31], immunocaptured-PCR (IC-PCR) [32], and LAMP [8,33]. The disadvantages of PCR and IC-PCR detection methods are that they require instruments, professional personnel, bacterial enrichment, and genome or plasmid extraction [15,34]. LAMP is a novel amplification approach developed by Notomi et al. [35], which is rapid and highly specific, and has been applied for various pathogens, including parasites [36,37], fungi [38], bacteria [20], and viruses [39,40]. The biggest disadvantage of LAMP is that it produces aerosol during detection, leading to many false positives.
Conversely, immunodiagnostic approaches are more rapid, sensitive, and stable than nucleic acid assays, especially for samples such as milk [41], whole blood [42], and saliva [43]. Many novel immunosensors have been developed, which combine immunology with magnetic nanoparticles [22], platinum nanoparticles [44], or Pt nanomotors [45]. Instrument-free and mobile diagnostic technologies could transform the current foodborne pathogen detection systems, particularly in resource-limited settings.
In this study, we developed an instrument-free, sensitive immunosensor to effectively, rapidly, and sensitively detect Salmonella based on a pair of mAbs recognizing different antigenic determinants on the surface of Salmonella ( Figure 4B) and sensitive probes (Figure 3). The AIB system reported in this manuscript was developed using a mAb pair, NHS beads, and HRP. In this system, mAb 1B12 was conjugated with HRP, forming a sandwich structure (Figure 3). The mAb 2F1 was coated on the surface of the NHS beads used to enrich Salmonella ( Figure 5). We demonstrated the feasibility and practicability of the AIB system for detecting Salmonella using INBs and HRP probes. The sensitivity of the AIB system was 9 CFU mL −1 and higher than that of PCR (10 5 CFU) [13], real-time PCR coupled with immunomagnetic separation or centrifugation (2 × 10 4 CFU) [42], LAMP (1.3 to 28 CFU) [43], conventional culture-based methods, and antibody or aptamer-based assay (10 1 to 10 3 CFU) ( Table 1). However, the nucleic-acid based detection methods still take several hours to enrich the Salmonella, followed by extraction of the plasmid or genome using a commercial kit [13,31,46]. Besides, special equipment, including the thermal cycling instrument, an electrophoresis apparatus, and a gel imaging system, are essential for PCR [13] and real-time PCR [31,32] methods. This special equipment is expensive and requires professional technical assistance, which largely limits their applications, especially for areas with poor resources. LAMP is a simple detection method that does not require special equipment (unlike PCR or real-time PCR), but it cannot enrich the target substances from the samples, and a commercial kit is also needed to extract the plasmid or genome [37,38]. Moreover, for some special samples, such as sputum, blood, and feces, it is very difficult to accurately and quickly enrich the target substances, and much time and materials are needed during the process [47]. Additionally, the false positive results caused by LAMP can be deadly. Therefore, the above drawbacks make LAMP inappropriate for fast, sensitive on-site detection. For antibody or aptamer-based ELISA assay, the sensitivity of the related ELISA assay ranged from 10 1 to 10 3 CFU for artificially samples, but the process takes several hours. In addition, similar INB methods, ELISA assay lacks the ability to rapidly enrich targets from the environment, especially for special samples such as milk, whole blood, and saliva, meaning antibody or aptamer-based ELISA assays are insufficiently sensitive, which largely limits their applications. Besides, professional experimental skills are essential, as well as a microplate reader, which is expensive and not available anywhere, particularly in resource-poor areas. Compared with culture-based techniques and biochemical assays, PCR, real-time PCR, and normal LAMP, and antibody or aptamer-based ELISA assay, the AIB assay only needs a portable magnetic frame and TMB buffer. It does not require a long pre-enrichment step followed by genome or plasmid extraction, any special equipment, or professional skills. The AIB system developed here has two advantages over normal detection methods. The first one is that the surfaces of the NHS beads are coated with abundant mAb 2F1 that specifically recognize Salmonella ( Figure 4A), which means the prepared INBs can efficiently and accurately enrich the Salmonella from different biological samples within 30 min, as shown in Figure 6A,B. Compared with the carboxylic modified magnetic beads used in our previous study and in other research articles [22], the NHS beads are coupled with antibodies with high efficiency and short timeframes, without needing EDC or glutaraldehyde for activation. The other advantage is that the mAb probe, having high affinity, can rapidly form a sandwich structure within 20 min, as shown in Figure 6C,D. Additionally, the HRP coupled with mAb 1B12 is efficiently catalyzed by TMB buffer, showing a strong blue color ( Figure 1C). The optical density ( Figure 1D) changed within 10 min. Accordingly, the INBs showing rapid and efficient enrichment together with the sensitive mAb probe allow the AIB system to rapidly and sensitively detect Salmonella within 50 min. Figure 7 shows that the AIB system has good sensitivity and specificity based on the INBs and HRP mAb probes. We evaluated the practical application of the AIB system using artificially contaminated samples, including water ( Figure 7B,D), milk ( Figure 8A), and eggs ( Figure 8B). The results indicated that the AIB system enabled accurate and rapid screening of Salmonella and could potentially be used in the food industry and hospitals. The prepared samples were also identified by PCR and LAMP to evaluate the accuracy of the AIB assay. In Figure 8C,D, we can clearly see that the minimum detection limits for PCR and LAMP were 9 × 10 2 CFU and 9 × 10 1 CFU, respectively, in both milk and eggs samples. The sensitivity of the PCR and LAMP methods was lower than the AIB assay for the same samples. Having high sensitivity and specificity, the AIB system can rapidly and accurately detect Salmonella without requiring special equipment or professional skills, making the AIB system more applicable in various environments. Table 2 shows the eight bacterial strains used in this study.

Production of mAbs Against Salmonella
The antigens were prepared using 10 8 CFU of Salmonella dissolved in PBS buffer and inactivated for 30 min at 80 • C [20]. The prepared antigens were used to immunize BALB/c mice, with 10 7 CFU being mixed with FIA for the first immunization and 10 7 CFU mixed with FCA for the second and third immunizations. Three days after the boosted immunization, spleen cells were collected and fused with SP2/0 cells via PEG at 37 • C [79]. The fused cells were maintained in HAT medium for 7 days. One week later, the fused cells were cultured in HT medium until the first round of screening. The positive hybridomas in each plate were screened by ELISA.

Purification of Ascites
After three subcloning cycles, we successfully obtained two positive hybridoma cell lines that stably secreted antibodies against Salmonella. Two hybridomas were constructed to prepare the ascites, which were purified using protein A-sepharose. Briefly, all ascites were filtered using a 0.2-micron filter, and the pH value of the ascites was adjusted to 8.0 using 1.0 M Tris-Cl (pH 8.0). Next, the prepared ascites were incubated with protein A-sepharose for 30 min at 25 • C. After incubation, the ascites were collected, and the column was washed with 100 mM Tris-Cl (pH 8.0) and 10 mM Tris-Cl (pH 8.0), respectively. Lastly, the column was washed with 50 mM glycine (pH 3.0), and the purified immunoglobulins from each antibody were determined via bicinchoninic acid assay and SDS-PAGE.

ELISA and Western Blot Assay
For the ELISA, each well was coated with 10 6 CFU of Salmonella and incubated for 2 h at 37 • C. All wells were then washed with PBS-T and blocked with 5% BSA. Next, mAb 1B12 and 2F1 (1:1000) were added to each well and incubated for 1 h at 37 • C. All wells were washed with PBS-T, then the goat anti-mouse IgG (H+L) HRP (1:5000) was added and incubated for 1 h at 37 • C. One hour later, the supernatant was removed and washed with PBS-T. Lastly, soluble TMB substrate solution (TIANGEN, Beijing, China) was added to detect the immunoreaction.
For the Western blot assay, 10 6 CFU of Salmonella was transferred to a nitrocellulose (NC) membrane, which was blocked with 5% BSA for 1 h at 37 • C. Next, mAb 1B12 and 2F1 were added (1:1000) and incubated for 2 h at 37 • C. Two hours later, the antibodies were collected and the NC membranes were washed with PBS-T. Goat anti-mouse IgG (H+ L) HRP (1:5000) was then added and incubated for 1 h at 37 • C. Lastly, the NC membranes were washed, and the Western blot kit (BioBest, Anhui, China) was used to detect the immunoreaction.

HRP mAb Probe Preparation and Characterization
The HRP mAb probes were synthesized using mAbs coupled with HRP. All probes were prepared as described in our previous study [22]. Briefly, 10 mg of HRP was dissolved in 1 mL of 0.1 M NaHCO 3 and then oxidized for 2 h with 1 mL of 10 mM NaIO 4 . Then, the HRP liquid was mixed with 1.5 mL of 0.1 M Na 2 CO 3 , 25 mg of 1B12, 2F1, and reacted in a 5 mL tube containing 0.6 g of Sephadex G25 for 3 h at 25 • C. Three hours later, the liquid was collected and terminated with 0.225 mL of 0.132 M NaBH 4 for 0.5 h, followed by 0.675 mL of 0.132 M NaBH 4 for 1 h at 25 • C. The process was performed in the dark. Lastly, the prepared HRP mAb probes were stored in PBS buffer at 4 • C. The activities and titers of the HRP mAb probes were determined using ELISA. Briefly, two HRP mAb probes at different dilutions (from 1:100 to 1:51200) were added to each well, which had been coated with Salmonella and blocked with BSA, and incubated for 1 h at 37 • C. Next, the supernatant was removed from each well and washed with PBS-T. The immune responses were then detected using soluble TMB substrate solution.

INB Preparation and Characterization
The INBs were prepared using NHS beads conjugated with mAb 2F1 using a MAg25K/NHS kit. First, the NHS beads were mixed evenly, and 1 mL of NHS bead solution (10%, v/v) was added to a 2 mL tube. The supernatant was removed by magnetic separation, and the NHS beads were washed twice with 2 mL of absolute ethanol. Second, 1.2 mg of mAb 2F1 dissolved in a coupling buffer was added and mixed with NHS beads for 2 h at room temperature. Next, 1 mL of blocking buffer was added to block the beads for 2 h, while being rotated and mixed at room temperature. After blocking, the supernatant was removed, and 1 mL of wash buffer was added to wash the beads five times. Lastly, the supernatant was removed, and the INBs were dissolved in PBS buffer (pH 7.4) for further use. The prepared INBs were evaluated via SDS-PAGE, Western blot, and the AIB system.

Performance of the AIB System
The sensitivity of the AIB system was evaluated using serially diluted Salmonella, with concentrations ranging from 10 7 to 10 0 CFU mL −1 . One microliter of each Salmonella concentration was tested using the AIB system following the testing protocol. The specificity of the developed AIB system was confirmed using Salmonella and similar microbial strains (E. coli, S. aureus, K. pneumoniae, Shigella, A. baumannii, P. aeruginosa, and Streptococcus).

System Evaluation using Contaminated Samples
To evaluate the AIB system, we used artificial milk and egg samples containing Salmonella concentrations ranging from 10 5 to 10 0 CFU mL −1 . Each sample was tested using the AIB system according to the testing protocol.

Conclusions
In this study, we developed a novel immunosensor based on magnetic NHS beads and antibody probes. The novel AIB system enabled visual, rapid, and sensitive Salmonella detection, without requiring specialized equipment or skills. This system can potentially be widely used to diagnose infectious diseases caused by Salmonella spp.