Development of a mu3ABC-Based Lateral Flow Immunochromatographic Strip for Rapid DIVA-Compatible Detection of Antibodies Specific to Foot-and-Mouth Disease Virus

1. Introduction

Foot-and-mouth disease (FMD) is a rapidly spread and highly contagious vesicular disease of coven-hoofed animals, including swine, cattle, buffalo, goat and sheep. The disease distributes in Asia, Africa and East Europe which causes high economic impact on endemic countries. The incubation period of FMD depends on the species of animals, exposure dose and the route of infection [1]. Cattle and swine are highly susceptible to foot-and-mouth disease virus (FMDV) while clinical outcome in small ruminants is mild or asymptomatic [2]. Asymptomatic carrier animals can spread the virus leading to infection for susceptible animals. The virus remains in epithelial cells and lymphoid tissues in oropharyngeal areas of cattle for 28 days [3]. Deep sequence analysis revealed the presence of FMDV subpopulations including antigenic escape mutants in nasopharyngeal mucosa of persistent infected cattle [4].

FMDV comprises seven distinct serotypes: O, A, C, Asia 1, SAT 1, SAT 2, and SAT 3. Each serotype includes various subtypes and topotypes, with no cross-protection among serotypes [5,6]. FMDV distributes globally which is classified based on VP1 sequences into seven endemic pools of FMDV serotypes and topotypes circulating in specific regions [7]. Pools 1, 2, and 3 encompass distinct geographical areas across Asia and Eurasia (Southeast/Central/East Asia, South Asia, and West Eurasia/Middle East, respectively), which serotypes O, A, and Asia 1 are predominantly circulating in all three regions. Serotype O is the most prevalent in Asia and present in all countries [8]. However, the circulating FMDV topotypes or lineages in each pool may be different, such as O/SEA/Mya98 in pool 1 and O/ME-SA in pool 2 with occasional incursion from pool 2 to pools 1 and 3 [9]. Regarding Africa, the virus distribution comprises three endemic pools: Pool 4 (Eastern and North Africa), which is characterized by the circulation of serotypes O, A, SAT 1, SAT 2, and SAT 3 in Eastern Africa, whereas only serotypes A and O are present in North Africa; Pool 5 (West and Central Africa), which includes serotypes O, A, SAT 1, and SAT 2; and Pool 6 (Southern Africa), which is restricted to serotypes SAT 1, SAT 2, and SAT 3. Lastly, Pool 7 is confined to South America, where serotypes O and A circulate exclusively in Venezuela. Recently, cross-pool migrations of FMDV have been observed across multiple regions [7]. For instance, the O/ME-SA/Ind-2001e lineage (Middle East-South Asia topotype) introduced from Pool 2 to Pool 1 between 2016 and 2021 subsequently became the predominant lineage circulating in Southeast Asia [7]. More recently, the co-circulation of SAT 1 topotype I (originating from pool 4 in East Africa) and a seed-vaccine-related SAT 1 topotype III strain was detected in the Middle East in 2025 [10,11]. Subsequently, in March 2026, SAT 1 topotype III incursion was reported in Xinjiang and Gansu provinces, China, situated 2,000 km apart [12]. This transboundary dissemination across endemic pools and the coexistence of multiple serotypes underscore the critical role of livestock movements in widespread disease transmission.

Animal quarantine and the screening of infected livestock for slaughter are core control strategies that have proven effective in maintaining FMD-free status [13]. However, in FMD endemic areas where vaccination is implemented as a control measure, a significant challenge in these programs is the presence of persistently infected animals, which can act as silent reservoirs of the virus [4,14]. To address this, a diagnostic tool that can differentiate infected from vaccinated animals (DIVA) has become essential [15]. The foundation of the DIVA strategy lies in detecting antibodies against non-structural proteins (NSPs) of the virus, which are produced during viral replication and are typically absent in purified antigen containing vaccines. Therefore, they act as a specific marker to distinguish immunity raised by natural infection from vaccination. Among the various NSPs evaluated for the DIVA strategy, the 3ABC polyprotein stands out due to its high immunogenicity [16]. Antibodies against 3ABC are detectable shortly after infection and tend to persist longer than those targeting other NSPs [17]. Consequently, the 3ABC-based assay is widely considered as the most reliable diagnostic tool for the DIVA strategy [16,17,18,19].

Recently, immunochromatographic strips have been widely adopted across various fields due to their specificity, rapid turnaround, and suitability for field-based detection. In this system, colloidal gold nanoparticles are commonly used as a label to provide a visible signal. These nanoparticles are conjugated with specific antibodies or antigens to form a detection complex. As the sample flows across the nitrocellulose membrane via capillary action, the target analytes are captured at the test line, resulting in a visible color change. Several immunochromatographic assays have been developed for the detection and differentiation of FMDV antibodies in pig serum samples [20,21]. Previously, we developed FMDV diagnostic ELISA based on 3ABC for antibody detection and DIVA discrimination [22]. In this study, to enhance user convenience and field adaptability, we developed a 3ABC-based lateral flow immunoassay (LFA) using colloidal gold nanoparticles for rapid visibility. The performance of this assay was demonstrated across three different livestock species using both reference and field sera. This assay provides a practical and rapid tool for FMD surveillance, facilitating the efficient movement control of animals both domestically and internationally.

2. Materials and Methods

2.1. Serum Samples and Reagents

A total of 566 serum samples from cattle, pigs, and goats were used to evaluate the performance of the mu3ABC strip test. Among these, 200 samples were reference cattle sera (100 positive and 100 negative standard antisera) kindly provided by the Regional Reference Laboratory for Foot and Mouth Disease in South East Asia (RRL). The remaining samples were field sera collected from various livestock species. Mouse IgG, bovine serum albumin (BSA), casein, sucrose and Tween-20 were purchased from Sigma–Aldrich Chemical Corporation (St. Louis, Mo, USA). All solvents and other chemicals were of analytical reagent grade.

2.2. Production of 3ABC Polyprotein

mu3ABC is a genetically engineered form of FMDV 3ABC gene by substituting cysteine at position 142 and 163 with serine and glycine, respectively [23]. In this study, mu3ABC cloned in pQE80L (Qiagen, Germantown, MD, USA) was expressed in E. coli cultures, as described previously [24]. Briefly, BL21 DE3 E. coli competent cells (New England Biolabs, Ipswich, MA, USA) were transformed with pQEmu3ABC and induced in early-log phase culture by 0.2 mM IPTG for 4 h. Cultures were collected and the E. coli cell disruption was achieved by chemical and physical methods. In addition, a recombinant Autographa Californica Multiple Nucleopolyhedrovirus (AcMNPV) containing mu3ABC gene was produced and inoculated into High Five insect cells (Thermo Fisher Scientific, Waltham, MA, USA) to express mu3ABC as described elsewhere [22]. Both insoluble 3ABC proteins obtained from E. coli and insect cells were dissolved in binding buffer (50 mM NaH₂PO₄, 8 M Urea, pH 7.5) and loaded to HiTrap SPFF column (Cytiva, Marlborough, MA, USA), following the manufacturer’s protocol. The eluate fractions with mu3ABC band in 10% SDS-PAGE were pooled and loaded to HisTrap FF column (Cytiva, Marlborough, MA, USA). The presence of mu3ABC in the eluates were confirmed by 10% SDS-PAGE and western blot. The concentration of the purified mu3ABC was measured using Pierce™ 660nm protein assay (Thermo Fisher Scientific, MA, USA). The protein was stored at -80°C until used.

2.3. SDS-PAGE and Western Blot

mu3ABC profiles were studied by SDS-PAGE and western blot following the previous published protocol [24]. The purified mu3ABC were electrophoresed through SDS-PAGE (Thermo Scientific, Waltham, MA, USA) with 10% resolving gel and 4% stacking gel at 140 V and 300 mA for 65 minutes. The gel was subsequently blotted onto a nitrocellulose membrane (Bio-Rad Laboratories, Hercules, CA, USA) in Tris-Glycine transfer buffer (25 mM Tris, 192 mM glycine and 20% Methanol, pH 9.9) using the wet transfer method. After blocking with BlockPRO™ 1 Min Protein-Free Blocking Buffer (Energenesis Biomedical, Taipei, Taiwan), the blotted membrane was incubated with anti-FMDV serum (1:200) at room temperature for 1 h, followed by incubating with Protein G conjugated with HRP (Sigma-Aldrich, St. Louis, MO, USA) diluted at 1:2,000 in TBST. After each step, the membrane was washed with TBST (50 mM Tris-HCl pH 7.5, 100 mM NaCl, 1 mM EDTA, 0.1% Tween 20) for three times. Finally, the membrane was incubated in TMB peroxidase substrate (Sera Care, Milford, MA, USA) until the color was developed. The reaction was stopped by thoroughly rinsing the membrane with deionized water.

2.4. Optimization of Essential Parameters

A simple dot blot was performed to optimize conditions for strip test development. The assay performance was evaluated using serum samples diluted at 1:10 and 1:5 in sample buffer. For the capture zones, recombinant mu3ABC protein at concentrations of 0.5, 1, and 2 mg/mL was immobilized on the test (T) zone, while mouse IgG (1 mg/mL) served as the control (C) zone.

Furthermore, the protein G-conjugated gold nanoparticles (AuNPs) were evaluated at optical densities (OD520) of 2.8 and 4.2 (2× and 3× concentrations, respectively) to determine the optimal signaling intensity.

2.5. Preparation and Assembly of the Strip

The conjugate pad (K Biosciences, Pathumthani, Thailand) was prepared by coating it with protein G-conjugated colloidal gold nanoparticles using an XYZ3000 Dispense system (BioDot, Irvine, CA, USA) at a jetting rate of 10 µL/cm, followed by drying at 37°C for 2 h.

For the nitrocellulose (NC) membrane (CN140; Sartorius, Gottingen, Germany), mouse IgG (1 mg/mL) and recombinant mu3ABC protein (1.5 mg/mL) were dispensed at 1 µL/cm to form the control (C) and test (T) lines, respectively, using the same XYZ3000 system. The membranes were then immersed in a blocking solution (1% BSA, 0.05% Tween 20, and 3% sucrose in 0.15 M PBS) for 30 min, and subsequently dried at 37°C for 1 h.

Finally, the processed NC membrane, conjugate pad, sample pad (Ahlstrom, Helsinki, Finland), and absorbent pad (Cytiva, Marlborough, MA, USA) were assembled onto adhesive backing cards (Kenosha, Amstelveen, the Netherlands) with a 2-mm overlap between adjacent components. The fully assembled cards were then precision-cut into 4-mm wide strips using a CM4000 Guillotine cutter (BioDot, Irvine, CA, USA).

2.6. Assay Procedures

Ten-fold serum samples diluted with sample buffer were applied to the sample pad. After the sample had migrated to the absorbent pad, an additional 70 µL of sample buffer was added to the sample pad as a chaser. The results were visually observed within 20-30 minutes.

2.7. Comparison Study of mu3ABC Strip with an FMDV ELISA Kit

To evaluate diagnostic performance, 566 serum samples from cattle, pigs, and goats were tested. Each sample was analyzed using both the mu3ABC strip test and the ID Screen® FMD NSP ELISA kit (Innovative Diagnostics, France), the latter being performed according to the manufacturer’s instructions. For the ELISA, a sample-to-negative control percentage (S/N%) threshold of ≤ 50% was used to define positive results. This value represents the ratio of the optical density (OD) of the test sample to that of the negative control, multiplied by 100 to yield a percentage.

2.8. Statistical Analysis

Diagnostic sensitivity and specificity were calculated using 2×2 contingency tables. The agreement between the mu3ABC strip test and the commercial ELISA was evaluated using Cohen’s kappa coefficient [25], available in the GraphPad QuickCalcs online calculator (GraphPad Software, San Diego, CA, USA). Kappa values were interpreted as: 0.00–0.20 (slight), 0.21–0.40 (fair), 0.41–0.60 (moderate), 0.61–0.80 (substantial), and 0.81–1.00 (almost perfect).

3. Results

3.1. Comparison of mu3ABC Produced in E. coli and Insect Cells

Previously, we and other studies demonstrated that mu3ABC protein expressed in insect cells was a suitable antigen for development of ELISA to differentiate infected from vaccinated animals [17,19,22]. In this study, we attempted to decrease costs while increasing robust protein production by changing the mu3ABC expression host from insect cells to E. coli. We found that the recombinant mu3ABC expressed in both E. coli and insect cells had a molecular weight of approximately 55 kDa. Analysis by SDS-PAGE and western blot revealed that both E. coli and insect cell derived mu3ABC were expressed mainly as inclusion bodies within the insoluble fraction. Thus, the solubilization required urea at the concentration of 8 M. Western blot verified the specificity of the purified E. coli and insect cell derived mu3ABC against anti-FMDV antibodies in a convalescent bovine serum (Figure 1). The results confirmed that mu3ABC produced in insect cells could be replaced with that from E. coli.

3.2. Optimization of Different Parameters

To maximize diagnostic sensitivity and minimize background interference, key parameters were systematically optimized. Preliminary evaluations were conducted using hand-assembled test strips to determine the appropriate conditions for subsequent strip production. The optimized factors included the serum sample volume, the concentration of mu3ABC, and the amount of protein G-conjugated gold nanoparticles. Regarding the serum sample volume, a 1:10 dilution provided a clearer signal compared to a 1:5 dilution, which exhibited higher background interference. For the test line, mu3ABC was evaluated at concentrations of 0.5, 1, and 2 mg/mL; while 1 mg/mL provided a significantly clearer signal than 0.5 mg/mL, no further improvement in signal intensity was observed at 2 mg/mL (Figure 2). Therefore, 1 mg/mL was identified as the optimal concentration. Additionally, the amount of protein G-gold conjugate was adjusted to an OD₅₂₀ of 4.2, which resulted in optimal visual clarity.

3.3. Diagnostic Performance Across Species

The diagnostic performance of the mu3ABC strip test was evaluated across three animal species, using the commercial NSP ELISA as a reference (Table 1). In swine samples, the strip test demonstrated 100% sensitivity and specificity. A similarly high sensitivity of 100% was observed in goats, although the specificity was slightly lower at 91.51% due to nine ELISA-negative samples yielding positive results on the strip. In cattle, the strip test achieved 85.19% sensitivity and 88.61% specificity, compared to the NSP ELISA.

A total of 566 serum samples from various cloven-hoofed species, including cattle, pigs, and goats, were tested using the mu3ABC strip tests. FMDV-positive sera containing NSP-specific antibodies result in visible red bands at both the test and control lines, while negative sera show a visible band only at the control line (Figure 3). In parallel with the strip test, the sera were tested with the commercial ELISA kit. The sensitivity and specificity of the strip test, relative to the ELISA were 88.18% (179/203) and 92.56% (336/363), respectively. The mu3ABC strip test showed substantial agreement with commercial ELISA, yielding a Cohen’s kappa value of 0.805 with 95% CI of 0.754-0.856.

3.4. Sensitivity and Specificity of Immunochromatographic Strip

To further evaluate the diagnostic performance, the strip test was additionally tested using a panel of 200 reference sera from the RRL, yielding a sensitivity of 78% and a specificity of 91% (Table 2). It is important to note that variations in performance among serological platforms can occur due to differences in target epitopes and cut-off thresholds. This was evident as the commercial ELISA used in this study also classified only 69 out of the 100 positive reference samples as positive according to the manufacturer’s instructions. The identical Cohen’s kappa coefficient of 0.690 achieved by both the strip test and the commercial ELISA represents a substantial level of agreement with the reference standard, demonstrating that the mu3ABC strip test provides diagnostic reliability comparable to established commercial ELISA kits.

4. Discussion

Lateral flow immunochromatographic assays are widely used as rapid, cost-effective diagnostic tools for point-of-care testing. The LFA platform consists of three functional elements: an immobilized protein at the control line for assay validation, a specific antigen at the test line for target analyte capture, and a colloidal gold-labeled detector that facilitates visual observation. In a previous study, our team developed a recombinant 3ABC protein in which two cysteine residues at the 3C region (positions 142 and 163) were substituted with serine and glycine, respectively [24]. It was initially expressed using an insect cell system and successfully utilized as a diagnostic antigen in a 3ABC-based ELISA to differentiate infected from vaccinated animals [22]. Consequently, this modified 3ABC protein was selected as the immobilized antigen at the test line in the current study. While the initial production in insect cells took approximately 96 hours for expression, the E. coli expression system required an induction period of only 4 hours. Evaluation by SDS-PAGE and western blot confirmed that the quality and antigenicity of the recombinant proteins were comparable. Furthermore, bacterial culture media are much cheaper and simpler than the media used for insect cell cultivation. Accordingly, the E. coli system was adopted for mu3ABC production due to its significantly lower cultivation time and cost-effectiveness.

Maximizing the diagnostic performance in this study was achieved through the optimization of physicochemical parameters. Our results demonstrate that adjusting the antigen loading density on the test line, the concentration of the colloidal gold-mu3ABC conjugate, and the optimal sample volume was critical for maximizing binding efficiency. Furthermore, modifying the ionic strength and pH of the running buffer, along with adding surfactants such as Tween-20, significantly reduced non-specific binding and improved the signal-to-noise ratio. These practical optimizations provided a cost-effective route to increasing sensitivity, especially in challenging samples. In addition to these physical parameters, the assay could be further improved at the molecular level, such as employing computational epitope mapping to identify and exclude cross-reactive epitopes within the 3ABC polyprotein [26]. The optimization of protein orientation via site-specific biotinylation could also improve antigen-antibody binding efficiency [27]. With these fundamental parameters established, future integration of digital image-based quantification could serve as a robust form of signal amplification [28,29]. This would further enhance the detection of low-titer samples, providing a sensitive and objective DIVA diagnostic tool for large-scale FMD surveillance across diverse livestock species.

The versatility of the developed LFA is largely attributed to the use of Protein G-conjugated gold nanoparticles as a universal detector. Unlike conventional assays that rely on species-specific secondary antibodies, Protein G exhibits a high and broad-spectrum affinity for the Fc region of IgG across various species. By specifically targeting this portion, the antigen-binding sites remain accessible for efficient binding to the 3ABC antigen. This structural advantage ensures that the formation of the immunocomplex is not compromised, leading to high diagnostic sensitivity and providing the basis for a multi-species diagnostic platform. In this regard, while previous studies also developed NSP-based LFAs [20,21], their applications were primarily focused on FMDV antibody detection in pig sera. However, our mu3ABC-based LFA provides a broader application across three different livestock species, including cattle, swine, and goats. As a result, this species-independent platform offers a more practical and convenient way to monitor FMDV across different types of livestock.

The observed variations in sensitivity between the mu3ABC strip test and the NSP ELISA could be linked to the different physical environments of the antigen. In the strip test, the antigen is adsorbed and dehydrated on a nitrocellulose membrane matrix, whereas in the ELISA, it remains bound to a polystyrene surface within a buffered aqueous environment. These surface-dependent conformational changes, along with the shorter reaction times inherent to lateral flow assays, may affect epitope accessibility and binding affinity [30,31].

While NSP-based ELISA remains the gold standard for DIVA due to its high sensitivity, it requires specialized laboratory facilities, trained personnel, and time-consuming procedures. In contrast, lateral flow assays (LFAs) provide a rapid, point-of-care solution for field diagnostics. These assays enable immediate screening in resource-limited areas where laboratory access is constrained. In practice, the LFA can be implemented as an initial diagnostic tool for onsite decision-making at animal checkpoints or during suspected outbreaks. This facilitates real-time quarantine measures while samples are being transported for laboratory confirmation via ELISA. By combining the speed of the LFA with the high sensitivity of the ELISA, more effective outbreak management can be achieved.

The strategic advantage of the developed mu3ABC strip test lies in its ability to overcome the limitations of molecular diagnostics in long-term surveillance. While RT-qPCR is inherently capable of identifying infected animals, its application in DIVA is limited by the transient nature of viral shedding [32]. In contrast, our LFA identifies long-lasting NSP-specific antibodies, providing a broader diagnostic window even after molecular markers are no longer detectable. Furthermore, when compared to emerging point-of-care (POC) platforms such as RT-LAMP [33,34] CRISPR-based assays [35,36], or electrochemical biosensors [28,37] the mu3ABC LFA remains the most viable option for large-scale screening. Although isothermal and CRISPR methods offer high sensitivity for active viral detection, and biosensors provide quantification, their high manufacturing costs and logistical complexities remain significant barriers in endemic settings. Consequently, the mu3ABC LFA serves as proof that a cost-effective and instrumentation-free platform is still the most practical solution for FMD screening and surveillance.

5. Conclusions

The developed mu3ABC strip test demonstrates reliable diagnostic performance for the detection of FMDV NSP-specific antibodies in field conditions. With a diagnostic sensitivity of 88.18% and specificity of 92.56% relative to a commercial ELISA, and a substantial agreement (Kappa = 0.690) with the RRL reference standard, the test provides results comparable to established laboratory-based assays. The simplicity, rapid turnaround time, and lack of requirement for specialized equipment make the mu3ABC strip test a highly accessible and effective screening platform. By facilitating the differentiation of infected from vaccinated animals (DIVA), this test enables the rapid identification of infected individuals in the field, thereby significantly supporting FMD surveillance and control efforts.

6. Patents

A Thai petty patent application based on the results of this study has been submitted (Application No. 2603000415; Filing date: [30 January 2026]).