To develop and validate the Pictor PictVet™ Mycoplasma bovis Multiplex ELISA serum assay, we used a total of 180 heat-treated bovine serum samples, previously characterized using the commercially available ID Screen® Mycoplasma bovis Indirect ELISA (IDvet), which served as the reference test and was performed according to the manufacturer’s instructions. Fifty-six of these samples were collected in late 2022 from spring-calving herds located on a South Island farm in New Zealand. The cohort included both male and female Jersey and Friesian cattle spanning a range of ages, from calves to adults. Animals were sampled during routine surveillance testing under the National M. bovis Eradication Programme. All animals appeared clinically healthy at the time of collection, suggesting subclinical infections. These 56 samples tested positive on the IDvet ELISA and were subsequently evaluated on Pictor’s multiplex platform between January and February 2023. The remaining 124 samples, which tested negative on the IDvet assay, were randomly selected from archived serum specimens. These samples were sourced from M. bovis-free herds across various geographic regions in New Zealand. The cattle represented a mix of sexes and ages, similar to the positive cohort. Both collection and testing of these samples occurred in January 2023.

All samples were acquired via Assure Quality, a New Zealand government-owned national laboratory involved in diagnostic testing under the National M. bovis Eradication Programme. The samples were stored at the Physical Containment Level 2 (PC2) facility of SVS Laboratories in Hamilton, New Zealand. All serum samples were heat-treated and stored at −80 °C until testing to preserve antibody integrity and mitigate any risk of antibody decay. The timeframe between sample collection and testing ranged from 1 to 2 months for both cohorts, ensuring minimal degradation and maintaining sample quality for accurate assay performance evaluation.

The laboratory staff conducting the Pictor PictVet™ ELISA were not formally blinded to the IDvet comparator test results during assay evaluation.

The assay incorporated two distinct surface proteins of M. bovis – MilA (MBOVPG45_0710), which has previously been shown to be useful as the basis for highly sensitive and specific conventional indirect ELISAs (12–18), and a novel antigen, K310 (MBOVGP45_310). Recombinant glutathione S-transferase (GST) fusion proteins were expressed in and purified from E. coli carrying pGEX expression plasmids encoding the mature M. bovis MilA and M. bovis K310 proteins (17). The manufacturer (GenScript) confirmed the identity of the proteins by SDS-PAGE and Western blotting using anti-GST monoclonal antibodies (mAb). The identity of each of the proteins was further confirmed using antigen-specific monoclonal antibodies in the Pictor PictVet™ M. bovis Multiplex ELISA.

Monoclonal antibodies specific to Mycoplasma bovis antigens MilA and K310 were custom-developed by GenScript. Purified MilA and K310 antigens, expressed in E. coli BL21 Star™ (DE3) after cloning into the pGEX-4 T-1 vector, were used to immunize New Zealand White rabbits up to three times. Peripheral blood mononuclear cells were then isolated, and antigen-specific B cells were identified and selected. The variable regions of the selected monoclonal antibodies (VH and VL) were sequenced and retained. These rabbit-derived variable regions were genetically fused to bovine IgG constant regions (CH and CL) to generate rabbit–bovine chimeric antibodies, ensuring compatibility with bovine-specific secondary detection antibodies. Rabbit VH/VL and bovine CH/CL sequences were cloned into pcDNA3.4 expression vectors incorporating optimized signal peptides and Kozak consensus sequences. These constructs were co-transfected into Chinese Hamster Ovary (CHO) cells using GenScript’s TurboCHO™ system, employing an optimized heavy-chain to light-chain plasmid ratio to maximize antibody expression. Recombinant antibodies were produced in CHO cells, and the purified supernatant was used for assay development and specificity testing. Anti-bovine IgG secondary antibodies targeting the Fc region were used to avoid interference with antigen-binding domains during detection.

PictArray™ Multiplex ELISA plates were prepared by dispensing 10 nL of each of the M. bovis antigens, MilA and K310, onto a 96-well ELISA plate, as shown in Figure 1A. Additionally, bovine IgG was included as part of the panel to serve as an internal positive control. A print buffer control was included to demonstrate the background signal intensity due to the print buffer and/or antigen carry over. The spots were printed in duplicate. PictArray™ Multiplex ELISA plates were then incubated with a blocking solution to reduce non-specific binding of antibodies to the ELISA plate.

A conventional indirect ELISA was performed using PictArray™ Multiplex ELISA plates. Briefly, 100 μL of a 1:100 dilution of each of the serum samples was added to each of the wells and the plate incubated at 37 °C for 30 min. The plates were then washed 5 times. After washing, 100 μL of horseradish peroxidase-conjugated sheep anti-bovine IgG heavy and light chain antibody was added and the plate was incubated at 37 °C for 30 min. The plates were washed 5 times, and 100 μL of TMB substrate was added and the plate incubated at room temperature for 20 min. The plates were then read using the PictImager™ (sciREADER CL2 colorimetric plate reader), which captures high-resolution images of the spots. The spot signal intensities were then measured using Pictorial™ software.

As the ID Screen® Mycoplasma bovis Indirect ELISA (IDvet) was used as a comparator test, positive agreement, negative agreement, and overall agreement were calculated to assess concordance with the comparator test. The positive percent agreement (PPA) was defined as 100% x a / (a + c), and the negative percent agreement (NPA) as 100% x d / (b + d), and overall agreement as 100% x (a + d) / (a + b + c + d), where a = the number of samples positive in both tests, b = the number of samples positive in the PictVet test but negative in the IDvet test, c = the number of samples negative in the PictVet test but positive in the IDvet test, and d = the number of samples negative in both tests. These metrics were calculated to evaluate the test’s diagnostic performance in the absence of a recognised reference standard.

The limit of detection (LoD) was determined empirically following the CLSI EP17-A guidelines (19, 20). First, the Limit of Blank (LoB) was established by analyzing blank samples and determining the 95th percentile of the blank signal distribution. To estimate the LoD, a low concentration sample was measured in 20 independent replicates over three non-consecutive days. The LoD was defined as the lowest concentration at which ≥95% (at least 19 out of 20) of the replicate measurements produced a signal greater than the LoB. This approach does not assume a normal distribution. The LoD was determined using two monoclonal antibodies, one specific for MilA and one specific for K310. Based on initial two-fold serial dilutions, three concentrations close to the LoB were selected for the analytical sensitivity study. Each of the three concentrations was assayed in 20 replicates, and the LoD was defined as the concentration at which 95% of the 20 samples had a signal intensity greater than the LoB. This assay was repeated three times on different days.

Pictor PictVet™ Mycoplasma bovis Multiplex ELISA results were reported both as quantitative spot signal intensities and as qualitative outcomes (positive or negative). The assay includes two antigens, MilA and K310—each printed as a discrete spot within the same well. Cut-off values were determined based on the median spot colour intensity (unitless values generated by image analysis software) and set at 56.75 for MilA and 41.59 for K310. A sample was considered positive if the median spot intensity exceeded the cut-off for either antigen. If both values fell below their respective thresholds, the sample was interpreted as negative.

Analyses were performed using GraphPad Prism version 10.5.0 (GraphPad Software, San Diego, CA). The software was used to calculate summary statistics such as means, standard deviations, and confidence intervals, where applicable.

To develop a highly sensitive and specific assay for screening for antibodies against M. bovis, we designed a multiplex ELISA assay, incorporating two different M. bovis antigens, MilA and K310, using Pictor’s multiplex ELISA platform (Figure 1A). The figure shows where each antigen was dispensed into each well of a 96-well ELISA plate to create the multiplex M. bovis test panel. Specific binding was visualized as a stained spot on the multiplex ELISA plate. We hypothesized that incorporating two distinct M. bovis antigens would improve the assay performance, as the use of two antigens was likely to increase the likelihood of detecting M. bovis-specific IgG antibodies in each animal, as antibody responses against two distinct surface proteins of M. bovis could be detectable in each sample. To confirm the specificity of the Pictor PictVet™ Mycoplasma bovis Multiplex ELISA, monoclonal antibodies against MilA (Figure 1B) and K310 (Figure 1C) were first used to validate the presence of these antigens on the assay platform. Polyclonal bovine IgG spots were also included to demonstrate the absence of any specific binding in wells where no M. bovis-specific antibodies were present (Figure 1D). This negative control established the baseline for background signal intensity, ensuring that any signal detected in the assay was due to specific antigen–antibody interactions and not due to non-specific binding.

To test our hypothesis, the diagnostic performance of the Pictor PictVet™ Mycoplasma bovis Multiplex ELISA was compared to the IDvet ELISA kit. There was a distinct separation between the IDvet+ and IDvet- samples when the results from assaying them in the Pictor PictVet™ Mycoplasma bovis Multiplex ELISA were plotted based on spot signal intensities (Figures 1E,F). Reactions with both the MilA and the K310 antigens in the assay differentiated between positive and negative serum samples. The positive samples had higher signal intensities for both the MilA and K310 antigens than the negative samples. The well-defined separation of these two groups indicated that the assay could distinguish between infected and uninfected animals.

The diagnostic performance of the Pictor PictVet™ Mycoplasma bovis Multiplex ELISA was evaluated by comparing it to the commercially available IDvet ELISA, which served as the comparator method (Table 1). A total of 180 serum samples were tested. The Pictor PictVet™ ELISA correctly identified all 56 IDvet-positive samples (PPA of 100%) and 122 of 124 of the IDvet-negative samples (NPA of 98.4%), yielding an overall agreement of 98.9%. These results demonstrated strong agreement with the IDvet ELISA and support the potential of the Pictor PictVet™ Multiplex ELISA as a reliable alternative for detecting antibodies against M. bovis in cattle.

To assess individual antigen performance within the Pictor PictVet™ Mycoplasma bovis Multiplex ELISA, the two target antigens, MilA and K310, were evaluated using the IDvet ELISA as the comparator test (Table 2). The MilA antigen alone identified 50 of 56 positive samples (PPA of 89.3%), with only one false positive (NPA of 99.2%). The K310 antigen alone identified 54 of 56 positive samples (PPA of 96.4%) with only one false negative (NPA of 99.2%). However, the MilA/K310 combination yielded a higher overall performance, detecting all 56 positive samples (PPA of 100%), with only two false positives (NPA of 98.4%). These findings indicated that, while both individual antigens perform well, the MilA/K310 combination provides the best diagnostic accuracy for the detection of Mycoplasma bovis specific antibodies, supporting the inclusion of both antigens as core components of the multiplex assay.

To compare the analytical sensitivity of the Pictor PictVet™ Mycoplasma bovis Multiplex ELISA with the commercially available IDvet ELISA, monoclonal antibody (mAb) titration experiments were performed targeting the MilA and K310 antigens. As shown in Figure 2A, both assays yielded similar titration curves across a wide range of anti-MilA mAb concentrations. However, at higher antibody concentrations, particularly at 1 μg/mL, a decreased signal was seen in the IDvet ELISA compared to that seen at 0.5 μg/mL, suggesting a possible prozone effect. In contrast, in the Pictor PictVet™ ELISA, the signal was saturated at 1 μg/mL, but there was no reduction in signal intensity, suggesting it was more robust at high antibody concentrations. At the lower end of the titration curve, the Pictor platform appeared to detect signals at lower mAb concentrations than the IDvet assay, suggestive of improved sensitivity and an enhanced dynamic range.

To determine the LoD, twenty replicate measurements were conducted at mAb concentrations near the LoB and assayed over three independent days (Figures 2B–D). The LoD was defined as the lowest concentration at which ≥ 95% of replicates yielded a signal above the LoB across all assay runs. The Pictor PictVet™ ELISA had LoDs of 0.97 ng/mL for the MilA antigen (Figure 2B) and 0.48 ng/mL for the K310 antigen (Figure 2C). In contrast, the IDvet ELISA had a higher LoD of 1.95 ng/mL (for the MilA monoclonal antibody) (Figure 2C). These results indicate that the Pictor PictVet™ Multiplex ELISA may have superior analytical sensitivity, an extended dynamic range, and greater assay stability at both high and low antibody concentrations, supporting its utility as a robust and reliable diagnostic platform for detecting Mycoplasma bovis-specific antibodies in cattle.