Influence of Gallic Acid-Containing Mouth Spray on Dental Health and Oral Microbiota of Healthy Dogs: A Pilot Study

Simple Summary Dogs frequently have oral diseases, especially periodontal diseases. Medications for oral health are being used in dogs to prevent infections. Oral microbiota plays a major role in dental health. The present study displayed the effect of gallic acid-containing mouth spray (GAMS) on healthy dogs’ oral health and microbiota. The mouth spray improved the gingival and calculus indexes in healthy dogs. GAMS altered the dental microbiota. Further studies are obligatory to claim that the GAMS could be used to maintain the oral health of healthy dogs. Abstract The pilot study aimed to investigate the effects of GAMS on oral microbiota in healthy dog subjects. Thirty-eight dogs were recruited and randomly allocated to the placebo (n = 19) and treatment groups (n = 19). The dogs were treated with mouth spray once daily for 42 days. The changes in the gingival index (GI), plaque index (PI), and calculus index (CI) were measured at baseline (day 0) and end of the study (42nd day). The changes in the oral microbial composition of representative dogs (placebo, n = 7; and treatment, n = 7) were also evaluated at baseline and end of the study. Oral microbial composition was assessed by sequencing. The sequences were annotated using the QIIME 2.0TM. The GI, PI, and CI indexes were reduced after the GAMS usage. The abundance of the commensal bacterial phylum Actinobacteria and Chloroflexi, genera Frederiksenia, and Bergeyella was improved after six weeks of GAMS usage. GAMS reduced the pathogenic bacterial species, including Neisseria sp., Desulfobulbus sp., Capnocytophaga canis, and Corynebacterium mustelae. Moreover, some pathogenic bacterial abundances were increased at the end of the study. All the microbial variations were observed within the group. The inter-group analysis revealed that the changes were unrelated to GAMS usage. Further studies need to be carried out using more experimental subjects to confirm the effectiveness of GAMS. More metagenomic data are required to evidence the GMAS impact on the oral microbiome of healthy dogs.


Introduction
The oral cavity is a habitat for several microbial species. Several factors, including food matter, care regimens, and oral care, influence the composition and functional contributions of the microbial population [1]. Extreme microbial dysbiosis promotes the risk of developing periodontal diseases (PD), an inflammatory condition marked by a complicated interplay between the immune function and oral microorganisms. Biofilm forms and multiplies on the tooth surface (supragingival plaque) before spreading under the gingiva

Preparation of GAMS
A detailed composition of the GAMS was reported in our previous study [34]. Briefly, gallic acid, ZnCl 2 , poloxamer188, poloxamer407, carrageenan, hesperidin methyl chalcone, sodium metabisulfite, and water were mixed in the reported ratio [32].

Experimental Animals
Thirty-eight healthy dogs were recruited for the study. The veterinarian confirmed the dogs' health through physical examination, thoracic radiography, complete blood count (CBC), and blood biochemistry (blood urea nitrogen, creatinine, alanine aminotransferase (ALT), and alkaline phosphatase (ALP)). The mesocephalic skull dogs aged over 2 years with complete investigating teeth (104, 108, 204, 208, 304, 309, 404, and 409) were included in the study. Dogs on medications (steroids, non-steroidal anti-inflammatory drugs, or antibiotics) or allergic to oral spray were excluded from the study. The animal care and use committee, Faculty of Veterinary Medicine, Chiang Mai University, Chiang Mai, Thailand, approved the experiment (Ref. No. R2/2564). Dogs were randomly divided into experimental groups (G1; n = 19) and a control group (G2; n = 19).

Dental Procedure and Mouth Spray Administration
Before experimentation, a general physical examination was performed for dogs. Amounts of 0.2 mg/kg diazepam (Diazepam ® ; Government Pharmaceutical Organization (GPO), Bangkok, Thailand) (intravenous administration) and 4 mg/kg tramadol (Tramadol ® ; T.P. Drug Laboratories (1969) Co., Ltd., Bangkok, Thailand) (subcutaneous administration) were used for sedation. After sedation, dogs received 4-6 mg/kg propofol (Pofol ® ; Dongkook Pharmaceutical Co., Ltd., Chungcheongbuk-do, Republic of Korea) (intravenous administration) to stimulate anesthesia. An endotracheal tube was then intubated, and the cuff was filled. The anesthesia stage was maintained using isoflurane (Attane ® ; Piramal Critical Care, Inc., Pennsylvania, United States of America) and oxygen during the dental procedure. Periodontal probing, oral evaluation, and dental charting were performed at baseline (week 0) and after 6 weeks of study, and the changes were recorded per the modified Triadan system. The dental indices were measured on (bilaterally) maxilla: canine and fourth premolar (104, 204, 108, 208), and mandible: canine and first molar (304,404,309,409). G1 was treated with mouth spray (1 mL of spray = 10 puffs; Dose: 4 puffs per day) before bed and withheld food and water for half an hour after application for 6 weeks. The control group (G2) was not subjected to any treatment.

Next-Generation Sequencing (NGS)
NGS evaluated the oral microbiome of the experimental dogs. The saliva samples were collected at baseline and after six weeks of study. QIAamp UCP DNA Micro Kit (QIAGEN, Hilden, Germany) was used to isolate the total genomic DNA from the saliva samples. The sequencing was performed at the Omics Sciences and Bioinformatics Center, Faculty of Science, Chulalongkorn University, Bangkok, Thailand, as described in our previous study [34].

Statistical Analysis
Fisher's exact, Mann-Whitney U, and Wilcoxon signed-rank tests were performed to determine the experimental subjects' significant changes in the studied parameters and characteristics. The changes were considered significant if the p < 0.05. The values were presented as median ± interquartile ranges. The oral microbial sequences were analyzed Vet. Sci. 2023, 10, 424 4 of 20 using the QIIME2.0 TM . Alpha diversity was analyzed using the metric pielou_e, Simpson, and Shannon entropy to measure the relative evenness of species richness, abundance, and taxonomical diversity with statistical significance (p ≤ 0.05). The beta diversity was measured for the control (Cpre vs. Cpost) and treatment (Tpre vs. Tpost) groups using the unweighted UniFrac distances matrix with a 3D PCoA plot. The raw operational taxonomic units (OTUs) were used to determine phylum to species and compared the phylum in the control and treatment group as a heat map. The relative frequency was calculated using the observed OTUs for the phylum, genera, and species. The taxonomical differences between the Cpre, Cpost, Tpre, and Tpost samples were estimated using the Wilcoxon signed-rank test with a statistical significance of p ≤ 0.05.

Demographic Data and Changes in GI, PI, and CI
The gender, weight, and age of the dogs were not significantly different during the study (Table 1). GI, PI, and CI were measured at baseline and after six weeks of the study. Eight teeth and their three (GI and CI) and two (PI) positions were utilized for the measurements. The results were represented as median and interquartile (IQ) values. The  , whereas CI values of the tooth codes 108, 204, 208, 404, and 409 were also reduced after 6 weeks in the control group (Table 2).

Oral Microbiome Analysis
The baseline (Cpre) and sixth-week samples (Cpost) had 455,410 and 442,275 microbial sequence reads in the control group, respectively. The non-chimeric sequences of 350,430 and 321,204 were retrieved in Cpre and Cpost samples. Moreover, 343,924 and 351,717 microbial sequence reads were obtained from the treatment group's baseline (Tpre) and sixth-week samples (Tpost), respectively. The non-chimeric sequences 264,859 and 261,791 were retrieved in the Tpre and Tpost samples for further analysis (Table S4).

Alpha Diversity
The alpha diversity was estimated by Pielou's evenness, Simpson index, and Shannon index with the Kruskal-Wallis (pairwise) test. Pielou's evenness is an index that measures diversity along with species richness. The estimated Pielou's evenness for the control (p = 0.6547) and treatment (p = 0.7494) samples indicated that there are no significant differences in the species evenness, which meant the species are distributed evenly within the control (Cpre vs. Cpost) and treatment (Tpre vs. Tpost) groups ( Figure 1A,B). The Simpson index was used to calculate the relative abundance of the different species. The estimated Simpson index for the control (p = 0.8480) and treatment (p = 0.6547) groups were not significantly different, indicating that species richness was not significantly different among the groups ( Figure 1C,D). However, the evenness and species richness were higher in the treatment than in the control group. The diversity richness of the oral microbiome was estimated using the Shannon index. The Kruskal-Wallis (pairwise) test identified the differences in the control (p = 0.5653) and treatment (p = 0.1797) groups. The results showed that even though the statistical differences were not observed in the control and treatment groups, there could be a difference in diversity for the treatment group compared to the control group ( Figure 1E,F).

Beta-Diversity
The similarity and variations in the microbiome were performed using PCoA, and unweighted UniFrac distances between the groups were represented. Cpre and Cpost samples were not completely separated from each other in 3D space. PCoA plot axis 1, 2, and 3 explained the variances in the microbial abundances in the samples as 26.96%, 10.83%, and 9.074%. However, the microbial load of the Cpost samples Cpost-DG3V2-1, Cpost-DG3V2-4, and Cpost-DG3V2-8 were located near the Cpre samples. They formed a cluster, which indicated that there could be a microbial similarity between the Cpre and Cpost samples, but the samples Cpost-DG1V2-1, Cpost-DG1V2-2, Cpost-DG1V2-4, and Cpost-DG1V2-8 were completely separated in the 3D space, which described that there could be a possible microbial difference in the Cpre and Cpost samples ( Figure 2A).

Beta-Diversity
The similarity and variations in the microbiome were performed using PCoA, and unweighted UniFrac distances between the groups were represented. Cpre and Cpost samples were not completely separated from each other in 3D space. PCoA plot axis 1, 2, and 3 explained the variances in the microbial abundances in the samples as 26.96%, 10.83%, and 9.074%. However, the microbial load of the Cpost samples Cpost-DG3V2-1, Cpost-DG3V2-4, and Cpost-DG3V2-8 were located near the Cpre samples. They formed a cluster, which indicated that there could be a microbial similarity between the Cpre and Cpost samples, but the samples Cpost-DG1V2-1, Cpost-DG1V2-2, Cpost-DG1V2-4, and Cpost-DG1V2-8 were completely separated in the 3D space, which described that there could be a possible microbial difference in the Cpre and Cpost samples ( Figure 2A).

Taxonomical Assignment
The taxonomy (from phylum to species) was estimated and assigned for the control and treatment samples and visualized as the heatmap (Figure 3). separated from the Tpre samples described that the increased microbial dissimilarity may exist between the Tpre and Tpost samples in the treatment group. PCoA plot axis 1, axis 2, and axis 3 explained the variances in the microbial abundances in the samples as 25.18%, 14.46%, and 10.75% ( Figure 2B).

Taxonomical Assignment
The taxonomy (from phylum to species) was estimated and assigned for the control and treatment samples and visualized as the heatmap (Figure 3).

Taxonomical Assignment
The taxonomy (from phylum to species) was estimated and assigned for the control and treatment samples and visualized as the heatmap (Figure 3).

Clinical Parameters
At the baseline, the sex and weight of the dogs were not significantly different in both the control and treatment groups ( Table 1). The effect of GAMS on GI, PI, and CI values was presented (Table 2). In the control group, no significant changes were observed in GI, and PI values, after six weeks of study. CI values were improved at seven positions compared to the control baseline. Whereas GI, PI, and CI values were improved at seven, one,

Clinical Parameters
At the baseline, the sex and weight of the dogs were not significantly different in both the control and treatment groups ( Table 1). The effect of GAMS on GI, PI, and CI values was presented ( Table 2). In the control group, no significant changes were observed in GI, and PI values, after six weeks of study. CI values were improved at seven positions compared to the control baseline. Whereas GI, PI, and CI values were improved at seven, one, and sixteen positions, compared to the baseline (pre) values in the treatment group ( Table 2).
The effect of mouth spray on PI value was not enhanced significantly at several positions. The results indicated that the GAMS usage improved the GI and CI scores effectively.
It is necessary to develop and improve the therapeutics to control and manage oral diseases, especially PD, in dogs [38]. Our GAMS showed better results in reducing the GI and CI, but a more prevalent impact of the PI score was not noted. Therefore, further tuning is required in the formulation to control the plaques.

The Changes in the Oral Microbiome
The sequencing study showed that the sequence's reads were increased in the postsamples of control (Cpost) and treatment (Tpost) groups. Comparatively, the number of reads was less in Tpost samples than in Cpost samples, indicating that the oral bacterial abundances were decreased after GAMS treatment (Table S4).
The alpha diversity, such as Pielou's evenness, Simpson, and Shannon index, stated that species richness, evenness, and diversity did not significantly differ (p ≤ 0.05) within the control ( Figure 1A,C,E) and treatment ( Figure 1B,D,F) groups. The beta diversity of the microbial community was represented as the PCoA plot, which stated that the microbial load of the treatment group's samples was notably more varied than the control samples. The data indicated that the GAMS influences the oral microbial load in the experimental dogs (Figure 2A,B).
The changes in the oral microbiota of pre-and post-samples of the control and treatment groups were compared (Figures 4-9). The results indicated significant changes within the group comparisons (Cpre vs. Cpost; Tpre vs. Tpost) ( Table 3). No significant changes were observed between the control and treatment groups (Table S5).

Changes in the Bacterial Phylum
Synergistota members were not majorly detected in the healthy oral samples, but the members of phylum Synergistota have a role in periodontitis [39]. The changes in the abundances of Patescibacteria and Desulfobacterota were observed in association with inflammatory bowel diseases and Crohn's disease [40,41]. The 16S rDNA profiling stated that Actinobacteria was the oral inhabitant in healthy human and canine oral samples [42][43][44]. Dewhirst et al. stated that Chloroflexi was the dominant phylum in the canine subgingival plaques [45]. Spirochetes are related to periodontitis and could destroy gingival tissue [46]. Our study showed that the abundance of the above-mentioned phylum was altered in the control and treatment groups after the study period. The changes were significant within the group but insignificant between the groups (Figures 4 and 5; Table 3; Table S5), indicating that GAMS was not shown any significant impact on the oral microbiota of dogs.
Bergeyella is abundantly present in healthy felines [47]. The members of the genus Staphylococcus may transfer from animals to humans, and they are associated with several diseases and are found in the oral cavities of small animals [48,49]. The relative abundance of Desulfomicrobium was higher in the subgingival plaque of the healthy adult Beagle dogs [50]. Frederiksenia was commonly found in healthy dogs, and its relative abundance was increased after probiotic treatment [51]. Nises et al. reported that several phylotypes of oral Treponema species are common in dogs despite periodontal disease status [52]. In the present study, the relative frequency of the Bergeyella (p = 0.0425) and Staphylococcus (p = 0.018) was increased and decreased significantly in the Tpost samples, respectively.
The relative abundance of Frederiksenia (p = 0.028) was increased significantly in the Tpost samples. Treponema was detected only in the treatment group (Figures 6 and 7; Table 3 and Table S5).
Desulfovibrio genus was more abundantly present in the PD-affected dogs than in the healthy subjects [7]. Desulfobulbus genus is associated with a subgingival plaque of periodontitis [53]. The relative frequency of Desulfovibrio (p = 0.003) and Desulfobulbus (p = 0.01) was reduced significantly after GAMS usage (Figure 7).
Fretibacterium may involve periodontal pathogenesis in humans. The abundance of Fretibacterium was significantly higher in the periodontitis subjects than in the healthy humans [54]. In our study, the relative frequency of Fretibacterium was reduced in the Cpost (p = 0.018) and Tpost (p = 0.091) samples (Figures 6 and 7), indicating that the reduction in Fretibacterium was not associated with the GAMS usage.
Flexilinea was abundantly present in the periodontitis sites of the periodontal patients [55]. The relative frequency of the Flexilinea was greatly reduced in the Tpost (p = 0.0425) samples (Figure 7), suggesting that GAMS usage may act against Flexilinea.
The results showed that GAMS could improve the oral microbiome by enhancing the normal oral flora. However, positive changes in some of the genera were observed in both control and treatment groups, so additional studies are required to demonstrate the protective effect of GAMS against PD.
C. canis is more prevalent in dogs with periodontitis than in healthy dogs [56,57]. C. canis was also isolated from a patient's wound caused by a dog bite [58]. Desulfobulbus sp. also causes PD [59]. Neisseria sp. is a commensal of the canine and feline oral cavities. However, Neisseria sp. is a potential cause of cutaneous infections in people secondary to animal bites [60]. The relative frequency of C. canis and Desulfobulbus sp. was reduced in both the control and treatment groups at the end of the study (Figures 8 and 9). Neisseria sp. abundance was reduced significantly (p = 0.043) in the treatment group (Table 3).
The mouth spray containing gallic acid efficiently decreased the abundance of 22 species and increased the abundance of 17 species in cats [32]. The changes in some species were unexplainable, and their role in oral health diseases must be elucidated, and some species were found either in control or treatment samples. Brachymonas sp. may be the commensal oral bacteria that may prevent dental caries. Brachymonas sp. relative frequency was increased in the control and treatment groups (Figures 8 and 9). P. cangingivalis is the most prevalent canine oral bacterium [57], which was detected only in the control group, and its relative frequency was increased after six weeks ( Figure 8). D. orale was isolated from the subgingival plaque of a patient with periodontitis, which plays a potential etiopathogenetic role in PD [61,62]. The abundance of D. orale and Nicoletella species was found to be decreased in the post-treated experimental group when compared to its baseline values indicating that the mouth spray containing gallic acid could reduce the pathogen load in cats [32]. In the present study, D. orale relative frequency was reduced in the control and treatment groups (Figures 8 and 9). B. zoohelcum causes rare but severe human clinical diseases, mostly arising from animal bites. Older adults may contract Bergeyella infections following continuous contact with dogs or cats without being bitten [63]. B. zoohelcum abundance was increased in both control and treatment groups (Figures 8 and 9).
Like genus variations, species-level variations in the control and treatment groups need further studies to prove the impact of GAMS on the oral health of healthy dogs.

Conclusions
The study has limitations, such as fewer experimental subjects, short study duration, and metagenomic samples. The GAMS act well against the GI and CI than the PI. The microbiome analysis revealed that the GAMS usage altered the oral microbiota. Specifically, the abundance of the commensal bacterial phylum Actinobacteria and Chloroflexi, genera Frederiksenia, and Bergeyella was improved after six weeks of GAMS usage. The GAMS usage reduced the pathogenic bacterial species, including Neisseria sp., C. canis, Desulfobulbus sp., and C. mustelae. Moreover, some pathogenic bacterial abundances were increased at the end of the study. All the microbial variations were observed within the group. The inter-group analysis revealed that the changes were unrelated to GAMS usage. Further studies need to be carried out using more experimental subjects to confirm the effectiveness of GAMS. More metagenomic data are required to evidence the GMAS impact on the oral microbiome of healthy dogs.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/vetsci10070424/s1, Table S1: Logan and Boyce index for determining the plaque index; Table S2: Warrick and Gorrel method to determine the calculus index; Table  S3: Löe and Silness index to determine the gingivitis index; Table S4: The estimated sequences in the control and treatment samples; Table S5: The average and SD of the relative frequency of taxonomical parameters and their comparison between the groups.