Phenotypic and genotypic characteristics of Trueperella pyogenes isolated from ruminants

Abstract

Trueperella pyogenes is an opportunistic pathogen that causes suppurative infections in animals including humans. Data on phenotypic and genotypic properties of T. pyogenes isolated from ruminants, particularly goats and sheep, are lacking. We characterized, by phenotypic and genotypic means, T. pyogenes of caprine and ovine origin, and established their phylogenetic relationship with isolates from other ruminants. T. pyogenes isolates (n = 50) from diagnostic specimens of bovine (n = 25), caprine (n = 19), and ovine (n = 6) origin were analyzed. Overall, variable biochemical activities were observed among the T. pyogenes isolates. The fimbriae-encoding gene, fimE, and neuraminidase-encoding gene, nanH, were, respectively, more frequently detected in the large (p = 0.0006) and small (p = 0.0001) ruminant isolates. Moreover, genotype V (plo/nanH/nanP/fimA/fimC) was only detected in the caprine and ovine isolates, whereas genotype IX (plo/nanP/fimA/fimC/fimE) was solely present in the isolates of bovine origin (p = 0.0223). The 16S rRNA gene sequences of all T. pyogenes isolates were clustered with the reference T. pyogenes strain ATCC 19411 and displayed a high degree of identity to each other. Our results highlight phenotypic and genotypic diversity among ruminant isolates of T. pyogenes and reinforce the importance of characterization of more clinical isolates to better understand the pathogenesis of this bacterium in different animal species.

Introduction

Trueperella pyogenes (formerly known as Arcanobacterium pyogenes, Actinomyces pyogenes, and Corynebacterium pyogenes) commonly inhabits skin and mucous membranes of the upper respiratory, gastrointestinal, and urogenital tracts of animals.^7,10,19 In a number of animal species, including wildlife, this gram-positive, opportunistic bacterium causes a variety of suppurative infections including abscesses, arthritis, endocarditis, mastitis, metritis, osteomyelitis, pneumonia, and vasculitis.^2,3 In domestic animals, mastitis (45%), abscesses (18%), pneumonia (11%), and lymphadenitis (9%) are the most common clinical manifestations.¹¹

Previous work has genotypically characterized ruminant isolates of T. pyogenes recovered from European bison, cattle, and white-tailed deer.^{3,6,9,15–17} All reported isolates of T. pyogenes carried the pyolysin gene (plo),^3,15,17 which encodes an offensin that binds to cholesterol-containing rafts resulting in formation of pores in the host cell membrane followed by apoptosis.¹³ Screening for other virulence genes, for example, those coding for neuraminidase P (nanP), neuraminidase N (nanH), collagen-binding protein (cbpA), fimbrial subunits of type A, C, E, and G (fimA, fimC, fimE, and fimG, respectively), demonstrated that the genotypic profiles varied greatly among T. pyogenes isolates.^3,15,17

Despite earlier studies, data on phenotypic and genotypic characteristics of T. pyogenes isolated from cattle, goats, and sheep are still lacking. To date, very few T. pyogenes isolates recovered from small ruminants have been analyzed genotypically.^6,14,16 Therefore, we identified the phenotypic and genotypic characteristics of T. pyogenes isolates recovered from infections of small and large ruminants and established the phylogenetic relationship of these isolates with previously characterized T. pyogenes isolates from other ruminants.

Materials and methods

T. pyogenes isolates (n = 50) were cultured from diverse clinical diagnostic specimens derived from cattle (n = 25), goats (n = 19), and sheep (n = 6). All 50 isolates were recovered from epidemiologically unrelated specimens submitted to the Texas A&M Veterinary Medical Teaching Hospital (VMTH; College Station, TX) between 2010 and 2017. Samples were inoculated onto trypticase soy agar supplemented with 5% sheep blood (MilliporeSigma, St. Louis, MO) and incubated for 48 h at 37°C in 5% CO₂. Identification of 35 isolates (25 bovine, 8 caprine, and 2 ovine) collected in 2010–2015 was based on Gram stain, colony morphology, biochemical analyses (RapID CB PLUS system, Remel, Lenexa, KS), the Christie–Atkins–Munch–Peterson (CAMP) test, and PCR detection of the pyolysin gene (plo).⁷ T. pyogenes ATCC 19411 was used as a reference strain. The CAMP test was carried out on the 35 isolates using Staphylococcus aureus strain ATCC 25923. Identification of 15 isolates recovered from 11 caprine and 4 ovine specimens in 2015–2017 was performed via matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS; Microflex LT, Bruker Daltonics, Billerica, MA) and the plo-specific PCR. The mass spectrometer was calibrated for molecular weights with a range of 3,637–16,952 Da prior to sample testing utilizing the bacterial test standard (Bruker Daltonics) according to the manufacturer’s recommendations.

Genomic DNA was extracted (QIAprep spin miniprep kit, Qiagen, Valencia, CA) from 24-h cultures of T. pyogenes isolates grown in brain–heart infusion broth (MilliporeSigma) according to the manufacturer’s instructions. Bacterial isolates were first screened for the presence of the Arcanobacterium haemolyticum phospholipase D (pld) gene via conventional PCR, using previously developed primers⁵: forward AhF (5’-ATGTACGACGATGAAGACGCG-3’) and reverse AhR (5’-GCTTCCTTGTCGTTGAGATTATTAGC-3’). The PCR program was carried out as follows: denaturation at 95°C for 4 min followed by 30 cycles of 1 min at 95°C, 1 min at 60°C, and 1 min at 72°C; and 1 step of 7 min at 72°C. The pld-negative isolates were examined for 8 known and putative virulence genes, utilizing previously developed PCR protocols and primers.^7,17 The PCR program was performed as follows¹⁵: denaturation at 95°C for 4 min; 35 cycles of 94°C for 1 min, annealing for 30 s at different temperatures (60°C for nanH-, nanP-, plo-, cbpA-, and fimC-specific PCRs; and 57°C for fimA- and fimG-specific PCRs),^7,17 72°C for 3 min; and a final extension at 72°C for 7 min. All amplifications were carried out in 50 µL of the following reaction mixture: 10 pmol of each primer (0.5 µL each), 10 mM of each dNTP (N0447S, New England BioLabs, Ipswich, MA), 1× Taq polymerase buffer (B9004S, New England BioLabs), 1 U of Taq DNA polymerase (M0273L, New England BioLabs), and ~50 ng of genomic DNA. The PCR products were separated by electrophoresis through a 2% agarose gel in Tris–acetate–EDTA buffer (40 mM Tris, 20 mM acetic acid, 1 mM EDTA; Thermo Fisher Scientific, Lenexa, KS), stained with ethidium bromide (MilliporeSigma), and visualized using an Alpha Imager HP (Alpha-Innotech, San Leandro, CA).

A fragment of the 16S ribosomal (r)RNA gene was amplified utilizing the universal primers UNF (5’-GAGTTTGATCCTGGCTCAG-3’) and UNR (5’-GGACTACCAGGGTATCTAAT-3’) targeting bases 9–27 and 805–786 of the 16S rRNA gene of Escherichia coli, respectively.¹ PCR was performed in 50 µL of a reaction mixture containing 10 pmol of each primer, 10 mM each dNTP (N0447S, New England BioLabs), 1x Taq polymerase buffer (B9004S, New England BioLabs), 1 U of Taq DNA polymerase (M0273L, New England BioLabs), and ~50 ng of genomic DNA. The program was composed of an initial denaturation at 95°C for 4 min followed by 35 cycles of 94°C for 30 s, 56°C for 30 s, 72°C for 30 s, and a final extension at 72°C for 7 min. The amplicons of expected sizes were further sequenced (Molecular Cloning Laboratories, San Francisco, CA), and the sequence results were verified via BLASTn (http://blast.ncbi.nlm.nih.gov). The sequences of the present and previous study¹⁵ were imported into CLC Genomics Workbench 8.0.2 (https://goo.gl/urWqpy). Multiple sequence alignment (https://goo.gl/oh67Fm) was then performed on the sequences, and the resulting alignments were used to construct a neighbor-joining tree using Kimura substitution model with 2,000 bootstrap replicates. The 2-tailed Fisher exact test was used for comparison of virulence gene distribution between isolates from large and small ruminants. A p value of <0.05 was considered significantly different.

Results

Identification of 35 isolates collected from cattle (n = 25), goats (n = 8), and sheep (n = 2) in 2010–2015 required both biochemical and molecular testing. The use of the RapID CB PLUS system demonstrated that all 35 isolates were positive for glucose and ribose utilization. In contrast, sucrose and maltose were utilized by 26% and 86% of the isolates, respectively. The 35 isolates were positive for α-glucosidase and negative for β-glucosidase activity. Moreover, all of the isolates showed positive results for β-D-glucosaminidase, glycosidase, and β-galactosidase, as well as testing positive for proline, leucyl, and leucine aminopeptidase. One of 35 isolates had a negative result for pyrrolidine aminopeptidase (Supplementary Table 1). All 35 isolates were catalase-negative. Only a single isolate produced yellow pigment. A portion of isolates showed fatty acid esterase (91%), tryptophan aminopeptidase (57%), and phosphatase (54%) activity. Nitrate reductase and urease were present in only 3% of the isolates (Supplementary Table 1). As a result, the RapID CB PLUS system identified 35 isolates as A. haemolyticum with the probability of >99%. However, the reverse CAMP test results were negative for the 35 isolates. Furthermore, an amplicon of 528 bp observed for A. haemolyticum ATCC BAA-1784 was not detected in any of the isolates. In contrast, the 35 clinical isolates, as well as T. pyogenes ATCC 19411, were positive by PCR for the plo gene. The plo-specific PCR was negative for A. haemolyticum strain ATCC BAA-1784. Thus, despite the biochemical test results, the 35 isolates were identified as T. pyogenes. The 11 caprine and 4 ovine isolates collected in 2015–2017 were identified as T. pyogenes by MALDI-TOF MS and positive plo-specific PCR results.

The clinical isolates displayed variable profiles of virulence factors (Table 1); 20 different genotypes were identified. Genotype III (plo/nanH/nanP/fimA/fimC/fimE) was the most dominant genotype and was detected in 18% of the isolates. Genotypes IV (plo/nanH/nanP/fimA/fimE), V (plo/nanH/nanP/fimA/fimC) and IX (plo/nanP/fimA/fimC/fimE) were equally represented, containing 12% of respective T. pyogenes isolates. Most genotypes, including the genotype that contained all 8 genes (genotype I), were singly represented (genotypes VII, VIII, X, XII, XIV–XVI, XVIII–XX; Table 1).

Table 1.

Frequency of detection of known and putative virulence genes in 50 isolates of Trueperella pyogenes.

Genotype number	Genotype	No. of isolates	Isolates
I	plo nanH nanP cbpA fimA fimC fimE fimG	1 (2)	B53-006
II	plo nanH nanP fimA fimC fimE fimG	3 (6)	B28-001, C43-039, C62-037
III	plo nanH nanP fimA fimC fimE	9 (18)	B32-070, B45-059, B46-008, B57-062, B57-063, C42-002, C45-040, C49-008, C66-086
IV	plo nanH nanP fimA fimE	6 (12)	B41-056, B44-047, C49-012, C64-018, C66-081, C67-061
V	plo nanH nanP fimA fimC	6 (12)	C55-097, C62-055, O49-050, O57-017, O66-023, O67-048
VI	plo nanH fimA fimC fimE	2 (4)	B24-017, C45-041
VII	plo nanP cbpA fimA fimC fimE	1 (2)	B45-061
VIII	plo nanP fimA fimC fimE fimG	1 (2)	B48-049
IX	plo nanP fimA fimC fimE	6 (12)	B24-034, B31-037, B44-039, B44-100, B45-100, B57-070
X	plo nanP fimA fimE	1 (2)	C40-018
XI	plo nanP fimC fimE	2 (4)	B55-062, B56-098
XII	plo cbpA fimA fimE	1 (2)	B45-008
XIII	plo fimA fimC fimE fimG	3 (6)	B36-077, B40-004, B51-078
XIV	plo fimA fimC fimE	1 (2)	B54-012
XV	plo nanH nanP fimA fimC fimG	1 (2)	O60-018
XVI	plo nanH nanP fimA	1 (2)	C62-045
XVII	plo nanH nanP fimC fimE	2 (4)	C63-010, C63-011
XVIII	plo nanH nanP fimA fimG	1 (2)	C63-017
XIX	plo nanH fimA fimC	1 (2)	O67-036
XX	plo nanH nanP fimE	1 (2)	C67-077

Numbers in parentheses are percentages.

Phylogenetic analysis of partial sequences of the 16S rRNA gene was performed on 50 sequences of T. pyogenes isolates, 46 obtained from our study and 4 from a prior study¹⁵ (Fig. 1). The 4 isolates (808, 836, 868, and 886) from the prior characterization study represented a total of 25 T. pyogenes isolates recovered from European bison.¹⁵ The analysis also included sequences of other Trueperella and Arcanobacterium species as additional references. As a result, the T. pyogenes isolates were clustered with the reference strain, T. pyogenes ATCC 19411, which further supported the identification of the clinical isolates as T. pyogenes (Fig. 1). All 16S rRNA sequences of T. pyogenes isolates tested displayed 99–100% identity to each other. Eight sequences of T. pyogenes isolates from our study that represented detectable variations were deposited in the GenBank database under the following accessions: KX592201 (O57-017), KX592202 (O49-050), KX592203 (C49-012), KX592204 (C42-002), KX592205 (B57-062), KX592206 (B53-006), KX592207 (B45-059), and KX592208 (B31-037).

Figure 1.

Phylogenetic analysis of 16S ribosomal (r)RNA gene partial sequences of Trueperella pyogenes and other related species. Phylogenetic analysis of partial sequences of 16S rRNA gene was performed on 46 Trueperella pyogenes isolates in our study. The analysis also included various Trueperella and Arcanobacterium species as references; as well as 4 T. pyogenes isolates, previously isolated from European bison (isolates 808, 836, 868, and 886).¹⁵

Discussion

Initially, the 35 clinical isolates of T. pyogenes collected in 2010–2015 were incorrectly identified as A. haemolyticum by the RapID CB PLUS system. The isolates did not display a reverse positive CAMP reaction, which is consistently observed for A. haemolyticum. Moreover, the synergistic CAMP test results were negative for all 35 isolates and T. pyogenes ATCC 19411. The negative CAMP test is a typical feature of T. pyogenes,¹⁹ despite the fact that enhancement of staphylococcal hemolysis by T. pyogenes is occasionally observed.^6,15,18 The isolates were further examined for the presence of pld, a phospholipase D–encoding gene that is specific for A. haemolyticum.⁵ An amplicon of 528 bp observed for A. haemolyticum ATCC BAA-1784 was not detected in any of the isolates. In contrast, the 35 clinical isolates and T. pyogenes strain ATCC 19411 tested positive for plo, the gene that was previously detected in all T. pyogenes isolates of various origin.^6,15–17,21 Identification of the 15 isolates recovered in 2015–2017 was performed via MALDI-TOF MS, with confirmation by positive plo-specific PCR results. As a result, a total of 50 clinical isolates were identified as T. pyogenes.

Adhesion of T. pyogenes to host cells is a complex process that involves multiple molecules. CbpA is a collagen-binding protein A that allows T. pyogenes to adhere to collagen types I, II, and IV.⁴ Despite an important role of CbpA in cell adhesion, the frequency of the cbpA gene is highly variable among T. pyogenes isolated from various ruminants. Two prior studies demonstrated a high prevalence of the cbpA gene among T. pyogenes isolates of ruminant origin with ~48% (n = 45) and 100% (n = 57) of bovine isolates carrying the cbpA gene.^4,17 However, in our work, cbpA was the least represented gene and was harbored by only 3 bovine isolates (Table 1). The low carriage among ruminant isolates was also noticed previously, being detected in only 1.4% (n = 72), 1.5% (n = 66), and 7% (n = 29) of large ruminant isolates.^3,6,16 Interestingly, the lack of cbpA among caprine and ovine isolates in our study is consistent with findings of prior studies in which all tested isolates from goats and sheep tested negative for cbpA.^6,12,14

In addition to CbpA, T. pyogenes isolates can express neuraminidases H (NanH) and P (NanP), molecules that are also involved in bacterial adhesion.⁸ The occurrence of genes encoding NanH and NanP have been shown to vary 40% and 64%, respectively, to 100% in bovine T. pyogenes isolates.^6,8,15,17 Our study demonstrated that nanH and nanP were present in 68% and 84% of the isolates, respectively (Table 1). Either one or both genes were detected in 38% and 62% of the isolates, respectively. Interestingly, most of the isolates from small ruminants (88%) contained both genes. In contrast, the T. pyogenes isolates negative for the 2 genes were only of bovine origin. Two previous studies showed that, out of 21 isolates recovered from small ruminants, only a single isolate of ovine origin did not possess either nanH or nanP genes.^10,12 The lack of the 2 neuraminidase-encoding genes was also observed in at least 9 European bison isolates, among which the genotype lacking both genes, plo/fimA/fimC, was found to be predominant.¹⁵ Furthermore, T. pyogenes isolates negative for both nanH and nanP were identified in 5% of the isolates (n = 61) whose origin was not specified.⁶ Among T. pyogenes isolates from mastitic milk of dairy cows, ~36% of the isolates lacked both genes.²⁰ Yet, in another study, no neuraminidase-deficient genotypes were identified in any of 65 T. pyogenes isolates cultured from cranial abscesses of white-tailed deer³; the plo/nanH/nanP/fimA/fimC/fimE/fimG genotype that contained both nanH and nanP genes was detected in 48% of these isolates.³ Thus, the prevalence of nanH and nanP varies greatly among clinical isolates of ruminant origin. Overall, identification of nanH/nanP/cbpA-deficient genotypes suggests that molecules other than neuraminidase and CbpA may contribute to adhesion properties of T. pyogenes.

Fimbriae are filamentous surface appendages that are involved in cell-to-cell or cell-to-surface adherence of various bacteria. Fimbriae-encoding genes (fimA, fimE, fimC, and fimG) are consistently detected in T. pyogenes isolates.^3,15,17 In our study, fimA was among the most frequently detected genes and was found in 90% of the isolates (Table 1). Interestingly, 2 bovine and 3 caprine isolates lacked the fimA gene. The fimA-negative genotype was also earlier observed in ~9% and 36% of T. pyogenes isolates from metritis and non-metritis cows, respectively.¹⁶ The significance of the differences between the 2 groups remains uncertain.⁶ In contrast, all other genetically characterized isolates recovered from small and large ruminants including goats (n = 12, total number of isolates tested), sheep (n = 9), European bison (n = 25), dairy cows (n = 57), and deer (n = 66) contained the fimA gene.^{3,12,14,15,17}

The fimE gene was harbored by 80% of the isolates. The high prevalence of the fimE gene (98%) was also noted in other studies.^3,17 The carriage frequency of fimC is generally lower than that of fimA and fimE and may range from 67% to 88% (78% in our study).^3,15,17 The fimG gene was least represented among the fimbrial genes tested and was present in only 20% of our isolates. The low occurrence of fimG is supported by a prior study, in which the gene was detected in 24% of T. pyogenes isolated from European bison.¹⁵ However, it was also shown that 67% and 68% of Trueperella isolates recovered from dairy cows and white-tailed deer, respectively, carried the fimG gene.^3,17

In our study, the fimbriae-encoding gene, fimE, was detected more frequently in bovine isolates (p = 0.0006) than in caprine and ovine isolates. In contrast, the nanH gene was detected with higher frequency in caprine and ovine isolates (n = 24) than bovine isolates (n = 10; p = 0.0001). However, the data from a 2017 study that analyzed 57 and 7 isolates from large (cattle, n = 57) and small (n = 7) ruminants, respectively, does not support the identified correlation for either fimE or nanH (p = 1.0000).¹² In our study, genotype V (plo/nanH/nanP/fimA/fimC) was only detected in caprine and ovine isolates (n = 6); whereas genotype IX (plo/nanP/fimA/fimC/fimE) was solely present in the isolates of bovine origin (n = 6; p = 0.0223). Further studies involving a greater number of T. pyogenes isolates from small and large ruminants are warranted to confirm the identified association of the 2 genotypes to the respective animal species.

Studies involving virulence factors are important for investigation of molecular epidemiology and pathogenicity of various pathogens.¹² Our study contributes to the understanding of phenotypic and genotypic diversity that exist among T. pyogenes isolates recovered from large and small ruminants and, overall, reinforces the importance of characterization of more clinical isolates in order to better understand the pathogenesis of this bacterium in different animal species.