Differences in the gut microbiota of dogs (Canis lupus familiaris) fed a natural diet or a commercial feed revealed by the Illumina MiSeq platform

The gut microbiota is the collection of living microorganisms inhabiting the gastrointestinal (GI) tract of the host. It has been estimated that the gut microbiota of humans and animals consists of 10¹⁰–10¹⁴ microbial cells, a number roughly similar or 10 times higher than the number of host cells [1, 2].

In the past, studies on the microbiota were performed using culture-based methods, and research in the field was limited. However, recent advances in next-generation sequencing (NGS) technologies have allowed a more comprehensive analysis of the complex and diverse gut microbial communities [3]. Therefore, the number of studies on the gut microbiota using NGS system, which could provide a broad and deep understanding of the microbiota, is on the rise. These studies have revealed that the gut microbiota is closely linked with the host’s health and disease status, including maintenance of the GI health, stimulation of the immune system, development of obesity, and various GI disorders, including inflammatory bowel disease [4, 5]. Concomitant with the analysis of the relationship between the microbiota and health, several studies are aiming to identify the various factors affecting the microbiota [6]. Some studies have suggested that among such diverse factors, the diet greatly influences the composition of the gut microbiota; hence, recent studies have largely focused on the relationship between the diet and gut microbiota. The results revealed that animal-based diets might cause an increase in the abundance of Alistipes, Bilophila, and Bacteroides at the genus level, and a decrease in the abundance of Firmicutes at the phylum level in the human gut microbiota; furthermore, the human gut microbiota has been divided into a Prevotella enterotype and a Bacteroides enterotype, according to long-term diet [7, 8].

The dog (Canis lupus familiaris) is one of the closest companion animals to humans; over the years, as the quality of life improved, the number of people who raised dogs increased. In addition, people began treating their dogs as family members rather than pets, and began focusing on their health [9]. However, gut microbiota studies were mainly focused on human and human-oriented mouse models [4‐8]. Several studies of the dog were conducted, including comparisons of the microbiota of obese and lean dogs; comparisons of the microbiota in the presence or absence of GI diseases; and comparisons of the microbiota according to the presence or absence of dietary fiber or boiled meat in the diet [10‐13]. Nevertheless, most of these studies of dog gut microbiota were performed using the 454 pyrosequencing techniques, which are rarely used now-a-days; furthermore, they were not nearly as numerous as human-based studies, and did not match the increase in the numbers of pet owners, and their increased interest in dog health.

In recent years, the number of pet owners who feed their dogs a natural diet, i.e., one consisting of bones, raw meat (such as chicken and beef), and vegetables, instead of commercial feed, has increased [14]. This increase was fueled by the 2007 pet food recalls, because of melamine contamination [15]. In parallel with this trend, the advantages and disadvantages of a natural diet were discussed. Feeding dogs a natural diet was associated with some health benefits, such as fresher breath, healthier coat and skin, alleviation of arthritis, and improved immune response [16]. In contrast, some studies have provided evidences that discourage the use of natural diet because of nutritional imbalance and bacterial contamination. Feeding dogs a natural diet led to a pronounced nutritional imbalance and increased the risk of exposure to zoonotic pathogenic bacteria, including Salmonella spp., Campylobacter spp., and pathogenic Escherichia coli, which threaten the dog and public health [14, 15, 17, 18]. These studies, however, were only based on nutrition, pathogen detection, and clinical experience, and there were few studies about changes in microbiota composition associated with natural diet and commercial feed in dogs [19].

Given the above, factors that can influence the gut microbiota, especially diet type, in dogs, need to be studied. The current study was performed to investigate the effect of long-term diet, i.e., natural diet and commercial feed, on the gut microbiota. Specifically, (1) we identified the core microbiota of dogs fed a natural diet or a commercial feed up to the species level; and (2) compared the differences in alpha diversity, beta diversity, and the composition of gut microbiota between animals fed the two different types of diet.

To the best of our knowledge, the current study is the first to investigate the effect of long-term diet on the gut microbiota of dogs fed a natural diet compared with ones fed a commercial feed that have been actually applicated by dog owners by identifying the core microbiota up to the species level and comparing the differences in the gut microbiota between the two diet groups using NGS technology.

In the current study, the Illumina MiSeq platform and EzBioCloud database were used to analyze the gut microbiota of the dog. Among the various NGS systems, the Illumina MiSeq platform generates long and high-quality sequence reads, with the lowest error rates; it is also the most cost-effective platform, and hence suitable for small investigations [29‐33]. The Illumina MiSeq platform was therefore chosen from the available NGS systems for the current study. In general, it is known that classification up to the species level may not be possible in MiSeq, because the species-level classification system of the reference database is not sufficient, rather than limitations of sequencing. In the case of SILVA and RDP, which are widely used, classification information is provided only up to the genus level, and the database is not updated periodically. However, the EzBioCloud database has a total of 78,870 taxa information (a total of 17,903 species with valid species names), which is more systematic, accurate, and periodically updated [34]. Therefore, data analysis using the EzBioCloud database allowed us to classify correctly up to the species level. The results of this study were similar to some but not all previous studies that employed the Illumina MiSeq platform to investigate the dog gut microbiota. In a study by Sandri et al. [19], the gut microbiota of dog was found to be composed of Firmicutes (43%), Bacteroidetes (19.8–26.9%), Fusobacteria (4.7–11%), and Proteobacteria (1.3–4.3%); in another study [35], it was found to be composed of Firmicutes (84.4% of all sequences), Bacteroidetes (2.9%), Fusobacteria (3.2%), Proteobacteria (7.8%), and Actinobacteria (1.7%). These differences might be due to individual variation of gut microbiota and differences in the animals. In particular, the gut microbiota is highly affected by the host genotype and environmental exposure, i.e., conditions that are difficult to standardize [36]. The inconsistency might also be attributable to the different DNA extraction kits employed. Previous studies have revealed that DNA yield, quality, and integrity, and the microbial community results vary depending on the DNA extraction kits used [37, 38]. The number of OTUs identified in the current study was nonetheless similar to that of other studies (range 129–242); furthermore, the Good’s library coverage in the current study was higher than that in other studies, which suggested that our results might reflect the actual bacterial gut community of dogs enrolled in this study.

In this study, diets were found to have a greater extent on the gut microbiota than other factors. No specific tendencies were observed when each sample was categorized by other factors (breed, gender, age, and weight of the dogs); however, there were pronounced differences in beta diversity, i.e., the measure of group comparison according to dietary types. Specifically, the samples mostly clustered together between the two diet groups based on PCA analysis (Fig. 3). In addition, significant differences were observed between the two diet groups in the number of OTUs, species richness, and evenness (Fig. 2). According to alpha diversity, i.e., species diversity, the richness estimates, and diversity indices of the microbiota of the natural diet group were significantly higher (p < 0.05) than those in the commercial feed group. Differences in the core microbiota at the phylum, family, and species levels were also observed between the two groups. The core microbiota comprised shared organisms found in the majority of individuals [39]. In the current study, core microbiota accounted for more than 99 percent at the phylum level, more than 90 percent at the family level, and more than 66 percent at the species level and could have a different impact on host health. Thus, the diet might indeed be responsible for the differences in alpha diversity, beta diversity, and the core microbiota. These differences might be due to differences in the way the two diet types were manufactured and differences in the main ingredients of the two diets. Generally, commercial feeds contain controlled nutrients and controlled microorganisms, because they undergo formal manufacturing processes, including compression through high temperature and high pressure and microbial monitoring. However, natural diets do not go through any manufacturing process and are fed to dogs as raw, so that more nutrients and microorganisms in the natural habitat are absorbed into the gut of the dog. In addition, these two diet types had differences in the main ingredients. The commercial feeds given to dogs recruited for the current study contained crude protein (18–21% of total content), crude fat (8–10%), crude fiber (3–5%), and crude ash (7%), with 10% of moisture. Based on these values, the main ingredients of the commercial feed were carbohydrates (up to ca. 50%). On the other hand, the main ingredients of the natural diet, which consisted of bones and raw meat, were crude protein (30–52%) and fat (11–50%), regardless of the meat type [40]. A previous study of the human gut microbiota revealed that there was a difference in gut microbiota between individuals with protein/fat-based eating habits and individuals with carbohydrate-based eating habits, because of the differences in microorganisms required for the digestion of carbohydrates, proteins, and fats [7]. By analogy, in the current study, the microbiota of dog would have changed depending on the most frequently consumed ingredients, i.e., the carbohydrate-based commercial feed and the protein- and fat-based natural diet. Moreover, the differences in beta diversity and the core microbiota observed herein were consistent with the results of other studies that demonstrated that long- or short-term diets play a substantial role in shaping human gut microbiota [7, 8].

From the perspective of microbial infection, this study suggests that the potential risk of opportunistic infection could be higher in dogs fed a natural diet than in dogs fed a commercial feed. Previous studies have revealed that dogs fed a natural diet are more likely to be exposed to bacterial contamination (ca. 30–50% of Salmonella spp. and 50% of E. coli group) and could be at a greater risk of foodborne illness than dogs given a commercial feed, because there was no regulation for microbial monitoring in raw meat and dogs fed a natural diet was more likely to be exposed to contaminated raw meat [14, 17, 18]. Furthermore, the presence of these opportunistic microorganism could pose a threat to public health, through dissemination of infectious agents to the owner or other animals whose immune system is weakened [15]. In the current study, the proportion of C. perfringens and F. varium, which are opportunistic microorganisms, was higher in the natural diet group than in the commercial feed group (Fig. 4c). Clostridium perfringens is usually found in the gut of human and animals, as a member of their normal flora [41]. Nevertheless, C. perfringens, which may infect the host after ingestion of contaminated raw beef or chicken, could cause necrotic enteritis, diarrhea, and gas gangrene [42, 43]. Furthermore, in the past, C. perfringens was the third most common bacterial foodborne illness in England and Wales; C. perfringens is still recognized as a foodborne pathogen [44, 45]. F. varium is also found in the gut of dogs as a member of the normal gut flora; however, under certain circumstances, i.e., if the microbial composition of F. varium was altered, or if contaminated with soil or feces, it might cause colon cancer, intra-ocular infections, and conjunctivitis [46, 47]. Specifically, in a mouse inoculation test, F. varium was found to be associated with ulcerative colitis [48]. However, in the current study, Salmonella enterica and pathogenic E. coli, which were previously detected in culture-based studies, were not evident from the NGS data [14, 18]. This discrepancy is probably because of the detection limit of the NGS system [49]. Culture-based protocols enrich the targeted microorganisms and control the growth of other microorganisms; on the other hand, NGS-based methods reflect the current distribution of gut microbiota without the need for any microbiological processing.

This study provides the basis for gut microbiota studies based on dietary type in dogs. However, because this study was conducted with healthy dogs, we cannot elaborate on the direct relationship between the health status and differences in microbiota according to diet type; we only discuss the risk of opportunistic infection. Therefore, to elucidate the correlation between health and diet type, follow-up studies need to be conducted with diseased dogs.

This study is the first to analyze the association between dog gut microbiota and long-term diet (i.e., natural diet and commercial feed) using the Illumina MiSeq platform. Pronounced differences were detected in the microbiota of the two diet groups. Differences in the core microbiota at the phylum, family, and species levels were observed between the two groups. The microbiota of the natural diet group was characterized by higher richness and diversity compared with the commercial feed group. Samples from each group mostly clustered together according to the diet type. In addition, dogs fed a natural diet could be at a higher potential risk of opportunistic infection than dogs fed a commercial feed. Collectively, these results indicate that diet likely affects the microbiota, thereby playing a key role in animal health. This study provides the basis for gut microbiota studies based on dietary type in dogs; furthermore, this study, together with follow-up studies, could be used to ultimately devise an appropriate diet for dogs.