Phenotypic and Molecular Characterization of the Capsular Serotypes of Pasteurella Multocida Isolates From Bovine Respiratory Disease Cases in Ethiopia

Background: Pasteurella multocida is a heterogeneous species and opportunistic pathogen that causes bovine respiratory disease. This disease is one of an economically important disease in Ethiopia. Losses due to mortality and associated expenses are estimated to be higher in the country. Studies revealed that limited information is available regarding the capsular types, genotypes, and antimicrobial sensitivity of P. multocida isolates circulating in the country. This suggests, further molecular advances to understand the etiological diversity of the pathogens representing severe threats to the cattle population. Results: Bacteriological analysis of nasopharyngeal swab and pneumonic lung tissue samples collected from a total of 400 samples revealed isolation of 61 (15.25%) P. multocida subspecies multocida. 35 (20.59%) were isolated from calves and 26 (11.30%) from adult cattle. Molecular analysis using PCR assay targeting KMT1 gene (~460 bp) amplication was shown in all presumptive isolates. Capsular typing also conrmed the presence of serogroup A (hyaD-hyaC) gene (~1044 bp) and serogroup D (dcbF) gene (~657 bp) from 56 (91.80%) and 5 (8.20%) isolates, respectively. The biovar typing identied 48 (78.69%) isolates of biovar 3. Phylogenetic analysis based on the hyaD-hyaC gene nucleotide sequences showed that P. multocida serogroup A are genetically closely related to the different P. multocida isolates retrieved from the GenBank. Antibiogram of isolates revealed high sensitivity towards chloramphenicol (100%), amoxicillin (95.08%), and amikacin (90.16%). Resistance was observed in cloxacillin (88.52%) and Penicillin-G (77.05%). Conclusions: In the present study, P. capsular type A biovar 3 predominantly isolated from pneumonic cases in cattle useful information research Hence, P. multocida isolates were found to be highly susceptible to chloramphenicol (100%), amoxicillin (95.08%), and (90.14%). Susceptibility was also observed in oxytetracycline (88.52%), streptomycin (83.61%), tetracycline (78.69%), and clindamycin (77.05%). Isolates revealed an intermediate resistance to erythromycin (61.30%) and ampicillin (55.74%). Isolates were resistance to Penicillin-G (77.05%) and cloxacillin (88.52%). Similar results were by Musteria et al. (2017), where susceptibility of chloramphenicol (89.4%) and tetracycline (80.9%) were recorded. Ampicillin (53.2%) was found as an intermediate drug. Penicillin-G (10.6%) and streptomycin (14.9%) were inecient, while vancomycin was inactive against isolates. study showed isolates of P. multocida were susceptible to amoxicillin, chloramphenicol, cephalexin, polymyxin-B, and orfenicol; however, resistance was observed to tetracycline, erythromycin and Penicillin-G.


Background
The genus Pasteurella is a member of the Pasteurellaceae family, which includes a large and diverse group of gram-negative Gammaproteobacteria. Pasteurella species are highly prevalent and cause several economically important endemic and epizootic diseases in a wide range of animals worldwide [1,2,3]. Characteristically, the organism is small (0.2 µm up to 2.0 µm), rods/coccobacilli, capsulated, non-spore-forming, non-motile and bipolar in stain. The bacterium is facultative anaerobic, fermentative and grow best with media supplemented in serum or blood [4,5,6]. Among members of this species Pasteurella multocida (P. multocida) is the most frequently reported heterogeneous species that causes respiratory disease and hemorrhagic septicemia in cattle [7,6]. P. multocidais a commensal and opportunistic pathogen that resides in the upper respiratory tracts of cattle [8,9]. It is a common bacterial pathogen implicated with bovine respiratory disease (BRD), or "shipping fever", non-septicemic pneumonia (10,11). Clinical manifestations range from asymptomatic or mild chronic upper respiratory in ammation to acute pneumonic and/or disseminated disease [3]. Pathogenesis of this organism emerges in terms of complex interaction with host speci c factors such as age, diet, immune status, environment and bacterial virulence factors [4]. The major virulence factors of P. multocida are capsular protein and lipopolysaccharide (LPS) [1]. The other factors may include genes encoding structures such as mbriae and bacterial adhesins or outer membrane proteins [12,13]. However, the presence of P. multocida in the upper respiratory tract of animals is not always associated with disease [14,15].
Strains of P. multocida are classi ed into ve capsular serogroups (A, B, D, E, and F) according to Carter [16] using indirect hemagglutination test and into 16 somatic or LPS serotypes by Heddleston gel diffusion precipitation assay [17]. Polymerase chain reaction (PCR) typing has been applied as a rapid and sensitive molecular method for capsular genotyping using primers designed for species and type speci c detection that was unique to all P. multocida strains [18]. Moreover, multiplex PCR assay was employed as an alternative technique for capsular typing with primers designed following identi cation, sequence determination, and analysis of the capsular biosynthetic loci of each capsular serogroup [19]. Serogroups were further classi ed into eight LPS genotypes (L1 -L8) according to Harper et al. [20].
Despite the extensive research conducted over several years on BRD, it continues to result in great economic impact to the cattle industry [9]. Previous studies revealed the extent of respiratory disease problems in Ethiopia and losses due to mortality, morbidity and associated expenses are estimated to be higher in the country [21]. In those studies, P. multocida was isolated at species level as one of the major bacterial pathogens from nasal, trans-tracheal swab, and pneumonic lung samples using conventional bacteriological methods [22,23,24,25]. However, limited information is available regarding the capsular types and genotypes of P. multocida isolates circulating in the country due to lower sensitivity of detection based on phenotypic characterization. Hence, molecular advances are indispensable to understand the etiological diversity of the pathogens representing severe threats to the cattle population and to compile adequate molecular epidemiology records. Therefore, the present study was designed to determine the phenotypic and genotypic characteristics of P. multocida associated with BRD and to evaluate the antimicrobial susceptibility of isolates from Ethiopian cattle.

Sample size and sample collection
A cross sectional survey with purposive sampling method was employed to collect samples. Nasopharyngeal swabs and pneumonic lung tissue samples were collected from veterinary clinics and abattoir, respectively. A total of 400 samples (n = 170) from calves and (n = 230) from adult cattle were collected during the study period from April 2018 to January 2019.

Nasopharyngeal swab sample
Clinical cases of respiratory infection were inspected, and nasopharyngeal swab samples were collected from clinically sick calves and adult cattle using sterile laryngeal swab. Brie y, swab was directed via the ventral nasal meatus into nasopharynx, rotated vigorously for 30 sec at the contralateral side. The swab was retracted and inserted into sterile screw capped test tube containing transport medium of modi ed Cary-Blair Medium (Park Scienti c, UK).

Pneumonic lung sample
Abattoir survey was carried out on cattle slaughtered and lung with pneumonic cases were inspected for irregularity in shape, cranioventral reddening, marbling, non-friable foci, or brinous pleuritis [26]. Approximately a 3x3 mm piece of lung tissue was taken aseptically from the edge of the lesion and placed in sterile universal bottle. All samples were maintained in cold chain and transported to the Research and Development laboratory of the National Veterinary Institute (NVI) of Ethiopia.
Isolation and phenotypic characterization P. multocida strains were isolated employing standard bacteriological assay. Brie y, pneumonic lung tissue samples were minced and suspended in 4 ml sterile physiological saline (pH 7.0 ± 0.2) and centrifuged at 3260 x g for 3 min and supernatant was discarded. The sediment was reconstituted with 100 µl sterile physiological saline. Ten µl of the suspension and nasopharyngeal swabs were streaked comparably onto blood agar base (HiMedia, India) with 5% sheep blood and MacConkey agar (HiMedia, India).
Plates were incubated at 37 o C for 24 -48 hrs aerobically and phenotypic characterization was carried out on Pasteurella presumptive colonies.
Antimicrobial susceptibility test All P. multocida isolates were tested for antimicrobial susceptibility against twelve types of antibiotics using the disc diffusion method (Quinn et al., 2002). Ampicillin (AMP, 10 µg), Penicillin-G (P 10 units), and Cloxacillin (CX 5 µg) disks (HiMedia, India) were evaluated. Clinical and laboratory standard institute breakpoint [27] was used for interpretation of the results.

Molecular Characterization
Bacterial culture preparation Presumptive isolates of P. multocida were streaked on brain heart infusion (BHI) agar (Oxoid, UK) enriched with 10% horse serum. Plates were incubated at 37ºC for 24 -48 hrs aerobically. Loopful colonies of each of the pure isolates were transferred into a sterile 1.5 ml Eppendorf tubes and washed twice in nuclease free water. Cell suspension was centrifuged at 16,100 x g for 3 min and 1 ml of cell suspension was heated at 100ºC for 10 min to optimize whole-cell DNA templates for PCR assay.

DNA extraction
Genomic DNA was extracted using DNeasy ® Blood and Tissue kit (QIAGEN GmbH, Germany) following the manufacturer's instructions. Brie y, 200µl sample template was transferred into 1.5 ml microfuge tubes. 20 µl proteinase K and 200 µl buffer AL (lysis buffer) was added, mixed and incubated at 56 o C for 10 min. 200 µl 96% ethanol was added, transferred into DNeasy mini spin column with 2 ml collection tube, and centrifuged at 6,000 x g for 1 min. Then, 500 µl buffer AW1 (washing buffer) was added and centrifuged at 6,000 x g for 1 min. Again 500 µl buffer AW2 (washing buffer) was added and centrifuged at 20,000 x g for 3 min.
Finally, samples were transferred into 1.5 ml Eppendorf tubes and 50 µl buffer AE (elusion buffer) was added, centrifuged at 6,000 x g for 1 min, and the eluted DNA yield was stored at -20ºC until PCR assay.
PCR detection of P. multocida PCR assay for P. multocida was carried out using species-speci c primers. All ampli cation and sequencing primers were synthesized by Euro ns Genomics, Austria. Primer for the identi cation of P. multocida species was based on previous report by Townsend et al. [18] and Jakeen et al. [28] as described in Table 1. PCR reaction mix (20 µl) consisted of 2 µl 5 pmol of each primer, 10 µl IQ super mix (Bio-Rad), 3 µl RNase free water and 3 µl DNA template. Ampli cation was carried out using thermal cycler (PCRmax™ Alpha Cycler 2, AC296, UK). Brie y, initial denaturation at 95 o C for 5 min, followed by 35 cycles of denaturation at 95 o C for 1 min, annealing at 55 o C for 1 min, extension at 72 o C for 1:30 min, and nal extension at 72 o C for 7 min.

PCR for capsular typing
The capsular antigens of P. multocida was assayed in multiplex PCR using serogroup speci c primers targeting serogroups A, B, D, and E as described previously [19]. PCR reaction mix was prepared in 40 µl nal volume of 6 µl 5 pmol of each primer, 20 µl of IQ super mix, 2 µl RNase free water, and 6 µl template DNA. Ampli cation was carried out at initial denaturation at 95 o C for 5 min,  Ten µl of each PCR product was mixed with 6x gel loading dye and loaded into separate wells of gel. Ten µl of DNA ladder (100 bp or 1 kb plus, Fermentas) was added into the last lane. Gel electrophoresis was conducted at 120V for 60 min and PCR products were visualized under gel documentation system (UVI TEC, UK) stained with GelRed (Biotium, Inc).
Nucleotide sequencing PCR ampli ed P. multocida speci c DNA band was puri ed using Wizard ® SV Gel and PCR product puri cation kit (Promega, Germany) following the manufacturer's instructions. Concentration of the puri ed PCR product was quanti ed using Nanodrop 2000c spectrometer (Thermo Scienti c, USA). The concentration was adjusted according to the instruction recommended by the sequencing company (LGC Genomics, Germany). The resulting sequences were edited, and contig was formed using Vector NTI 11.5 software (Invitrogen). Consensus sequences were generated and aligned using BioEdit 7.1.3.0. Degree of sequence similarity search was conducted using BLAST program (NCBI).

Ethical statement
Requirement compliance for animal ethics approval from University and Institution was not required for this study. Due to this reason, samples were collected from animals without experimental intervention and lung samples from abattoir slaughtered animals. Consent was rst obtained from the animal owners before sampling.

Data Analysis
Data were analyzed using Graph Pad Prism version 8.4.1 and statistical analysis was considered at P< 0.05. Molecular evolutionary genetic analysis (MEGA) version 7 was used for sequence analysis and ClustalW multiple for sequence alignments [32].
Phylogenetic relationships among the current isolates with isolates from other geographical areas were determined by phylogenetic trees construction using the Neighbor-Joining algorithm.

PCR assay
All presumptive P. multocida isolates were con rmed by PCR assay and revealed ampli cation product of approximately 460 bp band size for species speci c detection (Fig. 4). P. multocida capsular typing con rmed the presence of hyaD-hyaC gene of serogroup A speci c and ampli ed product showed approximate band size around 1044 bp in 56 (91.80%) isolates (Fig. 5).

Sequence analysis
The hyaD-hyaC gene nucleotide sequences of three Ethiopian outbreak isolates (Assosa, Bale Robe and Bishoftu) and 26 sequences retrieved from the GenBank were used. The Neighbor-Joining method with the maximum composite likelihood nucleotide substitution model and the pairwise deletion option was computed using MEGA version 7 software. The percentage bootstrap scores above 50% (out of 1000 replicates) are shown next to the branches (Fig. 6).

Discussion
Bovine respiratory disease (BRD) is an infectious respiratory disease of cattle with multifactorial causes that encompasses multiple viral and bacterial pathogens, a multitude of environmental and management factors and their interactions are responsible for the outbreak of disease [33,34]. The bacterial pathogens associated with BRD include P. multocida, Mannheimia haemolytica, Mycoplasma species, Histophilus somni, and Trueperella Pyogens [35,36]. Bibersteinia trehalosi is also considered as an emerging pathogen to cause BRD but not a primary pathogen to cause BRD [36]. Though, bacterial pathogenic diversity has yet to be further investigated; the current study emphasized on isolation and molecular characterization of P. multocida from BRD cases to show the prevalence and its contribution among all other pathogens in the study areas.
The present study revealed remarkable evidence in the distribution and occurrence of P. multocida infection in different agroecological regions of Ethiopia. Clinical ndings and lung lesion are consistent with clinical features associated with BRD as reported in previous studies [8,36]. In the current study, P. multocida was recovered from 61 (15.25%) of BRD cases. Isolation based on age showed 20.59% and 11.30% was identi ed from calves and adult cattle. This nding is higher than previous report of 3.34% [22] and lower than Musteria et al., [24] reported 39.2%. Similar nding 13.29% was reported by Gebremeskel et al. [25]. As is evident from the study, clinically sick calves showed (16.80%, 31.11%) and adult cattle (10.67%, 11.61%) P. multocida strains from nasopharyngeal and pneumonic lungs, respectively. Calves are more prone to P. multocida infection than adult cattle (P< 0.05) and this may be due to prior exposure of adult cattle to the pathogen [8,9].
Isolation and identi cation of P. multocida based on the cultural and biochemical characters revealed the preponderance of P. multocida subspecies multocida from pneumonic cases of cattle. Higher incidence was identi ed in Bishoftu (25.0%) followed by Bale-Robe (17.50%), Yabelo (16.25%), Mekele (8.75%), and Asosa (8.75%). Though, the overall incidence of P. multocida in the current study was 15.25%, individual cases revealed an incidence rate as high as 44.4% from calves' pneumonic lung tissue in Bishoftu. Analysis of P. multocida infection in recent years indicated that the frequency of this pathogen ranges from 3.4% to 39.2% [22,24,25] in different study areas of Ethiopia. P. multocida could be found in mixed infections of the lung along with M. haemolytica, H. somni, T. pyogenes, B. trehalosi, Mycoplasma species, or various viruses of cattle [36]. Thus, the current nding suggests further investigation to isolate and identify other pathogens associated with bovine pneumonic cases.
In previous studies, in spite of the distribution of P. multocida species [22,24,25] it is not much known about the capsular types and serotypes of P. multocida strains associated with BRD in Ethiopia. Capsular types and serotypes of P. multocida differ based on the type of disease and host species affected. In this study, Capsular typing revealed that P. multocida serogroup A (91.80%) is the predominant strain followed by serogroup D (8.20%). This nding is consistent with Kong et al. [37] who identi ed serogroup A as a predominant isolate from cattle. In similar study, Katsuda et al. [38] detected capsular type A (93.7%), D (6.3%), and Ewers et al., [39] identi ed capsular type A (93.2%) and D (3.3%) from cattle. Higher incidence of capsular type A (99.3%) has also been reported by Davies et al. [40] and lower prevalence of serogroup A (68.19%) and D (22.72%) has been identi ed from cattle and buffalo [41].
The current nding showed that P. multocida capsular type A biovar 3 was identi ed as the predominant isolate in the study areas, which account for 78.69% followed by biovar A:1, A:2, and A:12 at rate of 11.47%, 6.56%, and 3.28%, respectively. Phylogenetic analysis based on the hyaD-hyaC gene nucleotide sequences showed that the three P. multocida serogroup A current isolates from Ethiopia (Assosa, Bale Robe and Bishoftu) are genetically closely related to the different P. multocida isolates retrieved from the GenBank. The nucleotide sequences of the current isolates submitted to the GenBank to get gene accession number.
In this study, P. multocida isolates were found to be highly susceptible to chloramphenicol (100%), amoxicillin (95.08%), and amikacin (90.14%  [22] showed isolates of P. multocida were susceptible to amoxicillin, chloramphenicol, cephalexin, polymyxin-B, kanamycin and orfenicol; however, moderate resistance was observed to tetracycline, erythromycin and Penicillin-G. Isolation and distribution of P. multocida serotypes vary in different areas over time. Therefore, subsequent monitoring on emerging serotypes of P. multocida is essential for the development of effective control strategy. In Ethiopia, a monovalent vaccine prepared from inactivated P. multocida capsular type B at the National Veterinary Institute (NVI) has been used for long period of time. Despite the use of this vaccine, there is high rate of morbidity and mortality following respiratory infection in different parts of the country [21]. This may be due to limited protection of the vaccine for heterologous serotypes of P. multocida strains [42,43]. Although the ndings in the current study which depicts the frequent isolation of P. multocida serogroup A is by no means conclusive, it highlights useful information vis-à-vis the use of monovalent vaccine comprising P. multocida serogroup B for the control of pneumonic pasteurellosis in Ethiopia. Hence, serotyping and genotyping of isolates and genetic relatedness is required to compile information on molecular epidemiology of P. multocida and other pathogens causing BRD in the country for the improvement of the existing prevention and control strategy.

Conclusion
Although many pathogens are responsible for majority of the cases which are yet to be determined, the current study revealed that P. multocida capsular type A biovar 3 is the most common serogroup associated with bovine respiratory cases in the study areas. The current nding suggests further comprehensive study on the possible pathogens involved in precipitating BRD at national level. The current nding provides useful information for further research, improvement of currently used vaccine, and choice of antimicrobials for treatment. This study has been approved by the doctoral committee of Koneru Lakshmaiah Education Foundation, department of biotechnology for requirement compliance of ethical standards in handling and specimen collection from animals. Consent was rst obtained from the animal owners before sampling.

Consent for Publication
Not applicable Availability of data and materials All data supporting the ndings of this study can be obtained from the corresponding author upon formal request. Nucleotide sequence data generated in this study is submitted to GenBank Database to get accession numbers.

Competing Interest
Authors declare that they have no competing interest.

Funding
Laboratory analysis of this research was supported nancially by National Veterinary Institute (NVI) of Ethiopia.

Authors' contributions
All authors participated in the conception and design of the study; MA conducted all laboratory works and drafted the manuscript. BV, TA, and EG analyzed the data and rigorously revised the manuscript. AB, GD, and TD participated in laboratory work. MY coordinated and revised the paper. All authors read and approved the nal manuscript.  Antimicrobial susceptibility pattern of P. multocida isolates.

Figure 4
Agarose gel electrophoresis of PCR ampli ed product of KMT1 gene (~460 bp) speci c for P. multocida. Lane M: 100 bp DNA ladder; lane P: Positive control; lanes 1-13: tested samples; lane E: extraction control; lane N: negative control.