Background

Trueperella pyogenes is a well-known pathogen of domestic ruminants and pigs causing mastitis, abortion and a variety of pyogenic infections [1]. As summarised by Jost and Billington [2] this bacterial pathogen is also able to cause diseases in a large number of different animal species including antelopes, bisons, camels, chickens, deer, elephants, gazelles, horses, macaws, reindeer, turkeys and wildebeest and also in companion animals such as dogs and cats. In 2010, Ülbegi-Mohyla et al. [3] characterised two T. pyogenes isolated from septicaemia of a gecko and a bearded dragon both phenotypically and genotypically. In 2012, Oikonomou et al. [4] reported that T. pyogenes is one of mainly frequently isolated pathogens from the investigated mastitis cases. The T. pyogenes acts as an opportunistic pathogen, causing endometritis in dairy cattle once the protective epithelium has been lost after parturition [5]. In later research, the complete genome sequencing of T. pyogenes was undertaken from a field isolate of a dairy cow suffering from metritis [6, 7] and from infected goats [8]. Grey Slender Lorises (Loris lydekkerianus nordicus) are a primate species from the family Lorisideae whose taxonomy is currently under revision. Their habitat is East and South India as well as Sri Lanka [9]. Grey Slender Lorises are primarily insectivorous nocturnal animals with loose social interactions. They forage on trees in dry zone forests where they also sleep in aggregations of several animals. In 2012, a T. pyogenes 11-07-D-03394 was isolated from a facial abscess of a six-year-old Grey Slender Loris kept in a terrarium at Frankfurt Zoo [10].

Case presentation

The Grey Slender Lorises in the present study originated from a European Association of Zoo and Aquaria (EAZA) breeding programme and were living at Frankfurt Zoo. The Lorises were kept in pairs in the nocturnal animal house. In October 2011, T. pyogenes 11-07-D-03394 was isolated, as previously described [10], from a facial abscess of a six-year-old male Grey Slender Loris. In May 2012, a second T. pyogenes 121,008,157 was isolated from a nasal swab of a ten-year-old male Grey Slender Loris which was living in the same terrarium and was suffering from purulent rhinitis. The T. pyogenes 121,008,157 was isolated together with Enterococcus spp., Pasteurella spp., Pseudomonas aeruginosa and coliform bacteria. After two weeks, this Grey Slender Loris died and T. pyogenes was isolated postmortem together with Enterococcus faecalis, Klebsiella pneumoniae and Escherichia coli from the skull of this animal. E. faecalis, K. pneumoniae and E. coli were additionally isolated from the liver, kidney, lung, spleen, tongue, brain and nasal mucosa. The post mortem analysis of the animal revealed a catarrhal purulent rhinitis, a catarrhal purulent sinusitis associated with an extended purulent catarrhal pneumonia, an acute fibrinous-purulent pericarditis as well as an interstitial nephritis. The animal had also lost two upper left molars and three top right molars. Later on, in December 2012, a third Grey Slender Loris (9.5 year old female) died in the same terrarium also from bacterial septicaemia following a dental abscess. In postmortem microbiological examinations, the T. pyogenes was isolated together with E. faecalis from the liver, kidney, lung, intestine, vagina and orbita and in high numbers from paranasal sinuses and dental cavities. The T. pyogenes isolate 121,018,522 used for further studies was isolated from the paranasal sinuses of this animal. Post mortem analysis of the animal also revealed a chronic nephritis of both kidneys. The animal had similar teeth problems as the two previously mentioned lorises. The bacteriological carcass inspection of the deceased lorises found T. pyogenes to be the predominating bacteria. In the present study the first isolate T. pyogenes 11-07-D-03394 obtained from the first case and the two T. pyogenes isolates (121,008,157 and 121,008,522) obtained from the second and third cases were identified both phenotypically and genotypically. Additionally, all three isolates were investigated for epidemiological relationships by means of various genotypical tests. For control purposes the strains T. pyogenes DSM 20630T, T. pyogenes DSM 20594 and Arcanobacterium haemolyticum DSM 20595T were included.

Phenotypic and genotypic identification

Both newly isolated T. pyogenes were identified phenotypically as described by Eisenberg et al. [10]. A genotypic identification of all three T. pyogenes isolates was subsequently performed by sequencing the 16S rRNA gene [10,11,12,13] and the glyceraldehyde-3-phosphate dehydrogenase encoding gene gap [13]. Furthermore, the three T. pyogenes isolates were genotypically identified by amplifying the T. pyogenes species-specific region of the 16S-23S rRNA intergenic spacer region (ISR) and the T. pyogenes species-specific region of the superoxide dismutase A encoding gene sodA. The three T. pyogenes isolates were additionally characterised by amplifying of several known and putative virulence factor encoding genes. These virulence genes included gene plo encoding pyolysin, gene cbpA encoding a collagen-binding protein, gene nanH encoding neuraminidase H, gene nanP encoding neuraminidase P and the fimbriae encoding genes fimA, fimC and fimE [3, 11, 12]. In addition, tetracycline resistance encoding gene tet(W) was amplified as described by Zastempowska and Lassa [14].

Genomic fingerprinting

The genomic fingerprinting of the three T. pyogenes isolates originating from the lorises was performed using four genomic DNA fingerprinting methods. These included three repetitive element primed (rep)-PCRs [ERIC-PCR, BOX-PCR and (GTG)5-PCR] and random amplification polymorphic DNA (RAPD-PCR) analysis as described in detail by Glaeser et al. [15]. The sequences of the oligonucleotide primers and the thermocycler programs are summarised in Table 1. Cluster analysis of genomic fingerprinting patterns was performed in GelCompar II version 4.5 (Applied Maths) using the unweighted pair-group method using the arithmetic average (UPGMA) clustering method based on the Pearson correlation (0.5% optimisation; 1% position tolerance), which considers the presence/absence and the intensity of DNA bands. A consensus matrix was calculated and a composite clustering was performed.

Table 1 The Oligonucleotide primer sequences and PCR conditions used in the present study
Full size table

Multilocus sequence analysis (MLSA)

This represents the first MLSA-based study applied to field isolates of the species T. pyogenes. The MLSA was performed with the target genes translation elongation factor G encoding gene fusA, translation elongation factor Tu encoding gene tuf, methionyl-tRNA synthetase encoding gene metG and DNA gyrase, subunit A encoding gene gyrA. The sequences of the oligonucleotide primers for amplifying the four housekeeping genes were designed with the sequence data of the A. haemolyticum DSM 20595T genome project [16]. The target genes, the sequences of the oligonucleotide primers and the thermocycler programs are summarised in Table 1. Sequences of partial genes were concatenated in the following order: fusA (746 nt), tuf (795 nt), metG (836 nt) and gyrA (937 nt). Analyses were performed at the nucleotide and amino acid sequence level with T. pyogenes DSM 20630T, T. pyogenes DSM 20594 and A. haemolyticum DSM 20595T as controls. MLSA analysis was performed using the MEGA version 6 [17]. Full-length gene sequences from the genome of A. haemolyticum DSM 20595T (CP002045) were used as reference sequences to obtain the correct open reading frame (ORF) for translation into amino acid sequences. Alignments of nucleotide and the amino acid sequences were performed with MUSCLE implemented in MEGA version 6 [18]. The phylogenetic trees in single gene base analysis of the four target genes, respectively, were constructed with the neighbour-joining method [19] using the Kimura-2-parameter model for nucleotide sequences [20] or the JTT matrix-based method for amino acid sequences [21]. The phylogenetic trees of concatenated sequences were constructed using the maximum-likelihood method based on evolutionary distances calculated with the general time reversible model for nucleotide sequences [22] and again with the JTT matrix-based method for amino acid sequences. A discrete gamma distribution was used to model evolutionary rate differences among sites (5 categories; +G) and a rate variation model allowed for some sites to be evolutionarily invariable (+I). Nucleotide and amino acid sequence similarities of single and concatenated genes were determined based on p-distances calculated in MEGA version 6.

Results and discussion

Compared to previously identified T. pyogenes [23], the already described T. pyogenes 11-07-D-03394 [10] and the newly investigated T. pyogenes 121,008,157 and T. pyogenes 121,018,522 in the present study, could be identified by determining hemolysis and CAMP-like haemolytic reactions, by various other phenotypical and genotypical tests, together with T. pyogenes 11-07-D-03394, by sequencing 16S rRNA and gap genes (Fig. 1), and by amplifying of T. pyogenes species-specific regions of ISR and superoxide dismutase A encoding gene sodA as molecular targets, respectively (Table 2). The three loris isolates and the type strain T. pyogenes DSM 20630T shared 100% 16S rRNA gene sequence similarity and 95% to 99% 16S rRNA gene sequence similarity with type strains of other species of the genera Trueperella and Arcanobacterium (98.8% similarity with T. abortisuis DSM 19515T, 94.6% similarity with A. haemolyticum DSM 20595T). In addition, as shown in Fig. 1, the three loris isolates and T. pyogenes DSM 20594 shared identical gap gene sequences, with 99% sequence similarity to T. pyogenes DSM 20630T and 70% to 88% sequence similarity to the type strains of other species of the genera Trueperella and Arcanobacterium (88% similarity to T. abortisuis DSM 19515T, 73% similarity to A. haemolyticum DSM 20595T).

Fig. 1
figure 1

Phylogenetic trees based on sequences of 16S rRNA (a) and glyceraldehyde-3-phosphate dehydrogenase encoding gene gap (b) of the three T. pyogenes strains of the present study and reference strains of the genra Trueperella and Arcanobacterium obtained from GenBank (NCBI). Trees were constructed using the maximum-likelihood method based on evolutionary distances calculated with the general time reversible model. Numbers at branch nodes represent the percentage of replicate trees in which the associated taxa clustered together in bootstrap tests (1000 replicates). Only bootstrap values ≥70% are shown

Full size image
Table 2 Phenotypical and genotypical properties of the investigated T. pyogenes
Full size table

To demonstrate the clonal lineage and the epidemiological relationship, the T. pyogenes isolates from this study were further found to possess known and putative virulence factor encoding genes plo, nanH, fimA, fimC, fimE and tet(W). The genes cbpA and nanP could not be detected by the amplification with the applied primer systems (Table 2). The possible importance of the known and putative virulence factors has been discussed previously [2, 10, 11].

The T. pyogenes were also investigated for an epidemiological relationship by four genomic fingerprinting methods, three (rep)-PCRs and one RAPD-PCR, respectively. (rep)-PCRs were introduced to differentiate microbes by combining the advantage of DNA amplification with the application of repetitive sequence based oligonucleotide primers. RAPD fingerprinting has also been described as a useful tool for analysing the genetic structure of closely related bacteria. Since RAPD fingerprints target the whole genome, this method is far more sensitive in terms of detecting genetic diversity than, for example, 16S rRNA gene sequencing [24]. The genomic fingerprint analysis undertaken in the present study yielded for all three T. pyogenes isolates of loris origin uniform patterns in ERIC-PCR, BOX-PCR, (GTG)5-PCR, and in RAPD-PCR analysis. The patterns differed from those of T. pyogenes DSM 20630T and T. pyogenes DSM 20594 (Fig. 2) indicating that the three isolates are genetically identical and therefore of same clonal origin, while the other strains are somehow genetically different. All four techniques had already been used successfully for molecular typing of various bacterial species [25,26,27,28,29,30,31]. BOX-PCR typing was used for DNA fingerprinting of a variety of bacteria including T. pyogenes from musk deer [32].

Fig. 2
figure 2

Genomic fingerprint pattern of the three Grey Slender Lorises strains in comparison to T. pyogenes reference strains with three different (REP)-PCRs (ERIC-PCR, BOX-PCR, and (GTG)5–PCR) and random amplification polymorphic DNA (RAPD-PCR)

Full size image

The clonal relationship of the three T. pyogenes of loris origin could also be demonstrated by MLSA, investigating the target genes fusA, tuf, metG and gyrA. All four partially sequenced nucleotide sequences and amino acid sequences of the housekeeping genes were concatenated in the following order: fusA-tuf-metG-gyrA and FusA-Tuf-MetG-GyrA with a nucleotide sequence of 3314 bp and an amino acid sequence of 1103 aa. The targeted amplified DNA fragments were double sequenced on both strands and the sequences were deposited in the GenBank (National Center for Biotechnology Information) (Table 3). The phylogenetic analysis was based on the combined utilisation of the fusA-tuf-metG-gyrA partial gene sequences. The GenBank accession numbers of locus sequences obtained in this study are provided in Table 3. The cluster analyses of the phylogenetic trees of concatenated sequences succeeded in subdividing the three investigated T. pyogenes isolates originating from the three Grey Slender Lorises and the reference strains (T. pyogenes DSM 20594, T. pyogenes DSM 20630T and A. haemolyticum DSM 20595T). The concatenated sequence of the four loci on the nucleotide level among the 6 investigated isolates and strains with the corresponding 3314 bp resulted in 2468 conserved, 846 variable and 28 parsimony-informative sites being confirmed. The concatenated tree built with the nucleotide sequence clustered the investigated three isolates originating from the three Grey Slender Lorises and shared 100% sequence similarity (with 0.00 genetic distance) and bootstrap supports of 100%. However, the reference strains T. pyogenes DSM 20594 and T. pyogenes DSM 20630T shared with 98.9% and 99%, respectively the pairwise nucleotide sequence similarity (with 0.012 and 0.010 genetic distance) with the three isolates originating from the three Grey Slender Lorises (Fig. 3a – Table 4). The nucleotide composition (GC content) of the four investigated target genes (fusA-tuf-metG-gyrA) of the three T. pyogenes from the Grey Slender Lorises, T. pyogenes DSM 20594 and T. pyogenes DSM 20630T exhibited an identical GC content of 61.3 mol% and 54.7 mol% for the reference strain A. haemolyticum DSM 20595T. However, the percentage in the content of Guanine and Cytosine were identical for the three T. pyogenes isolates originating from the Grey Slender Lorises (29.7 mol%, 31.6 mol% respectively) and for the two reference strains T. pyogenes DSM 20594 and T. pyogenes DSM 20630T (29.8 mol% and 31.5 mol% respectively) (Table 5). The concatenated sequence of the four loci on the nucleotide level shows clearly on 15 different sites that the three T. pyogenes isolates originating from the Grey Slender Lorises were identical and differed from the other three reference strains (Table 6). On the other hand, the concatenated sequence of the four loci on the amino acid level among the 6 investigated isolates and strains revealed 1103 sites with 570 conserved, 217 variables sites and 1 parsimony-informative site. The concatenated tree built with the amino acid sequence succeeded in clustering the three investigated T. pyogenes isolates originating from the three Grey Slender Lorises and separating them from the other reference strains (T. pyogenes DSM 20594, T. pyogenes DSM 20630T and A. haemolyticum DSM 20595T). The three isolates originating from the three Grey Slender Lorises shared 100% amino acid sequence similarity and bootstrap supports of 87%, while the reference strains T. pyogenes DSM 20594 and T. pyogenes DSM 20630T shared 99.7% amino acid sequence similarity with the three isolates originating from the three Grey Slender Lorises (Fig. 3b). The percentage of amino acid similarities among T. pyogenes species and A. haemolyticum DSM 20595T ranged from 80.3% to 80.4%, while within the T. pyogenes species it ranged from 99.6% to 100%. Moreover, the concatenated sequence of the four loci on the amino acid level clearly showed on the site 248 that the three T. pyogenes isolates of Grey Slender Loris origin were identical (Aspartic acid) and differed from the other three reference strains (Glutamic acid).

Table 3 GenBank accession numbers of locus sequences of T. pyogenes strains obtained in this study
Full size table
Fig. 3
figure 3

Phylogenetic analysis based on concatenated partial fusA-tuf-metG-gyrA nucleotide sequences of a total of 3314 nucleotide positions (a) and FusA-Tuf-MetG-GyrA amino acid sequences of a total of 1103 amino acid positions (b) of the three investigated target genes of the three T. pyogenes of Grey Slender Loris origin, T. pyogenes DSM 20594, T. pyogenes DSM 20630T and A. haemolyticum DSM 20595T. Trees were constructed using the maximum-likelihood method based on evolutionary distances calculated with the general time reversible model (for nucleotide sequences) or the JTT matrix-based method (for amino acid sequences). Numbers at branch nodes represent the percentage of replicate trees in which the associated taxa clustered together in bootstrap tests (1000 replicates). Only bootstrap values ≥70% are shown

Full size image
Table 4 Average pairwise distances of the concatenated sequences of the investigated strains
Full size table
Table 5 Nucleotide percentage of the concatenated sequences for the four loci sequences
Full size table
Table 6 Matrix of the variable nucleotide positions of the concatenated sequences among the investigated strains
Full size table

Conclusion

The several genetic markers of the presented MLSA approach showed clearly that the three T. pyogenes originating from Grey Slender Lorises and the two T. pyogenes reference strains belong to three different clonal complexes, respectively. The results of the present investigation which represent the first detailed epidemiological study of T. pyogenes of this origin clearly indicated that all three T. pyogenes which contributed with other potentially pathogenic bacteria to the septicemia of the three lorises, respectively shared a clonal origin. However, whether the cross infection between the three animals with T. pyogenes isolates, which was present in the lorises’ terrarium over a certain period of time, occurred because of direct contact or a lack of disinfection of the animal facility after detecting the first or the second case remains unclear.