Development and Validation of PCR Diagnostic Assays for Detection of Avibacterium paragallinarum and Ornithobacterium rhinotracheale

Simple Summary Respiratory diseases cause severe economic damage to poultry farming. Common respiratory diseases in poultry include ornithobacteriosis and infectious coryza. The causative agents of these diseases, Ornithobacterium rhinotracheale and Avibacterium paragallinarum, are difficult-to-cultivate bacteria and are often found in mixed infections, which complicates the diagnosis of these diseases. PCR is a rapid, sensitive, and specific method that does not require special sample preparation; therefore, it is often used to detect infectious agents of animal diseases. We developed and validated two efficient and sensitive diagnostic assays for the rapid and accurate detection of A. paragallinarum and O. rhinotracheale DNA in bacterial isolates and clinical samples using real-time PCR. These diagnostic assays are expected to improve the differential diagnosis of respiratory diseases in poultry. PCR-based diagnostic assays can provide an alternative to microbiological cultivation and significantly reduce the time required to obtain results. Abstract PCR is the most effective method for detecting difficult-to-cultivate pathogens and pathogens that are part of mixed infections in animals, such as Ornithobacterium rhinotracheale, which causes bird ornithobacteriosis, or Avibacterium paragallinarum, which causes infectious coryza. In this work, we developed and validated two efficient and sensitive diagnostic assays for the rapid and accurate detection of A. paragallinarum and O. rhinotracheale DNA in bacterial isolates and clinical samples using real-time PCR with TaqMan-like probes. When designing the PCR assays, we performed in silico analysis, optimized DNA isolation methods and PCR conditions, and assessed the analytical and diagnostic performance of PCR. We designed primers and probes that have no mismatches with published whole-genome sequences of bacteria. The optimization of conditions showed that the PCR assays are sufficiently robust to changes in temperature and oligonucleotide concentration. The validation showed that the developed assays have high analytical and diagnostic sensitivity and specificity. These assays are expected to improve the differential diagnosis of respiratory diseases in chickens and turkeys.


Introduction
Respiratory diseases are among poultry's most frequent infectious pathologies and cause serious economic damage to aviculture.They can be caused by various bacterial and viral pathogens, either independently or in association.Among the bacterial infections of the respiratory tract in poultry, infectious coryza and ornithobacteriosis occupy the leading places.The causative agent of infectious coryza in chickens, Avibacterium paragallinarum (formerly Haemophilus paragallinarum), is a Gram-negative bacterium of the Pasteurellaceae family; it is divided into three serogroups (A, B, and C) that comprise nine serotypes.In different countries, different serogroups and serotypes of the pathogen are predominant [1][2][3][4].Infectious coryza of chickens is common in countries with developed aviculture, including Russia [5][6][7]; it leads to a slowdown in the growth rate of broilers and a decrease in the egg production of chickens [8].Chickens and turkeys of all ages are susceptible to infectious coryza, especially chickens older than 4 weeks.
The causative agent of ornithobacteriosis, Ornithobacterium rhinotracheale (ORT), is a Gram-negative, sedentary, non-spore-forming bacterium of the Flavobacteriaceae family with 18 known serotypes (A-L) that are not directly associated with virulence [9,10].Meat poultry, such as turkeys and broiler chickens, are most often affected [11,12].Since nonspecific and variable clinical manifestations characterize the disease, it is essential to have a specific and sensitive diagnostic test [13].Phenotypic and immunological methods for the diagnosis of ornithobacteriosis still need to be developed or improved [14].
A. paragallinarum and O. rhinotracheale are difficult to cultivate in vitro, and their identification by classical microbiological methods is arduous.Moreover, isolates can be obtained from clinical material only during the acute period of the disease.These microorganisms are often accompanied by a co-infection with other pathogens such as E. coli, P. multocida, Bordetella avium, Mycoplasma gallisepticum, etc., which also complicates the isolation of pure cultures and diagnosis of the diseases [15,16].An alternative approach to rapid diagnostics of infectious coryza and ornithobacteriosis is the use of PCR techniques.This approach does not require a large amount of material and does not require the isolation of a pure bacterial culture.
Currently, there are no validated and approved diagnostic PCR assays in the Russian Federation for detecting the DNA of A. paragallinarum and O. rhinotracheale in clinical material.Moreover, no guidelines for the PCR diagnostics of infectious coryza and ornithobacteriosis have yet been developed by the World Organization for Animal Health (WOAH, founded as OIE).Considering the growing size of the poultry industry and the economic losses from these diseases, the importance of developing PCR assays is undeniable.Thus, our work aimed to develop and validate PCR assays for the detection of pathogens in infectious coryza and ornithobacteriosis.

The Panel of Samples: Bacterial Isolates, Viruses, Hosts, and Clinical Samples
The panel of samples was formed, comprising isolates of O. rhinotracheale and A. paragallinarum (serogroups A, B, and C), strains of causative agents of bacterial diseases of animals and birds that are taxonomically close to the target ones, as well as causative agents of diseases with symptoms similar to infectious coryza and ornithobacteriosis, including viral ones.The strains of O. rhinotracheale and A. paragallinarum were obtained from the research collection of the Federal Center for Animal Health (ARRIAH).Table S1 presents more information about the strains.Samples of viral cDNA and genomic DNA from vaccine strains, field isolates, ATCC isolates, and animals were obtained from the research collection, which was formed in the Department of Biotechnology of the Russian State Center for Quality and Standardization of Veterinary Drugs and Feed (VGNKI).In addition, the panel included DNA samples from five avian species (Gallus gallus, Meleagris gallopavo, Anas platyrhynchos, Coturnix coturnix, and Columba livia), which are the primary hosts of the studied pathogens (Table S2).
Additionally, 120 clinical samples from chickens were obtained from ARRIAH.The type and number of samples are shown in Table 1.Sixty samples of pathological material were obtained from sick chickens.The diagnosis of ornithobacteriosis and infectious coryza was established based on clinical signs.Subsequent confirmation of these 60 positive clinical samples was performed through bacteriological isolation of O. rhinotracheale and A. paragallinarum, respectively.Sixty samples were also collected from healthy birds without clinical manifestations of any disease.The selection of target genes was carried out using whole-genome sequences of A. paragallinarum and O. rhinotracheale deposited at NCBI (Tables S3 and S4).The selection and quality assessment of oligonucleotide primers and probes for targets was carried out using the online resources "Eurofinsgenomics PCR Primer Design" [17], "IDT OligoAnalyzer" [18], "PCR Primer Stats" [19], and "Eurofinsgenomics Oligo Analysis Tool" [20].The optimal annealing temperature and the possibility of forming homo-and heterodimers and "hairpins" were assessed.The sequences and characteristics of the oligonucleotides are provided in Tables 2 and 3.An internal control (IC) whose nucleotide sequence was not homologous to any of the analyzed organisms was designed to provide confidence in successful DNA isolation and amplification.The IC sequence is presented in the Suppl.Materials.Recombinant phage λ with an inserted IC sequence was produced in GenTerra JSC (Russia).
PCR optimization was carried out on concentrations of oligonucleotides, concentration of magnesium chloride, and temperature parameters on a Rotor-Gene Q thermocycler (Qiagen, Hilden, Germany), with 10 µL of DNA per reaction volume of 25 µL (see Supplementary Materials).Optimized conditions for PCR assays for the detection of A. paragallinarum and O. rhinotracheale DNA are provided in Table 4.DNA extraction was conducted from 100 µL of a bacterial suspension of pure cultures of isolates of O. rhinotracheale and A. paragallinarum (Section 2.1) using the AmpliSens ® RIBO-prep kit according to the manufacturer's instructions (CRIE, Moscow, Russia).
DNA was extracted directly from clinical specimens.Preliminarily, tissue fragments were minced using scissors, diluted 1:10 in sterile 0.9% NaCl solution, homogenized for 4 min using the automated homogenizer TissueLyser LT (Qiagen, Hilden, Germany) according to the manufacturer's instructions, and centrifuged at 1500× g for 2 min to obtain the supernatant.DNA was extracted from the supernatant using the AmpliSens ® RIBO-prep kit.
Nucleic acids from vaccine strains, field isolates, and ATCC isolates (Section 2.1) were extracted previously and stored at −40 • C in the VGNKI research collection.
Host DNA extraction was conducted from bird muscle tissue fragments using the AmpliSens ® DNA-sorb-C-M kit following the manufacturer's instructions (CRIE, Moscow, Russia).
Before DNA extraction, 10 µL of IC containing 2 × 10 5 copies/mL recombinant phages λ was added to each test tube containing the tested samples.

Validation of PCR Assays
Validation of diagnostic assays was carried out for compliance with the decisionmaking criteria for PCR methods according to a standard scheme, assessing the analytical and diagnostic characteristics of PCR [21][22][23].

Analytical Specificity
To confirm the specificity of the PCR assays, we used the panel of samples of bacterial isolates, viruses, and hosts described in Section 2.1.The studies were carried out in three independent rounds of PCR, each of which contained two replicates of each sample.

Analytical Sensitivity (Limit of Detection), Efficiency, and Coefficient of Determination (R 2 )
To determine the PCR analytical characteristics, we used DNA of plasmids containing cloned PCR fragments of the O. rhinotracheale rnaP and A. paragallinarum lysS genes and genomic DNA of A. paragallinarum and O. rhinotracheale.The recombinant pAL2-TA plasmids were produced in Evrogen (Moscow, Russia).The plasmids were sequenced to confirm the presence of the intended insert.The concentration of plasmids and genomic DNA was measured on a Quantus fluorimeter (Promega, Madison, WI, USA) following the manufacturer's instructions.The genomic and plasmid DNA copy numbers were calculated using the following equation: Number of copies = (X ng/µL × 6.0221 × 10 23 molecules/mol)/ (N × 660 g/mol × 1 × 10 9 ng/g), where X = Quantus read (ng/µL); N = length of the plasmid or bacterial genome size in bp; and 660 g/mol = average mass of one bp dsDNA.Tenfold serial dilutions of plasmid DNA in the concentration range of 10 8 -10 2 copies/mL were prepared.Dilutions of genomic DNA from A. paragallinarum and O. rhinotracheale in the concentration range of 10 8 (10 7 )-10 2 copies/mL, respectively, were prepared.To generate a standard dose-dependent curve, two replicates of each tenfold serial dilution of plasmid or genomic DNA were used, and the mean threshold cycle (Ct) value and standard deviation (SD) were calculated.The mean Ct values were plotted against log10 of DNA copy number per reaction, and linear curves were generated to estimate the curve slope, PCR efficiency, and coefficient of determination (R 2 ).
PCR efficiency (E) was assessed using the standard curve slope, as shown in the following equation: Repeatability and reproducibility were assessed by the coefficient of variability of Ct values, CV(Ct), and the coefficient of variability of the number of genomic copies per reaction, CV(N).

Diagnostic Specificity and Diagnostic Sensitivity
The diagnostic characteristics of the PCR assays were determined based on the analysis of the pathological tissues of chickens, in which the presence or absence of pathogens of infectious coryza and ornithobacteriosis was clinically confirmed.The type and number of specimens are shown in Section 2.1 (Table 1).The studies were performed in three independent rounds of PCR, each containing two replicates of each sample.Indices were calculated according to Formulas (3) and (4).

Design of Real-Time PCR Assays
We chose housekeeping genes with conservative sequences within species to design the PCR assays.Specifically, we chose genes that participate in translation since translation is one of the most essential processes in the cell [24].As of February 2023, the NCBI refseq database had whole-genome data for 45 isolates of A. paragallinarum (Table S3).Approximately one-third of the genomes were assembled completely; the rest were at the contig or scaffold level.Among the sequenced whole-genome isolates, there were representatives of A. paragallinarum from all three serogroups (A, B, and C).The primers and probe used for detecting A. paragallinarum DNA (Table 3) were matched to the housekeeping gene lysS, which encodes a lysine-tRNA ligase involved in the translation process.Forward and reverse primers were designed to amplify a 139 bp segment from nt number 699,152 to 699,290 (numbering according to accession number NZ_CP050316.1).The primers were selected based on the analysis of multiple sequence alignments of the lys gene of A. paragallinarum isolates and strains from the NCBI refseq database.Figure S1A presents the annealing regions of the primers and probe, showing 100% homology between the selected oligonucleotides and all A. paragallinarum sequences from the NCBI refseq database.
The specificity of the oligonucleotides was assessed by aligning the selected primers and probe to the lysS gene sequence of A. paragallinarum-related bacteria (Figure S1B).The oligonucleotides have from two to six mismatches with the sequence of the lysS gene of the closest relatives.Thus, PCR with the selected primers and probe should specifically detect the DNA of A. paragallinarum.
For the PCR detection of O. rhinotracheale, we selected primers and a probe (Table 3) for a region of the rnaP gene (ribozyme RNase P gene), which is highly conserved in this species.The target fragment is 101 bp long, from nt number 72,116 to 72,216 (numbering according to accession number NZ_CP006828.1).The rnaP gene is a housekeeping gene that belongs to the category of non-coding RNA genes.The sequence of this gene consists of sub-sequences that are highly conservative among bacteria and sub-sequences that are specific to more narrow taxonomic groups [25].The primers were selected by analyzing multiple sequence alignments of the rnaP gene of all isolates and strains of O. rhinotracheale, whose whole-genome sequences were deposited in the NCBI refseq database.Figure S2A shows the annealing region of the selected primers and probe.The forward primer at positions 5 and 17 has mismatches in the sequence with the genome of the isolate FARPER-174b (NZ_CP CP035107.1).We decided to ignore this since this isolate was isolated from the discharge of the paranasal sinuses of laying hens in Peru, i.e., in an agrarian region remote from Russia.
In silico, the primer specificity was assessed using the Primer-BLAST online resource.Figure S2B shows multiple alignment in the region of primer and probe annealing in the rnaP gene sequence taken from the representative genome of O. rhinotracheale and representative genomes of some related species with a high similarity in the rnaP gene (Ornithobacterium hominis, Riemerella anatipestifer, Bergeyella cardium, and Capnocytophaga gingivalis).The primers and probe have many mismatches with the rnaP gene sequence of other bacterial species.Thus, PCR with the selected primers and probe should effectively detect the DNA of O. rhinotracheale.

Analytical Specificity
To confirm the specificity of the PCR assays, a panel of DNA samples was used.The PCR results are presented in Table 5.In our sample panel, the PCR assays detected the pathogens of infectious coryza and ornithobacteriosis with 100% specificity.DNA amplification was observed only for the target microorganisms.

Analytical Sensitivity (Limit of Detection), Efficiency, and Coefficient of Determination (R 2 )
In the experiments to determine analytical characteristics, it became apparent that the LOD of detection for A. paragallinarum was 7 × 10 3 copies/mL, and for O. rhinotracheale it was 4 × 10 3 copies/mL.The efficiency, slope, and coefficient of determination for PCR detection of A. paragallinarum DNA and O. rhinotracheale DNA corresponded to the accepted criteria (efficiency > 90%, slope 3.3-3.6,R 2 > 0.98) (Tables 6 and 7).

Repeatability
We determined repeatability using PCR with plasmid and genomic DNA, whose copy numbers per reaction were close to the limit of PCR sensitivity.PCR with 70-170 copies of DNA from A. paragallinarum per reaction (N) was performed in 10 replicates along with PCR with calibration samples of plasmid or genomic DNA serially diluted from 10 8 to 10 4 copies/mL.PCR with 40-105 copies of DNA from O. rhinotracheale per reaction (N) was performed in 10 replicates along with PCR with calibration samples of plasmid DNA serially diluted from 10 8 to 10 4 copies/mL or genomic DNA serially diluted from 10 7 to 10 4 copies/mL.After that, CV(Ct) and CV(N) were calculated.The results are presented in Table 8.For both PCR assays, good values of Ct repeatability were obtained; the coefficient of variability CV(Ct) ranged from 0.5% to 1.4%.The coefficient of variability for DNA copies per reaction CV(N) ranged from 10.7% to 25%, which corresponds to the accepted criteria.
Thus, ~70 copies of the A. paragallinarum lysS gene and ~40 copies of the O. rhinotracheale rnaP gene were reproducibly detected in the respective PCR assays.In other words, the LOD of the PCR assay for A. paragallinarum was 70 genomic copies per reaction (=7 × 10 3 copies/mL), and for O. rhinotracheale it was 42 genomic copies per reaction (=4.2 × 10 3 copies/mL) (Tables 6 and 7).
To obtain reproducible results in laboratory practice, samples with target DNA content above the LOD should be considered "positive".This requires setting the cutoff limit of the cycle.We considered the cutoff cycle limit at Ct ≤ 33 for the PCR assay for A. paragallinarum detection and at Ct ≤ 32 for the PCR assay for O. rhinotracheale detection.

Reproducibility
We assessed PCR reproducibility using three independent experiments performed on different days by different operators using different thermal cyclers.In each experiment, we used two replicates of each tenfold serial dilution of plasmid DNA and genomic DNA, as well as DNA samples diluted close to the PCR sensitivity limit.The results are presented in Tables 9 and 10.For the A. paragallinarum PCR assay, the coefficient of variability CV(Ct) ranged from 0.3 to 1.4% and the CV(N) from 1.9% to 20.7%.For the O. rhinotracheale PCR assay, the coefficient of variability CV(Ct) ranged from 0.4% to 1.9% and the CV(N) from 2.2% to 22.7%.These values indicate acceptable reproducibility for both PCR assays.

Diagnostic Specificity and Sensitivity
To assess the diagnostic characteristics of the PCR assays, 30 known positive and known negative clinical samples from chickens were analyzed for each pathogen separately.The results are presented in Tables S5 and S6.All known positive samples containing A. paragallinarum or O. rhinotracheale were found to contain DNA of A. paragallinarum or O. rhinotracheale.All known negative samples that did not contain A. paragallinarum or O. rhinotracheale tested negative.The indices of diagnostic sensitivity and specificity calculated from the obtained results with a 95% confidence level are presented in Table 11.

Discussion
The common avian bacterial pathogens, A. paragallinarum and O. rhinotracheale, are challenging to diagnose.Not surprisingly, the first PCR-based test for A. paragallinarum identification was developed in the 1990s.To identify A. paragallinarum DNA, Chen et al. developed a PCR-based assay with detection using gel electrophoresis in 1996.The primers were selected for the region of the pyrG gene that encodes a glutamine-hydrolyzing enzyme [26].In 2008, Corney et al. developed another assay for the detection of A. paragallinarum DNA based on real-time PCR with primers for the pyrG gene [27].Clothier et al. successfully validated this assay as a diagnostic one in 2019 [28].In 2021, Kuchipudi et al. [29] developed an assay based on real-time PCR with primers for a region of the A. paragallinarum recN gene encoding a repair protein.During validation, the assay was more sensitive compared to the assay of Corney et al. [27].Bogomazova et al. [30] found that the primers perfectly matched recN gene sequences from all 50 A. paragallinarum genomes deposited in the NCBI refseq database.However, the recN gene is not part of the minimal set of genes required for the bacterial cell [24], and its loss may be compatible with the preservation of viability.
To detect O. rhinotracheale DNA, a real-time PCR assay with primers for the 16S ribosomal RNA gene (16S rRNA) was developed in 2013 [31].In 2022, Hashish et al. improved this technique by changing the probe position and proposing their own unique set of primers and probe for the same target [32].It should be noted that the 16S rRNA gene is traditionally used for bacterial identification and phylogenic studies; it belongs to the category of non-coding RNA genes, is conservative, and has multiple copies [33].The genome of O. rhinotracheale contains three copies of the 16S rRNA gene.We tested the design of Hashish et al. [32] on all 16 O.rhinotracheale whole genomes deposited in the NCBI database and found that in 14 out of 16 cases (88%), the primers and probe annealed perfectly to the 16S rRNA gene sequence.However, in 2 strains out of 16, we found 1-2 mismatches between the 16S rRNA gene sequence and the probe sequence, which can significantly reduce PCR efficiency.
Bogomazova A. et al. [30] described several principles for in silico analysis in the design of PCR assays for the diagnosis of infectious animal diseases.Namely, it is necessary to (1) take into account the intraspecies genetic polymorphism of the bacterial pathogen; (2) indicate the NCBI database (nt/nr or refseq genomes or wgs) for BLAST searches and the number of bacterial gene sequences available for analysis; and (3) indicate the presence or absence of polymorphism at the sites of primer and probe annealing.The design of the assays for identifying pathogens of infectious coryza and ornithobacteriosis in this work was carried out considering these principles.As target genes for PCR, we chose genes participating in translation since these genes are part of a minimal gene set essential for bacterial viability [24].This approach should provide the necessary robustness to the PCR assays developed.

Conclusions
In this work, we developed and validated two efficient and sensitive diagnostic assays for the rapid and accurate detection of A. paragallinarum and O. rhinotracheale DNA, both in bacterial isolates and in clinical samples, using real-time PCR with TaqMan-like probes.The use of these diagnostic assays will improve the differential diagnosis of respiratory diseases in chickens and turkeys.PCR-based diagnostic assays can be considered an alternative to microbiological culturing and significantly reduce the time required to obtain results.

Table 1 .
Clinical samples of A. paragallinarum and O. rhinotracheale.

Table 2 .
Characteristics of targets and oligonucleotides for the PCR assays for detecting A. paragallinarum DNA.

Table 3 .
Characteristics of targets and oligonucleotides for the PCR assays for detecting O. rhinotracheale DNA.
2.3.Nucleic Acids ExtractionIn this work, we tested two different methods for extracting DNA: precipitation and DNA sorption on silica gel (see Supplementary Materials).

Table 5 .
Results of PCR for detection of A. paragallinarum lysS and O. rhinotracheale rnaP gene fragments.

Table 6 .
Limit of detection and standard curve parameters of PCR assay for A. paragallinarum.

Table 7 .
Limit of detection and standard curve parameters of PCR assay for O. rhinotracheale.

Table 8 .
Assessment of repeatability of PCR assays.

Table 9 .
Assessment of reproducibility of the PCR assay for A. paragallinarum detection.

Table 10 .
Assessment of reproducibility of the PCR assay for O. rhinotracheale detection.

Table 11 .
Diagnostic specificity and diagnostic sensitivity of the PCR assays for A. paragallinarum and O. rhinotracheale DNA detection.