Donkeys have unique digestive characteristics; the biggest difference from ruminants is that donkeys are hindgut fermentation. The donkey has a well-developed hindgut structure with a length of more than 4.5 m and a volume of more than 110 l, which is approximately 16 times the volume of its foregut (Liu et al., 2019).

The animal gut tract has a complex microbial ecosystem, and the microorganisms in the gut were closely related to the life activities of the host, which may affect growth and metabolism, nutritional digestion, immunity, and the ability to resist invading pathogens (Egert et al., 2006; Edgar, 2013). Studies have shown that the region (He et al., 2018) and environment (Rothschild et al., 2018) were the main factors affecting gut microbes. The environment is one of the factors that cause and maintain the differences in and diversity of gut microbes in the hosts of different regions. The proportion of shared flora decreases exponentially with an increase in the distance from the host distribution area, resulting in a gradual increase in the difference in flora (Moeller et al., 2017). Monaghan et al. (2020) studied the structure, function, antibiotic resistance, and resistance to Clostridioides difficile infectious diarrhea of gut microbes in 105 urban and rural populations in central India and found that the composition of bacteria and viruses in the gut has obvious urban–rural differences, with geography having the greatest impact. Studies had shown that geographical conditions played a key role in shaping the diversity of the gut flora of iguanas, bats, and European fireflies (Lankau et al., 2012; Phillips et al., 2012; Sudakaran et al., 2012). Additionally, the research on factors influencing the gut flora of mice and fish had also found that hosts’ geographic conditions and diet affected the composition and structure of the gut flora (Sullam et al., 2012). Lorenc et al. (2014) used 454 pyrosequencing technology to sequence the 16S rRNA of 121 house mice in eight regions of Western Europe and found that geographical conditions were the main factors that determined the colonization pattern of the gut microbial community. These results indicate that microbiome interventions to improve clinical treatment should focus on geographical specificity.

The diversification of ecological types in China has created abundant donkey resources. More than 30 donkey species have been observed, such as the Tibetan Plateau Qinghai donkey (QH, in a plateau continental climate) and the Shandong Dezhou donkey (DZ, in a temperate monsoon climate in the North China Plain). Different regions often have different environmental conditions, and how host and symbiotic bacteria adapt to different environments has been investigated (Zhou et al., 2016). At this stage, although attention to the research on gut microbes has been increasing, few studies have investigated donkey gut microbes. In this study, a high-throughput 16S rRNA sequencing method combined with metagenomics was used to study the composition and structure of donkey gut microbes, as well as the commonality and characteristics of donkey gut microbes in different regions (Qinghai, Dezhou), laying a foundation for further research.

The elevation of the Qinghai was higher than that of Dezhou, and the temperature and oxygen concentration decreased with an increase in elevation. The species composition of the gut flora was susceptible to environmental temperature, and decreased environmental temperature leads to a decrease in the alpha diversity of the flora (Chevalier et al., 2015; Moreno-Navarrete and Fernandez-Real, 2019). The alpha diversity of the gut flora of mice living at 12°C for a long time was significantly lower than that of mice living at 29°C (Zietak et al., 2016). However, other studies had found that the structure of the gut flora of mice living in a warm environment (22°C) was similar to that of mice living in a cold environment (4°C) (Li et al., 2019a,b). In addition, the community structure of the host gut flora remains stable in long-term hypoxic environments (Lucking et al., 2018; Khanna et al., 2021). Liu et al. (2020) analyzed the gut of Tibetan wild ass and African wild ass in the Qinghai region by using 16S rRNA genes sequencing; the results showed no significant difference in alpha diversity between the two groups, but the difference in beta diversity was significant. Tibetan wild ass had a relatively more complex bacterial network and a stronger dry matter digestion ability than the African wild ass. In this study, the diversity and richness of the flora of the Qinghai donkey were higher than those of the Dezhou donkey (Figure 1), indicating that the Qinghai donkey may have unique gut flora that can adapt to the hypoxic environment of the Tibetan Plateau.

At the phylum level (Figure 2C), the dominant flora of the Qinghai donkey were Firmicutes, Proteobacteria, and Cyanobacteria in the stomach; Firmicutes and Proteobacteria in the small intestine; and Firmicutes and Bacteroidetes in the cecum and rectum. We concluded that Firmicutes and Proteobacteria were the dominant phyla in the foregut of QH, and Firmicutes and Bacteroidetes were the dominant phyla in the hindgut of QH. The dominant flora of the Dezhou donkey were Firmicutes and Bacteroidetes in the stomach, Firmicutes in the small intestine, and Firmicutes and Bacteroidetes in the cecum and rectum. The abundance of Firmicutes in the Dezhou donkey foregut was higher than that in the hindgut, and that of Bacteroidetes in the hindgut was higher than that in the foregut. In equine animals, the foregut was mainly responsible for the digestion and absorption of food, with slight fermentation, and the hindgut was related to microbial fermentation, especially the large intestine (Argenzio et al., 1974, 1977; DiBaise et al., 2008). Firmicutes mainly uses carbohydrates in herbivores (Brulc et al., 2009), which can improve animal immunity and enhance gut function. Bacteroides have abundant genes encoding carbohydrate-active enzymes and can easily switch according to the availability of energy source types in the gut (Flint et al., 2012). Bacteroides were significantly enriched in the digestion-related microbiota of the large intestine and were not affected by the location of the gut in the mucosal-related microbiota. Firmicutes and Bacteroides were the two most dominant phyla among several other herbivores, such as goats (Li et al., 2019a,b), cattle (Mao et al., 2015), and horses (Al and Andrews, 2009). Previous research and this study suggest that the accumulation of Firmicutes and Bacteroidetes in the large intestine may help the host adapt to the complex internal environment.

At the genus level, Lactobacillus dominates the foregut, and Streptococcus dominates the hindgut. Lactobacillus can degrade fibrous carbohydrates (e.g., pentose, hexose, and starch), participate in the uncoupling of bile salts (Bao et al., 2012), and produce antimicrobial substances (e.g., bacteriocins and lactate) or compete with pathogens for mucosal adhesion sites and nutrients to inhibit the proliferation of pathogens (Umu et al., 2017). Lactate, the fermentation product of Lactobacillus, can acidify the gut mucosa. Research on horses found that Lactobacillus species dominated the stomach, whereas Streptococcus was significantly increased in the duodenum (Costa et al., 2015). In another study of DZ, the foregut microflora was dominated by Lactobacillus, and the hindgut microflora was dominated by Streptococcus, and Lactobacillus and Streptococcus were beneficial probiotics for equine animals (Liu et al., 2019).

In this study, the abundance of Firmicutes, Fibrobacteres, Verrucomicrobia, and Spirochaetes in the rectum was higher in QH than in DZ, and Bacteroidetes and Clostridiales were more abundant in the rectum of DZ than that of QH. Most bacteria in Firmicutes were butyrate-producing, and butyrate was one of the final metabolites of polysaccharides (Pryde et al., 2002). Polysaccharides entering the gut were crucial factors affecting the physiological state and composition of gut commensal bacteria. Butyric acid plays an important role in maintaining the integrity of rectal tissues and can prevent colon disease. Studies have shown that more than 35% of the enzymes required for animal digestion and metabolism were produced in the gut flora, and 25% of these enzymes were involved in carbohydrate metabolism (Gill et al., 2006). Firmicutes encode few carbohydrate-degrading enzymes but more ABC transporters (ATP-binding transporters) to transport carbohydrates (Mahowald et al., 2009). The most important carbohydrate transport system is the ATP-ABC-type transport system (Turroni et al., 2018). Bacteroides, the second most dominant group in the intestine, can degrade carbohydrates and provide the host with 10–15% of energy from food. Due to the efficient polysaccharide degradation system of Bacteroidetes and its ability to produce large amounts of short-chain fatty acids, Bacteroidetes has become the most studied strain in the transportation and utilization of polysaccharides in the intestinal flora, and approximately 20% of the genes in their genomes were used to complete the decomposition of sugar, also a possible reason why it had become a dominant strain (Singh, 2019). Jena et al. (2016) fed rats with a high-sugar diet for 60 days: the proportion of Escherichia coli and Clostridium in the intestinal tract increased, and the content of Lactobacillus decreased. They also investigated the composition of fecal microbiota and pro-inflammatory cytokines in serum by denaturing gradient gel electrophoresis. They found that the expression of genes such as TLR2, TLR4, and NF-kB increased in various tissues, and the inflammatory response in blood and tissues was enhanced, which significantly affected their metabolic status.

Besides, Proteobacteria, Cyanobacteria, Actinobacteria, and chloroplasts exist almost exclusively in the foregut of donkeys in the QH region. Actinobacteria affect various metabolic and physiological activities, including the production of extracellular enzymes, antibacterial activity, and formation of other secondary metabolites (Schrempf, 2001; Zhao et al., 2018). Cyanobacteria and Chloroplasts were involved in pathways related to photosynthesis (Crespo-Piazuelo et al., 2018). These flora became the potential biological markers to distinguish the two donkey breeds of QH and DZ.

The results of the present study suggest that QH had more pathways enriched in the glycolysis and gluconeogenesis pathways of carbohydrate metabolism than DZ did. The metabolic pathway results demonstrated that pyruvate ferredoxin oxidoreductase and acetyl-CoA synthetase participate in the citrate cycle, and PGK and 2,3-bisphosphoglycerate-independent phosphoglycerate mutase participate in the carbon fixation in photosynthetic organisms. In Methanosarcina species, ferredoxin was shown to be involved in methanogenesis from acetate (Ferry, 1993). H₂ formation from pyruvate was stimulated by ferredoxin; thus, it may act as an electron carrier between pyruvate oxidoreductase and pyruvate hydrogenase (Hatchikian et al., 1982). All the tested archaea contained pyruvate ferredoxin oxidoreductase, which played a role in catabolism and anabolism. Thus, pyruvate ferredoxin oxidoreductase appears to represent the only mechanism for pyruvate acetyl-CoA conversion in the archaeal domain (Ikeda et al., 2006). Notably, 2,3-Diphosphoglycerate-independent phosphoglycerate mutase (IPGAM) catalyzes the reversible conversion of 3-phosphoglycerate (3-PGA) to 2-phosphoglycerate (2-PGA) during glycolysis key enzyme (Johnsen and Schonheit, 2007). There were two isoforms of PGK in the human genome, PGK1 and PGK2, with similar structures and functions, and more than 80% of the amino acid sequences were similar (McCarrey et al., 1992). PGK1 was the first key enzyme that generates ATP in the glycolytic pathway, and its participation in glycolysis was the main function of PGK1, which was important for the continuously production of cellular energy under hypoxic conditions (Sun et al., 2015). Acetyl-CoA was a central metabolite in carbon and energy metabolism. In mammalian cells, carbohydrates were converted into various biomolecules via many processes. Carbohydrates were first decomposed into acetyl-CoA, and then acetyl-CoA was used as a precursor for anabolism. This pathway was an important for the conversion of sugars into other biomolecules. The main synthesis pathways of mitochondrial acetyl-CoA comprise oxidative decarboxylation of the glycolysis product pyruvate, fatty acid β-oxidation, and branched-chain amino acid decomposition. Mitochondrial acetyl-CoA normally enters the tricarboxylic acid cycle for further metabolism (Boroughs and DeBerardinis, 2015). Although the available energy produced by gluconeogenesis and glycolysis was limited, this approach can supplement the much-needed energy for muscle hypoxia during heavy physical labor or long-term strenuous exercise, which was consistent with the environmental and service conditions of QH. The aforementioned analysis demonstrated that the main enriched metabolic pathways of QH donkeys were carbohydrate metabolism, and they were mainly enriched in glycolysis and gluconeogenesis pathways.

In this study, we investigated the differences in gut microbes in QH from the Tibetan Plateau and DZ by using 16S rRNA gene high-throughput sequencing and metagenomic sequencing. The results showed that the flora diversity and richness of QH were higher than those of DZ, and the flora diversity and richness of the hindgut were higher than those of the foregut, with no sex difference. The major pathways associated with the Qinghai donkey were signal transduction mechanisms and carbohydrate transport and metabolism, and Bacteroidetes were the major contributor to these functions.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary material.

The animal study was reviewed and approved by Animal Care and Use Committee of Qingdao Agricultural University.

SL and YS conceived of and designed the experiments. SL, GZ, TX, SZ, JC, DA, JD and JS collected the samples. RG conducted the experiments. Samples were processed by RG, TX, JW and FZ. RG analyzed the data. RG wrote the original manuscript, and SL, YS, WS and QP contributed to writing the manuscript. All authors contributed to the article and approved the submitted version.

This research was funded by the Shandong Province Science Major Project on Improved Agricultural Varieties, grant numbers 2013lz016 and 2017LZN022; Donkey Innovation Team of Shandong Modern Agricultural Industry Technology System, grant number SDAIT-27; Major Agricultural Application Technology Innovation Projects of Shandong Province, grant number SD2019 XM 008; Ministry of Agriculture and Rural Affairs “Construction of technical route for phenotypic identification of donkey skin and meat traits” (grant number 19211183); Experimental Technology Research Programme of Qingdao Agriculture University(20210021).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.