Quadruplex PCR assay for identification of Corynebacterium pseudotuberculosis differentiating biovar Ovis and Equi

Background Corynebacterium pseudotuberculosis is classified into two biovars, nitrate-negative biovar Ovis which is the etiologic agent of caseous lymphadenitis in small ruminants and nitrate-positive biovar Equi, which causes abscesses and ulcerative lymphangitis in equines. The aim of this study was to develop a quadruplex PCR assay that would allow simultaneous detection and biovar-typing of C. pseudotuberculosis. Methods In the present study, genomes of C. pseudotuberculosis strains were used to identify the genes involved in the nitrate reduction pathway to improve a species identification three-primer multiplex PCR assay. The nitrate reductase gene (narG) was included in the PCR assay along with the 16S, rpoB and pld genes to enhance the diagnosis of the multiplex PCR at biovar level. Results A novel quadruplex PCR assay for C. pseudotuberculosis species and biovar identification was developed. The results of the quadruplex PCR of 348 strains, 346 previously well-characterized clinical isolates of C. pseudotuberculosis from different hosts (goats, sheep, horse, cattle, buffalo, llamas and humans), the vaccine strain 1002 and the type strain ATCC 19410T, were compared to the results of nitrate reductase identification by biochemical test. The McNemar’s Chi-squared test used to compare the two methods used for C. pseudotuberculosis biovar identification showed no significant difference (P = 0.75) [95% CI for odds ratio (0.16–6.14)] between the quadruplex PCR and the nitrate biochemical test. Concordant results were observed for 97.13% (338 / 348) of the tested strains and the kappa value was 0.94 [95% CI (0.90–0.98)]. Conclusions The ability of the quadruplex assay to discriminate between C. pseudotuberculosis biovar Ovis and Equi strains enhances its usefulness in the clinical microbiology laboratory. Electronic supplementary material The online version of this article (10.1186/s12917-017-1210-5) contains supplementary material, which is available to authorized users.


Background
Corynebacterium pseudotuberculosis is a facultative intracellular bacterium that is the causative agent of caseous lymphadenitis (CLA) in goats and sheep, abscesses and ulcerative lymphangitis in horses and oedematous skin disease (OSD) in Buffalo. It also causes sporadic infections in other species including cattle, camels, llamas, and humans [1][2][3][4]. C. pseudotuberculosis can be classified in two biovars, based on their ability to convert nitrate to nitrite, nitrate-positive strains are classified as biovar Equi and the nitrate negative ones as biovar Ovis [5]. In sheep and goats, CLA is predominantly caused by biovar Ovis strains, whereas horses and buffalos are mostly infected by biovar Equi strains [6][7][8]. Infection by C. pseudotuberculosis is distributed worldwide, causing significant disease in horse, sheep and goat herds [8,9]. The main economic losses attributed to C. pseudotuberculosis infection include decreased milk production, decreased weight gain, reduced value of hides due to scarring, and the cost of the drugs and labor needed to treat disease [9].
Direct and indirect tests to detect C. pseudotuberculosis have already been proposed, such as complement fixation test [10], synergistic hemolysis inhibition test [11], microagglutination assay [12], phospholipase D (PLD) antigen-based ELISA [13] and a multiplex PCR developed by our research group [14]. While these tests are useful for clinical diagnosis in diseased animals, none can differentiate the two biovars of C. pseudotuberculosis, which is currently only accomplished by biochemical tests. Differences between biovars are relevant for host and tissue specificity and appear to be associated with virulence [15,16].
Disease caused by C. pseudotuberculosis biovars has different clinical manifestations in the susceptible hosts [6,8,17,18], and biovar identification is important for understanding the epidemiology of infection, and consequently for disease control. Moreover, biovar identification can also have clinical implications. Since cattle can be infected by strains of both biovars, which may have different tissue preferences: biovar Ovis infects chiefly the mammary gland [17] and skin [18] and biovar Equi causes ulcerative lymphangitis and coronet lesions [9,19,20].
A dominant genetic characteristic that differentiates the biovars of C. pseudotuberculosis is the presence of the nitrate reduction operon in the biovar Equi strains [21]. Additionally, biovar Equi strains have 15 genes that are absent in biovar Ovis strains, including the narKGHJI operon, and a gene cluster encoding the molybdopterins moeB, moaE, molB, molA, moeY, moaC, moeA, and moaA and two hypothetical proteins [21].
Currently, only nitrate reduction test distinguishes C. pseudotuberculosis biovars Ovis and Equi [5,9]. The available phenotypic tests, performed in a laboratory or commercially available, for C. pseudotuberculosis identification are usually effective. However, the phenotypic tests can be expensive and unavailable for some laboratories that prefer to use molecular techniques. Currently, clinical microbiology laboratories are experiencing a change from classical to new diagnostic tools as PCR, real-time PCR, sequencing and MALDI-TOF mass spectrometry [22,23], which, due to the decrease in costs, may prefer to use molecular over biochemical tests. Moreover, the molecular tests are usually faster, easier and have less subjectivity in interpretation [24].
Our collaborative group has sequenced more than 60 C. pseudotuberculosis strains isolated from different hosts around the world, performing genomic, proteomic and clinical studies to not only to understand the pathogen but also try to find a way to control the spread of bacteria. Thus, the aim of this study was to develop a quadruplex PCR assay that would allow simultaneous detection and biovar-typing of C. pseudotuberculosis strains.

Nitrate reductase genes
Presence or absence of nitrate reductase genes were analyzed in nineteen C. pseudotuberculosis genomes (Table 1) in our previous work [21].

Bacterial strains and culture conditions
A total of 348 C. pseudotuberculosis strains, 346 field isolates [25][26][27], C. pseudotuberculosis ATCC 19410 T type strain, and C. pseudotuberculosis 1002 vaccine strain, were used in this study. These strains were obtained from the repository of the Laboratório de Genética Celular e Molecular, Instituto de Ciências Biológicas and of the Laboratório de Bacteriologia Aplicada, Escola de Veterinária of the Universidade Federal de Minas Gerais. C. pseudotuberculosis biovars Ovis and Equi were aerobically grown in brain heart infusion (BHI) (Acumedia Manufacturers, Baltimore, USA) agar plates at 37°C for 48 h. The C. pseudotuberculosis isolates were identified by standard biochemical tests [9,28,29]. Nitrate reduction was assessed using nitrate broth (Merck, Billerica, USA) and further reduction beyond nitrite was tested by addition of zinc dust (Sigma-Aldrich, St Louis, USA) [28].
Genomic DNA extraction, primers, and quadruplex PCR Genomic DNA extraction were performed according to the previously described protocol [30]. The oligonucleotide primers used in this study are listed in Table 2. Primers used to target 16S rRNA, rpoB, and pld genes of C. pseudotuberculosis were previously described [14,31,32]. Primers targeting the narG gene were designed by aligning the narG nucleotide sequences of C. pseudotuberculosis biovar Equi strains available from the whole genome sequenced strains (Table 1). Quadruplex PCR were carried out in a final volume of 50 μL, containing 20 ng of genomic DNA, 1 μM of each primer, 0.25 mM dNTPs, 1 units of Taq DNA polymerase (Life Technologies, Carlsbad, USA), 2 mM MgCl 2 , and 1X buffer (200 mM Tris-HCl pH 8.4, 500 mM KCl) (Life Technologies, Carlsbad, USA). Amplification was performed using the thermal cycler (PTC-100, MJ Research, Hercules, USA) as follows: the first denaturation at 95°C for 4 min; followed by 30 cycles of denaturation at 95°C for 30 s, annealing at 58°C for the 30s, and extension at 72°C for 1.5 min. The amplified products were submitted to electrophoresis in 1.0% agarose gel (w/v) in Tris-borate-EDTA (TBE) buffer (89 mM Tris Base, 89 mM Boric Acid and 2 mM EDTA pH 8.0), stained with 0.5 mg / mL ethidium bromide and visualized under UV light.

Sequencing of singleplex PCR products
In order to confirm the quadruplex PCR results, ten randomly chosen isolates were tested further in singleplex PCR assays with the four C. pseudotuberculosis-specific primer pairs. PCR products were purified using Agencourt AMPure XP (Beckman Coulter Company, Beverly, Massachusetts, USA) according to the manufacturer's instructions, and each product was sequenced in both directions using primers targeting the 16S rRNA, rpoB, pld and narG gene and the Big Dye V3.1 Terminator Kit (Applied Biosystems, USA) using an ABI 3500 DNA analyzer (Applied Biosystems, California, USA). Sequences were analyzed on the Geneious suite of molecular biology (http://www.geneious.com) with 16S rRNA (GenBank accession nos X81916, X81907, and X84255), rpoB (Gen-Bank accession no. AY492239), pld (GenBank accession nos L16586 and L16587) and narG (GenBank accession no AJF93840.1) as the reference genes.

Statistical analysis
Comparison between nitrate reduction test and quadruplex PCR was performed by McNemar's Chi-squared test, and the agreement was calculated using the kappa statistic. Statistical analysis were performed using the packages psych [33] and epibasix [34] on R software version 3.2.3 [35].

Results
Comparative genome analysis showed that C. pseudotuberculosis biovar Equi strains (258, 31, 262, MB20, E19 and CCUG27541) had narKGHJI gene clusters, however strains 1/06-A, 316, 162, and CIP52.97, although showing positive results in the nitrate reduction test, did not exhibit narKGHJI operon in their genome. On the other hand, genomic sequence analysis identified partial genes molB, narJ, moeA, and moeB in the strains 1/06-A, 162 and CIP52.97. Since the strains 1/06-A, 316, 162, and CIP52.97 were nitrate reductase positive in biochemical test and the genes were not identified in their genomes, primers to target narKGHJI cluster and narT gene were designed ( Table 2). The PCR tests (Additional file 1: Figure S1), sequencing and the optical map showed that the genes for narKGHJI and narT are present in the genomes of those strains (data not shown).
The multiplex PCR assay that targets 16S rRNA, rpoB and pld genes [14] was improved by the inclusion of C. pseudotuberculosis biovar-specific primers for the narG gene (narG - Table 2), in a novel quadruplex PCR assay (Fig. 1). The assessment of our quadruplex PCR assay was performed in a double-blind fashion. The results of the quadruplex PCR of the 348 previously well-characterized strains of C. pseudotuberculosis from different hosts (goats, sheep, horse, cattle, buffalo, llamas and humans) [21,[25][26][27] were compared to the results of nitrate reductase identification by biochemical test, and are shown in Table 3. The McNemar's Chi-squared test used to compare the two methods employed for C. pseudotuberculosis biovar identification showed no significant difference (P = 0.75), with an odds ratio of 1 (95% CI for the odds ratio: 0.16-6.14) between the quadruplex PCR and the nitrate biochemical test. Concordant results were observed for 97.13% (338/348) of the strains (Table 3), and the kappa statistic value was 0.94 [95% CI (0.90-0.98)], denoting excellent concordance between biochemical and molecular tests for nitrate reductase identification. The limit of detection of the new quadruplex PCR was 200 ng of DNA from C. pseudotuberculosis biovar Equi, which corresponds to approximately 100 bacteria.

Discussion
Previously, identification of C. pseudotuberculosis biovars was only possible only through the established procedures that included isolation and identification of the agent using biochemical tests such as the nitrate reduction test, which separates the nitrate-positive biovar Equi from nitrate negative biovar Ovis strains [5]. Herein, we developed, by the addition of a new oligonucleotide primer pair targeting the narG gene to the former multiplex PCR assay [14], a robust new assay for identification of C. pseudotuberculosis at species and biovar levels.
The comparative genome analysis showed in C. pseudotuberculosis biovar Equi strains (258, 31, 262, MB20, E19 and CCUG27541) narKGHJI gene clusters that participate via the respiratory anaerobic process of the nitrate reduction similar to Escherichia coli [21,36]. The C. pseudotuberculosis narKGHJI gene cluster showed significant similarity with the protein sequences found in other Actinomycetes, such as C. diphtheriae, C. glutamicum, and Mycobacterium tuberculosis. All C. pseudotuberculosis biovar Ovis strains do not present any gene of the narKGHJI operon in their genomes [21].
The nitrate locus in C. pseudotuberculosis is composed of the narKGHJI operon and by a cluster of genes encoding the molybdopterin moeB, moaE, molB, molA, moeY, moaC, moeA, and moaA (Fig. 2) [21]. Molybdopterin is a cofactor that is indispensable for the activity of nitrate reductase. In the narGHI complex, the narG gene is a member of a superfamily of enzymes that use a Molybdopterin-guanine-dinucleotide (Mo-bisMGD) cofactor (bisMGD) for their catalytic activity.
Our results showed that among the 348 C. pseudotuberculosis tested, only 10 strains (2.87%) showed differences between the biovar classification provided by molecular and biochemical tests (Table 3), which was observed even after repeating the assays. Furthermore, the kappa coefficient, which is a robust statistic that measures interrater agreement for qualitative items, confirms that regardless of the technique used for biovar classification the results obtained were very similar. It is also important to consider that kappa values range from −1 to +1, where 1 represents a perfect agreement between the raters, and 0.81 to 1.00 represents almost perfect agreement, such as the observed in our data (0.94) [28,37]. Discordance between both phenotypic and genotypic methods can be explained by the effect of environmental factors on gene expression [29,38] amino-acid substitution, the genetic background of the strain (that can vary according to geographical locations) and mutations. Considering discordances due to mutations genotypic tests have proven to be more reliable and sensitive as diagnostic tool than phenotypic tests [30][31][32][39][40][41][42]. Others studies also showed discordant results between genotypic versus phenotypic methods [30-32, 39-41, 43, 44].
To our acknowledgement, this is the first molecular approach able to clearly differentiate between C. pseudotuberculosis biovar Ovis and Equi, although different restriction patterns, ribotypes and ERIC-PCR clustering pattern have been associated to biovars [26,36,[45][46][47]. The advantages of multiplex PCR assay over biochemical tests are the speed, performance and reproducibility, and the ability to test large numbers of isolates simultaneously [26,45,46]. Identification is based upon the number and sizes of four products amplified by PCR. Moreover, the use of molecular techniques reduces the manipulation of viable bacteria in the laboratory and consequently the risk of accidental infection, as C. pseudotuberculosis can eventually be a zoonotic agent [2]. Moreover, this new diagnostic tool, the quadruplex PCR assay for identification and Fig. 2 Nitrate locus from C. pseudotuberculosis biovar Equi.This locus contains: the genes encoding the molybdopterin moeB, moaE, molB, molA, moeY, moaC, moeA and moaA and the genes encoding the nitrate reductase narK, narG, narH, narJ, narI. Insertion show between ansA and rpsH genes is lacking in nitrate negative C. pseudotuberculosis biovar Ovis strains. Arrows represent open reading frames and their orientations. Blue and pink: common genes shared between C. pseudotuberculosis biovar Ovis and biovar Equi strains. Pink: ribosomal proteins. Hatched: additional or different genes. Red: narKGHJI operon and grey: genes encoding the molybdopterin moeB, moaE, molB, molA, moeY, moaC, moeA and moaA biotyping of C. pseudotuberculosis follows the new trends on clinical microbiology laboratory that is currently incorporating more molecular biology tools in its routine [23,48]. In addition, despite not having been tested in the present study, due to its analytical sensitivity of 100 bacteria. |It is likely that this quadruplex PCR can also be applied to direct testing from clinical samples, as it has been done successfully for the three-primer (16S rRNA, rpoB, and pld) multiplex PCR [14]. The quadruplex PCR proposed in this study facilitates and deepens the level of identification of C. pseudotuberculosis strains at clinical microbiology laboratory, and thereby improves the diagnosis of infection by providing more information for decision making. These results are especially significant considering that C. pseudotuberculosis infects a wide range of hosts and produce different clinical manifestations. Furthermore, it was recently suggested that C. pseudotuberculosis biovars have differences at the molecular phylogenetic level, indicating an anagenesis process within the species [37,49]. The evolutionary analysis of conserved genes (rpoB, gapA, fusA, and rsmE) suggesting a gradual anagenesis of C. pseudotuberculosis in that study [49] substantially increases the importance of a molecular technique capable of efficiently separating the biovars Ovis and Equi. In fact, a pan-genome analysis of fifteen C. pseudotuberculosis strains showed a significant number of genes not shared by both biovars, including remarkable differences in the 16 detected pathogenicity islands [50].
In this study, it was also observed that some C. pseudotuberculosis strains (1/06-A, 316, 162 and CIP52.97) were able to reduce nitrate when tested by the biochemical approach and were also positive in the quadruplex PCR assay, albeit did not show in their genomes genes associated with nitrate reduction. The genomic analysis of nitrate locus identified that partial genes encoding the molybdopterin and narKGHJI operon of these strains was absent [21]. These results may have been due to low overall coverage, poor capture efficiency of certain regions, genomic regions that were previously not assembled or poorly assembled, including unambiguously aligning repetitive regions, such as transposons, and difficulty in unambiguously aligning repetitive regions [41,51]. Then, after resequencing of the narKGHJI operon region and optical mapping of these strains, it was observed that these strains have the nitrate locus in their genome and corrections on their information on GenBank are under way.

Conclusions
A novel quadruplex PCR assay for C. pseudotuberculosis species and biovar identification was developed. The nitrate reductase gene narG was included in the assay along with the 16S, rpoB and pld genes to improve the diagnosis of the multiplex PCR at biovar level. There was a significant concordance between the biovar classification provided by the molecular and biochemical test. The ability of the expanded quadruplex PCR assay to discriminate between C. pseudotuberculosis biovar Ovis and Equi strains enhances its value.