A p37-based ELISA used to monitor anti–Mycoplasma hyorhinis IgG in serum from pigs immunized with inactivated M. hyorhinis vaccines

Abstract

Mycoplasma hyorhinis is an important pathogen of swine that can often occur as a respiratory coinfection with viral pathogens, but can also cause arthritis and polyserositis in infected animals. To date, no assay is available to assess the serologic response to M. hyorhinis vaccines, to our knowledge. We used recombinantly expressed M. hyorhinis p37 protein to monitor the magnitude of the IgG response in vaccinated animals. The assay was able to distinguish animals vaccinated with M. hyorhinis from those vaccinated with the other important Mycoplasma species: M. hyopneumoniae and M. hyosynoviae. When formulated with an ideal adjuvant, inactivated vaccines designed to protect animals against M. hyorhinis induced a measurable and dose-dependent antibody response against the p37 protein. Additionally, the protein appears to be highly conserved between strains of M. hyorhinis isolated in the United States. The specificity of the assay as well as the conservation and immunogenicity of the p37 protein make it an ideal candidate antigen for use in measuring the immune response against M. hyorhinis after vaccination in weaned pigs.

Mycoplasma hyorhinis is an important pathogen in the U.S. swine industry. Each year, infections caused by these pathogen impact producers through increased veterinary costs, decreased productivity, and increased mortality. M. hyorhinis can be found as a commensal resident of the nasal cavity and respiratory tract of swine,³ and virulence and pathogenicity have been shown to vary among strains.² Occasionally, animals may also develop otitis and conjunctivitis as a result of infection.²

M. hyorhinis p37 is a 43.5 kD membrane-spanning protein that is thought to function in thiamine pyrophosphate binding⁹ and in adherence to host cells.⁵ An assay similar to the one that we report herein has been developed¹¹ for use with human sera to determine whether a link exists between M. hyorhinis and prostate cancer. In our study, the p37 protein was used as an antigen to measure the immunoglobulin G (IgG) response of swine to vaccination with inactivated, whole-cell M. hyorhinis vaccines. The assay was further validated to ensure that results of the test would not be confounded by vaccination with either M. hyopneumoniae or M. hyosynoviae.

In order to determine the suitability of the p37 protein as an antigen for use in monitoring antibody to multiple M. hyorhinis strains, the p37 gene from 16 strains contained within our collection was sequenced. Strains were chosen for sequencing in order to include multiple geographic locations over several years of sampling. We amplified the p37 gene from genomic DNA, and sequenced via the Sanger di-deoxy chain termination method. M. hyorhinis strains used either for sequencing or for vaccine production (Table 1) were grown in Friis medium as described previously⁶ and inactivated with binary ethylenimine. The p37 locus was amplified from 25 µL of culture fluid using p37Loc F (CAGAATCTATATCTAAGTTTAGGTTC) and p37Loc R (GTTGATCTAGATAATGCTAGCGTAAC) primers in 2× EconoTaq master mix (Lucigen, Madison, WI) and 50 nM of each primer in a 100-µL reaction. The PCR was performed using an initial incubation of 5 min at 94°C followed by 40 cycles of 94°C for 30 s, 50°C for 30 s, and 72°C for 1.5 min. Final extension of products was carried out at 72°C for 10 min. PCR products were sent to a core facility (Eurofins MWG Operon, Huntsville, AL) for sequencing using the same primers used in the PCR reaction. Translated p37 sequences were aligned using Clustal W. Regardless of time or place of origin, sequence homology within isolates was quite high, containing only one amino acid substitution among all strains tested. Relative to the sequence used to generate the recombinant protein (GenBank accession M37339.1), this change occurred at position 166 of the amino acid sequence where the majority of strains encode threonine, and strains NPL8 and NPL9 encode lysine. The sequence of the recombinant p37 protein used to coat ELISA plates differs from that of the field isolates reported herein at 6 amino acid residues, excluding the 6× histidine tag and enterokinase site introduced by the vector. These positions are 5F>L, 61D>G, 62K>E, 63E>K, 106S>F, and 256S>F in the field strains relative to the previously cited sequence. However, this did not appear to affect the outcome of testing given that serum from immunized animals was able to recognize the recombinant protein in the ELISA.

Table 1.

Summary of location and year of isolation for Mycoplasma hyorhinis strains.

NPL M. hyorhinis strain	Year isolated	Location isolated
1	2009	Saint Peter, MN
2	2010	Farmland, IN
3	2010	Jackson, MN
4	2010	Britt, IA
5	2010	Rose Hill, NC
6	2011	Guymon, OK
7	2011	Sleepy Eye, MN
8	2011	Clarks Grove, MN
9	2012	Guymon, OK
10†	2012	Jackson, MN
11	2012	Laurel, NE
12	2012	Guymon, OK
13	2012	Jackson, MN
14	2012	Jackson, MN
15	2012	Saint Peter, MN
16	2013	Saint Peter, MN

NPL = Newport Laboratories, Worthington, MN.

^†Strain used to formulate vaccine for our M. hyorhinis study.

A BLASTP¹⁰ search for proteins homologous to the M. hyorhinis p37 amino acid sequence² showed low homology with other functional protein sequences in the NCBI nr (non-redundant) database. Among the homologs found in the search, one originated from M. conjunctivae and the other from M. ovipneumoniae, which have not been isolated previously from swine. However, M. conjunctivae appears to be genetically related to other Mycoplasma species that infect swine.⁷ Homologous proteins from M. hyopneumoniae, M. hyosynoviae, and M. flocculare, a commensal strain commonly found in swine,¹ were also identified in the search. Nonetheless, the overall homology of all proteins identified in the database was low and lacked significant contiguous amino acid sequence identity. This indicated that p37 might serve as a good antigen for serosurveillance in pigs given that no conserved regions appear to be present in the protein when comparing p37 sequences between species.

Sequence information for M. hyorhinis p37 was derived from the NCBI database (GenBank accession M37339.1). This sequence was modified in order to optimize codon bias for expression in Escherichia coli and to eliminate opal stop codons that are translated as tryptophan in Mycoplasma species. The optimized sequence was synthesized by a commercial supplier (GenScript, Piscataway, NJ) and provided in a pUC57 vector. The sequence was amplified from the vector with primers p37 F (GACGACGACAAGATGCAGGCCTCTGCAGTCGAC) and p37 R (GAGGAGAAGCCCGGTTATTTAATCGCCTTTTCGTAG) using 2× EconoTaq master mix (Lucigen) and 50 nM of each primer in a 100-µL PCR reaction. PCR was performed using an initial incubation of 5 min at 94°C followed by 40 cycles of 94°C for 30 s, 50°C for 30 s, and 72°C for 1.5 min. Final extension of products was carried out at 72°C for 10 min. Once amplified, the PCR product was purified from the reaction mixture and ligated into a pET30-Ek/LIC expression vector (EMD Biosciences, San Diego, CA). The ligated vector was transformed into One Shot Top 10 E. coli cells (Thermo Fisher, Waltham, MA) by electroporation, and transformed clones were selected on lysogeny broth (LB) agar with 50 µg/mL of kanamycin. Clones were grown overnight in a selective medium (LB with 50 µg/mL of kanamycin), and purified plasmids were screened for the presence of insert. Plasmids from positive clones were sequenced at a core laboratory (Eurofins MWG Operon) in order to verify the integrity of the construct. A sequence-verified plasmid was then transformed into E. coli BL-21 (DE3) cells. Recombinant p37 protein was expressed using auto-induction media and purified from inclusion bodies. The purity of the recombinant protein was verified by western blot prior to developing the ELISA.

The ELISA was performed on porcine serum samples by diluting the purified p37 protein in carbonate–bicarbonate buffer and coating it onto the surface of 96-well plates (Immulon 2HB, Corning, Corning, NY) at a concentration of 100 µg/mL. Plates were allowed to incubate overnight at 4°C, and the coating solution was removed prior to blocking plates. Plates were blocked with 300 µL/well of Neptune Block (Immunochemistry Technologies, Bloomington, MN) and incubated for 2 h at 37°C before washing once with 0.05% polysorbate 20 in phosphate-buffered saline (PBST). Next, serum samples and controls were diluted 1:1,000 in Super Block (Thermo-Pierce, Rockford, IL), added to the plate in duplicate in 100-µL volumes, and incubated for 1 h at 37°C. Plates were washed 3 times in PBST. Goat anti-swine IgG horseradish peroxidase conjugate (Kirkegaard & Perry Laboratories, Gaithersburg, MD) was added to the plate at a 1:5,000 dilution and allowed to incubate for 1 h at 37°C. Plates were then washed 4 times in PBST and developed using a 2-component 3,3’,5,5’-tetramethylbenzidine (TMB) substrate (Kirkegaard & Perry Laboratories). The substrate was added at 100 µL/well and allowed to develop for 10 min at room temperature. The reaction was stopped by adding 100 µL of 1 N sulfuric acid to each well. Plates were read at a wavelength of 450 nm, and absorbance data were collected. The mean absorbance of the 2 assay control wells was determined and subtracted from that of every other well on the plate. Next, the mean absorbance of the duplicate samples and positive and negative serum controls was determined. Finally, the mean absorbance of every sample and the negative serum control was divided by that obtained from the positive control serum to obtain sample-to-positive (S/P) absorbance ratios that were used to evaluate the immune response in swine.

In order to determine whether the ELISA could be used to monitor the immune response against M. hyorhinis in swine, we obtained serum from a study that compared the performance of 2 different adjuvants. In that study, 119 high-health animals were obtained from a commercial supplier and housed in a clean facility until they reached 10 wk of age. These animals received 2 vaccinations of inactivated M. hyorhinis vaccine (Newport Laboratories, Worthington, MN) formulated at different dose levels using commercial adjuvants A (Newport Laboratories) and B (Merial, Athens, GA; Fig. 1). Three groups (J–L) were included for comparison to groups A–C to help control for any effects that concentrating antigen on a 100 kD ultra-filter may have had on the immunogenicity of the antigen. Groups A–F had 10 animals each, G–L had 9, and 5 animals were included as non-treated controls. Vaccinations were given at days 0 and 21 and delivered via the intramuscular route in a 2-mL dose. Serum samples were obtained from animals at days 0, 21, and 35 for testing in the ELISA.

Figure 1.

ELISA results for Mycoplasma hyorhinis vaccination study. White bars, cross-hatched bars, and black bars depict the mean sample-to-positive (S/P) ratio for each treatment group at days 0, 21, and 35 of the study, respectively. Error bars depict the standard error of mean of the associated data set. Columns are significantly different from one another if they do not share a letter. Asterisk (*) denotes the only group that was significantly different than the control group at day 21 of the study based on the results of the Dunnett post-hoc test. Groups included in the study were as follows: A = high dose, ~4× concentrated, 10% v/v adjuvant A; B = medium dose, ~4× concentrated, 10% v/v adjuvant A; C = low dose, ~4× concentrated, 10% v/v adjuvant A; D = high dose, 10% v/v adjuvant B; E = medium dose, 10% v/v adjuvant B; F = low dose, 10% v/v adjuvant B; G = high dose, 66% v/v adjuvant B; H = medium dose, 66% v/v adjuvant B; I = low dose, 66% v/v adjuvant B; J = high dose, not concentrated, 10% v/v adjuvant A; K = medium dose, not concentrated, 10% v/v adjuvant A; L = low dose, not concentrated, 10% v/v adjuvant A; and untreated controls.

A significant immune response was detected against the p37 protein (Fig. 1). The response in the ELISA diminished as the dose of the antigen was decreased, indicating that the assay may have value as a means of measuring the antibody response against M. hyorhinis semiquantitatively. We established an ELISA cutoff of 0.15 S/P; only one false-positive result was obtained from the controls. The false-negative rate (vaccinates that tested negative) was 11%. However, all vaccinated animals that tested negative received vaccine formulated with adjuvant A or received a reduced dose of adjuvant B at the lowest antigen dose.

It is common practice in the field to vaccinate pigs against M. hyopneumoniae. Therefore, sera from animals vaccinated with a commercial M. hyopneumoniae vaccine (Merial) were tested to ensure that vaccination status would not interfere with the M. hyorhinis assay results. Serum from 39 animals that received either a porcine circovirus 2 (PCV-2) vaccine (Merial) or an inactivated M. hyopneumoniae vaccine was tested in the assay. Animals were obtained from a high-health herd and vaccinated via the intramuscular route at days 0 and 21 using a 2-mL dose of the inactivated M. hyopneumoniae vaccine. Nineteen animals that received the PCV-2 vaccine were considered non-vaccinates. Serum taken when the animals were 8 wk of age at day 35 of the study protocol was tested using both the M. hyopneumoniae antibody ELISA (IDEXX Laboratories, Westbrook, ME) and the M. hyorhinis ELISA.

Similarly, given that M. hyosynoviae commonly colonizes the tonsils of pigs and also is a significant pathogen of swine,⁸ testing was conducted to determine if antibodies to this pathogen would be detected by ELISA. Fifty high-health animals were obtained from a commercial supplier and housed in a clean facility until they had reached 4 wk of age. The animals received 2 vaccinations of inactivated M. hyosynoviae vaccine (Newport Laboratories) formulated at different dose levels using 2 different adjuvants. Treatment groups A and B received undiluted inactivated antigen formulated with commercial adjuvant A (Newport Laboratories) or B (Merial), respectively. Groups C and D received antigen diluted in PBS to 25% of that delivered to groups A and B. Group C received antigen formulated with adjuvant A as above. Group D received antigen formulated with adjuvant B as above. Immunizations were given at days 0 and 21 via the intramuscular route in a 2-mL dose. Serum samples were obtained from animals at day 35 for testing in the M. hyorhinis ELISA as well as a M. hyosynoviae ELISA.⁸

Despite a strong antibody response against both M. hyopneumoniae and M. hyosynoviae in vaccinated animals, no significant increase in antibody titer was detected in the p37 ELISA as a result of vaccination (Fig. 2). On the contrary, p37 ELISA titers were significantly lower in M. hyopneumoniae–vaccinated animals. This indicates that antibodies made in response to vaccines formulated with either of these pathogens will not recognize the recombinant protein. However, we did observe an overall 77% seropositive rate for M. hyorhinis p37 antibody in the M. hyopneumoniae study and an 18% seropositive rate in the M. hyosynoviae study at day 35. Colonization with M. hyorhinis typically begins in swine at 6–10 wk of age, with higher colonization rates occurring after 10 wk.³ The pigs in our studies were 8 wk of age at the time serum was drawn for testing. Regrettably, we did not attempt to isolate M. hyorhinis from these pigs in order to confirm their colonization status. However, we hypothesize that this may be the result of natural colonization. Regardless of M. hyopneumoniae or M. hyosynoviae vaccination status, M. hyorhinis p37 ELISA titers were significantly lower than observed in M. hyorhinis–vaccinated animals, particularly when an optimal antigen dose and adjuvant was used.

Figure 2.

A. Results of Mycoplasma hyopneumoniae vaccination study. White bars depict the results of the M. hyorhinis p37 ELISA. Black bars depict the results of the M. hyopneumoniae ELISA. Results from the non-vaccinates and vaccinates from each assay were compared using the Student t-test and were significantly different (α = 0.05). B. Results of Mycoplasma hyosynoviae vaccination study. Groups included in the study were as follows: A = full dose, 10% v/v adjuvant A; B = full dose, 66% v/v adjuvant B; C = 25% v/v antigen dose, 10% v/v adjuvant A; D = 25% v/v antigen dose, 66% v/v adjuvant B. White bars depict the results of the M. hyorhinis p37 ELISA. Black bars depict the results of the M. hyosynoviae ELISA. Asterisks (*) indicate groups that were significantly different than the controls based upon the results of the Dunnett post-hoc test (α = 0.05). Error bars in each panel depict the standard error of mean of the associated dataset.

We were able to measure the magnitude of the immune response using the p37 ELISA and demonstrate a significant increase in the humoral response against p37 protein when comparing 2 adjuvants. Clearly, adjuvant B resulted in a superior immune response. This result should enable a significant improvement in measurement of vaccine performance during future development. The vaccines used in our study were inactivated whole-cell vaccines, and one might expect that an increase in titer against an antigen such as p37 might correlate to an increase in titer against other virulence factors present on the surface of Mycoplasma cells. Our assay relies on a very strong positive control, commonly having an absorbance of ~2.0 in the test. Therefore, it could likely be made to be more sensitive if the control were replaced with weaker serum allowing for higher S/P values in the test samples.

Interestingly, the results of the M. hyopneumoniae trial revealed a decrease in antibody response against the p37 protein when animals had been immunized with an inactivated M. hyopneumoniae vaccine. Despite the low background titer present, the results were statistically significant (α = 0.05). Using the established S/P value, vaccination with the inactivated M. hyopneumoniae vaccine reduced seroprevalence in the M. hyorhinis ELISA from 84% positive in non-vaccinates to 70% in vaccinates. The overall seroconversion rate was higher in the M. hyopneumoniae study than in any of the other studies that we conducted, suggesting that these animals may have been colonized by M. hyorhinis by the end of the study. Therefore, it is tempting to suggest that antibodies made in response to the M. hyopneumoniae vaccine may have been able to help prevent colonization with M. hyorhinis strains present in the pig’s environment. Indeed, a study conducted using serum from M. hyopneumoniae challenged animals in an ELISA that used Tween20-extracted M. hyorhinis antigen demonstrated that antibodies made in response to challenge may cross-react with M. hyorhinis.⁴ Further research will need to be done to determine whether this response is protective when animals are presented with a biologically meaningful M. hyorhinis challenge and what antigens this response may be targeting in M. hyorhinis. It is also possible that the seroconversion that we observed is the result of low-level cross-reactivity against other species of Mycoplasma present in the pig’s environment given that we did not attempt to isolate M. hyorhinis from these pigs. However, the fact that M. hyorhinis p37 antibody titers are higher in the controls than in animals vaccinated with either M. hyopneumoniae or M. hyosynoviae implies that this antibody does not originate from exposure to either of these species.