First isolation and whole-genome sequencing of Trueperella abortisuis from a goat in Canada

Abstract

Objective

The present study reports the first isolation and whole-genome sequencing of a Trueperella abortisuis bacterium from a goat.

Animals and sample

The T. abortisuis was isolated from the uterus of a goat following an abortion.

Procedure

The T. abortisuis was identified by pure culture phenotype and MALDI-TOF analysis and further characterized by whole-genome sequencing.

Results

This isolate was reliably identified as T. abortisuis and showed similar properties to type strain T. abortisuis DSM 19515^T, which was recovered from a sow following an abortion. The assembled genome of this isolate was 2 564 866 bp long with a GC content of 63.9%. A total of 30 virulence-related genes were determined, suggesting the pathogenic potential of this organism.

Conclusion and clinical relevance

This study details the first isolation of T. abortisuis from goats. The genotypic findings of this isolate will serve as a baseline description for any similar future studies.

Trueperella abortisuis is a gram-positive bacterium, formerly known as Arcanobacterium abortisuis. It was first isolated from the placenta of an aborted sow in Japan in 2009 (1). Since then, this bacterium has been reported in several pigs, cows, dogs, and cats worldwide (1–7). To date, there are no reports of isolation of this organism from goats. In this article, we report the first isolation of T. abortisuis from a domestic goat in Canada and further characterize this isolate with whole-genome sequencing.

One euthanized 3.6-year-old female Nigerian dwarf goat was submitted to the Animal Health Centre, Abbotsford, British Columbia, for postmortem examination in February 2021. The goat had a 1-month history of weight loss, lethargy, and anemia that was partially responsive to empirical treatment with copper and albendazole. She was discovered recumbent and aborted triplets at ~130 d of gestation, 7 d before being euthanized. She was treated with ceftiofur, toltrazuril, hemostam, thiamine, meloxicam, and ivermectin over the next 7 d but remained lethargic, recumbent, and inappetent, and was ultimately euthanized with intravenous pentobarbital. On postmortem examination, a severe necrosuppurative endometritis was noted. Histopathology revealed large numbers of intralesional, small, Gram-positive bacilli and coccobacilli and fewer Gram-positive cocci in short chains. A uterine sample was cultured on 5% sheep blood agar plates (Oxoid, Nepean, Ontario) and incubated aerobically and anaerobically at 35°C in an atmosphere containing 5% CO₂. After 24 and 48 h of incubation, bacteria were identified based on colony morphology, basic phenotypical properties (Gram staining, catalase, and oxidase test), and MALDI-TOF MS analysis using MALDI Biotyper (Bruker, Milton, Ontario). High numbers of Staphylococcus aureus (+++) were isolated along with Vagococcus lutrae (+++), Streptococcus spp. (++), and Escherichia coli (++) in aerobic culture. Tiny, β-hemolytic colonies were isolated in high numbers (+++) in anaerobic culture, which was later identified as T. abortisuis with a MALDI score of 2.31 (MALDI score > 2.0 indicates reliable species level identification). Colony morphology of this isolate was similar to that in other previously published studies, which described this isolate as a strictly anaerobic, Gram-positive diphtheroid showing small, beta-hemolytic colonies in blood agar plate and testing negative for catalase and oxidase (1,3). As T. abortisuis was reported in pigs and cows with abortion and reproductive issues in the past, we conducted whole-genome sequencing with both Oxford Nanopore and Illumina MiSeq sequencing platforms to further support our phenotypic identification and characterize this isolate genotypically.

For Oxford Nanopore sequencing, bacterial DNA was extracted from pure colonies using the MasterPure Complete DNA and RNA Purification Kit (LGC, Twickenham, UK), following the manufacturer’s instructions. The DNA concentration and purity were checked with a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, Massachusetts, USA). The DNA was prepared for sequencing using a 1D genomic sequencing kit from Oxford Nanopore Technologies (Oxford, UK), following the manufacturer’s protocol. The DNA was treated with FFPE Repair Mix and Ultra II End Repair/dA-Tailing Module (New England Biolabs, Ipswich, Massachusetts, USA) to repair any damage done during DNA extraction and purification. AMPure XP magnetic beads and ethanol washes were used to clean the DNA of remaining enzymes and reagents before sequencing adapters were ligated to the DNA with Blunt/TA Ligation Master Mix. The DNA library was mixed with running buffer and library loading beads, then loaded onto a flow cell (version 9.4.1) with a minimum of 800 available pores for sequencing on a GridION Mk1 instrument. The library was run on the sequencer using ONT MinKNOW software for up to 12 h.

For Illumina Miseq sequencing, genomic DNA was extracted from pure colonies using the DNeasy Blood & Tissue Kit (Qiagen, Toronto, Ontario), following the manufacturer’s instructions. The DNA concentration and purity were determined by NanoDrop and Quant-iT fluorescent assay (Thermo Fisher Scientific). Libraries were constructed, according to British Columbia Cancer’s Genome Sciences Centre plate-based and paired-end library protocols, on a Microlab NIMBUS liquid handling robot (Hamilton, Reno, Nevada, USA). Briefly, 50 ng of genomic DNA was sonicated (Covaris LE220) in 62-microliter volume to 250 to 350 bp. Sonicated DNA was purified with PCRClean DX magnetic beads (Aline Biosciences, Woburn, Massachusetts, USA). The DNA fragments were end-repaired, phosphorylated, and bead-purified in preparation for A-tailing using a custom NEB Paired-End Sample Prep Premix Kit (New England Biolabs). Illumina TruSeq adapters were ligated 15 minutes at 20°C and adapter-ligated products were bead-purified and enriched with 8 cycles of PCR using dual index primers that enabled library pooling. Libraries were pooled and sequenced with paired-end 150 bp reads on an Illumina HiSeqX instrument.

NanoPlot (v 1.33.0) and chopper (v 0.6.0) were used for quality check and adaptor trimming for Nanopore sequencing reads. For Illumina reads, sequence reads were quality-checked using FastQC (v 0.12.1). All the sequencing reads were uploaded in Galaxy (https://usegalaxy.org) and a hybrid assembly was completed using Unicycler pipelines (Galaxy v 0.5.0+galazy1). The quality of the assembled genome was assessed using Quast (v 5.2.0). Genome annotation of the isolate was done using National Center for Biotechnology Information (NCBI) Prokaryotic Genome Annotation Pipeline (v 6.6). Several previously published studies evaluated and described the use of molecular targets 16S rDNA, gene gap encoding glyceraldehyde 3-phosphate dehydrogenase, and gene tuf encoding the elongation factor Tu for identification of T. abortisuis (1,3,4,6). We extracted these genes of interest from the assembled genome and identified their sequence similarity with type strain T. abortisuis DSM 19515^T using NCBI BLAST. Further, we constructed a phylogenetic tree based on the core genome alignment of our isolate with 5 different closely related Trueperella species available in NCBI GenBank. All these genomes were reannotated by prokka (v 1.14.5) and GFF files were used in conjunction with MAFFT (v 7.520), as part of the Roary pipeline (v 3.13.0), to generate the core genome alignment. The tree was visualized using FigTree software (v 1.4.4). Arcanobacterium pluranimalium, a type strain (NCBI accession number: GCF_016907845.1; https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_016907845.1/) was used as an out-group to root the tree. Default parameters were used for all software unless otherwise specified.

At present, there are no known virulence factors for T. abortisuis. We screened our isolate for some of the known virulence genes of a closely related species, T. pyogenes, namely pyolysin (plo), collagen-binding protein A (cbpA), neuraminidases (nanH and nanP), and fimbriae (fimA, fimC, fimE, and fimG) (8). We extracted nucleotide sequences of all available genes of interest from NCBI GenBank and determined their presence in our isolate using MyDbFinder 2.0, Center for Genomic Epidemiology, by selecting a minimum of 70% identity threshold and a minimum of 60% coverage in length. Later, we confirmed them by BLASTN and BLASTX searches. Further, we screened our isolate for all known and putative virulence factors of various functionalities (e.g., adherence, invasion, exotoxin, immunomodulation, etc.) associated with 32 bacterial genera listed in the virulence factor database (VFDB) using VFanalyzer (9). In addition, antimicrobial-resistant genes were screened using Resfinder and CARD databases.

As shown by previous studies, MALDI-TOF appears to be a powerful tool for species-level identification of T. abortisuis. The isolate investigated in the current study could reliably be identified as T. abortisuis by the culture phenotype and MALDI-TOF MS analysis (1,3,5,6,10). The MALDI results for the strain in this study showed a close match to type strain T. abortisuis DSM 19515^T. Following identification, the strain was subjected to whole-genome sequencing analysis. Illumina HiSeq generated 2758323 paired reads with a genome coverage of ≈161X. Nanopore sequencing generated 26681 reads with a mean length of 6048 bp. The assembled genome resulted in 1 circular contig of 2564866 bp long with a mean GC content of 63.9%, which is similar to the previously reported GC content of 63.8% for type strain T. abortisuis DSM 19515^T (1). We submitted the assembled genome to GenBank (GCF_031885465.1; https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_031885465.1/). Annotation by Prokaryotic Genome Annotation Pipeline for this isolate from the present study revealed a total of 2315 genes, of which 2242 genes were with coding sequences, 6 rRNAs, 47 tRNAs, 3 ncRNAs; and 17 pseudo genes. NCBI BLAST results for 16S rDNA, gap, and tuf genes for this isolate showed 99.7%, 100%, and 99% sequence similarity, respectively, to type strain T. abortisuis DSM 19515 (NCBI accession numbers: 16s rDNA, NR_041607; gap, HF947288; tuf, LN875778). We deposited the targeted gene sequences for the study isolate in NCBI GenBank (16S rRNA: OR689031, https://www.ncbi.nlm.nih.gov/nuccore/OR689031; gap: OR707087, https://www.ncbi.nlm.nih.gov/nuccore/OR707087; tuf: OR707088, https://www.ncbi.nlm.nih.gov/nuccore/OR707088). Overall, our studied genome was closest to type strain T. abortisuis DSM 19515 (GCA_030811095.1: https://www.ncbi.nlm.nih.gov/datasets/genome/GCA_030811095.1/), with an average nucleotide identity value of 98.47% based on NCBI taxonomy check results. Further, a typical dendrogram was generated based on core-genome alignment of the studied strain along with reference genomes of a few closely related species collected from NCBI GenBank (Figure 1).

FIGURE 1.

Phylogenetic tree generated based on core-genome alignment of the studied strain along with reference genomes of closely related species collected from the National Center for Biotechnology Information (NCBI) GenBank.

Neuraminidases (nanH and nanP), which are important enzymes for adherence to epithelial cells, weakening the host immune response, and sialic acid catabolism, were detected in our isolate (Table S1, available online from: www.canadianveterinarians.net). Previously, these enzymes were also reported in various frequencies in T. pyogenes from goats and sheep (8). However, other T. pyogenes virulence genes of interest (plo, cbpA, fimA, fimC, fimE, and fimG) were absent in our isolate. Further, we screened our isolate for virulence genes listed in the VFDB database (9) and identified homologues of 28 genes of various virulence classes, including genes related to stress survival (sodA), immune evasion (rmlB, pgi, galE3), secretion system (T6SS-II), immunomodulation (rmlA), anti-apoptosis factor (nuoG), nutritional/metabolic factor (narH and glnA1), and adherence (srtD), among others (Table S2, available online from: www.canadianveterinarians.net). ResFinder and CARD results detected no specific antibiotic resistant genes for this isolate.

The clinical significance of this bacterium in this case is unclear. Involvement in the endometritis and abortion of this goat is considered possible, as T. abortisuis has been associated with abortion and reproductive losses in other species (1,3,11). In addition, the bacteria in the uterine lesion had a morphology consistent with T. abortisuis.

Following the first description of T. abortisuis recovered from a placenta of porcine abortion in Japan, several studies reported isolation of this organism from porcine, bovine, and other animal species (1,2). Nine strains of T. abortisuis were isolated from the urogenital tract of pigs with various clinical signs over a period of 9 y in Germany (5). Further, T. abortisuis was isolated from an umbilical swab, 2 anal swabs, and the placenta of 4 pigs in Germany following abortion (10). This bacteria was isolated during routine microbiological screening of extended boar semen in the United States (7). Six distinct T. abortisuis strains were recovered over a period of 5 mo from aborted fetal material of 6 pigs at a single farm in Mecklenburg-West Pomerania federal state, Germany (4). In Scotland, T. abortisuis was isolated from pooled fetal stomach contents of aborted swine fetuses (3). Beyond pigs, this isolate has been reported in a male dog with an anal sac abscess, in the urine of a cat with urolithiasis and nephrolithiasis, and in a cat with perianal abscess (6). It has not been previously isolated from caprine tissue. Trueperella pyogenes, a closely related isolate of T. abortisuis, was previously reported in caprine abortion (11,12).

Similar to our findings, T. abortisuis was previously often recovered in mixed growth with other bacteria (3–5,13). However, none of those other bacteria were consistently isolated and most were considered commensal or contaminant organisms (5,10,13). In a series of abortions and placentitis cases reported in domestic pigs in the United States between 2017 and 2020, T. abortisuis was isolated consistently, supporting its potential role in porcine abortion and reproductive failure as either a primary cause or a co-pathogen (13). In the present study, because T. abortisuis was isolated with a mix of bacteria from the uterus, it was difficult to determine which of these bacteria was the primary cause of abortion, or whether it was due to a coinfection. In addition, antemortem treatment with antibiotics (ceftiofur) might have affected bacterial culture results by eliminating or reducing the load of other potential bacterial pathogens. Twort’s Gram stain was used to further identify the bacteria on histopathology. The most common bacteria within the uterine lesions were Gram-positive short rods, consistent with T. abortisuis, suggesting its potential role in endometritis in this goat. We also detected 30 different virulence-associated genes, suggesting the pathogenic potential of this isolate.

As the pathogenic importance of T. abortisuis remains unclear, further work is required to determine the clinical importance of this bacterium as a potential cause of abortion in goats and to define the molecular mechanisms of virulence. To our knowledge, this study provides the first genotypic description of T. abortisuis isolated from a caprine species in Canada.