Oral flora of stray dogs and cats in Algeria: Pasteurella and other zoonotic bacteria

Background and Aim: Knowledge of potentially pathogenic bacteria presents in the oral cavity of dogs and cats may be helpful in determining appropriate treatment for infected bite wounds. About 120.000 people are exposed to dog and cat bites every year in Algeria, but little is known about the dog and cat oral flora causing bite wound complications. The purpose of this study was to identify potential zoonotic bacteria from oral cavity of dogs and cats and to determine their susceptibility to antibiotics to contribute to the treatment of bite wound infection. Materials and Methods: Oral swabs from 100 stray dogs and 100 stray cats were collected and cultured in several media: Chocolate agar, MacConkey agar, and Mannitol Salt Agar. Bacterial isolates were identified using several commercial kits of the analytical profile index and tested for antibiotic susceptibility by disk diffusion method. Results: Overall, 185/200 (92.5%) dogs and cats carried zoonotic bacteria in their mouths, of which 55.13% (102/185) had at least two bacterial pathogens. 374 pathogenic strains belonging to 15 genera were isolated: Eleven were Gram-negative (Proteus, Pasteurella, Escherichia, Moraxella, Klebsiella, Acinetobacter, Enterobacter, Pseudomonas, Aeromonas, and Neisseria Haemophilus) and four were Gram-positive (Staphylococcus, Streptococcus, and Corynebacterium, Bacillus). Fifty-one strains of Pasteurella were isolated from 44 carriers of Pasteurella (21 Pasteurella multocida, 21 Pasteurella pneumotropica, and 9 Pasteurella spp.). Pasteurella strains were tested for antibiotic resistance. Resistance to at least one drug was observed in 8 (15.68%) of Pasteurella isolates and two strains (3.92%) were found to be multidrug-resistant (to two or more drugs). Erythromycin, penicillin, and ampicillin were the antimicrobials to which the isolates showed greater resistance (7.84%, 5.88%, and 3.92%, respectively). Conclusion: To the best of our knowledge, this study is the first in Algeria to detect potential human pathogenic bacteria in the oral cavity of dogs and cats. It reveals that these animals have multiple zoonotic bacteria in their mouths including Pasteurella species, which may be multidrug-resistant.

Available at www.veterinaryworld.org/Vol. 13/December-2020/29.pdf In Algeria, the number of stray dogs and cats wandering freely through the streets and urban areas is constantly increasing over the last few years, placing people at risk of bites [15,16].
Furthermore, around 120.000 people are exposed to animal bites each year, 80% of whom are bitten by stray dogs [15,16].
Despite the high frequency of annual visits to the emergency departments, in our country, due to animal bites [15,16], the oral flora of dogs and cats causing bite wound complications is still relatively unstudied.
The purpose of this study was to identify potential zoonotic bacteria from oral cavity of dogs and cats and to determine their susceptibility to antibiotics to contribute to the treatment of bite wound infection.

Ethical approval
The study protocol was approved by the ethics committee and decision board (number 01/2018) of P.I.C.C-U.H.E.P. of Algiers.

Study area
The study was conducted in the Department of Algies which is situated on the central coast of Algeria between 3°2'31.09'' east longitude and 36°45'9'' north latitude. It covers an area of 1190 km 2 and comprises 57 districts with a population of over 2.9 million inhabitants. Algiers is bounded from the north by the Mediterranean Sea, from the south by the Blida department, from the east by the Boumerdes department, and from the west by the Tipaza department. Public Industrial and Commercial Company-Urban Hygiene and Environmental Protection (P.I.C.C-U.H.E.P) is affiliated with the Algerian Ministry of Water Resources and Environment that controls zoonosis and vector-borne diseases such as rabies and leishmaniosis.
In the context of the National Rabies Prevention Program, P.I.C.C-U.H.E.P. Catches stray dogs and cats in the 57 districts of the Algiers department. The captured animals are subsequently sheltered in the dog-pound of El-Harrach during the legal period (7 days) before euthanasia to allow for owners to claim their pets in compliance with the Algerian legislation on the protection of animals. The geographical position of the dog-pound and the animal catching radius in Algiers is represented in Figure-1.

Sampling
From January 2018 to July 2019, 200 buccal swabs (tongue, gum, palate, cheeks, and teeth) were collected at irregular intervals from 100 stray dogs and 100 stray cats which were randomly selected without distinction of age, sex, or breed. All the sampled animals were caught by the P.I.C.C-U.H.E.P. from the 57 municipalities of Algiers during the study period that was apparently healthy and did not receive any previous treatment.

Bacterial analysis of samples
The specimens were transported in an icebox (+4°C) within 60 min after collection to the microbiology laboratory for further analysis. Following a control quality step, Escherichia coli American Type Culture Collection (ATCC) 25922, Staphylococcus aureus resistant Type Culture Collection (ATCC) 43300, S. aureus sensitive Type Culture Collection (ATCC) 25923, and Pseudomonas aeruginosa Type Culture Collection (ATCC) 27853 were used.
Each swab was directly plated on chocolate agar (Columbia agar medium supplemented with 5% of blood sheep and slowly heating to 80°C) (Pasteur Institute, Algeria), MacConkey agar (Pasteur institute, Algeria), and Mannitol salt agar "Chapman Medium" (Pasteur institute, Algeria). Chocolate agar plates were incubated for 48-72 h at 37°C in a 5% CO 2 atmospheric chamber, and both MacConkey and Mannitol salt agar plates were incubated aerobically at 37°C for 24 h.
After 24, 48, and 72 h of incubation, the plates were examined and different types of colonies were then subcultured separately to the appropriate media for the processing of pure cultures. When pure cultures were obtained, each isolate was identified according to the morphology of the colonies, the type of hemolysis, the Gram staining microscopy, and several biochemical tests for characterizing isolates.
Biochemical tests included: Catalase and oxidase activities, substrate utilization as determined by commercial biochemical test kits, and coagulase testing using rabbit plasma for Staphylococci identification. Pasteurella spp. were identified on the basis of Gram staining (Gram-negative coccobacilli), absence of hemolysis, absence of growth on MacConkey, and positive biochemical reactions: Catalase and oxidase activities, and indole test. Final identification of strains was done using API20NE commercial kit (Biomérieux, Marcy l'Etoile, France).

Antimicrobial susceptibility testing of Pasteurella
A subset of Pasteurella strains were tested for susceptibility to four antibiotics which are commonly prescribed for the treatment of animal bites. Resistance to amoxicillin/clavulanic acid 30 µg; penicillin 10 IU; ampicillin 10 µg; tetracycline 30 µg; and erythromycin 5 µg was determined on Muller-Hinton agar supplemented with 5% sheep blood (Pasteur Institute, Algeria) using the standard Kirby-Bauer disk diffusion method.

Statistical analysis and map conception
Pearson's Chi-square test (α=5%) was applied to compare Pasteurella prevalence, using Microsoft Excel world 2007. Differences were statistically significant when p<0.05.
The map showing the geographic location of the dog-pound and the animal catching area was conceived using Microsoft PowerPoint 2007, as available in Microsoft Office software (Microsoft, USA), while the quality of the artwork was enhanced by Adob Photoshop software.7.0.1 (Adobe, USA).

Prevalence and bacterial isolation
During the study period, a total of 100 stray dogs and 100 stray cats were swabbed. Bacteriological culture showed that 92.5% (185/200) of the swabbed animals had at least one zoonotic bacterium in their mouths. The carrier rates for zoonotic bacteria were similar in dogs (92%, 92/100) and cats (93%, 93/100). From 185 carriers of zoonotic bacteria, 374 pathogenic strains were isolated (174 and 200 in dogs and cats, respectively) ( Table-1).
Out of the 200 bacteriologically cultured swabs, 185 showed at least one type of colony in which 374 zoonotic bacterial species were isolated after identification. Eighty-three (44.86%) animals, including 44 dogs and 39 cats carried only one bacterial species, while the remaining 102 animals (48 dogs and 54 cats) carried at least two different bacterial species within a maximum of eight bacterial species for dogs and six bacterial species for cats (Table-3  Available at www.veterinaryworld.org/Vol.13/December-2020/29.pdf    There were some differences in the distribution of Pasteurella species between dogs and cats: P. multocida was the predominant species in cats (55% of feline isolates), while P. pneumotropica was the predominant species in dogs (55% of canine isolates).
In fact, there was no difference between dogs (10%) and cats (11%) in carrying P. multocida, but there was a difference in carrying P. pneumotropica: Of the 100 dogs sampled, 17 (17%) carried P. pneumotropica, although it was only found in four cats (4%) ( Table-4).

Susceptibility to antibiotics
Antibiotic susceptibility results showed that 18 (69.23%) isolates were susceptible to all antimicrobial drugs tested, while 30.67% of the strains were resistant to one or two antibiotics. Of the eight antibiotic-resistant isolates, five distinct antibiotic-resistant patterns were observed: Bi-drug-resistance was observed in P. pneumotropica carried by a dog, while tri-drug resistance was observed in P. multocida isolated from a cat (Tables-6 and 7).
From the effectiveness therapy point of view, erythromycin, penicillin, and ampicillin were antimicrobials for which isolates showed higher resistance rates with 7.84%, 5.88%, and 3.92% of isolates, respectively (Table-6).

Discussion
Human infected-wounds caused by dog and cat bites are usually polymicrobial, containing a mixture of aerobic and anaerobic bacteria [2,3], mainly through the oral flora of biting animals [5,10,11].
In this study, most of the dogs and cats sampled (92.5% and 185/200) were carriers of zoonotic bacteria in their oral cavities: In 83 animals, a single species was found, 50% of the animals sampled carried two species, and the rest had three or more species. These findings are in accordance with earlier studies [4,11,12,17,18] that reported various opportunistic and potentially pathogenic human bacteria are found in the oral cavity of dog and cat.
Bacteria which are not pathogenic to humans were not included in this work, as well as anaerobic    Several studies [19][20][21][22][23] described Gramnegative bacteria as the most common bacteria involved in animal bite complications, such as Pasteurella spp., Neisseria spp., Moraxella spp., and Enterobacteriaceae.

Antimicrobial agents R (%) S (%) R (%) S (%) R (%) S (%)
It is common knowledge that Gram-negative bacteria are often pathogenic; in addition to, septicemia and bacteremia that contain endotoxins whose release into soft-tissue and blood resulting in acute shock and death [24].
The high proportion of Gram-negative bacteria in dogs' and cats' mouths highlights the potential risk of complicating bite wounds infected with these microorganisms. Among Gram-negative bacteria, the predominant family was Enterobacteriaceae, followed by Pasteurellaceae, while Moraxellaceae, Pseudomonadaceae, Aeromonadaceae, and Neisseriaceae were the less abundant families.
Members of the Enterobacteriaceae family, which includes E. coli and P. mirabilis, are known to be natural inhabitants of the gastrointestinal tract of humans and animals [20][21][22][23][24][25]. These bacteria are also found in the esophagus and/or the mouth of dogs and cats [23,25,26].
In this report, the prevalence of E. coli in dogs and cats was relatively high (6.32% and 8% oral carriers, respectively) and similar to the 6.12% (dog) and 6.52% (cat) rates reported in South Africa and India, respectively [27,28]. However, P. mirabilis isolation rates in our study (16.67% in dogs and 13.5% in cats) appeared to be higher than those reported in South Africa (2.04% in dogs) and India (4.34% in cats) [27,28]. Our abundance of P. mirabilis isolation may be attributed to the contamination of food, water, soil, and the environment by fecal flora and licking between animals [25].
In the current study, Pasteurella was the most popular genus (17.81% of all isolates) in dogs, while in cats; it was the second most prevalent genus (10% of all isolates). Several previous studies have described Pasteurellaceae (mainly Pasteurella) as the most predominant genus in the oral cavities of healthy dogs and cats [4-11-27].
The isolation rate of Pasteurella in dogs (17.81%) appeared to be close to the 18.36% reported by Almansa Ruiz et al. [28] in South Africa. Whyte et al. [12] revealed a lower prevalence of Pasteurella (7.52%) in cats. Interestingly, the prevalence of Pasteurella in dogs (27%) was higher than that of cats (17%), which conflicts with other reports stating that the main carriers of Pasteurella are cats (up to 90%) [4,12,13,17].
Pasteurella organisms grow in culture on a variety of commercial media, including sheep blood and chocolate agar [14], but, they are fastidious and can be difficult to isolate and identify from oral flora containing many and various bacteria [29].
In humans, P. multocida is the most common pathogens isolated from soft-tissue infections resulting from 50% of dog bites, 75% of cat bites or scratches, and less frequently from licks on skin abrasions by pets [9,13,14,31].
Available at www.veterinaryworld.org/Vol. 13/December-2020/29.pdf In this study, the isolation rate of P. pneumotropica was 17% in dogs and 4% in cats, but P. dagmatis, P. stomatis, and P. canis were not isolated P. pneumotropica is known to be more common in laboratory mouse and rodent [32], although, some studies have reported isolation of P. pneumotropica from other species, like dogs [33].
Based on the biochemical characteristics, the strains initially identified as P. pneumotropica were divided into three biotypes named Jawetz, Heyl, and Henriksen. Both the biotypes Jawetz and Heyl are associated with rodents, while, the biotype Henriksen which was reclassified as P. dagmatis was primarily associated with cats and dogs [33].
One of the limitations of this study is that the API20NE commercial kits used to identify Pasteurella strains did not include many species of Pasteurella other than P. multocida and P. pneumotropica in their database. Thus, often, P. dagmatis, P. stomatis, and P. canis may often be misidentified as P. multocida or P. pneumotropica [33].
In addition, Benga et al. [33] reported that P. dagmatis was formerly designed as P. pneumotropica bioptype Henriksen, thus explaining the numerous early reports of clinical diseases caused by P. pneumotropica resulting from bites by dogs or cats.
P. pneumotropica may occasionally cause disease in both immunodeficient and immunocompetent individuals [32]. Other common Gram-negative bacteria implicated in bite complications were isolated at low frequency, including, Moraxella Moraxella and Neisseria were reported as part of the normal oral flora of dogs and cats and as an important pathogens in human wounds secondary to animal bites [4], causing severe disease in people including skin and tendon infections, as well as septicemia .
A. hydrophila has been shown to cause serious fatal infections in humans following bite wound infections [20].
In agreement with other investigations, culture-based study of the oral flora of dogs and cats showed that Staphylococci and Streptococci were among the most prevalent bacteria in pets [35].
In this study, 31% and 17% of sampled animals were carriers of Staphylococcus spp. and Streptococcus spp. in their mouths, respectively. These bacteria were the second most common species isolated from dog and cat bite-wounds, respectively [36].
The prevalence of these two species was more common in non-purulent cellulitis/lymphangitis wounds than in abscesses or purulent wounds [36]. Staphylococci-and Streptococci-related cellulitis is more diffuse and typically less severe than that seen with P. multocida infections [21].
Although, Pasteurella species, in particular P. multocida, is one of the most common opportunistic pathogens found in the oral cavity of dogs and cats leading to human infection after bites or scratches [13], no data are available on the resistance patterns of Pasteurella originating from dogs and cats in Algeria.
The majority of the Pasteurella isolates in this analysis were susceptible to all drugs tested, with only eight (15.68%) isolates demonstrating resistance to at least one antibiotic, of which six belonged to dogs.
These results are consistent with published studies reported high resistance of this bacterium to some antimicrobials [37,38]. From a therapeutic point of view, erythromycin, penicillin, and ampicillin were the less effective antimicrobials in the treatment of bite wounds, while multidrug resistance profiles were reported in two isolates.
Typically, Pasteurella species are susceptible to penicillin, ampicillin, and amoxicillin-clavulanate which are the standard therapy of human Pasteurellaassociated bite wound infection . Tetracycline, in the case of penicillin allergy, is an effective alternative [14].
In a recent study, by Ujvári et al. [31], all Pasteurella strains isolated from cats and humans were susceptible to ampicillin and tetracycline.
It should also be noted that none of the 72 P. multocida isolated from infections of the respiratory tract in dogs and cats investigated in the BfT-Germ Vet program demonstrated resistance to any of the tested antimicrobials (penicillin, ampicillin, erythromycin, and tetracycline) [40].
A study conducted by Ferreira et al. [37] in Brazil investigated P. multocida isolates from cat gingiva and dog oropharynx for their antimicrobial susceptibility. This study showed low levels of resistance to penicillin (7.3%); however, lower susceptibilities to erythromycin in isolates from both dogs and cats were observed (21.27%). In same study, resistance to tetracycline was absent.
Although, resistance of Pasteurella to β-lactam antibiotics has been described, no resistance of P. multocida to tetracycline and doxycycline has been reported in humans or pet animals. Furthermore, resistance to tetracycline has been described in pig and ruminant strains [31]. Interestingly, we found a P. multocida strain resistant to tetracycline in dog.
At least nine tetracycline resistance genes (TET genes) have been detected in Pasteurella species [40]. Many of them are related to plasmids or transposons and can, therefore, be exchanged horizontally, not only within the Pasteurellaceae family, but also with other Gram-negative bacteria [40].
In addition, experiments on the molecular basis of antimicrobial resistance indicate that Pasteurella has obviously acquired a number of resistance genes from other Gram-negative or Gram-positive bacteria [40].
Erythromycin is a member of the antibiotic macrolide class; recently, macrolide resistance has emerged in P. multocida [31]. There are no descriptions of resistance to this antibiotic in companion animals [37], but lower susceptibility to erythromycin in Pasteurella isolates from both dogs and cats was common [38].
The exchange of resistance genes between different organisms in the oral cavity as well as between oral bacteria and bacteria from other environments has already been described [8].

Conclusion
To the best of our knowledge, this study is the first in our country to detect zoonotic bacteria found in dogs and cats' oral cavities. It demonstrates that stray dogs and cats in Algeria carry many opportunistic and/or potentially pathogenic bacteria in their mouths, including Pasteurella spp., leading to infection of bite wounds and scratches.
It was shown that dogs were more carriers of Pasteurella spp. than cats, some of them were multidrug-resistant, a risk to public health.
Amoxicillin-clavulanate and tetracycline were the most sensitive drugs against Pasteurella species. Thus, these antimicrobials may be sufficient in humans to treat infected bites caused by Pasteurella.