Surveillance of Contagious Equine Metritis: Results of the First 5-Year Period of French Proficiency Tests for Taylorella equigenitalis Detection by Real-Time PCR.

Contagious equine metritis (CEM) detection by PCR is recognized by the European Union according to Commission Implementing Regulation (EU) No 846/2014, and real-time PCR is now recommended by the World Organisation for Animal Health Terrestrial Manual at the same level as the culture method. The present study highlights the creation of an efficient network of approved laboratories in France in 2017 for CEM detection by real-time PCR. The network currently consists of 20 laboratories. A first proficiency test (PT) was organized by the national reference laboratory for CEM in 2017 to evaluate the performance of the early network, followed by annual proficiency tests organized for ongoing periodic assessment of network performance. Results of the 5 PTs organized from 2017 to 2021 are presented, during which 5 real-time PCRs and 3 DNA extraction methods were used. Overall, 99.20% of the qualitative data corresponded to expected results and the R-squared of global DNA amplification calculated for each PT varied from 0.728 to 0.899. DNA extraction is also an important step in the analytical process, and results were more favorable with direct lysis compared to column extraction. Focusing on the most commonly used PCR (PCR 1: 86.4% of results) showed lowest cycle threshold values with direct lysis compared to column and magnetic bead extractions, and with magnetic bead extraction compared to column extraction, but neither of these differences were statistically significant.

Whole genome sequence analysis of the 2018 Persian onager isolate suggests sublineages within the Taylorella asinigenitalis species.

In 2018, a T. asinigenitalis strain (MCE663) was isolated in a Persian onager tested for contagious equine metritis (CEM) in a United Kingdom (UK) zoo. This bacterium had never been reported in the UK and Multilocus Sequence Typing described a new atypically divergent ST (ST60). Although the causative agent of CEM is the bacterium Taylorella equigenitalis, a first natural outbreak of endometritis caused by T. asinigenitalis ST70 was reported in 2019, putting its pathogenic potential into question. In this context, we aimed to further sequence the T. asinigenitalis MCE663 genome and characterize the strain using phenotypical and genetic approaches. Results showed that it gathered all identification characteristics of T. asinigenitalis with smaller colonies and it was susceptible to all tested antibiotics. Genome-level phylogeny showed that the genome MCE663 formed a distinct phylogroup, and only shared ≈ 96.1% of average nucleotide identity (ANI) with the three published T. asinigenitalis genomes, which together shared ≈ 98.3% ANI. According to current cut-offs consensus for species and subspecies delineation (95% and 98%, respectively), our results support the first insights of a sublineage delineation within the T. asinigenitalis species.

Comparison of five basal compositions of selective chocolate agar media for isolation of Taylorella equigenitalis.

The gold standard method to isolate and identify Taylorella equigenitalis, the contagious agent of equine metritis, is the culture method according to the World Organisation for Animal Health Terrestrial Manual. No selective T. equigenitalis chocolate agar medium has been developed since the 1980s and the existing media show limited performances due to the fastidious nature of T. equigenitalis and the presence of interfering bacteria in the genital tract of equines. Here, the growth rates of 6 T. equigenitalis strains and 7 non-T. equigenitalis strains were compared on Timoney's selective medium formulated with 5 different basal agars (Columbia, Eugon, Blood, Mueller-Hinton and Tryptose Blood) provided by 2 to 4 suppliers per basal agar. The impact of glucose and/or Vitox supplementation was also investigated. Overall, the performance of selective T. equigenitalis media could be improved by substituting Eugon or Columbia agar with Blood, Mueller-Hinton or Tryptose Blood agar. It is nevertheless essential to validate the basal agar/supplier pair using a panel of T. equigenitalis strains. Furthermore, our findings confirm the need to supplement the selective media with a mixture of amino acids, nucleotides, and organic, mineral and vitamin compounds, translated here by Vitox supplementation.

Genome sequence of Taylorella equigenitalis MCE9, the causative agent of contagious equine metritis.

Taylorella equigenitalis is the causative agent of contagious equine metritis (CEM), a sexually transmitted infection of horses. We herein report the genome sequence of T. equigenitalis strain MCE9, isolated in 2005 from the urethral fossa of a 4-year-old stallion in France.

Preservation of viable Taylorella equigenitalis in different commercially available transport systems.

The cultural diagnosis of the causal agent of contagious equine metritis (Taylorella equigenitalis) using transport swabs is challenging. Swabs must be placed in Amies charcoal medium, refrigerated during transport, and plated out at the laboratory no later than 48 h after sampling. In this study, the viability of T. equigenitalis strain CIP 79.7T in 11 commercial swab transport systems was initially compared at 1 day and 2 days of storage at ambient (20 ± 3 °C) or refrigerated (5 ± 3 °C) temperature. The four best swab transport systems, systems B, E, F (used as the reference) and K, were then compared at 0, 2, 3, 4, 7 and 10 days at refrigerated temperatures. Statistically significant differences were observed after 10 days only for system K compared to the reference, with approximately 95% viable T. equigenitalis recovered in system K compared to approximately 77% in system F. System K is thus promising for preservation and transport of viable T. equigenitalis for culture.

Relationship between rifampicin resistance and RpoB substitutions of Rhodococcus equi strains isolated in France.

OBJECTIVES: Study of the rifampicin resistance of Rhodococcus equi strains isolated from French horses over a 20-year period. METHODS: Rifampicin susceptibility was tested by disk diffusion (DD) and broth macrodilution methods, and rpoB gene sequencing and MLST were performed on 40 R. equi strains, 50.0% of which were non-susceptible to rifampicin. RESULTS: Consistency of results was observed between rifampicin susceptibility testing and rpoB sequencing. Strains non-susceptible to rifampicin by DD had a substitution at one of the sites (Asp516, His526 and Ser531) frequently encountered and conferring rifampicin resistance. High-level resistance was correlated with His526Asp or Ser531Leu substitutions; low-level resistance was correlated with Asp516Tyr substitution, a novel substitution for R. equi. Strains susceptible to rifampicin by DD showed no substitution in the three sites, except for two strains carrying, respectively, the His526Asn and Asp516Val substitutions (previously correlated with low-level rifampicin resistance). Both strains were isolated from an animal from which ten other strains were also isolated and found to be rifampicin-non-susceptible by DD. MLST showed the presence of 10 STs (including the novel ST43), but no association was observed with rifampicin resistance. CONCLUSIONS: This study confirms that certain substitutions in RpoB are more likely to confer high- or low-level rifampicin resistance, describes a new substitution conferring rifampicin resistance in R. equi and suggests non-clonal dissemination of rifampicin-resistant strains in France. Standard DD may miss strains with a low-level rifampicin-resistant substitution; further studies are needed to remedy the absence of R. equi-specific clinical breakpoints.

Overview of spatio-temporal distribution inferred by multi-locus sequence typing of Taylorella equigenitalis isolated worldwide from 1977 to 2018 in equidae.

The accurate identification of Taylorella equigenitalis strains is essential to improve worldwide prevention and control strategies for contagious equine metritis (CEM). This study compared 367 worldwide equine strains using multilocus sequence typing according to the geographical origin, isolation year and equine breed. The strains were divided into 49 sequence types (STs), including 10 described for the first time. Three major and three minor clonal complexes (CCs), and 11 singletons, were identified. The genetic heterogeneity was low (0.13 STs/strain) despite the wide diversity of geographical origins (n = 16), isolation years (1977-2018) and equine breeds (n = 18). It was highest outside Europe and in the 1977-1997 period; current major STs and CCs already existed before 1998. Previous data associated the major CC1 with the first CEM outbreaks in 1977-1978 in the United Kingdom, Australia and the United States, and revealed its circulation in France. Our study confirms its circulation in France over a longer period of time (1992-2018) and its distribution in Spain and Germany but not throughout Europe. In addition to CC1, relationships between non-European and European countries were observed only through ST4, ST17 and ST30. Within Europe, several STs emerged with cross-border circulation, in particular ST16 and ST46 from the major complexes CC2 and CC8. These results constitute a baseline for monitoring the spread of CEM outbreaks. A retrospective analysis of a higher number of strains isolated worldwide between 1977 and the early 2000s would be helpful to obtain an exhaustive picture of the original CEM situation.

Acute Endometritis due to Taylorella equigenitalis Transmission by Insemination of Cryopreserved Stallion Semen.

Taylorella equigenitalis can be transmitted during artificial insemination. This report describes clinical T. equigenitalis transmission by cryopreserved stallion semen. T. equigenitalis isolates from a mare's vaginal discharge and semen from the same batch of the cryopreserved semen used for the insemination gave identical API ZYM, antibiotic susceptibility, and multilocus sequence typing results (ST-46); furthermore, the multilocus sequence typing lineage ST-46 is known to circulate in the country of semen collection. These results support the need for strict contagious equine metritis screening of processed semen before use for artificial insemination.

Towards European harmonisation of contagious equine metritis diagnosis through interlaboratory trials.

The performance of culture and PCR methods routinely used to diagnose contagious equine metritis (CEM) was evaluated and compared by two interlaboratory trials involving a total of 24 European laboratories, including 22 National Reference Laboratories for CEM. Samples were swab specimens artificially contaminated with bacteria present in the genital tract of Equidae, some with and some without Taylorella equigenitalis, the causative agent of CEM, and T asinigenitalis, responsible for possible misidentification as T equigenitalis Throughout both interlaboratory trials, PCR performed better in terms of specificity and sensitivity than the culture method, supporting the assertion that PCR should be accepted for CEM diagnosis. However, the culture performance during the second interlaboratory trial was better than during the first one, suggesting that the expertise of participants improved. This reveals the advantage of regular interlaboratory trials to constantly improve the expertise of laboratories. It also shows the need to develop new culture media that are more selective and/or better geared to the metabolism of T equigenitalis in order to improve the bacteriological diagnosis of CEM.

Draft Genome Sequence of Taylorella equigenitalis Strain MCE529, Isolated from a Belgian Warmblood Horse.

Taylorella equigenitalis is the causative agent of contagious equine metritis (CEM), a sexually transmitted infection of horses. We herein report the genome sequence of T. equigenitalis strain MCE529, isolated in 2009 from the urethral fossa of a 15-year-old Belgian Warmblood horse in France.

Development of a single multi-locus sequence typing scheme for Taylorella equigenitalis and Taylorella asinigenitalis.

We describe here the development of a multilocus sequence typing (MLST) scheme for Taylorella equigenitalis, the causative agent of contagious equine metritis (CEM), and Taylorella asinigenitalis, a nonpathogenic bacterium. MLST was performed on a set of 163 strains collected in several countries over 35 years (1977-2012). The MLST data were analyzed using START2, MEGA 5.05 and eBURST, and can be accessed at http://pubmlst.org/taylorella/. Our results revealed a clonal population with 39 sequence types (ST) and no common ST between the two Taylorella species. The eBURST analysis grouped the 27 T. equigenitalis STs into four clonal complexes (CC1-4) and five unlinked STs. The 12 T. asinigenitalis STs were grouped into three clonal complexes (CC5-7) and five unlinked STs, among which CC1 (68.1% of the 113 T. equigenitalis) and CC5 (58.0% of the 50 T. asinigenitalis) were dominants. The CC1, still in circulation in France, contains isolates from the first CEM outbreaks that simultaneously emerged in several countries in the late 1970s. The emergence in different countries (e.g. France, Japan, and United Arab Emirates) of STs without any genetic relationship to CC1 suggests the existence of a natural worldwide reservoir that remains to be identified. T. asinigenitalis appears to behave same way since the American, Swedish and French isolates have unrelated STs. This first Taylorella sp. MLST is a powerful tool for further epidemiological investigations and population biology studies of the Taylorella genus.

Genomic characterization of the Taylorella genus.

The Taylorella genus comprises two species: Taylorella equigenitalis, which causes contagious equine metritis, and Taylorella asinigenitalis, a closely-related species mainly found in donkeys. We herein report on the first genome sequence of T. asinigenitalis, analyzing and comparing it with the recently-sequenced T. equigenitalis genome. The T. asinigenitalis genome contains a single circular chromosome of 1,638,559 bp with a 38.3% GC content and 1,534 coding sequences (CDS). While 212 CDSs were T. asinigenitalis-specific, 1,322 had orthologs in T. equigenitalis. Two hundred and thirty-four T. equigenitalis CDSs had no orthologs in T. asinigenitalis. Analysis of the basic nutrition metabolism of both Taylorella species showed that malate, glutamate and alpha-ketoglutarate may be their main carbon and energy sources. For both species, we identified four different secretion systems and several proteins potentially involved in binding and colonization of host cells, suggesting a strong potential for interaction with their host. T. equigenitalis seems better-equipped than T. asinigenitalis in terms of virulence since we identified numerous proteins potentially involved in pathogenicity, including hemagluttinin-related proteins, a type IV secretion system, TonB-dependent lactoferrin and transferrin receptors, and YadA and Hep_Hag domains containing proteins. This is the first molecular characterization of Taylorella genus members, and the first molecular identification of factors potentially involved in T. asinigenitalis and T. equigenitalis pathogenicity and host colonization. This study facilitates a genetic understanding of growth phenotypes, animal host preference and pathogenic capacity, paving the way for future functional investigations into this largely unknown genus.

Analysis of plasmid diversity in 96 Rhodococcus equi strains isolated in Normandy (France) and sequencing of the 87-kb type I virulence plasmid.

To characterize the potential epidemiological relationship between the origin of Rhodococcus equi strains and the type of their virulence plasmids, we performed a comparative analysis of virulence plasmid types encountered in 96 R. equi strains isolated from (1) autopsied horses, (2) organic samples (horse faeces, manure and straw) and (3) environmental samples. Our results revealed no clear epidemiological link between virulence plasmid type and the origin of R. equi strains isolated from horse-related environments. To understand this result, we determined the nucleotide sequence of the second most frequently isolated virulence plasmid type: a 87-kb type I (pVAPA116) plasmid and compared it with the previously sequenced (and most commonly encountered) 85-kb type I (pVAPA1037) plasmid. Our results show that the divergence between these two plasmids is mainly due to the presence of three allelic exchange loci, resulting in the deletion of two genes and the insertion of three genes in pVAPA116 compared with pVAPA1037. In conclusion, it appears that the divergence between the two sequenced rhodococcal virulence plasmids is not associated with the vap pathogenicity island and may result from an evolutionary process driven by a mobility-related invertase/resolvase invA-like gene.