Gut microbiome of native Arab Kuwaitis

The human gastrointestinal tract is home to trillions of bacteria [1]. Gut microbes play an important role in digestion of nutrients [2], metabolism of drugs and other compounds [3, 4], and synthesis of micronutrients [5], neurotransmitters and other metabolites [6, 7], as well as in the development of gut-specific immune system [8]. The gut microbiome also has an important role in maintaining health, and dysbiosis of the gut microbiome has been associated with various health disorders including metabolic syndrome, diabetes mellitus, obesity, inflammatory bowel disease, colorectal cancer, rheumatoid arthritis, atopy, eczema, autoimmune diseases and psychiatric disorders [9, 10].

There is a high degree of variation in the gut microbiome composition between individuals and within individuals over time, and people living in different geographical regions tend to have different gut microbiome [11, 12]. The diversity and composition of the gut microbiome is influenced by a variety of different factors including diet, environment, antibiotic use, proximity of hosts and host genetics [13, 14] and it is thought that globally, the gut microbiome of healthy individuals segregates into two or three distinct enterotypes based on the abundance of key bacteria—Bacteroides, Prevotella and members of the Clostridiales order [14, 15].

Characterising the ‘normal’ gut microbiome in different geographical regions provides a comprehensive picture of the microbiome from a global perspective. This is an important baseline information for understanding dysbiosis of the gut microbiome and how it relates to different populations. There are numerous studies on the gut microbiome of populations from various parts of the world [12, 16‐22], but only two studies from the Arabian Peninsula [23, 24], a major geographical region. These two studies are from Saudi Arabia. There are no studies from Kuwait, another country in the Arabian Peninsula. There is genetic variation [25] and variation in food consumption patterns across the Arabian Peninsula [26], which would influence the gut microbiome. Therefore, we studied the gut microbiome of native adult Arab Kuwaitis.

Twenty-five participants (10 males and 15 females, Table 1), aged 24 to 57 years were recruited from March 2017 to May 2017 (Table 1). The body mass index (BMI) value was not available for one participant. Ten (41.6%) and five (20.8%) of the remaining 24 participants were overweight and obese respectively. One participant was diagnosed with Crohn’s disease post specimen collection and informed the study team of the diagnosis, but did not report gastrointestinal symptoms at the time of specimen collection. All participants reported consuming a typical Kuwaiti diet.

A total of 9,420,376 2 × 300 base pair reads were generated from sequencing. Following filtering of the ASV table, 9,360,113 sequences (median of 385,526 reads per sample [interquartile range (IQR) = 128,030–537,590]) representing 1621 ASVs were included in the final analysis. Sequencing data were not rarefied as the rarefaction curves (Additional file 1: Figure S1) which confirmed that specimens had been adequately sampled.

Eleven bacterial phyla were detected (Fig. 1). The dominant phyla were Firmicutes and Bacteroidetes (representing 48% and 46% of total sequencing reads respectively), and Bacteroidetes was the most abundant phylum detected in the majority of participants (n = 18/25, 72%). Firmicutes was the most abundant phylum in seven other participants. Interestingly, participants with a gut microbiome abundant in Bacteroidetes had a lower bacterial diversity compared to participants with a microbiome abundant in Firmicutes (median Shannon diversity index in Bacteroidetes abundant gut = 3.80 [IQR = 3.36–3.95] vs. Firmicutes abundant = 4.58 [IQR = 4.34–4.83]; Z = − 3.15, P = 0.002; Fig. 2).

At the genus level, specimens clustered into two distinct groups driven by the abundance of Bacteroides (Fig. 3), with specimens high in Bacteroides clustering together and specimens with lower Bacteroides abundance clustering together. Bacteroides was detected in all specimens and was the most abundant genus in the majority of participants (n = 22/25, 88%, median relative abundance of 37% [IQR 20–41%]). Of the remaining three participants, one had a gut microbiome dominated by Prevotella_9 (52% relative abundance), one had a highly diverse microbiome with Prevotella_9 being the most abundant genus (10% relative abundance) and the third had a high abundance of Alistipes relative to other participants (23% vs. median abundance of 4% [range 0–13%]). Interestingly, this participant was the individual diagnosed with Crohn’s disease.

The core gut microbiome of participants was represented by 18 ASVs (Additional file 2: Table S1). Fifteen ASVs representing the core gut microbiome were from the Clostridiales order, primarily the Lachnospiraceae and Ruminococcaceae families. ASVs representing Bacteroides fragilis/xylanisolvens (ASV46), Bifidobacterium longum (ASV50) and Escherichia/Shigella (ASV53) were also identified as members of the core microbiome.

Bacteroides ASVs comprised 30% of all sequence reads. Prior to filtering of the ASV table we identified a total of 223 ASVs assigned to the Bacteroides genus (113 remained post filtering of the ASV table), eleven of which were present in 50% or more of study participants, indicating a high diversity in Bacteroides spp. across individuals. Twenty-three of the 223 identified Bacteroides ASVs had an average relative abundance greater than 0.5% and represented more than 80% of the relative abundance of all the Bacteroides ASVs (Additional file 3: Table S2). Given the importance of Bacteroides spp. in the gut microbiome and the high diversity within this genus between individuals, we investigated this genus further. We produced a phylogeny from a multiple alignment of the 223 Bacteroides ASVs and published 16S rRNA gene sequence for each type strain in the Bacteroides genus (Additional file 4: Figure S2). The relationships shown in the tree were used to group related ASVs and classify them into phylogeny groups based on the type strain present in the group. Based on the tree, ASV46-Bacteroides fragilis/xylanisolvens (which is a part of the core gut microbiome) was more closely related to B. xylanisolvens/acidofaciens/caecimuris type strains than B. fragilis type strain. The Bacteroides dorei/vulgatus group was the most abundant phylogeny group (representing 11.91% of all sequence reads) and was detected in all samples, whereas the Bacteroides fragilis group represented 0.73% of total reads and was detected in only 12 of the 25 samples (Additional file 5: Table S3).

Information about BMI was available for 24 participants; nine participants had a BMI of normal, ten had a BMI of overweight and five were considered obese (Table 1). There was no observed difference in the ratio of Firmicutes: Bacteroidetes between individuals with a normal BMI and overweight or obese individuals (median Firmicutes: Bacteroidetes ratio in normal BMI = 0.83 [IQR = 0.72–1.47] vs. overweight/obese = 0.78 [IQR = 0.55–1.27]; P = 0.1440).

It is well established that Firmicutes and Bacteroidetes typically dominate the gut microbiome of healthy adults across the world [37]. This is consistent with our finding that the gut microbiome of adult native Arab Kuwaitis is dominated by either Firmicutes or Bacteroidetes. Individuals with a Firmicutes dominated gut had an increased bacterial diversity compared to those with a Bacteroidetes dominated gut, and Bacteroides was the most abundant genus in all but three individuals.

The limited data describing the gut microbiome of people in the Arabian Peninsula has focused on individuals from Saudi Arabia [23, 24], and describes the dominant phyla to be Firmicutes and Actinobacteria. Diet is thought to be a major driver of differences in the gut microbiome composition seen globally [14, 38], and the high abundance of Bacteroidetes in our cohort compared to previous studies of Arabian Peninsula inhabitants could be a result of differences in diet between Kuwaitis and Saudi Arabians. The Kuwaiti diet is generally considered a blend of Arabian, Persian, Indian and Mediterranean cuisines. However, Arabian and Western fast foods rich in meat, sugar and fat are becoming increasingly popular with a consequent rise in the prevalence of obesity, diabetes and hypertension [39]. Bacteroides abundance has been associated with Westernized diets high in animal fat and protein [14], and a previous study demonstrated that introduction of a four-day animal based diet resulted in decreased abundance of Firmicutes in the gut microbiome and a subsequent increase in the abundance of spp. from the Bacteroides, Alistipes and Bilophilia genera [40]. It is possible that the high relative abundance of Bacteroidetes observed in our cohort is due to the increasing popularity of Westernized diets among Kuwaitis; however larger studies with more detailed dietary information are needed to investigate this further. It is worth comparing the Kuwaiti flora with the floras in low- and middle-income countries. The remote hunter-gatherer populations such as the Hadza from Tanzania, Pygmies from Central Africa, Mases from Peru and Amerindians from Venezuela have a high microbial diversity and an enriched taxa consisting of Prevotella, Succinovibrio, Treponema, Cyanobacteria, Tenericutes, Clostridium, Catenibacterium, Eubacterium, Lachnospira and Salmonella. Prevotella is thought to enhance the ability to digest and extract valuable nutrition from fibrous plant foods. People living in traditional farming or fishing communities like the Bantus of Africa, the Tunapuco population of the Andean highlands or the rural Malawian communities have a high microbial diversity and an enriched taxa consisting of Prevotella, Succinovibrio, Treponema, Ruminococcus and Bacteroides. This microbiome in traditional communities is thought to exhibit enrichment in carbohydrate- and xenobiotic-processing due to their access to more digestible sugars and therapeutic drugs (summarized in 38). In India, the gut microbiome is dominated by Firmicutes followed by Bacteroidetes, Actinobacteria and Proteobacteria and is enriched with microbial xenobiotic degradation pathways [41].

In China, the abundant taxa are Bacteroidetes, Firmicutes, Proteobacteria and Actinobacteria. However, there was a rural–urban divide with a significantly higher abundance of Proteobacteria in rural population. Urbanization was associated with loss of microbial diversity. Gene diversity increased with urbanization, along with an increase in antibiotic resistance and virulence genes, which were strongly correlated with the presence of Escherichia and Shigella [42]. In Argentina, Bacteroides was the predominant flora [43].

The gut microbiome is thought to differ between lean and obese individuals, particularly with respect to the relative abundance of Firmicutes and Bacteroidetes [44]. In a recent cross-sectional study of nearly 4000 adult Kuwaitis, 37% were overweight and 40.3% were obese [45]. Thus, our study participants reflected this general population, as a majority of the participants were overweight/obese. A recent systematic review noted that while several studies have identified obese individuals have a high Firmicutes: Bacteroidetes ratio compared to non-obese individuals, other studies have found contradictory results or no difference in the abundance of these phyla between obese and non-obese individuals [46]. Even though, a majority of our participants were overweight/obese, we did not observe any relationship between BMI and Firmicutes: Bacteroidetes ratio, which may be due to the small sample size. Clavel and Ecker [47] noted that a sample size of at least 500 specimens may be needed to examine associations between BMI and gut microbiome composition. Additionally, comparisons of the Firmicutes: Bacteroidetes ratio offers limited resolution given the numerous species present in these phyla.

Bacteroidetes is a complex phylum, and the class Bacteroidia is made up of five bacterial families (Bacteroidaceae, Marinilabiliaceae, Porphyromonadaceae, Prevotellaceae and Rickenellaceae), each of which is commonly found in the human gut. Bacteroides is one of the six genera under Bacteroidaceae and is a highly diverse genus, made up of numerous species (bacterio.net, List of Prokaryotic names with Standing in Nomenclature [35]). Bacteroides was the most prevalent and abundant genus identified in our cohort. Bacteroides spp. exist in a mutualistic relationship with the human host [48] and have an important role in fermentation of complex sugars [49], metabolism of proteins [50] and deconjugation of bile salts [51]. Bacteroidetes are enriched in patients suffering from type 1 and type 2 diabetes mellitus [52]. A recent study examining the bacterial richness of the gut microbiome of obese and lean individuals found reduced richness to be associated with adiposity, insulin resistance, dyslipidemia and an inflammatory phenotype [53]. Furthermore, participants with reduced bacterial richness had enriched Bacteroides spp. [52]. An enriched Bacteroides microbiome may be a marker of these characteristics in the Kuwaiti population.

Our phylogenetic analysis of Bacteroides revealed a high genetic diversity within individual species and suggests that species from the Bacteroides dorei/vulgatus phylogeny group may represent core gut microbiome in this Kuwaiti population. The high diversity of Bacteroides spp. across individuals has been observed in other populations [54] and may reflect differences in diets between participants; however, our small sample size prevented us from investigating this further. It is also possible that different Bacteroides ASVs present in a phylogeny group may have differing metabolic capabilities.

This study has limitations. It was difficult to recruit participants as obtaining a stool sample from a healthy person is a cultural taboo, thus the sample size was small. However, despite the small sample size, this study presents the first description of the gut microbiome composition of native Arab Kuwaiti adults. While all participants reported consuming a typical Kuwaiti diet, we did not collect detailed dietary or behavioural information leading up to specimen collection, which limits our ability to associate unique gut microbiome profiles with specific diets or behaviours. Finally, 16S rRNA gene studies are subject to primer bias, which may limit detection of relevant taxa, and provides low resolution at the species level. However, using an ASV approach provides consistent labels and allows comparison of ASVs across studies.

Our results showed that Firmicutes and Bacteroidetes were the dominant phyla in the gut microbiome of Arab Kuwaitis. Bacteroides was the most abundant genus with Bacteroides dorei/vulgatus forming the predominant phylogenetic group. While the predominance of Bacteroidetes may reflect the increasing Westernization of the Kuwaiti diet, larger studies examining the impact of diet and geography on the gut microbiome in the Arabian Peninsula are needed. As a majority of the study participants was overweight/obese, this also, would have contributed to the microbiome composition.