Identification of Trueperella bernardiae isolated from peking ducks (Anas platyrhynchos domesticus) by phenotypical and genotypical investigations and by a newly developed loop-mediated isothermal amplification (LAMP) assay.

Trueperella (T.) bernardiae is a well-known bacterial pathogen in infections of humans, rarely in animals. In the present study, five T. bernardiae isolates, isolated from five Peking ducks of four different farms, were identified by phenotypic properties, by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) analysis, and genotypically by sequencing the 16S ribosomal RNA (rRNA) gene, the superoxide dismutase A encoding gene sodA, and the glyceraldehyde-3-phosphate dehydrogenase encoding gene gap. In addition, the T. bernardiae isolates could be identified with a newly developed loop-mediated isothermal amplification (LAMP) assay based on the gyrase encoding housekeeping gene gyrA. All these tests clearly identified the T. bernardiae isolates to the species level. However, the detection of the specific gene gyrA with the newly designed LAMP assay appeared with a high sensitivity and specificity, and could help to identify this bacterial species in human and animal infections in future. The importance of the T. bernardiae isolates for the clinical condition of the ducks and for the problems at farm level remains unclear.

Introduction

Trueperella (T.) bernardiae is a gram-positive, non-motile, and facultatively anaerobic coccobacillus which was originally identified as a coryneform group 2 bacterium (Na’Was et al. 1987). Eight years later, this bacterium was recognized as Actinomyces bernardiae (Funke et al. 1995) and then reclassified as part of the genus Arcanobacterium as Arcanobacterium bernardiae (Ramos et al. 1997). According to a proposal of Yassin et al. (2011), Arcanobacterium bernardiae was finally classified to the newly described genus Trueperella as T. bernardiae, together with T. pyogenes, T. bialowiezensis, T. bonasi, and T. abortisuis.

In humans, T. bernardiae was first described as an opportunistic pathogen and later could be identified alone or in combination with other bacteria as causing joint infections (Gilarranz et al. 2016; Gowe et al. 2018), urinary tract infections (Lepargneur et al. 1998), abscesses (Parha et al. 2015; VanGorder et al. 2016; Calatrava et al. 2019; Pan et al. 2019), wound infections (Weitzel et al. 2011; Rattes et al. 2016; Cobo et al. 2017), a diabetic foot infection (Schneider et al. 2015), bacteremia (Otto et al. 2013; Roh et al. 2019), and septic thrombophlebitis (Lawrence et al. 2018).

The first characterization of T. bernardiae of animal origin (3-day-old piglet) was made by Hijazin et al. (2012c). In a second case, Arnafia et al. (2017) described a T. bernardiae strain recovered from a purulent thelitis of a 12-year-old male dog.

These previously isolated T. bernardiae isolates of animal origin were identified and further characterized by matrix-assisted desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) and by sequencing various genomic targets (Hijazin et al. 2012c; Arnafia et al. 2017).

At present, no further data are available concerning the isolation of T. bernardiae from animals or animal infections.

In the present investigation, five T. bernardiae isolates recovered from post mortem samples of Peking ducks (Anas platyrhynchos domesticus) could be identified phenotypically and genotypically, and with a newly described loop-mediated isothermal amplification (LAMP) assay based on the housekeeping gene gyrA.

Material and methods

Bacterial strains

The five T. bernardiae isolates investigated in the present study (T. bernardiae D12-0613-1-4-3, T. bernardiae D13-1622-5-3-2, T. bernardiae D14-1577-4-8-1, T. bernardiae D13-1772-748-1-2, T. bernardiae D14-1481-2029-1-1) were isolated from five Peking ducks (Anas platyrhynchos domesticus) from three different farms in Germany (n = 3) and one in Thailand (n = 2). The Geman samples were collected from the heart, lung, and joint, while the Thailand samples were delivered as swabs after post-mortem examination; unfortunately, the source organs were not defined. All five isolates were collected during post-mortem examination in the period between 2012 and 2014. Further details about the five isolates are given in Table 1. The bacterial cultivation and a preliminary identification of the bacteria were performed at Ripac-Labor GmbH, Potsdam, Germany. The bacterial culturing of the T. bernardiae isolates was carried out on sheep blood agar plates (Oxoid GmbH, Wesel, Germany) for 48 h at 37 °C under a microaerophilic gas atmosphere using a candle jar.

Table 1

Data on the five T. bernardiae isolates recovered from Peking ducks investigated in the present study

Strain code	Farm/country	Sample drawing	Sample source	Further information
T. bernardiae D12-0613-1-4-3	A/G	07/05/2012	Post-mortem/heart; no pathological findings	Accompanying bacteria: Corynebacterium spp., Aerococcus viridans, Escherichia (E.) coli
T. bernardiae D13-1622-5-3-2	B/G	02/10/2013	Post-mortem/lung edema	Accompanying bacteria: Aspergillus fumigatus, Corynebacterium confusum, Coenonia anatina; increased mortality and joint infections at farm level
T. bernardiae D13-1772-748-1-2	C/T	25/10/2013	Post-mortem	Accompanying bacteria: Globicatella sulfidifaciens
T. bernardiae D14-1481- 2029-1-1	C/T	20/08/2014	Post-mortem	n.d
T. bernardiae D14-1577-4-8-1	D/G	05/09/2014	Post-mortem/joint infection	Accompanying bacteria: Trueperella pyogenes, E. coli; Increased mortality at farm level

G Germany, T Thailand, n.d. no data available

Phenotypic identification

A phenotypic identification was performed using conventional cultural and biochemical assays as previously shown (Hassan et al. 2009; Ülbegi-Mohyla et al. 2009) and with the API-Coryne test system (BioMérieux Deutschland GmbH, Nürtingen, Germany) in accordance with the manufacturer’s instructions. Furthermore, the bacterial isolates were identified by MALDI-TOF MS using a Microflex LT (Bruker Daltonik GmbH, Bremen, Germany) instrument following the manufacturer’s instructions using the direct transfer method. Briefly, one microbial colony was first smeared in duplicate onto spots of the MALDI MSP 96 target plate (MicroScout Target plate; Bruker Daltonik GmbH) with sterile toothpicks. The air-dried bacteria were overlaid with 1µL of an α-cyan 4-hydroxycinnamic acid matrix solution (HCCA, in 50% acetonitrile and 2.5% trifluoroacetic acid in pure water), followed by drying and loading into the mass spectrometer. The analysis of the spectra was carried out by MBT Compass Explorer 4.1 software (Bruker Daltonik GmbH).

DNA extraction

The genomic DNA of the five isolates, type strain T. bernardiae DSM 9152 T, and various other strains of genus Trueperella and genus Arcanobacterium (A.) were extracted using the DNeasy Blood and Tissue kit (Qiagen GmbH, Hilden, Germany), in conformance with the manufacturer’s instructions. The concentration and purity of DNA were measured by means of a NanoDrop spectrophotometer (ND1000; Thermo Fisher Scientific GmbH, Dreieich, Germany).

Sequencing the molecular targets

The five T. bernardiae isolates were also investigated by sequencing the following molecular targets: 16S rRNA gene, superoxide dismutase A encoding gene sodA, and glyceraldehyde-3-phosphate dehydrogenase encoding gene gap. The sequences of the oligonucleotide primers and the PCR conditions were used as previously described for 16S rRNA gene (Hassan et al. 2009), gene sodA (Hijazin et al. 2011, 2012c), and gene gap (Wickhorst et al. 2019). The PCR products were purified and sequenced by Eurofins Genomics Germany GmbH (Ebersberg, Germany). The obtained sequences of the different genes of the T. bernardiae isolates were aligned and further analyzed using the clustal w method of the MegAlign program version 15 (DNASTAR, Inc., Madison, WI, USA) and compared with the nucleotide sequences of the targets 16S rRNA gene, sodA, and gap of type strain T. bernardiae DSM 9152 T, and type strain of T. pyogenes DSM 20630 T obtained from the NCBI GenBank, and for control purposes from A. haemolyticum DSM 20595 T also obtained from the NCBI GenBank.

LAMP assay

Design of oligonucleotide primers for LAMP assay

Oligonucleotide primers for the T. bernardiae-specific LAMP assay were developed using the gyrase subunit A encoding gene gyrA of T. bernardiae (LNIZ01000002).

The LAMP primers (forward outer primer gyrA-F3, backward outer primer gyrA-B3, forward inner primer gyrA-FIP, backward inner primer gyrA-BIP, forward loop primer gyrA-LoopF, and backward loop primer gyrA-LoopB) were designed using the LAMP designer software (PREMIER Biosoft, San Francisco, CA, USA) (Table 2). The oligonucleotide primers were synthesized by Eurofins Genomics.

Table 2

Oligonucleotide primer sequences of gyrase subunit A encoding gene gyrA used for development of the T. bernardiae LAMP assay

Designation	Sequences 5′- 3′	Primer length (bp)	Melting temperature (°C)
gyrA-F3	CACCAGGTAGAGGTCATCA	19	56.7
gyrA -B3	TCCTCGACGATCTTCTGC	18	56.0
gyrA -FIP	GCCGGATGAGGGCAATGAGAAGAGCGCCTCATGATC	36	˃75
gyrA -BIP	CGGGCTCATCGAACTGCTCTGCATGGCGAGGATATG	36	˃75
gyrA -LoopF	CTCGTCCAGCATGTCGAG	18	58.2
gyrA -LoopB	CGCGATCAACGAGATCCA	18	56.0

LAMP reaction and amplification conditions

In accordance with the manufacturer’s instructions, the LAMP assay based on gene gyrA was carried out with the five T. bernardiae isolates, type strain T. bernardiae DSM 9152 T, and with control strains of genus Trueperella and closely related genus Arcanobacterium. A total volume of 25 µL for each reaction included 15 µL GspSSD isothermal master mix (ISO-001) (OptiGene Ltd., Horsham, UK) and 2.5 µL primer mix (ISO-001; OptiGene Ltd.), gyrA- F3 primer, and gyrA-B3 primer with a final concentration equivalent to 0.2 µmol/L, gyrA-FIP primer, and gyrA-BIP primer with final concentration equivalent to 0.8 µmol/L and gyrA-LoopF Primer and gyrA-LoopB Primer with a final concentration equivalent to 0.4 µmol/L. Subsequently, 5 µL DNA was added as a template. The LAMP assay was run at 70 °C for 20 min with a melting curve analysis step (annealing curve 98 to 80 °C ramping at 0.05 °C/s) in a real-time fluorometer GenieII® (OptiGene Ltd.).

Analytical sensitivity and specificity of the LAMP assay

Determination of the analytic sensitivity of the LAMP assay was performed seven times using a serially diluted DNA (10−1–10−6) isolated from type strain T. bernardiae DSM 9152 T in AE buffer (10 mM Tris–Cl, 0.5 mM EDTA; pH 9.0) with the conditions mentioned above. DNA isolation and concentration were performed as stated above. The amount of DNA ranged from 3.0 ng/µL (10−0) to 3.0 fg/µL (10−6) bacterial DNA. The colony-forming unit (cfu/mL) was subsequently estimated.

The specificity of the LAMP assay was determined using the DNA of T. bernardiae DSM 9152 T and closely related species of genus Trueperella and Arcanobacterium. These included T. pyogenes DSM 20630 T, T. pyogenes DSM 20594, T. pyogenes 59/11, T. abortisuis DSM 19515 T, T. bialowiezensis DSM 17162 T, T. bonasi DSM 17163 T, A. hippocoleae DSM 15539 T, A. pluranimalium DSM 13483 T, and A. phocae DSM 10002 T. The LAMP assay was performed with the optimized LAMP protocol with a run-time of 20 min.

Results and discussion

The phenotypic properties of the five T. bernardiae isolates of duck origin investigated in the present study were almost identical to those of type strain T. bernardiae DSM 9152 T, and to previously characterized T. bernardiae strains of pig and dog origin (Hijazin et al. 2012c; Arnafia et al. 2017). All T. bernardiae isolates gave positive reactions for pyrazinamidase, pyrrolidonyl arylamidase, and α-glucosidase, and reacted negatively in nitrate reduction and for alkaline phosphatase, β-glucuronidase, β-galactosidase, and N-acetyl-β-glucosaminidase. Also, all isolates did not hydrolyze esculin, urea, and gelatine. The isolates also fermented D-glucose, except T. bernardiae D14-1481-2029-1-1 and type strain T. bernardiae DSM 9152 T, D-ribose, D-maltose, and glycogen, but not D-xylose, D-mannitol, D-lactose, and D-saccharose. In addition, all isolates showed a negative catalase reaction (Table 3).

Table 3

Biochemical properties of the five T. bernardiae isolates investigated in the present study and T. bernardiae DSM 9152 T

Biochemical properties	T. bernardiae D12-0613–1-4–3	T. bernardiae D13-1622–5-3–2	T. bernardiae D13-1772–748-1–2	T. bernardiae D14-1481- 2029–1-1	T. bernardiae D14-1577–4-8–1	T. bernardiae DSM 9152 T
Nitrate reduction	−	−	−	−	−	−
Pyrazinamidase	+	+	+	+	+	+
Pyrrolidonyl Arylamidase	+	+	+	+	+	+
Alkaline phosphatase	−	−	−	−	−	−
α-Glucuronidase	−	−	−	−	−	−
β-Galactosidase	−	−	−	−	−	−
β-Glucosidase	+	+	+	+	+	+
N-Acetyl- β -glucosaminidase	−	−	−	−	−	−
Esculin	−	−	−	−	−	−
Urease	−	−	−	−	−	−
Gelatine	−	−	−	−	−	−
Fermentation
Glucose	+	+	+	−	+	−
Ribose	+	+	+	+	+	+
Xylose	−	−	−	−	−	−
Mannitol	−	−	−	−	−	−
Maltose	+	+	+	+	+	+
Lactose	−	−	−	−	−	−
Saccharose	−	−	−	−	−	−
Glycogen	+	+	+	+	+	+
Catalase	−	−	−	−	−	−
T. bernardiae identification (%) according to Api- Coryne test system	99.7	99.7	99.7	99.9	99.7	99.9

 + positive reaction, − negative reaction, type strain

With the additionally performed MALDI-TOF MS analysis, all five isolates were identified to the species level as T. bernardiae with log-score values varying between 1.87 and 2.2 (data not shown). MALDI-TOF MS appeared to be a fast, accurate, and less expensive tool for microbial identification of bacteria, viruses, and fungi (Singhal et al. 2015), also including T. bernardiae (Hijazin et al. 2012a) and various other species of genera Trueperella and Arcanobacterium (Hijazin et al. 2012a, b).

The five T. bernardiae isolates in the current study were additionally identified genotypically by amplification and sequencing of the 16S rRNA gene. The nucleotide sequence of T. bernardiae D12-0613-1-4-3 (GenBank accession number: MT364890), T. bernardiae D13-1622-5-3-2 (MT364891), T. bernardiae D13-1772-748-1-2 (MT364892), T. bernardiae D14-1481-2029-1-1 (MT364893), and T. bernardiae D14-1577-4-8-1 (MT364894) were compared with type strain T. bernardiae DSM 9152 T (HE653979), T. pyogenes DSM 20630 T (X79225), and Arcanobacterium (A.) haemolyticum DSM 20595 T (AJ234059). The nucleotide sequence data of T. bernardiae D12-0613-1-4-3, T. bernardiae D13-1622-5-3-2, T. bernardiae D13-1772-748-1-2, T. bernardiae D14-1481-2029-1-1, and T. bernardiae D14-1577–4-8–1 revealed a sequence homology of 99.7, 99.7, 99.2, 99.0, and 99.7% with type strain T. bernardiae DSM 9152 T, respectively (Fig. 1). The 16S rRNA gene sequence similarities of the five T. bernardiae to T. pyogenes DSM 20630 T and A. haemolyticum DSM 20595 T were equal or less than 98.1 and 94.8%, respectively (Fig. 1).

Fig. 1

Phylogenetic analysis based on nucleotide sequences of 16S rRNA gene of the five investigated T. bernardiae isolates isolated from Peking ducks, type strain T. bernardiae DSM 9152 T, and closely related T. pyogenes DSM 20630 T and A. haemolyticum DSM 20595 T obtained from NCBI GenBank

The five T. bernardiae isolates could be further characterized by PCR-mediated amplification of the genes sodA and gap. The sequences of gene sodA of T. bernardiae D12-0613–1-4–3 (MT410971), T. bernardiae D13-1622–5-3–2 (MT410972), T. bernardiae D13-1772–748-1–2 (MT410973), T. bernardiae D14-1481–2029-1–1 (MT410974), and T. bernardiae D14-1577–4-8–1 (MT410975) resulted in sequence similarities of 95.8, 94.7, 95.5, 93.8, and 95.0% with the sodA gene of type strain T. bernardiae DSM 9152 T (AM989465), respectively, while the similarity within the five isolates was between 99.0 and 100% (Fig. 2a).

Fig. 2

Dendrogram analysis of superoxide dismutase A encoding gene sodA (a) and glyceraldehyde-3-phosphate dehydrogenase encoding gene gap (b) of the five T. bernardiae isolates isolated from Peking ducks, type strain T. bernardiae DSM 9152 T, and closely related T. pyogenes DSM 20630 T and A. haemolyticum DSM 20595 T obtained from NCBI GenBank

The additionally investigated gap genes of T. bernardiae D12-0613–1-4–3 (MT410966), T. bernardiae D13-1622–5-3–2 (MT410967), T. bernardiae D13-1772–748-1–2 (MT410968), T. bernardiae D14-1481–2029-1–1 (MT410969), and T. bernardiae D14-1577–4-8–1 (MT410970) showed sequence similarities of 98.5, 98.8, 98.8, 99.0, and 99.0% with the gap gene of type strain T. bernardiae DSM 9152 T (HF947287), respectively, while the similarity within the five isolates was between 99.0 and 99.8% (Fig. 2b). The gene sequences of the sodA and gap genes of the five T. bernardiae isolates showed a clear difference to the control strains T. pyogenes DSM 20630 T (AM949566 and HF947285, respectively) and A. haemolyticum DSM 20595 T (AM983534 and CP002045, respectively) (Fig. 2a, b). All three mentioned genomic targets were already used to characterize various species of genus Trueperella, also including T. bernardiae (Hassan et al. 2009; Hijazin et al. 2011, 2012c; Arnafia et al. 2017).

The additionally used T. bernardiae gyrA-specific LAMP assay could successfully be used to identify the species-specific gene gyrA of all five T. bernardiae isolates in the present investigation. This newly established assay demonstrated a specificity for T. bernardiae DSM 9152 T with an annealing temperature between 91.9 and 92.4 °C. No cross-reactivity with any other related species of genus Trueperella or genus Arcanobacterium could be observed (Table 4).

Table 4

Specificity of the T. bernardiae LAMP assay based on gene gyrA for T. bernardiae DSM 9152 T, the five T. bernardiae isolates of duck origin, and other closely related species of genus Trueperella and Arcanobacterium

Species and strain number	Detection time mm:ss	Melting temperature (°C)
T. bernardiae DSM 9152 T	11:45	91.9
T. bernardiae D12-0613–1-4–3	14:15	92.0
T. bernardiae D13-1622–5-3–2	13:15	92.3
T. bernardiae D13-1772–748-1–2	13:30	92.3
T. bernardiae D14-1481- 2029–1-1	13:00	92.4
T. bernardiae D14-1577–4-8–1	13:15	92.4
T. pyogenes DSM 20630⊤	–	–
T. pyogenes DSM 20594	–	–
T. pyogenes 59/11	–	–
T. abortisuis DSM 19515 T	–	–
T. bialowiezensis DSM 17162 T	–	–
T. bonasi DSM 17163 T	–	–
A. hippocoleae DSM 15539 T	–	–
A. pluranimalium DSM 13483 T	–	–
A. phocae DSM 10002 T	–	–

DSM Deutsche Sammlung von Mikroorganismen und Zellkulturen

The developed LAMP assay provided an analytic sensitivity of 30 fg/μL with a mean detection time between 00:08:58 (3.0 ng/μL) and 00:18:35 min (30 fg/μL) (Table 5). For the DNA concentration at 30 fg/μL (10−5), the assay gave positive results in six of seven reactions (85.7%), whereas for DNA concentration at 3.0 fg/µL (10−6), the positive assay resulted in two of seven replicates (28.6%). The results of the T. bernardiae gyrA LAMP assay are shown in Fig. 3 and Table 4.

Table 5

Detection time and annealing temperature of the LAMP assay using bacterial serial dilutions of type strain T. bernardiae DSM 9152 T

T. bernardiae DSM 9152 T		Serial dilution
	cfu/mL	10−1	10−2	10−3	10−4	10−5	10−6
Detection time mean (mm:ss)	2.84 × 108	08:58	09:15	11:39	15:56	18:35	17:15
SD ( ±) detection time		00:49	02:52	01:29	04:25	08:32	06:00
Annealing temp. (°C) mean		91.9	91.8	91.9	91.8	91.8	91.8
SD ( ±) annealing		0.15	0.12	0.22	0.19	0.22	0.22

Fig. 3

Positive LAMP assay of the five T. bernardiae isolates T. bernardiae D12-0613–1-4–3, T. bernardiae D13-1622–5-3–2, T. bernardiae D13-1772–748-1–2, T. bernardiae D14-1481–2029-1–1, T. bernardiae D14-1577–4-8–1 obtained from Peking ducks, T. bernardiae DSM 9152 T, and as LAMP negative control T. pyogenes DSM 20630 T and nuclease free water as negative control

The application of LAMP assays as being a rapid and reliable method for detecting species of genus Trueperella and Arcanobacterium has been previously published. In 2013, Zhang et al. developed a LAMP assay using the gene encoding pyolysin, the plo gene, for a specific identification of T. pyogenes (Zhang et al. 2013). Furthermore, a pla LAMP assay was used for identifying A. pluranimalium (Abdulmawjood et al. 2015), and a cpn60 LAMP assay for identifying T. pyogenes from different animal origins (Abdulmawjood et al. 2016; Ahmed et al. 2020; Alssahen et al. 2020).

The present study gives a reliable phenotypic and genotypic characterization of T. bernardiae of duck origin, also including a newly developed LAMP assay. To our knowledge, the study gives the first detailed characterization of this bacterial species isolated from Peking ducks. However, the pathogenic importance of T. bernardiae, which was partly isolated together with various other bacteria from apparently healthy animals, for the high mortality rate or joint infections of the Peking ducks at farm level remains unclear. The described LAMP assay might help to identify this bacterial species in future and might elucidate the role this species plays in human and animal infections.