Establishment of a Real-Time PCR Assay for the Detection of Devriesea agamarum in Lizards

Simple Summary Bacterial infections can play an important role in dermatitis in lizards. The bacterial species Devriesea (D.) agamarum is a known cause of dermatitis, cheilitis and even fatal disease in lizards. Disease has most often been reported in Uromastyx species, but other lizards may also be affected. However, some are asymptomatic carriers, increasing the risk of spreading D. agamarum. Usually, D. agamarum is detected with culture-based methods. It was the aim of this study to establish a real-time PCR assay to expand diagnostic options in routine diagnostics. The presented assay is able to detect D. agamarum in clinical samples, decreasing laboratory turn-around time in comparison to conventional culture-based detection methods. This enables a fast therapeutic approach for affected animals and decreases the risk of spread. Abstract (1) Background: Devriesea (D.) agamarum is a potential cause of dermatitis and cheilitis in lizards. The aim of this study was to establish a real-time PCR assay for the detection of D. agamarum. (2) Methods: Primers and probe were selected targeting the 16S rRNA gene, using sequences of 16S rRNA genes of D. agamarum as well as of other bacterial species derived from GenBank. The PCR assay was tested with 14 positive controls of different D. agamarum cultures as well as with 34 negative controls of various non-D. agamarum bacterial cultures. Additionally, samples of 38 lizards, mostly Uromastyx spp. and Pogona spp., submitted to a commercial veterinary laboratory were tested for the presence of D. agamarum using the established protocol. (3) Results: Concentrations of as low as 2 × 104 colonies per mL were detectable using dilutions of bacterial cell culture (corresponding to approximately 200 CFU per PCR). The assay resulted in an intraassay percent of coefficient of variation (CV) of 1.31% and an interassay CV of 1.80%. (4) Conclusions: The presented assay is able to detect D. agamarum in clinical samples, decreasing laboratory turn-around time in comparison to conventional culture-based detection methods.


Introduction
Devriesea (D.) agamarum is a bacterial species known to cause dermatitis and cheilitis in lizards. Disease has most often been described in Uromastyx spp. [1][2][3][4][5]. However, D. agamarum can also infect other lizards [6,7]. It has been reported in captive [8,9] as well as in free-ranging lizards [7]. Clinical signs of disease generally include dermatitis or cheilitis, often described with a yellow crusty appearance [10]. Disease outbreaks with extensive mortality have also been reported, especially if lizards developed septicaemia. [7,11]. Bearded dragons have been described to asymptomatically carry D. agamarum in their oral cavities [2,3]. Treatment of affected animals usually includes debridement of dermal lesions and systemic use of antibiotics-especially cephalosporines are considered effective [4,12]and may also require disinfection of the enclosure [13]. Autovaccines have also been discussed as a treatment method [14].
Therefore, a fast and reliable diagnostic approach is an important consideration, both in clinical disease with suspected D. agamarum infection and in entry controls. D. agamarum is relatively easily cultured at 37 • C but also grows at temperatures of 25−42 • C on Columbia agar with 5% sheep blood [11]. Diagnosis can, however, be complicated in laboratories with limited experience with this pathogen. Matrix-Assisted Laser Desorption/Ionisation Time-of-Flight Mass Spectrometry (MALDI-TOF MS), one of the most frequently used standard techniques for the identification of bacteria in routine laboratory diagnostics, may not (yet) be able to identify D. agamarum when working with standardized databases [15]. However, this issue is likely to be overcome as databases expand. Currently, laboratories can improve D. agamarum identification by implementing and expanding their own MALDI-TOF MS databases or using 16S rRNA gene sequencing to identify cultured but unidentified isolates. Another option would be a specific PCR assay for D. agamarum, which might prove especially valuable if other infectious agents, such as viral or fungal pathogens, are also suspected, allowing concurrent testing from the same sample. The aim of this study was, therefore, to develop a PCR assay for the detection of D. agamarum.

Bacterial Isolates Used to Establish the qPCR
In total, 14 D. agamarum isolates were used in this study as positive controls. Three D. agamarum isolates (GenBank accession numbers: MT664091-93) were obtained from routine diagnostic submissions at Laboklin GmbH & Co. KG (Bad Kissingen Germany) in 2019 [15], while 11 were isolated between 2005 and 2009 at the Faculty of Veterinary Medicine, Ghent University (Table 1). Non-D. agamarum isolates (n = 34) were obtained from the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany). Some were included in order to determine the ability of the assay to exclude a broad spectrum of different bacterial species. Others, like Brachybacterium sp. or Dermabacter sp., were included as their sequences were described to be highly similar to D. agamarum [11] ( Table 2). These 34 isolates served to determine the specificity of the PCR.

DNA Preparation
Pure cultures of each strain were incubated in 750 µL lysis buffer (MagNA Pure DNA Tissue Lysis Buffer, Roche, Mannheim, Germany) and 75 µL proteinase K (proteinase K, lyophilisiert, ≥30 U/mg, Carl Roth GmbH & Co KG, Karlsruhe, Germany) for one hour at 65 • C. From this, 200 µL were utilized for automated nucleic acid (NA) extraction using the MagNA Pure 96 DNA and Viral NA Small Volume Kit (Roche, Mannheim, Germany) according to the manufacturer's instructions. The resulting NAs were eluted in a volume of 100 µL. The isolated NAs were kept at −18 • C until the PCR tests were performed. The DNA used for the dilution series was extracted manually using the QIAamp ® DNA MicroKit (50) (Qiagen, Hilden, Germany), and the DNA concentration was measured with a spectrophotometer (NanoDrop 2000, Thermo Fisher Scientific, Inc., Wilmington, NC, USA).

Design of Primers and Oligonucleotide Probe, PCR Protocol and Optimisation of Annealing Temperature
The 16S rRNA gene was selected as the target region based on the availability of sequence data from a variety of isolates. Sequences of D. agamarum (NZ_LN849456.1, LN849456.1, NR_044368.1, EU009865.1, KF647330.1) were retrieved from GenBank, and multiple sequence alignment was performed with other sequences of different bacterial species (e.g., Agromyces species, Arthobacter species, Brachybacterium species, Dermabacter species, Pseudomonas species) using MUSCLE (https://www.ebi.ac.uk/Tools/msa/muscle/ last accessed on 20 February 2023). The primers and probe were designed using primer3 (https://primer3.ut.ee/ last accessed on 20 February 2023).
Reactions included 1.0 µL of each primer (10 µM), 1.0 µL of the probe (2 µM), 4.0 µL DNA Process Control Detection Kit qPCR Reaction Mix and 5.0 µL template DNA in a total volume of 20.0 µL. Amplification was performed with a LightCycler 96 (Roche, Mannheim, Germany) in a 96-well format.
The following protocol was used: Preincubation at 95 • C for 30 s followed by 40 cycles of two-step-amplification (95 • C for 5 s and 60 • C for 30 s). PCR-grade water (Roche, Mannheim, Germany) served as a negative control. While all other bacterial samples were negative, Dermabacter hominis produced positive PCR results at an annealing temperature of 60 • C ( Figure 1). Therefore, DNA of Dermabacter hominis (DSM 30958) from the DSMZ as well as DNA of D. agamarum (GenBank Accession number: MT664092.1/0919Ur) [15], was tested in duplicate with different annealing temperatures (protocol 1: 65.0 • C, protocol 2: 66.0 • C, protocol 3: 67.0 • C) using a LightCycler 96 (Roche, Mannheim, Germany).

DNA Preparation
Pure cultures of each strain were incubated in 750 μL lysis buffer (MagNA Pure DNA Tissue Lysis Buffer, Roche, Mannheim, Germany) and 75 μL proteinase K (proteinase K, lyophilisiert, ≥30 U/mg, Carl Roth GmbH & Co KG, Karlsruhe, Germany) for one hour at 65 °C. From this, 200 μL were utilized for automated nucleic acid (NA) extraction using the MagNA Pure 96 DNA and Viral NA Small Volume Kit (Roche, Mannheim, Germany) according to the manufacturer's instructions. The resulting NAs were eluted in a volume of 100 μL. The isolated NAs were kept at −18 °C until the PCR tests were performed. The DNA used for the dilution series was extracted manually using the QIAamp ® DNA MicroKit (50) (Qiagen, Hilden, Germany), and the DNA concentration was measured with a spectrophotometer (NanoDrop 2000, Thermo Fisher Scientific, Inc., Wilmington, NC, USA).

Design of Primers and Oligonucleotide Probe, PCR Protocol and Optimisation of Annealing Temperature
The 16S rRNA gene was selected as the target region based on the availability of sequence data from a variety of isolates. Sequences of D. agamarum (NZ_LN849456.1, LN849456.1, NR_044368.1, EU009865.1, KF647330.1) were retrieved from GenBank, and multiple sequence alignment was performed with other sequences of different bacterial species (e.g., Agromyces species, Arthobacter species, Brachybacterium species, Dermabacter species, Pseudomonas species) using MUSCLE (https://www.ebi.ac.uk/Tools/msa/muscle/ last accessed on 20 February 2023). The primers and probe were designed using primer3 (https://primer3.ut.ee/ last accessed on 20 February 2023).
Reactions included 1.0 μL of each primer (10 μM), 1.0 μL of the probe (2 μM), 4.0 μL DNA Process Control Detection Kit qPCR Reaction Mix and 5.0 μL template DNA in a total volume of 20.0 μL. Amplification was performed with a LightCycler 96 (Roche, Mannheim, Germany) in a 96-well format.
The following protocol was used: Preincubation at 95 °C for 30 s followed by 40 cycles of two-step-amplification (95 °C for 5 s and 60 °C for 30 s). PCR-grade water (Roche, Mannheim, Germany) served as a negative control. While all other bacterial samples were negative, Dermabacter hominis produced positive PCR results at an annealing temperature of 60 °C ( Figure 1). Therefore, DNA of Dermabacter hominis (DSM 30958) from the DSMZ as well as DNA of D. agamarum (GenBank Accession number: MT664092.1/0919Ur) [15], was tested in duplicate with different annealing temperatures (protocol 1: 65.0 °C, protocol 2: 66.0 °C, protocol 3: 67.0 °C) using a LightCycler 96 (Roche, Mannheim, Germany).

Determination of Specificity at 66 • C and of Repeatability and Sensitivity of the Assay
Various non-D. agamarum bacterial species (Table 2) were tested in duplicate for positive reactions at 66 • C. The following protocol was used: Preincubation at 95 • C for 30 s followed by 40 cycles of two-step-amplification (95 • C for 5 s and 66 • C for 30 s). PCR-grade water (Roche, Mannheim, Germany) served as a negative control. DNA of 14 D. agamarum isolates (Table 1) was also tested.
To assess the intraassay repeatability of the PCR, standard deviations were calculated for 10-fold serial dilutions in triplicate on a single plate. For the interassay reproducibility, standard deviations were calculated for a 10-fold serial dilution series which was amplified three times daily for two days. The standard deviations of the CT values were used to calculate the percent of coefficient of variation (CV%).
Detection limit for bacterial cell dilutions: In order to determine the sensitivity of the assay, a culture of D. agamarum (isolate MT664091.1/0219Bf) was used, starting with a dilution (D0) of 0.5 McFarland (1.5 × 10 8 per mL). This suspension (D0) was 10-fold serially diluted (D1-D10) in duplicate, and 1 mL of each dilution was inoculated onto Columbia agar with defibrinated sheep blood (Becton Dickinson GmbH, Heidelberg, Germany/Oxoid GmbH, Wesel, Germany), incubated at 36 • C and checked for growth after 30 h. Colony forming units (CFU) were counted if the result was expected to be between 0 and 300 CFU. These results were then used to determine the limit of detection of the assay. DNA was extracted from 200 µL of each dilution (D1-D10) as described above, and PCR was carried out in duplicate. The detection of Dermabacter hominis (DSM 30958) was quantified in the same way.
Detection limit for DNA from pure culture: To evaluate the assay's sensitivity, PCR was carried out in triplicate using serial 10-fold dilutions of DNA prepared from colonies of D. agamarum (isolate MT664091.1/0219Bf).
Spiked-in matrix: The assay was also evaluated using a spiked-in matrix (D. agamarumnegative-tested lizard skin with a known concentration of target DNA). For these assays, 10 µL of the above-described dilutions D0 to D6 were inoculated onto D. agamarum-negativetested lizard skin. The skin was incubated in 500 µL lysis buffer, and 50 µL proteinase K and NA were extracted from 200 µL of this suspension as described above and eluted in a total volume of 100 µL NA. The PCR was carried out in duplicate as described above.

Testing of Samples Submitted to a Commercial Veterinary Laboratory
Clinical samples from lizards for which appropriate material (skin, crusts, dry swab) was submitted to a commercial veterinary laboratory between March 2022 and December 2022 and for which the submitting veterinarian indicated an interest in D. agamarum diagnostics were tested using the established protocol to evaluate the PCR for use with clinical samples. Some of the samples were derived from animals showing clinical signs, others from asymptomatic animals tested in the context of a health check-often when D. agamarum had been isolated from animals in the group previously. If suitable material was available (skin or swab in a transport medium), bacteriology was performed (as previously described [15]). Identification of isolates was based on growth characteristics on agar plates (Columbia Agar with defibrinated sheep blood and Endo Agar, Becton Dickinson GmbH, Heidelberg, Germany), biochemical parameters and MALDI-TOF MS. In doubt, colonies were also re-checked via PCR. If the results of the bacteriological culture were available, the results of the PCR and culture would be compared. Samples were considered PCR positive if the cycle threshold (CT) was <35.0 and equivocal if the CT was ≥35.0, but a signal was obtained. If swabs of different origins regarding the localisation (e.g., dermal and oral) of the same animal were received, all samples were tested separately. Amplicons from positive samples were sequenced (ABI PRISM 3130 XL Genetic Analyser, Applied Biosystems, Foster City, CA, USA) and sequences were analysed by BLAST (https: //blast.ncbi.nlm.nih.gov/Blast.cgi, last accessed on 20 February 2023).

Development of the PCR
The selected primer and probe sequences (Eurofins MWG Operon, Ebersberg, Germany) are shown in Table 3. The product size was expected to be 246 base pairs. Therefore, an annealing temperature of 66 • C was chosen for all further analyses as it was considered sufficient to discriminate between pure cultures of D. agamarum and Dermabacter hominis DNA without losing sensitivity for the detection of D. agamarum.

Specificity, Repeatability and Sensitivity of the Assay
Specificity of the assay using the protocol with 66 • C as an annealing temperature was determined using 14 D. agamarum isolates (Table 1) and 34 non-D. agamarum bacterial isolates (Table 2) in duplicate.
No signal was obtained from any bacterial DNA from isolates other than D. agamarum except for Dermabacter hominis. The CT value obtained using the DNA from pure cultures of Dermabacter hominis were high (34.55 and 34.48) in comparison to those reached using DNA from D. agamarum (12.08-18.14) but still below the threshold set for clinical samples. As these are the results for DNA extracted from pure culture, 66 • C was considered sufficient for further testing of samples without prior cultivation as samples without prior cultivation are expected to yield a lower pathogen level. The intraassay CV was calculated to be 1.31%, while the interassay CV was 1.80%.
Detection limit for bacterial cell dilutions: A positive PCR signal was detected for D. agamarum dilutions D0-D4. An equivocal signal was detected for D5 and D6. In culture, D4 corresponded to 2 × 10 4 colonies per mL. Therefore, the assay sensitivity was 2 × 10 4 colonies per mL with dilutions of bacterial cell culture serving as a template. This corresponds to approximately 200 CFU per PCR. For the serially diluted culture of Dermabacter hominis, no positive PCR signals were observed. Two dilutions (D0 and D1) resulted in equivocal CT values, with D0 being set to 0.5 McFarland (1.5 × 10 8 per mL).
Detection limit for DNA from pure culture: The DNA concentration was determined to be 42.5 ng/µL (A260/A280: 1.94). D. agamarum DNA was detectable in dilutions up to 1:10 5 . Therefore, DNA concentrations of as low as 425 fg/µL were detectable in the PCR.
Spiked-in matrix: Spiked-in matrixes produced clearly positive results up to skin spiked with 10 µL of D2 with D2 corresponding to 2 × 10 6 colonies per mL (approximately 360 CFU per PCR considering dilution during sample preparation).

Testing of Clinical Samples
In order to test the use of the developed PCR for clinical samples, a total of 48 samples from 38 lizards were tested for the presence of D. agamarum ( Table 4). The samples were derived from several species, mostly agamids (Pogona spp. and Uromastyx spp.), and were of different origins (zoological collection, animal rescue centre, private owner) from Germany and the Netherlands. Some of these animals were asymptomatic and were tested in the context of a health check. Others showed clinical signs such as skin lesions, hyperkeratosis or stomatitis (Table 4). Samples were mostly derived from the oral cavity or skin/crusts. Of the 38 animals tested, D. agamarum was detected by PCR in 16 animals (42.10%). A further five animals (13.16%) were considered to have equivocal results, and 17 animals (44.74%) were negative for D. agamarum. One animal (animal 3) that tested negative proved to be infected with a fungus of the family Onygenaceae. A bacteriological examination was performed for 33 of the 38 animals, but D. agamarum was not cultured. However, in most of these cases, various other bacterial species (Table 4) were cultured when bacteriology was performed.

Discussion
D. agamarum is an important pathogen causing skin lesions and, in some cases, systemic disease in lizards. Depending on the species, some animals can be inapparent carriers, while others may develop severe diseases. Diagnosis of the causative agent is therefore important in order to facilitate treatment as well as to prevent the spread of disease. Since animals may suffer when untreated and the risks of spreading increase with time, a fast diagnostic approach is important. The detection of D. agamarum is commonly achieved via culture, followed in some cases by 16S rRNA gene sequencing [11,16].
The PCR developed in this study provides a time-saving tool compared to culture and bacterial identification. Detection of D. agamarum and concurrent bacteriological examination was performed in 33 of the 38 animals, resulting in 13 of 33 clearly PCRpositive animals but no culture-positives. D. agamarum is expected to be abundantly present in symptomatic animals. Culturing of D. agamarum is not considered difficult and has been successfully performed in this laboratory before [15]. However, a successful culture depends on the quality of the submitted samples. Appropriate samples include affected tissue below hyperkeratotic crusts or inside of the crusts as well as organs in septicaemic lizards and subcutaneous granulomas. In asymptomatic animals as well as in symptomatic animals, isolation from the oral cavity, gastrointestinal tract or healthy skin may be challenging. D. agamarum was cultured in the laboratory in which the study was performed during the study period, but these samples were excluded from the study as no suitable material for concurrent PCR testing (e.g., dry swab, skin) was available. In this study, six animals (1, 2, 21 and 22-24) were known to have been symptomatic and had positive PCR results. Bacteriological culture was performed in three of these animals (22)(23)(24). In animal 24, a positive PCR result was only obtained from the skin sample, which was not tested by bacteriological culture. Animal 6 might have been symptomatic (no information was received, but due to the reported previous treatment, it seemed likely). It was treated with antibiotics prior to sampling, which might have influenced the bacteriology results. Possible reasons for the failure of culturing D. agamarum out of positive clinical samples in this study include previous antibiotic treatment, incorrect sampling techniques, contamination with (oral) microbiota, increased transport time, inappropriate transport conditions or overgrowth by other bacteria. The latter is especially important as in the presented cases, no selective media for gram-positive bacteria were used, and various different bacterial species are expected to be present on the skin [17,18].
PCR analysis is useful if the performance of bacterial culture is difficult, e.g., due to previous treatment with antibiotics, inadequate preanalytical conditions (such as increased or decreased temperature, increased transport time, inadequate transport medium), or in cases in which overgrowth by other bacteria make detection challenging or impossible. The detection of D. agamarum via PCR can also simplify concurrent PCR testing for other known pathogens, e.g., viral or fungal pathogens known to cause dermatitis [19,20], since the same extracted nucleic acids can be used. In general, PCR is advantageous when culturable samples are unavailable, for example, when older samples are tested or stored DNA is examined. However, bacterial DNA can persist in the environment [21], and D. agamarum has been shown to survive for several months in the environment, depending on the conditions [13]. A PCR could therefore detect bacteria even in cases in which these were not responsible for clinical signs or in which no replication-competent bacteria were present.
The PCR developed here was not 100% specific. Pure cultures of Dermabacter hominis did result in a weak positive signal. However, if diluted, only equivocal results were observed. Dermabacter hominis is genetically closely related to D. agamarum [2,11,22]. Dermabacter hominis is associated with the human microbiome [23]. It is occasionally described in human clinical samples such as abscesses or blood cultures [24][25][26] but is usually found to be of minor clinical significance [27]. So far, its clinical importance in reptiles is very unclear. Contamination during sampling or sample preparation should be considered a possible option leading to false equivocal results. However, clinical samples are expected to yield less bacterial DNA, making false equivocal results less likely. The 16S rRNA gene is known to be highly conserved between bacterial species, which makes it a useful target if the aim is to identify different bacterial species. It is a commonly used target for bacterial detection, and therefore a large amount of sequence data is available for a wide range of bacterial species. However, it may not be ideal for differentiating closely related bacteria. Currently, the availability of sequence data for D. agamarum other than the 16S rRNA gene is limited, but in the future, other targets may prove to be better options. In the meantime, especially equivocal CT values should be evaluated with caution in the face of clinical signs, sampling and sample preparation, and ideally, retesting is recommended. Possibly, skin samples might prove more useful than swabs as they yielded lower CT values in two of the three animals for which both sample types were available, but this might be highly dependent on the sampling method.
The PCR protocol developed in this study proved helpful for the detection of D. agamarum in clinical samples. D. agamarum was detected in oral swabs from clinically healthy Pogona species and serrated casquehead iguana (Laemanctus serratus), while the results in which equivocal results were obtained were also from clinically healthy Pogona species as well as from clinically healthy black hardun (Laudakia stellio picea). Pogona species have previously been shown to be possible inapparent carriers of D. agamarum and a possible source of infection for more sensitive species [2,3].
Therefore, this PCR protocol may not only be useful for clinical cases but also as a screening tool. However, the number of tested samples is still small, and testing of larger sample numbers is necessary in order to confirm the usefulness of this method for clinical practice.

Conclusions
A real-time PCR was developed that is able to detect D. agamarum in clinical samples. The assay provides a fast method for the detection of this important pathogen of lizards but should be evaluated with further samples in the clinical context.