The colon is a largely anaerobic ecosystem, although there are oxygen gradients across the colonic wall, with free oxygen rapidly metabolised by facultative anaerobes ^[1][28]. The prevailing environmental conditions mean that most colonic bacteria are strict anaerobes. For example, some of these anaerobes, particularly those derived from the Firmicutes phylum such as Roseburia species, are extremely sensitive to oxygen and cannot survive more than two minutes of exposure to air ^[2][29].

Bacteria are the most important contributors to microbiota biomass, with bacterial abundance progressively increasing along the intestinal tract. There are approximately 10³ cells per gram of contents in the duodenum of the small intestine, increasing up to approximately 10¹⁰ to 10¹¹ bacterial cells per gram (wet weight) of faecal material ^[3][19]. Archaea are comparatively much less abundant in many individuals, although can reach densities of up to 10⁹ cells per gram of stool in some people ^[1][28]. Viruses are present at a density of at least 10⁹ particles per gram of faeces ^[4][30] with the bowel mucosa possessing a high virus to bacterial cell ratio ^[5][31], and most are bacteriophages ^[6][32]. As with gut bacteria, the viral populations exhibit inter-individual variability and are also diet-responsive, likely because viral population dynamics are heavily influenced by their prevailing bacterial hosts ^[7][8][33,34]. Mycobiome studies suggest that Malassezia, Trichospora, and Candida species are typically the most dominant eukaryotes in the gut, although cell numbers are much smaller than the bacterial component ^[9][35]. The abundance of Candida species has been shown to significantly correlate with the consumption of carbohydrate-rich foods ^[10][11][36,37].

The dominant bacterial populations are mainly derived from five phyla; the Gram-positive Firmicutes and Actinobacteria, and the Gram-negative Bacteroidetes, Proteobacteria, and Verrucomicrobia ^[1][12][28,38]. Akkermansia muciniphila is a dominant and widespread species belonging to the latter and is a key mucin-degrading bacterium ^[13][39]. The strictly anaerobic Firmicutes and Bacteroidetes lineages predominate in most healthy individuals. This relatively simple composition at the phylum level, however, is in stark contrast to an extraordinary level of diversity at the species and strain levels, with many thousands of species now known to be capable of colonising the human colon.

Although there is a very broad range across the human population, Bacteroidetes spp. may comprise about 30% of the total microbiota on average, with some individuals tending to be either Bacteroides or Prevotella dominant ^[14][40]. Commonly occurring species include Bacteroides vulgatus, Bacteroides ovatus, Bacteroides uniformis, and Prevotella copri. The human colonic Firmicutes species mainly belong to two phylogenetic groups: the Lachnospiraceae that includes abundant genera such as Eubacterium, Roseburia, Blautia, and Coprococcus, and the Ruminococcaceae that encompasses Faecalibacterium and Ruminococcus species ^[15][41]. Other commonly reported genera that are typically found in lower abundance include Actinobacteria species such as Bifidobacterium adolescentis and Collinsella aerofaciens. Facultative anaerobes tend to be less dominant but typically include species such as Escherichia coli and sometimes the sulphate-reducing bacterial species Desulfovibrio piger. The most common and abundant archaeal species are methanogens, most often Methanobrevibacter smithii ^[1][28]. It is also worth noting that many of the most abundant species in the human colon are available in culture, which allows researchers to conduct detailed analysis of their physiology, although a larger proportion of the less dominant species remain either uncultured or are yet to be described ^{[16][17][18][19]}[42,43,44,45].

When considering the collective genome content (or “metagenome”) of the microbiota, the inherent and daunting complexity becomes even more apparent. The human gut metagenome is estimated to contain many millions of unique genes, many of which are currently uncharacterised, and is at least two orders of magnitude more complex than the human genome ^[20][21]. It is apparent, however, that a relatively small number of dominant species are widespread amongst the human population and often numerically abundant ^[20][21]. For example, in many healthy subjects, two of the most abundant species are the butyrate producers, Faecalibacterium prausnitzii and Eubacterium rectale ^[21][46]. There is also likely to be strong interplay between the species that comprise the densely populated and complex microbial ecosystem of the colon. Some of these interactions are competitive, including competition for available macro- and micro-nutrients, but there are also many types of non-competitive interactions, such as cross-feeding and/or co-operative degradation of recalcitrant dietary substrates ^[22][23][13,47].

The composition of the microbiota is largely stable in adulthood, but fluctuations in response to lifestyle factors such as dietary consumption patterns are common. Within individuals, there are also a multitude of gastrointestinal micro-environments available for microbial colonisation, such as in the mucus layer covering the gut mucosa, on the surface of food particles in the gut lumen, or liquid-phase luminal contents ^[24][25][26][48,49,50]. For bacterial species to persist in the human large intestine, cells must be incredibly competitive and have the capacity to adapt to a wide range of gut environmental factors such as fluctuations in pH ^[27][28][51,52], bile salt concentrations ^[29][53], and availability of micro-nutrients, including iron ^[30][54], which is an essential co-factor for some bacteria ^[31][55]. Many bacteria also have specific vitamin requirements ^[32][56]. This includes the B vitamins such as biotin [B7], cobalamin [B12], folate [B9], pyridoxine [B6], riboflavin [B2], and thiamine [B1] that are needed to help drive enzyme-catalysed reactions ^[33][57]. Bacteria that generate growth factors such as vitamins are prototrophs, whereas bacteria that cannot produce these are auxotrophs and need to obtain these from dietary sources or by cross-feeding from other bacteria. For example, certain dominant gut bacteria directly depend on the presence of thiamine and riboflavin to generate butyrate. Soto Martin et al. (2020) investigated the requirement of 15 human butyrate-producing gut bacterial strains for a range of B vitamins and revealed that one of the most abundant butyrate-producing species in the colon, Faecalibacterium prausnitzii, was auxotrophic for most of the vitamins tested and is therefore reliant on cross-feeding with vitamin-producing strains ^[34][58]. Vitamin-related co-factors are metabolically expensive to produce and therefore it should not be surprising that these growth factors are shared between gut bacteria.

As mentioned previously, the human gut metagenome possesses many more genes than the human genome, providing the host with a markedly increased functional capacity, particularly with respect to the breakdown of complex carbohydrates such as fibres ^[35][59]. A huge range of metabolites are therefore produced by the gut microbiota, and many of these can influence host physiology and health. Examples of health-related metabolites include short-chain fatty acids (SCFAs), lactate, secondary bile acids, lipids, vitamins, plant-derived flavonoids, amino acid and peptide metabolites, and trimethylamine ^[36][60]. Although the array of gut microbiota-derived metabolites is daunting, metabolomic and host-centric approaches are being used to investigate the impact of a range of bacterial metabolites on host health ^[26][37][38][6,50,61].

Despite the variation in microbiota composition between individuals, there are a number of conserved functional activities. Of key importance among these is bacterial fermentation, whereby energy sources such as complex dietary polysaccharides are broken down anaerobically into SCFAs plus gases, including hydrogen, carbon dioxide, and, in some individuals, methane ^{[39][40][41][42][43]}[62,63,64,65,66]. The major SCFAs detected in the large intestine are acetate, propionate, and butyrate ^[44][45][67,68]. Other organic acids such as lactate, succinate, formate, and valerate are also formed by gut bacteria but tend not to accumulate at high concentrations in healthy individuals, primarily due to bacterial cross-feeding and conversion to other SCFAs ^[28][52]. The total SCFA levels in faeces may reach levels of approximately 50–150 mM on average, with up to 95% of SCFAs absorbed by the colonic mucosa. Of the SCFAs, butyrate is absorbed preferentially by colonic epithelial cells ^[28][52], and the colonic epithelium derives as much as 60 to 70% of its energy needs from butyrate ^[46][69]. It has been estimated that individuals consuming a typical Western diet may produce a total of 0.5 to 0.6 M of SCFAs per day, which provides up to 5 to 10% of daily calorific requirements ^[47][48][70,71]. Bacterially produced SCFAs also exert substantial influence over host factors such as colonic pH ^[27][28][51,52], gut transit time, and the absorption of sodium, potassium, and water ^{[49][50][51][52]}[72,73,74,75].

The type of fermentation acids generated differs between species within the gut microbiota. Therefore, the concentration and type of SCFAs in the colon of a given individual will be impacted by the underlying species content of their microbiota. This may have implications for host health as different fermentation acids have varied impacts on the host. Butyrate, for example, is largely thought of as beneficial due to its putative anti-inflammatory and anti-carcinogenic effects, whereas lactate has a range of potential deleterious impacts.

Many bacterial species in the colon form acetate as a fermentation end product; therefore, it is often the most abundant SCFA ^[1][28]. Cross-feeding between formate-producing species, such as Ruminococcus bromii, and acetogens (which are bacteria that generate acetate as an end product of fermentation), including Blautia species, may be another substantial factor in the high rates of acetate production in the gut ^[53][76]. Dominant propionate producers include Bacteroides, Prevotella, and some Negativicutes species ^[1][54][55][28,77,78]. Key butyrate-producing bacterial lineages belong to the Firmicutes phylum and include Eubacterium rectale, Roseburia spp., Anaerobutyricum (formerly Eubacterium) spp., Anaerostipes spp., and Faecalibacterium prausnitzii ^[56][57][79,80]. Separately, lactate is formed by many different bacterial species in the colon, including Bifidobacterium species. However, this acid does not tend to accumulate in the healthy adult colon as it can be utilised as a growth substrate by cross-feeders ^[54][58][77,81]. Of these cross-feeders, some Firmicutes such as Anaerobutyricum (formerly Eubacterium) and Anaerostipes species are common inhabitants of the healthy human colon, and can metabolise lactate to form butyrate ^[38][39][61,62]. In vitro studies employing stable isotopes have revealed that lactate cross-feeding by species such as Anaerobutyricum and Anaerostipes spp. makes an important contribution (approximately 20%) to the butyrate pool ^[59][82].

Depending on dietary intakes, adults consume approximately 1 g per day of polyphenols, which are a diverse class of plant secondary metabolites with a range of postulated impacts on host health. Therefore, one of the other metabolic activities of gut bacteria that may be of importance for human health is that they are likely to have a major role in transforming these plant polyphenols (as reviewed in depth elsewhere ^[60][22]), as well as releasing these compounds during breakdown of fibrous plant material. The impact of these metabolites on host health are therefore intricately linked with the underlying composition of an individual’s gut microbiota. Certain phytochemicals may not be fully released from plant fibres if a given gut microbiota lacks primary degrading species that are able to breakdown the plant material in which the phytochemicals are bound up, and final metabolite pools will also be strongly influenced by the ability (or not) of the gut microbiota to transform these phytochemicals to other biologically active compounds, many of which are absorbed and enter systemic circulation ^[61][83]. Thus, it follows that the individual response to identical dietary interventions can vary greatly, depending on their baseline microbiota composition. This has potentially important ramifications for generic dietary advice guidelines and for the emerging field of personalised nutrition.

There are also further molecules produced by the gut microbiota with human health relevance, which include a multitude of small molecules such as ribosomally encoded and post translationally modified peptides (RiPPs), saccharides, amino acids, non-ribosomal peptides (NRPs), NRP-independent siderophores and polyketides (PKs), and PK–NRP hybrids. RiPPs include several subclasses such as bacteriocins, microcins, lantibiotics, and thiopeptides, which are antibacterial in nature and can therefore be important mediators modulating the interactions and competition between bacteria. By inhibiting the growth of certain types of bacteria, they may also help to protect the host from invading pathogens ^[62][84]. Examples of ribosomally synthesised antimicrobial products produced by bacteria include lantibiotics such as nisin O discovered from the human gut bacterium Blautia obeum A2-162 ^[63][85]. There is another category of small molecules known as polyketides (PKs) with potential to inhibit human pathogens. Metagenomic studies have revealed that gut bacteria may have the capacity to form polyketides (PK) and/or non-ribosomally synthesised peptides (NRP) and PK–NRP hybrids ^[64][86]. However, rather few of these small molecules have actually been isolated and characterised from human microbiota to date. Examples of previously described compounds include reutericyclins, which are a family of antimicrobial metabolites synthesised by PK–NRP enzymes such as in Lactobacillus reuteri that can kill the pathogen Clostridioides difficile at physiologically relevant concentrations ^[65][87]. In addition to putative antimicrobial activities, other secondary metabolites released by the gut microbiota, including polyketides, may possess anti-tumour and cholesterol-lowering properties ^[66][88]. The discovery of novel bacterial secondary metabolites in gut anaerobes is therefore an area of research worthy of further investigation and potential pharmaceutical development. However, this is an emerging area of research, with much still to be learned. It is still unclear, for example, to what extent changing host dietary consumption patterns might promote the growth of specific bacteria that can synthesise health-related small molecules or influence their production of these metabolically active metabolites.