Isolation of Oral Bacteria, Measurement of the C-Reactive Protein, and Blood Clinical Parameters in Dogs with Oral Tumor

Canine oral cancers have a poor prognosis and are related to chronic inflammation. This may pose a risk of secondary bacterial infection. This study aimed to compare the bacteria isolated from oral swab samples, values of C-reactive proteins (CRPs), and clinical blood profiles of dogs with and without oral mass. A total of 36 dogs were divided in three groups: no oral mass (n = 21), oral mass (n = 8), and metastasis groups (n = 7). Significantly, both the clinical groups (the oral mass group and metastasis group) showed anemia, a decrease in the albumin-to-globulin ratio (AGR), and an increase in the neutrophil-to-lymphocyte ratio (NLR), globulin-to-albumin ratio (GAR), CRP, and CRP-to-albumin ratio (CAR) compared to the normal group. CAR showed an increasing trend in the oral mass and metastasis groups (10 times and 100 times, respectively) compared to the no oral mass group (P < 0.001). Neisseria spp. (20.78%) was the main isolated bacteria in all groups. The main genera in the no oral mass group were Neisseria spp. (28.26%), Pasteurella spp. (19.57%), and Staphylococcus spp. (19.57%). Neisseria spp., Staphylococcus spp., Klebsiella spp., and Escherichia spp. were found equally (12.5%) in the oral mass group. Escherichia spp. (26.67%), Pseudomonas spp. (13.33%), and Staphylococcus spp. (13.33%) were the main genera in the metastasis group. Interestingly, Neisseria spp. decreased in the clinical groups (Fisher's exact = 6.39, P=0.048), and Escherichia spp. increased in the metastasis group (Fisher's exact = 14.00, P=0.002). The difference of oral bacteria in clinical dogs compared to healthy dogs may be related to microbiome alterations, and both the clinical groups showed the increment of inflammatory biomarkers. This suggested that further studies should be conducted on the correlation between the specific bacteria, CRP, blood clinical parameters, and type of canine oral mass.


Introduction
Nowadays, cancer is the most common diseases in both dogs and human. Te dogs and their owners share the same environment [1]. Many studies revealed that they response to the carcinogens from the environment in the same manner. For example, tobacco smoke was related to the increasing incidence of the upper respiratory tract cancer and oral cancer [2]. Head and neck cancers (HNC), especially oral cancers, are often classifed as serious diseases in dogs. Te most common type of oral cancer in dogs is oral melanoma which is followed by oral squamous cell carcinoma (SCC). Oral melanoma in human presents the high aggressive behavior. Canine oral melanoma shares the similar features and characteristics with human oral melanoma. It has been argued that the spontaneous oral cancer in dogs may be a good model and the comparative study for human cancer [1]. Benign tumors, such as fbromatous epulis, ossifying epulis, and acanthomatous ameloblastoma, show slow progress and are curable by excision surgery [3,4]. On the other hand, oral melanoma and SCC show aggressive behavior, a high metastasis rate, poor prognoses, and short survival times [5,6].
In human research, chronic infammation of oral cancer has been observed. Most of them trigger infammatory responses and induce chronic infammation status. Tumorpromoting infammation may play an important role in steps of cancer progression as mentioned by Hanahan and Weinberg [7]. An increase in clinical blood parameters related to infammation may indicate the existence of cancer or tumors. Te concentration of the C-reactive protein (CRP), an acute phase protein, is rapidly rising in response to trauma [8], infammation [9][10][11], infection [12][13][14][15][16], and several malignancies [8,[17][18][19]. Te elevated CRP levels are the essential information for diagnosis and prognosis not only in human medicine [20][21][22] but also in veterinary clinical medicine [16,17,[23][24][25][26][27]. A previous report identifed CRP and serum amyloid A (SAA) as diagnostic markers and prognostic indications after treatment of infammation for the bacterial pneumonia [15]. CRP, SAA, and haptoglobin were detected and a signifcant increase in the canine mammary cancer such as anaplastic carcinoma, complex adenocarcinoma, simple adenocarcinoma, and SCC with metastasis, which characterized in the clinical stage IV-V or by a mass diameter greater than 5 centimeters with ulceration and secondary infammation, was seen. Conversely, these dogs had the decreasing albumin concentration [17]. Moreover, canine mammary carcinoma showed a high concentration of CRP. Tis suggested that an infammatory response is associated with this type of cancer [27]. In addition to mammary cancer, elevated concentration of CRP was detected in hemangiosarcoma, nasal adenosarcoma, cholangiocellular carcinoma, acute lymphoblastic leukemia, malignant histiocytosis, lymphoma, malignant mesothelioma [26], and neuroendocrine carcinoma [28]. Te α 1 -acid glycoprotein (AAG) and CRP were increased after operations to remove malignant cancer from the dogs and these acute-phase proteins again increased further after recurrences and metastasis [8].
Canine oral microfora comprised numerous resident bacteria [29][30][31][32][33]. Tese bacteria communities can play an essential role in interactions with host immunity [34]. Te population and ecological system of the oral microbial agents can be changed as the host's health changes [34][35][36][37][38]. Many studies reported that the microbiome could either directly or indirectly increase the chances of developing cancers in human and animal specifc pathogen models. For example, the formation of mice colon cancer and hepatocellular carcinoma is promoted by Helicobacter hepaticus infection [39,40]. In addition, chronic infection of enteropathogenic Escherichia coli can result in colon cancer [41]. Te specifc bacterial pathogen promotes tumor formation and progression that is the principal mechanism of microbiota related to human SCC, gastric cancer, colorectal cancer, and melanoma [34,38,[41][42][43][44]. Recently, the impact of microbiome related to the process of cancer development was summarized as a new hallmark of cancer [45]. Canine oral cancers with progression and chronic infammation can break or decrease the host's oral mucosal defenses, and the failure of the host barrier may increase the risk of dysbiosis or the development of persistent or recurrent bacterial infection. Tis might be the result from persistent mucosal and/or epithelial cell colonization by microorganisms [46]. Te advancement in bacterial identifcation with polymerase chain reaction (PCR) amplifed 16S sequences. Dewhirst et al. found that the bacterial content in the dog mouth difers from that in the human mouth, and there were only 16.4% shared bacterial types in the dogs and human [47]. Ruparell et al. compared the microbiota from diferent areas within the canine oral cavity and found three niches (soft tissue surface, hard tissue surface, or dental plaque and saliva) in which the bacterial community profles difered [33]. deCarvalho et al. compared the oral swabs from the normal dogs and the canine oral melanoma dogs. Tey found that Tannerella forsythia and Porphyromonas gingivalis from the subgingival plaque samples were signifcantly increased in canine oral melanoma dogs compared to the control dogs. Tese bacterial species are related to the periodontal diseases and human esophageal cancer [48]. So far, there are few studies that compare the oral bacteria of healthy dogs and dogs with oral cancer. Te information of oral bacteria in canine oral tumors is limited and unclear. More information and knowledge about the bacterial population and the microbial alteration in the dog's mouth and the clinical blood profles might help us to increase the understanding about the role of oral bacteria and the blood profle in canine oral tumor and cancer.
For providing more information of canine oral cancer, this study investigated the oral swab bacteria using the culture dependent method with 16s rRNA gene sequencing identifcation and the clinical blood profles of the dogs with and without oral mass. We focused on identifying culturable bacteria living in the oral mucosal surface of dogs. Tus, the purposes of our study were to compare the bacteria isolated from canine oral swab samples, values of C-reactive proteins (CRPs), and clinical blood profles of dogs with and without oral mass.

Animals and Sample Collections.
Te thirty-six dogs in this observational prospective study were invited from the private animal hospitals in Bangkok and vicinity between June 2019 and March 2021. All the dog owners gave informed consent for sample collections from their dogs to be used for research purposes only (ACKU62-VTN-010). Te signalment information included sex, age, and breed that were recorded. Te dogs that had visited the small animal hospitals for the routine annual vaccination program and whose physical examination revealed no evidence of oral mass or serious clinical illness were included in this study as the "no oral mass" group or normal group (n � 21). In this group, there were 9 male and 12 female dogs. Teir ages were between 6 and 13 years. Te clinical group or the "oral tumor-bearing" group included dogs that presented with mass in their mouths during physical examination (n � 15) and/or had a later appointment for the surgical removal of an oral mass. Based on the defnitive diagnosis from histopathological examination and clinical staging of ffteen dogs in the clinical group, eight dogs were classifed in the oral mass group and seven dogs were classifed in the metastasis group. In both clinical groups, there were 8 male and 7 female dogs. Teir age was between 4 and 16 years (Table 1) (Figure 1).
All owners were asked to feed their dogs with commercial diets free of raw meat or milk products. Te dog's diets were managed to prevent oral contamination from raw food and fecal matter. An oral swab sample was collected from the tongue's dorsum mucosa and the mucosa of the hard palate of each dog. Te oral swab sampling was performed after 12 hours of fasting. Te swabs were then transported in the Stuart transport medium (Yangzhou Chuangxin medical device factory, Yangzhou city, Jiangsu, China) from the animal hospital to the faculty of Veterinary Technology, Kasetsart University under cold transport, 4°C. Ten, the bacterial culture was performed. Blood samples were collected and kept in the EDTA tube and Heparin tube for hematology and blood chemistry profle analysis, respectively.
In the oral mass group and the metastasis group, the oral swab, blood collection, and excision or incision biopsy were performed under anesthesia. Te tissue from the oral mass was sent to the veterinary pathologists for diagnosis, and fnally, a histopathological defnitive diagnosis was received.
Te CBC results were analyzed using an automatic analyzer (ProCyte Dx Hematology Analyzer, IDEXX Laboratories, Inc. USA). Te BUN, CRE, ALT, ALKP, TP, ALB, and GLOB levels were measured using automatic blood chemistry measurement equipment (Catalyst One Chemistry Analyzer, IDEXX Laboratories, Inc., USA).
Te plasma CRP concentration was measured using a commercial fuorescent immunoassay (Vcheck Canine CRP 2.0 Test Kit, Bionote, Gyeonggi-do, South Korea). In brief, fve microliters of each plasma sample from all dogs were diluted with the diluent bufer (to 100 μl), mixed thoroughly, and aliquoted to the test device (V200 Analyzer, Bionote, South Korea). Te plasma CRP concentrations were recorded for further statistical analysis. As part of routine diagnosis, CRP concentration results below the detection limit for the assay were set as less than 10 mg/L.

Bacterial Culture, Isolation, and
Identifcation. An oral swab was cultured on blood agar (BA) and MacConkey agar at 37°C for 24-48 hours. All diferent predominant colonies that grew in the last plane of agar were selected. Te predominant selected morphology colony obtained from each sample was further cultured on BA to obtain a pure culture. Genomic DNA from each pure isolate was extracted using E.Z.N.A. ® Bacterial DNA kit (Omega Bio-Tek, Doraville, GA, USA) following the manufacturer's instructions. Te extracted bacterial DNA of isolates was identifed using 16S rRNA gene sequencing by U2Bio Tailand (Bangkok, Tailand) and Bionics Co. Ltd. (Seoul, Korea). Te extracted bacterial DNA was prepared for the identifcation step using 16S rRNA gene sequencing. Te 16S rRNA gene was amplifed with the pair of primers 518F (5′-CCAGCAGCCGCGGTAATACG-3′) and 800R (5′-TACCAG GGTATCTAATCC-3′). Te thermocycler conditions were a predenaturation at 94°C for 4 min; 25 cycles of 94°C for 20 sec, 50°C for 40 sec, 72°C for 90 sec; a fnal extension step at 72°C for 3 min, and hold temperature at 4°C. PCR product was generated. Ten, sequencing was performed after purifcation with ABI3730XL Sequencer (Termo Fisher Scientifc, Massachusetts, USA) by the Sanger sequencing method.
Te nucleotide sequences were processed and assembled using BioEdit and the contig assembly program. Te percent identity of bacterial isolates was determined using the BLAST server (https://blast.ncbi.nlm.nih.gov/Blast.cgi) of the National Center for Biotechnology Information (NCBI). Te highest score and the closest relatives of the 16S rRNA gene were evaluated. Similarities to 16S rRNA gene sequences of the isolates that were ≥99% were used as criteria for identifcation.

Statistical Analysis.
In this study, a variable was considered as a normal distribution when the P value of the Shapiro-Wilk test was larger than 0.05. Te clinical blood profle data were showed as mean ± standard deviation (SD) when they were the normal distribution or showed as median with the interquartile range (IQR) when they were the non-normal distribution. Continuous variables in the normal distribution (ALB, AGR, and CRE) among three groups were compared using a one-way ANOVA test. Post hoc multiple comparisons were performed using the Bonferroni test. Te non-normal distribution of continuous variables (RBC, WBC, PLT, NLR, TP, GLOB, GAR, CRP, CAR, BUN, ALT, and ALKP) among three groups was analyzed using the Kruskal-Wallis test. Diferences in the two groups were analyzed by the Mann-Whitney U test. Te oral bacterial isolates were presented on percentage. Te Fisher's exact was used to assess the relationships of variable-bacterial profles to the groups of dogs. Multinomial logistic regression analysis was used to estimate the relative risk ratio (RRR) among the groups of variables. All data were facilitated by the STATA statistics analysis (StataCorp LLC, Texas, USA; serial number 401506228202), and P < 0.05 was considered signifcant. Te graphs of variables were created using the GraphPad prism program (GraphPad Software, San Diego, USA) and Microsoft Excel program (Microsoft 365, Washington, USA).

Result
3.1. Animals. Tirty-six dogs were included in this study. Twenty-one dogs (9 males and 12 females) were classifed into the group of no oral mass dogs. Tey were 6-13 years Veterinary Medicine International old. In the two clinical case groups, there were 15 oral tumor-bearing dogs (8 male and 7 female). Te age of dogs in this study was between 4 and 16 years old (Table 1). Te increase of age did not increase the risk ratio in the oral mass group (RRR � 1.11, 95% CI: 0.84-1.48) and the metastasis group (RRR � 1.38, 95% CI: 0.98-1.95) when evaluated based on the no oral mass group (P > 0.05) ( Table 2). Te proportion of female and male dogs that had oral masses or metastasis did not difer when evaluated based on the no oral mass group (P > 0.05) ( Table 2).
In the oral mass group, there were two cases of benign oral tumors in clinical stage I, which were acanthomatous ameloblastoma (n � 2), and six cases of malignant oral tumor with clinical stage II and III which were oral malignant melanotic melanoma (n � 1), oral SCC (n � 2), oral fbrosarcoma (n � 1), and oral malignant amelanotic melanoma (n � 2) ( Table 3) (Figures 2(a) and 2(b)). In the metastasis group with clinical stage IV (regional lymph node metastasis and/or pulmonary metastasis), there were oral malignant melanotic melanoma (n � 3), oral chondrosarcoma (n � 1), oral squamous cell carcinoma (n � 1), and oral fbrosarcoma (n � 2) ( Table 4) (Figures 2(c), 2(d), 3(a), and 3(b)). Te dogs in the clinical case groups showed the clinical signs related to the oral mass such as hypersalivation, halitosis, dysphagia, and bleeding from mass. Tere was a dog (Case 1) in the metastasis group that presented lethargy and panting sign.

Clinical Laboratory Results.
According to the CBC results, the median RBC in the oral mass group and metastasis group decreased when compared with the no oral mass group (P < 0.001) (Figure 4(a) and Table 5). Te dogs in both clinical groups exhibited anemia. Te proportion of dogs that had low RBC levels in the oral mass group (RRR � 12.74, 95% CI: 1.03-157.02) and the oral mass with the metastasis group (RRR � 12.74, 95% CI: 1.03-156.98) were more than the no oral mass group (P � 0.047, borderline signifcance) ( Table 6). Tere was a trend toward an increase in WBC in the clinical groups (P � 0.088). Moreover, the NLR in both the clinical groups showed a trend to increase when compared to the no oral mass group, and it was signifcantly increased in the metastasis group when compared to the no oral mass group (P � 0.001) (Figure 4(b) and Table 5). Te median of platelets in the metastasis group was signifcantly higher than in the no oral mass group (P � 0.012) (Figure 4(c) and Table 5). Only the metastasis group showed an increasing level of platelets compared to the normal reference range. Te clinical blood chemistry profles of the three groups were in the normal reference range. However, the median plasma ALKP in the metastasis group was signifcantly higher than those in the no oral mass group (P � 0.001) and the oral mass group (P � 0.037) ( Figure 4(d), Table 5).
Plasma TP, ALB, GLOB, CRP, AGR, GAR, and CAR are presented in Table 5 and Figure 5. Te protein level in all groups was in the normal reference range. Te TP was not diferent among the three groups of dogs (Table 5). ALB and GLOB showed the opposite trend. Te ALB showed decreased levels in both the clinical groups; conversely, the GLOB showed increased levels in both the clinical groups (Figures 5(a) and 5(b)). Te mean ALB and the median of the AGR in the two clinical groups were signifcantly lower than in the no oral mass group (P < 0.001) (Figures 5(a) and 5(c)). Moreover, AGR in the metastasis group was signifcantly lower than the oral mass group (P � 0.008) ( Figure 5(c)). Conversely, the median of plasma GLOB levels in the oral mass group (P � 0.003) and the metastasis group (P � 0.01) were signifcantly higher than the no oral mass  Figure 1: Te groups of animals in this study and techniques of sample collections. Tere was one normal group,the no oral mass group, and the two clinical groups, the oral mass group and the metastasis group.
group ( Figure 5(b)). GAR in the oral mass group (P < 0.001) and the metastasis group (P < 0.001) was signifcantly higher than the no oral mass group. In addition, GAR in the metastasis group was signifcantly higher than the oral mass group (P � 0.017) ( Figure 5(d)). All dogs in the no oral mass group had a CRP level of less than 10 mg/L, which is under the normal range or less than 20 mg/L. In the oral mass group, the median CRP level was 12.45 mg/L and the range was between less than 10 mg/L and28.2 mg/L. In clinical stage I and II, most dogs had CRP levels of less than 10 mg/L. Tey had the CRP level higher than 20 mg/L in clinical stage II. Most of the dogs in the metastasis group (clinical stage   Tis fgure presents the histopathology of acanthomatous ameloblastoma (20x,(a)), squamous cell carcinoma (SCC, 20x,(b)), malignant melanoma cells metastasis to the regional lymph node (40x,(c)), and the gross lesion of oral mass at the maxilla area (d).   Figure 4: Comparison of red blood cell (RBC, median (IQR))(a), the neutrophil/lymphocyte ratio (n/l ratio or NLR, median (IQR))(b), platelet counts (PLT, median (IQR))(c), and alkaline phosphatase (ALKP, median (IQR))(d) of dogs among the no oral mass group, the oral mass group, and the metastasis group. 6 Veterinary Medicine International  Veterinary Medicine International IV) had a plasma CRP level higher than 50 mg/L which is higher than the normal range. Te median plasma CRP concentrations of the oral mass group and the metastasis group were 12.45 and 112 mg/L, respectively, and were signifcantly higher than those of the no oral mass group (P < 0.001) ( Figure 5(e), Table 5). Also, the CRP levels of the metastasis group were signifcantly higher than those of the oral mass group (P � 0.002) ( Figure 5(e) and Table 5). However, one dog in the metastasis group received a nonsteroidal anti-infammatory drug before the blood collection, and the CRP level of this dog was less than 10 mg/L. So, the CRP level and CAR data of this dog were not included for the statistical analysis. Te median CRP and CAR in the oral mass group and the metastasis group were signifcantly higher than in the no oral mass group (P < 0.001) (Figures 5(e) and 5(f ) and Table 5). Te median CRP of the metastasis group was about 10 times higher than that of the no oral mass group. In addition, the median of the CAR of the oral mass and metastasis groups was higher than that of the no oral mass group about 10-15 times and 100 times, respectively.
Escherichia spp. showed signifcantly higher evidence in the metastasis group (Fisher's exact = 14.00, P � 0.002) compared with the no oral mass group. Finally, the summary fgure of the animal groups, the sample collections, and the interesting results of this study were provided in Figure 9.

Discussion
Te present study aimed to compare blood profles and plasma CRP levels among the no oral mass group and the two clinical groups. Both the clinical groups, the oral mass group and the metastasis group, showed a signifcant decrease in total RBC when compared with the no oral mass group. Tis fnding may imply that most of the severe tumor-bearing dogs showed the anemia status which is related to the mass bleeding. Moreover, the growing mass in the oral cavity may have interfered with the dog's nutrition consumption. Tis status relates to the anemia of infammatory disease (AID) which is the mild to moderate severity, nonregenerative anemia in chronic diseases including the neoplasia [49]. Previous reports have claimed that infammation plays a key role in this type of anemia. Shortening RBC life span, impaired erythropoietin-mediated erythropoiesis, and inhibition of iron metabolism are the clinicopathological features [49]. Te point is that if the oral mass cannot be treated, alternative treatments should be performed for these dogs such as blood transfusion in severe anemia cases, parenteral iron therapy, and the administration of recombinant human erythropoietin in mild to moderate anemic dogs. According to the CBC results, there was a trend toward increasing WBC and the NLR in the oral mass group and the metastasis group compared to the no oral mass group. Tis implies that the tumor-bearing group displayed the infammation feature. Many bacterial isolates in our study were the gram-negative bacteria. Gram-negative bacteria and their endotoxins, lipopolysaccharide (LPS), play the important role of infammatory stimuli that are activated through the infammatory cytokines [46,50]. In addition, lipoteichoic acid (LTA) of Gram-positive bacteria is also one of the antigens that activated the immune cells [34]. Te chronic infammation, loss of surface integrity in the oral mucosal epithelium, and structure of the oral cancer in patients may result from the persistent bacterial colonization of mucosa or cancer epithelial cells, which has been observed at various stages of oral cancer. Tus, chronic infammation could result from persistent mucosal or epithelial cell colonization by microorganisms. Tere is a lot of evidence of oral bacteria that are involved in the infammation [2,46,51].
In this study, the CRP level of both the clinical groups was signifcantly higher than those of the no oral mass group and the increasing of the CRP level was related to the severity clinical stage. Cancer and infammation are associated in   a way that some cancers arise at the sites whereas the cancer induces an infammatory microenvironment. We could support that CRP is a sensitive marker that has been shown to increase in response to malignancies [19,26] such as lymphoma [18,52], multiple myeloma [53], pancreatic cancer [54], and SCC [55]both in humans and dogs. In most studies, CRP levels were found to be highly elevated in patients with cancer and metastasis compared with the healthy control group or benign conditions [26,27,56]. Tese CRP-related infammatory evidence may be the potential role of the tissue injuries which related to the tumor consequence of infammatory response. Te role of infammation in the tumor development and progression was mentioned and added to the new version of hallmarks of cancer in human by Hanahan [45]. One dog with oral malignant melanoma that had regional lymph node and lung metastasis (clinical stage IV) and a plasma CRP level of 115 mg/L, died within seven days of the frst visit. Increasing CRP levels and blood profles showed a severe leukocytosis infammatory pattern. Te critical CRP value was mentioned in the systemic infammatory response or severe emergency cases [24,27]. According to our results, the CRP level in a high clinical stage presented as greater than 100 mg/L, as in previous studies. Elevated CRP is likely a response secondary to tumor necrosis and local tissue damage; this response may be attributable to bacterial translocation from the damaged oral mucosa, endothelium, and resulting septicemia that is associated with infammation in patients with malignancies. NSAIDs inhibit the cyclooxygenase enzymes and antiinfammation [57]. NSAIDs were the clinical factor for the dog that had a CRP level in the normal range.
In the oral mass with the metastasis group, the CRP and ALKP enzyme levels were signifcantly high. It is a fact that hepatocytes are the main cells that produce and release CRP to the systemic circulation. Tis process occurs in response to increased levels of circulating proinfammatory cytokines [58,59]. Terefore, the plasma CRP and ALKP levels may be correlated with proinfammatory mediators, especially interleukin-1 (IL-1) and interleukin-6 (IL-6) in liver cells [58][59][60]. Tere are three isoforms of ALKP: liver ALKP (LALKP), bone ALKP (BALKP), and corticosteroid-induced ALKP (CALKP). ALKP is the primary indicator of cholestatic liver disease. Tis enzyme also increases with severe bone destruction and steroid induction. Te increase in BALKP, which is shown in the total ALKP, is usually found in young puppies or during times of active osteoblast and bone development [61]. In our study, the metastasis dogs had increased ALKP levels. Teir abdominal ultrasound examination results did not show any alteration of the liver and/or bile duct structure. Te alteration of bone or bone lytic condition due to the cancer metastasis may be the cause of the increasing ALKP level in the malignant part with metastasis group compared to the no oral mass group and the oral mass without metastasis group. However, we cannot conclude that the total ALKP increase is due to bone destruction. In the future, a specifc method for BALKP detection should be performed for the exact BALKP level, and then we can know the source of ALKP that is increasing in metastasis cases [62].
We compared the clinical blood results between the no oral mass group and the tumor-bearing groups and found that the plasma ALB levels and the AGR in the two clinical groups, the oral mass group and the metastasis group, were signifcantly lower than in the no oral mass group. Conversely, the GLOB and the GAR in the clinical groups were signifcantly higher than in the normal group. In addition, the CAR was signifcantly higher in the clinical groups compared to the normal group. Tese results, like those in previous studies, showed that both hypoalbuminemia and  Veterinary Medicine International the elevated CRP concentration were associated with malignancy and related to survival time [26]. In our study, NLR was also evaluated as biomarker of a systemic infammatory response. Both clinical groups showed an increasing trend. Tis fnding was related with the previous study in dogs with gingivitis and dogs with oropharyngeal tumors [63]. Te predictive ability for patients with advanced cancers might be improved by combining CRP with other parameters, such as plasma albumin and NLR. In human medicine, the oral microbiomes of diseases and healthy people are diferent from those of dogs according to the sequencing method [30]. Terefore, there seem to be considered diferences between the oral microbiomes of the diferent species [31,33,47,64,65]. Tis study aimed to identify oral bacterial samples from dogs with no oral mass and clinical oral tumor-bearing dogs. We focused on the cultivable bacteria living in the oral mucosal surface of dogs. To this point, the assembly of the 16S rRNA gene of bacterial isolates provides a new view of the oral bacterial profle in the canine oral tumor-bearing group compared with the no oral mass group. Traditionally, bacterial identifcation in laboratories was performed using phenotypic tests, including Gram smear and biochemical tests, considering culture requirements and growth characteristics [66,67]. However, these methods of bacterial identifcation have major limitations due to the phenotypic tests and biochemical tests. Nowadays, the modern PCR, automated DNA sequencing, and work on 16S rDNA sequencing bacteria can make an accurate identifcation of bacterial isolates. From this point, the accurate identifcation is one of the most important functions of clinical microbiology laboratories and solves the problem related to the traditional method [66,[68][69][70].
According to our results, the phylum Proteobacteria was the most abundant bacteria, as reported in previous studies [30,33,64]. In our study, Pasteurellaceae and Neisseriaceae were the two main bacterial families in the no oral mass group. Te three majority genera of bacterial isolates were Neisseria spp., Pasteurella spp., and Staphylococcus spp. Tese three bacterial genera can usually be cultured from dog bite wounds in human [71,72]. Tis point is related to the previous studies stating that Neisseria spp., including N. canis, N. animaloris, N. dumasiana, N. zoodegmatis, and N. weaverii, are one of the normal foras of dogs [71]. In addition, Pasteurella spp. and Staphylococcus spp. were also isolated as cultivable oral microbial in domestic dogs, but they were related to dental plaque samples in previous studies [29,32]. In this study, Enterobacteriaceae was the majority bacterial family in both the clinical case groups. Escherichia spp. and Shigella spp. were not isolated from the dogs in the no oral mass group. Moreover, Neisseria spp. and Pasteurella spp. decreased in both oral mass groups. In addition, the bacteria in the genera Aeromonas spp., Acinetobacter spp., Elizabethkingia spp., Klebsiella spp., Enterococcus spp., Corynebacterium spp., and Escherichia spp. were found in the oral mass group. Tis alteration in oral bacterial isolation is may be due to the changes in the oral microenvironment that destroy the normal microbial population and conduct the opportunistic microorganisms to increase the population. Te unique microbial profle in humans with SCC is diferent from the normal humans [36]. Te discovery of bacterial replication in the tumor may be linked to the presence of bacterial nutrients and chemotactic compounds. Moreover, the alteration in the salivary microbiome of pancreatic cancer patients compared to the healthy control group empathized with those hypotheses. Te researchers attempted to select the human oral microbiome as the noninvasive method for the diagnosis of the oral and gastrointestinal cancer. Recently, the researchers discovered the relationship of the dysbiosis of bacteria in gut and oral bacteria in dogs. In that study, the Bacteroides bacteria were shared in the intratumoral, oral, and gut bacterium community of canine mammary tumor cases. Tis result confrmed that these bacteria might travel from the gastrointestinal tract to the tumor sites [73]. Tis was related to previous study in human oral SCC and the bacterial translocation. Tey isolated the enteric bacteria such as Klebsiella, Enterobacter, and Enterococcus from the tissue of oral mass and regional lymph node before surgery and the infection of sites after operation [74]. In addition, the patients with malignant cancers are more susceptible to get infected by the pathogenic bacteria in comparison to benign patients [75]. Tus, based on our results, the alteration of the oral microbial profle and the bacterial translocation in the canine oral tumor-bearing group is an interesting point for further study related to the diagnosis and prediction of canine oral cancer.
Oh et al. reported that Porphyromonas, Fusobacterium, Actinomyces, Neisseria and Pasteurella were the most abundant genera of the bacterial swab from the buccal area and the supragingival plaque [30,33]. According to the several distinct microbial habitats and the oral cavity area, the tongue is the most populated niche, and this area has an impact on other regions in the oral cavity. Te tongue and saliva facilitated the bacteria to travel around the oral cavity [50]. Ruparella et al. and team reported that saliva exhibited the lowest bacterial diversity, the buccal and the tongue dorsum mucosa had the most similar bacteria [33]. Moreover, the teeth area and the dental plaque were the selected area for the periodontal disease [31,32,67,70,[76][77][78]. So, the oral swab sample from the tongue's dorsum mucosa and the mucosa of the hard palate of each dog was taken to compare the microbes among the three groups of dogs in our study. Tis site of the sampling method showed the microbe profles that difer from the previous study that interested in the buccal site and supragingival plaque. According to our study, the bacteria from the supragingival plaque that is related to the periodontitis might not interfere with the results of bacterial isolates in our samples. Many oral bacteria are slow growing, require complex culture media, and specifc atmospheric requirements [69,70,79]. However, many bacterial species could not be cultured in this study because they are difcult to reproduce under laboratory conditions. Oral anaerobic bacteria such as Porphyromonas spp., Fusobacterium spp., Bacteriodes spp., Capnocytophaga spp., Prevotella spp., Tannerella spp., Treponema spp., and 14 Veterinary Medicine International Actinomyces spp. reported in the previous study [29,31,48,70,77,79,80] were not found in our study. Some bacterial isolates have been detected from the environment, as the channel of a dog's mouth is exposed to the outdoor environment. In veterinary medicine, there have been few reports about canine oral microbiomes with malignancies. Te present study investigated oral microbial isolates in canines bearing oral tumor and compared the oral microbial isolates and blood clinical profles among the no oral mass group and the clinical groups. Most frequently, isolates were normal oral fora which were not considered primary pathogens in the no oral mass group. Te alterations in canine oral bacteria with oral mass compared to healthy dogs in this study may be related to microbiome alteration and dysbiosis. Tis suggested that bacteria may play the same potential roles in the pathogenesis and cancer progression of oral cancer as infammatory pathogens in human. Tese were mentioned in the review article by Faden that there are many mechanisms including (a) chronic infection altered cell growth, (b) infection resulting in suppression of apoptosis, (c) chronic infection induced cell proliferation and DNA replication, (d) bacterial products caused the epithelial DNA damage and secondary hyper-proliferative epithelium, and (e) carcinogenic nitrosamine produced by E. coli linked to oral cancer development [51].
We focused on identifying cultivable bacteria living in the oral mucosal surface of dogs with the culture-based method and 16s rRNA gene sequencing. We obtained a lower number of bacterial colonies compared to recent reports using advance sequencing methods [30,33]. However, our results did not interfere by the DNA component of dead bacteria that can be detected by the advance sequencing methods in the previous studies [34]. Te most abundant bacterial phyla from the oral swab in this study were not similar as the study of McDonald et al. [81] and Dewhirst et al. [47]. Our data was similar to the study of Ruparell et al. [33]; this may be the result of the site of sample collection and the sample preparation for bacterial identifcation. Te direct oral swab samples for next generation sequencing (NGS) or the whole genome sequencing (WGS) provided the bacterial taxonomies more than the culture based and 16s RNA gene sequencing method and those methods could detect the uncultured bacteria [62][63][64]. Furthermore, the different settings of the disease groups, sample collection between awakening dogs and those under anesthesia, could be a confounding factor leading to diferent results [82]. Limitations of this study were the small numbers of dogs in each group, the diferent tumor types within the tumor bearing-dog groups and the cultured bacterial isolate-based method. Due to the limitations of this study, we suggest that the number of samples should be increased to provide more interesting and valuable information for the further study.

Conclusion
Tis study provides the knowledge of the oral bacterial population from the culture-based with 16s rRNA gene sequencing method that points toward the role of bacterial alteration and tumor-promoting infammation. Canine oral cancers have a poor prognosis and are related to the chronic status which may decrease the host oral mucosa immunity and increase the risk of secondary bacterial infection. We compared the bacterial isolates and blood profles of dogs in the three groups. Signifcantly, both clinical groups showed anemia, an increase in NLR, GAR, CRP, and CAR, and a decrease in AGR compared to the normal group. Neisseria spp. was the main genus of bacterial isolates in the oral swabs. Interestingly, Neisseria spp. decreased in the clinical groups, and Escherichia spp. increased in the clinical groups. Our results on the diferences of the oral bacteria among the groups of dogs and the increase in the CRP, CAR, and NLR suggest that the bacteria and the oral cancer may be related and play a potential role in cancer progression. Whether or not bacterial play the role in tumorigenesis and progression, it is an interesting point to further explore the efect that the bacteria may have on diferent phenotypes of cancer cells and their interactions with the cancer cells. Tus, further studies may provide the new knowledge in this feld and/or facilitate the new treatment options. Te studies should be conducted on the larger sample size and investigate the correlation between the specifc bacteria, the type of canine oral mass, the clinical blood profles, and the clinical stage of oral cancer.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.

Authors' Contributions
Setthawongsin C., Meekhanon N., and Rungsipipat A. contributed to conceptualization and methodology. Pisamai S., Rungsipipat A., and Setthawongsin C. were involved with sample and data collection. Khunbutsri D., Rungsipipat A., and Setthawongsin C. were involved in investigation and laboratory work. Jarudecha T. and Setthawongsin C. were involved with the data validation and statistical analysis. Setthawongsin C. was responsible for writing original draft. Ngamkala S., Raksajit W., Meekhanon N., Rungsipipat A., and Setthawongsin C., contributed to writing the original draft and graph presentation. Rungsipipat A. was involved in research supervision. All the authors reviewed and provided critical comments on the manuscript and read and approved the fnal version of the manuscript.

Supplementary Materials
Supplementary fle 1: the nucleotide sequences of bacterial isolates from the no oral mass dogs. Supplementary fle 2: the nucleotide sequences of bacterial isolates from the dogs with oral mass. (Supplementary Materials)