Bovine Genital Campylobacteriosis (BGC) is a venereal bacterial disease of cattle caused by Campylobacter fetus subsp. venerealis (Cfv) (OIE, 2021). Bulls act as reservoirs of the disease by carrying Cfv in the genital tract for long periods of time (Silveira et al., 2018). Infection of females occurs during natural breeding or artificial insemination, and causes endometritis, embryonic mortality and abortion, resulting in cow infertility, poor herd reproductive performance, and economic losses to the cattle industry (Mshelia et al., 2010; Michi et al., 2016).

Diagnosis of BGC requires accurate identification of the causative agent, which is challenging due to the two C. fetus subspecies that can be present in cattle, C. fetus subsp. fetus (Cff) and Cfv (Silveira et al., 2018). These subspecies have highly syntenic genomes and exhibit similar phenotypic traits, hampering their differentiation by molecular methods or phenotypic assays (Sprenger et al., 2012; Silveira et al., 2018). Microbiological culture followed by phenotypic identification is the classic approach for C. fetus identification and subspecies differentiation, as recommended by the Organization for Animal Health (OIE) (OIE, 2021). This differentiation relies on the 1% glycine tolerance test, in which Cfv is intolerant, while Cff is tolerant to glycine (OIE, 2021). Nevertheless, diagnosis of BGC by microbiological culture is challenging due to the fastidious growth and poor survival of the pathogen (Mshelia et al., 2010). On the other hand, the polymerase chain reaction (PCR) has emerged as a promising technique to differentiate C. fetus subspecies with the advantage of not relying on bacterial viability (McMillen et al., 2006; Silveira et al., 2018). Several assays have been developed targeting differences in genomic features such as the parA gene and the insertion element ISCfe1 (McMillen et al., 2006; Abril et al., 2007; McGoldrick et al., 2013; van der Graaf-van Bloois et al., 2013). However, these targets can be transferred horizontally, which can lead to lack of specificity when used for diagnostic purposes in clinical samples (Spence et al., 2011; Silva et al., 2020a; Polo et al., 2021). Recently, Cfv parA and ISCfe1 homologs were detected in another inhabitant of the bovine genital tract, Campylobacter portucalensis (Silva et al., 2020b), identifying this microorganism as a cause of false-positive results in molecular Cfv detection assays (Silva et al., 2020a). These reports highlight the importance of developing alternative molecular assays for reliable detection and differentiation of C. fetus subspecies.

Previous studies have shown that some single nucleotide polymorphisms (SNPs) in the core-genome of C. fetus differentiate Cff from Cfv (Abdel-glil et al., 2020). In this context, real-time PCR followed by High-Resolution Melting (HRM) analysis would allow the detection of such variations in amplicon sequences. This method is based on the amplification of a target of interest in the presence of a dsDNA-binding dye, which exhibits high fluorescence in the bounded state to dsDNA and low fluorescence when unbonded. The high-resolution melting follows the amplification step, with the gradual denaturation of the amplicons due to small increments in the temperature, which originates a melting profile specific of each product (Chua et al., 2015). The equipment captures changes in the fluorescence signal with high precision at different temperature points, detecting accurately differences in the melting behavior of sequences differentiated by only one SNP (Life Technologies Corporation, 2010). In the last years, this method has been employed as a tool for the identification and differentiation of pathogens (Chua et al., 2015; Zhang et al., 2021; Ghorbani et al., 2022; Pakbin et al., 2022). In this study, we developed two HRM assays to detect SNPs that identify and differentiate the C. fetus subspecies. These assays have the potential to be applied directly in the analysis of clinical samples.

The accurate identification of Cfv is crucial for the diagnosis of BGC since only subspecies venerealis is recognized as the etiologic agent of the disease (OIE, 2021). Misidentification of subspecies fetus as venerealis originates considerable economic costs related to testing, culling, and control strategies such as artificial insemination. On the other hand, misidentification of a Cfv as Cff perpetuates the disease in the herd with the associated costs related to decreased reproductive efficiency. In the last years, several real-time PCR assays have been developed to detect subspecies venerealis-specific sequences, such as the insertion element ISCfe1, parA, and virB11 genes (McMillen et al., 2006; McGoldrick et al., 2013; van der Graaf-van Bloois et al., 2013; Iraola et al., 2016). However, these sequences can be horizontally transferred and have been associated to specificity failures in real-time PCR assays (Spence et al., 2011; Silva et al., 2020a; Polo et al., 2021). Thus, accurate molecular diagnosis of BGC still requires the identification of molecular targets specific to Cfv.

A recent study based on whole-genome sequencing data identified SNPs differentiating ISCfe1 positive genomes, proposed as Cfv, from the remaining C. fetus strains (Abdel-glil et al., 2020). Real-time PCR coupled with HRM can differentiate SNPs and has emerged as a fast, easy to perform and cost-effective method for identification and differentiation of several bacterial pathogens (Zhang et al., 2021; Ghorbani et al., 2022; Pakbin et al., 2022).

In the present study, three of the SNPs proposed to differentiate the subspecies (Abdel-glil et al., 2020) were selected to develop real-time PCR assays coupled with HRM analysis to identify C. fetus subspecies. The most promising primer pairs target a Ppx/GppA family phosphatase (locus CFF8240_1016) and a Hit family protein (locus CFF8240_0641). Although these sequences differ by only one SNP, the melting behavior of the amplification products was significantly shifted, thus allowing subspecies differentiation. Both real-time PCR-HRM assays accurately identified the subspecies of 51 C. fetus strains, with unambiguously distinct melt curve profiles and melting temperature. Moreover, both assays revealed a good intra- and inter-assay reproducibility of the Tm values in all strains tested, evidenced by the low CV values. Although the Tm values showed slight differences between HRM runs, as observed in other studies (Naze et al., 2015; Ashrafi et al., 2017; Fehlberg et al., 2017), these differences were balanced by the inclusion of Cfv and Cff controls in each run. The subspecies classification was assigned based on the melting behavior of the Cfv and Cff controls included in each plate, whose inclusion is mandatory in all runs. We also evaluated strains with discrepant subspecies classification results in previous studies. Strains 98/v444 and BT 34/99 were here classified as Cfv, as indicated by Van Bergen et al. (2005), although they were typed as Cff in other studies (Wagenaar et al., 2001; Gorkiewicz et al., 2010). Strain 11800-21-2 was previously identified as Cfv (Gorkiewicz et al., 2010) but was in good agreement with other studies (Van Bergen et al., 2005; van der Graaf-Van Bloois et al., 2014, 2016a,b). The lack of standardized methods for subspecies classification and the absence of an explicit gold standard may be responsible for disagreeing classifications of C. fetus strains across studies. The developed HRM assays also have the potential to be implemented as an accurate method for the direct detection of Cfv in clinical samples. Both assays successfully detected Cfv in preputial samples spiked with 10³ CFU/mL. The suitability of the assays was also validated by the absence of non-specific amplification in preputial samples negative for C. fetus. Nevertheless, additional studies with samples from naturally infected animals and different matrixes, namely samples from aborted fetuses, should be tested to fully validate these assays for use in clinical samples. Additionally, the interlaboratory testing of these assays hereafter will be valuable to consider these assays as global diagnostic tools for the diagnosis of Bovine Genital Campylobacteriosis worldwide. Moreover, as we identified a polymorphism in the primer-binding site adjacent to the SNP in locus CFF8240_1016 in one bovine isolate (strain Cff 04/554), we cannot exclude specificity or sensitivity failures when other isolate collections or clinical samples are evaluated. This polymorphism may impact the amplification and/or melting temperature of the amplicons. In contrast, the assay targeting CFF8240_0641 proved to be effective without potential specificity issues, making it a preferential assay to be used for diagnosis.

This study also highlighted specificity failures of the glycine tolerance test, even when using standardized conditions such as inoculum size (10⁶ UFC/mL) and culture conditions. Although all Cfv were correctly identified by this phenotypic test, this would misidentify one Cff isolate. Previous studies already reported the occurrence of Cfv strains with tolerance to glycine (van der Graaf-Van Bloois et al., 2014, 2016a), which is acquired through mutation or transduction (Chang and Ogg, 1971), as well as Cff sensitivity to glycine (Wagenaar et al., 2001). Thus, this research also evidences the inconsistencies between the phenotypic analysis and the different molecular methods in the identification of C. fetus subspecies.

In conclusion, this study describes two real-time PCR-HRM assays for the highly specific and sensitive identification and differentiation of Cfv and Cff. Although exhibiting a similar performance in the present collection of strains, the assay targeting CFF8240_0641 is potentially more accurate due to possible, although presumably rare, polymorphisms in Cff strains. Importantly, the assays have the potential to be used for direct analysis of preputial samples and thus could prove to be a valuable tool for the diagnosis and control of BGC.

The original contributions presented in this study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

Ethical review and approval was not required for the animal study because bovine preputial samples were collected by certified veterinarians using the OIE recommended sampling method as part of the breeding soundness examination of bulls and as a clinical service requested by owners to the Faculty of Veterinary Medicine of the University of Lisbon. As samples were collected for diagnostic purposes, according to EU and national legislation (Directive 2010/63/EU and Decree-law no. 113/2013), no ethical approval from an Institutional Animal Care and Use Committee or other relevant ethics board was required. According to the publicly available regulation of the Veterinary Teaching Hospital of the Faculty of Veterinary Medicine of the University of Lisbon, all clinical and diagnostic procedures and records may be used for teaching and research purposes while maintaining confidentiality. Written informed consent was obtained from the owners for the participation of their animals in this study.

LL-D-C and ES: conceptualization, supervision, project administration, and funding acquisition. MS, LL-D-C, and ES: methodology and validation. MS, GP, LL-D-C, and ES: formal analysis. MS: investigation and writing—original draft preparation. SK, GP, LL-D-C, and ES: resources. SK, GP, LM, LL-D-C, and ES: writing—review and editing. MS, LM, LL-D-C, and ES: visualization. All authors have read and agreed to the published version of the manuscript.