A question of culture: bringing the gut microbiome to life in the -omics era

Introduction

The mammalian gut hosts a diverse microbial community (microbiota) that provides a range of ecological, nutritional and immunological functionalities, among others, that are relevant to the health and well-being of the host. For much of the twentieth century our understanding of the mammalian gut microbiota was largely developed using culture-dependent approaches with ruminant livestock providing the deepest mechanistic insights into their functional capacities mainly due to the long-recognised association between the rumen microbiota and animal health and productivity. Many of the seminal works in gut microbiology were originally developed for rumen microbial species and include the description of new media, culturing techniques and the isolation of diverse microbes in a rich scientific literature extending back to the 1940s. However, it was the development of culture-independent metagenomic approaches coupled with advances in DNA-sequencing technologies and computational methods that established the 'gut microbiome' as a discrete field of research. There have been several major metagenomic-based studies on the gut microbiomes of ruminants and other mammals that have provided unprecedented insights into their diversity and functional capacities, and transformed our awareness and appreciation of the extent of microbial 'dark

http://dx.doi.org/10.19103/AS.2020.0067.03 © Burleigh Dodds Science Publishing Limited. 2020. All rights reserved.

matter' in the form of uncultured microbes and their genes that reside in these communities.¹’^-41

Despite the insights afforded by these approaches, many studies have also highlighted the limitation of solely applying culture-independent approaches to characterise the microbiome. For instance, cultivation-based studies remain the best way to specifically link microbes with discrete biological activities and thereby provide new opportunities to move beyond the current observational paradigm and towards causation in gut microbiology research.^15-71 In addition, sequencing of individual gut-microbial strains improves mapping and assembly from metagenomic data.¹⁸-⁹* The human microbiome revolution in particular has led to a renewed interest in culturing approaches and the 'discovery' of culturing approaches and insights that have been well established for generations of rumen microbiologists. Together, these have given new life to culturing efforts and there are now multiple initiatives and related culture collections for mammals including humans (e.g. Culturable Genome Reference,¹⁹¹ Human Gastrointestinal Bacteria Culture Collection¹⁸¹), ruminants (e.g. HungatelOOO collection¹’⁰¹) and rodents (Mouse Intestinal Bacterial Collection¹”¹). Notably, several recent studies drawing upon 16S rRNA and metagenomic-based analyses have suggested the perceived unculturability of many gut microbes from mammalian gut environments may have been overstated.¹’^2-’⁶¹ These claims remain contentious, and, indeed, a recent study suggested that over 70% of human gut-microbial species lack a cultured representative.¹’⁷¹

The mammalian gut microbiota is highly diverse and typically includes viruses, bacteria, archaea, fungi and protozoa. While it is clear from numerous studies over the years that many fastidious gut microbes grow quite well under laboratory conditions there are others that remain resistant to cultivation as axenic cultures.¹’⁸¹ Our inability to culture these microbes suggests that there are fundamental aspects of their biology that we do not yet understand. There have been many culturing advances that continue to extend the repertoire of gut- microbial isolates¹’^9-2’¹ and meta-omic approaches nowalso offer unprecedented insights into the metabolic potential of these microbes that could be used to direct and support their growth under laboratory conditions. The integration of these culture-dependent and culture-independent approaches will provide new opportunities to dissect the dynamic host-microbiome relationship.

With that context, this chapter provides an overview of the existing and emerging isolation strategies that can be readily applied in laboratories with access to the types of facilities and equipment routinely used in gut microbiology. In particular, the influence of nutrients on the culturability of fastidious gut bacteria and archaea is discussed, as well as how emerging meta-omic, genetic and antibody-based strategies could be used to bring the gut microbiome to life.

Culturing methods and nutrient effects on microbial growth: an overview

This chapter focuses on advances in bacterial and archaeal cultivation but readers interested in the gut mycobiome are referred to the works by Theodorou et al.¹²²¹ and Haitjema et al.¹²³¹ which provide comprehensive overviews of the approaches used to isolate, cultivate and preserve anaerobic fungi. Despite significant advances in the cultivation of fastidious gut bacteria, archaea and fungi, the propagation of gut-derived phage and protozoa under laboratory conditions continues to remain challenging. Interested readers are referred to Gilbert and Klieve¹²⁴¹ and Newbold et al.^12S| respectively, for overviews of our current understanding of ruminal phage and protozoal biology.

The works published by McSweeney et al.¹²⁶¹ and Joblin^|27] provide well- structured and comprehensive overviews of the approaches and methods widely used to isolate and propagate fastidious bacteria and archaea from the rumen and other gut environments. The contemporary techniques widely used in anaerobic microbiology are still predominantly based on the approaches developed and described by Hungate, Bryant and their colleagues¹²⁸-²⁹¹ and summarised by Stewart et al.^130|The recipes described by McSweeney et al.¹²⁶¹ and their derivatives are still widely used by the research community and have consistently been shown to support the nutritional requirements of a diverse range of gut bacteria. Most microbiologists will be familiar with the methods used to prepare microbiological media and the use of anaerobic jars. Here, media is typically prepared, inoculated under oxic conditions and then transferred to an anaerobic jar where a reductant is used to generate anoxic conditions. In contrast, the methods and underlying theory used to prepare anoxic media for the more fastidious gut microbes will largely be unfamiliar to the uninitiated but can be quickly mastered with an experienced mentor. Many laboratories that routinely culture anaerobic gut microbes use anaerobic chambers to isolate and propagate anaerobic microbes. However, while anaerobic chambers provide increased flexibility and consistency between experiments it is important that strict work practices are adhered to so as to prevent oxygen contamination of the internal chamber environment. This includes where possible ensuring reagents and consumables are oxygen free before entry into the chamber.

The typical components of anoxic microbiological media include a source of nitrogen and carbon macronutrients (e.g. peptone, glucose), a source of essential micronutrients (e.g. yeast extract, trace metals solution, vitamin solution), a mineral salt solution to buffer the medium against pH changes, an oxygen indicator and a reductant. Media are usually prepared by mixing the individual components in a large autoclavable vessel with the exception of any heat-labile ingredients and the reductant. Since oxygen is poorly soluble in water, the medium is next rapidly deoxygenated by boiling and then cooled by bubbling with a constant stream of anoxic gas (typically, carbon dioxide or nitrogen), with aluminium foil wrapped around the mouth of the vessel enabling a build-up of positive pressure and preventing entry of atmospheric oxygen into the vessel. Once cooled, the pH is adjusted to the desired range and a reductant is added to titre the effects of any subsequent oxygen contamination. Finally, the medium is aliquoted as required in an anaerobic chamber and autoclaved. Sterile, heat-labile components can be added as required once cooled.

An alternative approach involves the preparation of select medium components separately before combining post-sterilisation in an anaerobic chamber to avoid the formation of Maillard and other toxic products that typically occur between sugars and amino acids during conditions of high heat and temperature.¹³¹¹ The phosphate salts typically used in buffering salts can also result in the formation of toxic products and inhibition of microbial growth, and media with reduced concentrations can support the recovery of greater microbial diversity including novel isolates.^132-341 Similarly, gelling agents can impact microbial growth and while bacteriological agar is typically used due to its low cost, other alternatives including gellan gum can also improve the recovery of novel microbial diversity.^134-361

Habitat-simulating media have been widely used to enable isolation of fastidious gut microbes.¹³³-³⁷¹ These media support growth of many diverse microbes however a disadvantage is that less abundant or slow-growing subdominant microbes are often out competed by fast-growing microbial 'weeds'.¹³⁸¹ Additional complicating factors are the challenges associated with producing microbial cultures in solid or liquid media. Many microbes, including from gut environments, are auxotrophic and their growth in liquid media is facilitated by cross-feeding with external nutrients that can satisfy their growth requirements.^139-411 In contrast, growth as distinct axenic colonies suggests prototrophy or that the medium can support their nutritional requirements. Taken together, there is much interestin rationally developing new formulations that are permissive for the growth of target microbes but inhibitory for nontarget microbes. In the following subsections, the effect of select macro- and micro-nutrients on microbial growth is briefly discussed.

Nitrogen sources

Much of the organic nitrogen in the rumen is sequestered in ammonium and protein biomass. Ammonia is the principal source of nitrogen for rumen microbes with many bacteria displaying specific amino acid auxotrophies or preferences, and are stimulated by the addition of specific amino acids (e.g. leucine, glycine) or peptides.^142-441 Rumen microbes have thus evolved specialised strategies to facilitate the release of this nitrogen and its assimilation in the growing cell. The main source of peptide breakdown in the rumen is via the activity of dipeptidyl peptidases which results in the production of dipeptides that are then further catabolised by dipeptidases.¹⁴⁴¹ It is well- recognised that rumen microbes exhibit specific preferences for nitrogen. For instance Pittman and Bryant¹⁴⁵¹ demonstrated that Prevotella ruminicola (then classified as Bacteroides ruminicola) exhibits a preference for peptide and ammonia nitrogen but does not utilise free amino acid nitrogen or nitrogen from a range of other low-molecular weight sources. Similarly, Synergistetes strain MFA1, Prevotella bryantii B14, Selenomonas and Streptococcus spp. exhibit a preference for peptides over amino acids.¹⁴⁶-⁴⁷¹ In contrast, other bacteria including Clostridium aminophilum, Peptostreptococcus anaerobius and Fibrobacter spp. exhibit a preference for amino acids.¹⁴⁸-⁴⁹¹ As expected, nitrogen sources can impact the growth of specific organisms with peptone sources affecting both the growth rate and cell characteristics of specific bacteria^150-521 while free amino acids can either promote or inhibit¹⁵³¹ the growth of select microbes with implications for their choice in media.

Carbon sources

Carbohydrates are the predominant source of energy for most gut microbes. Carbohydrates are structurally diverse and as expected gut microbes vary in their ability to utilise them to support growth with many microbes exhibiting specific carbohydrate preferences. McSweeney et al.¹²⁶¹ described a range of media that are selective for cellulolytic, xylanolytic, pectinolytic and amylolytic microbes and these or similar media are still widely used. In the gut, the influence of carbohydrates on microbial growth is influenced by host diet and the extent of host and microbial digestion. Microbes exhibit a hierarchy of carbohydrate utilisation with simpler carbohydrates often preferentially utilised. In the gut, the primary carbohydrate-degrading microbes catabolise the more complex carbohydrates releasing simpler oligosaccharides or monosaccharides to other microbes. These carbohydrate preferences and growth impacts have been widely used to isolate microbes. For instance, rapid transfer of enrichment cultures with complex carbohydrates facilitates enrichment of primary degraders at the expense of secondary degraders that are increasingly diluted. While carbohydrate-based enrichments were historically empirically determined, the advent of whole genome sequencing and custom databases (e.g. dbCAN2¹⁵⁴¹ and SACCHARIS¹⁵⁵¹) has enabled a more selective inclusion of carbohydrates to enrich for and isolate target microbes.

Many gut microbes can switch to asaccharolytic or proteolytic fermentation including clostridia, peptostreptococci and other microbes,¹⁴²-⁵⁶¹ and acquire their carbon from non-carbohydrate sources. The gut is also colonised by other microbes that are capable of carbohydrate-independent growth including reported asaccharolyitc microbes (e.g. Phascolarctobacterium spp.¹⁵⁷-⁵⁸¹ and Catenibacillus scindens^[S9]). Also, acetogens (e.g. Eubacterium limosum,^l60] tammar wallaby isolate TWA4¹⁶¹¹) and methanogenic archaea can use noncarbohydrate substrates with the latter capable of utilising a diverse array of substrates including carbon dioxide, formate, methanol and methylamines.¹⁶²¹ Thus, microbes can utilise a diverse array of carbon sources and their specific preferences can impact their cudurability in the laboratory.

Micronutrients

Many gut microbes display absolute nutritional requirements for specific micronutrients including vitamins, lipids, transition elements and signalling molecules.¹²⁰-⁶³¹ The importance of vitamins and trace metals as micronutrients is well recognised and where necessary these are typically supplied from standard stock solutions.¹²⁶¹ Many gut microbes exhibit complex micronutrient requirements that may at least partially explain their resistance to growth on laboratory media. In addition, micronutrients are typically provided in trace amounts and it can be difficult to predict the specific requirements of a microbe. Consequently, the inclusion of cell-based extracts (e.g. yeast extract, beef extract) and for many fastidious gut microbes, rumen fluid¹²⁶¹ and faecal waters,¹⁶⁴¹ act as a source of micro-nutrients and also help simulate a more physiologically relevant environment.

The ability of straight and branched chain-volatile fatty acids to stimulate growth of gut bacteria is also well known.¹⁶⁵-⁶⁶¹ Non-volatile fatty acids also stimulate growth with the growth and ability of Ruminococcus albus 8 to digest cellulose in a chemically defined medium enhanced by the addition of phenylpropanoic acid to the medium.¹⁶⁷¹ R. albus 8 also utilises phenylacetic acid to synthesise phenylalananine and the rate of cellulose digestion following its addition to defined medium is comparable to that in medium supplemented with rumen fluid.¹⁶⁸¹ Carbon dioxide also stimulates the growth of important gut bacterial species.¹⁶⁹-⁷⁰¹ While methanogens typically use hydrogen for methane production, there is also evidence of adaptations to low-hydrogen environments with Methanosphaera sp. WGK6 capable of utilising ethanol as an alternative to hydrogen to fuel methanogenesis.¹⁵¹ Orthologs of the alcohol and aldehyde dehydrogenase genes suspected of underpinning this capacity are found in other methanogenic species suggesting this may not be a unique phenomenon. Several studies have reported isolation of many environmental isolates in diffusion chambers likely due to the in situ production of essential growth factors. Using a similar approach, Lewis and colleagues¹⁷'¹ screened densely plated bacteria to identify slow-growing colonies whose growth was dependent on proximally located faster-growing colonies. Using this approach, they identified novel Faecalibacterium sp. and Sutterella sp. whose growth was dependent on the production of specific quinones by helper bacteria. Purified quinones enabled growth in the absence of helper cells although select strains exhibited quinone specificity. Capitalising on this approach, Lewis and colleagues¹⁷²¹ used a similar approach to identify an isolate that grew in the presence of y-aminobutyric acid (GABA) producing helper bacteria, or in the absence of the helper bacteria when supplied with GABA. Vartoukian et al.^!731 also showed that supplementation of medium with siderophores resulted in the cultivation of novel strains from the oral microbiota.

Micronutrients are typically supplied in trace amounts and Tramontano et al.¹⁵³¹ showed that specific medium components and microbial metabolites could also potentially inhibit growth in a study that examined 90 human gut bacteria. Thus, while ensuring that important nutrients are provided it is also critical that their concentrations do not inhibit growth.

Reductants

The vast majority of microbes that colonise the gut of ruminants and other mammals are strict anaerobes, and although some can tolerate exposure to atmospheric conditions (e.g. Bacteroides spp., Prevotella spp.), growth typically occurs only under anoxic conditions. Cysteine-HCI is the reductant of choice for many studies due to ease of use however titanium-based reductants are also widely used due to their greater redox potential.¹²⁶-⁷⁴-⁷⁵¹ In a challenge to this paradigm, Raoult and colleagues¹⁷⁶¹ demonstrated that the anaerobic bacteria Ruminococcus gnavus and Fusobacterium necrophorum could be grown under atmospheric condition when the medium was supplemented with the reductants ascorbic acid and/or glutathione. In a follow-up study, Dione et al.^!771 further extended this work and demonstrated that a diverse set of 623 bacteria including 82 strictly anaerobic and 9 microaerophilic species could be grown in a medium supplemented with ascorbic acid and glutathione.¹⁷⁸¹ The addition of uric acid, haemin and a-ketoglutarate further improved culturability of the microbes such that only the Mycobacteria failed to grow under the conditions examined. These observations have also been extended to extremely oxygen- sensitive methanogenic archaea.¹⁷⁹-⁸⁰¹ While these advancements have not been widely adopted they have been used to culture strict gut anaerobes including Dysosmobacter welbionis^[8'^] and they could be used to enable more routine processing of fastidious anaerobes under oxic environments.

In summary, there are a wide range of factors that influence the breadth and depth of microbes that can be cultured. The use of low nutrient medium and prolonged incubation times can support the growth of a broader range of fastidious bacteria potentially by preventing the overgrowth by competing

A question of culture: bringing the gut microbiome to life in the omics era

36

microbes or reducing the concentration of growth inhibitors.¹²⁰-^82-851 Culturomics has built on traditional culturing approaches by greatly extending the range of media and growth conditions typically tested resulting in the isolation of many novel gut microbes (Fig. 1a¹⁸⁶-⁸⁷¹). However, the impact of the macro- and micronutrients, and other variable growth conditions, on microbial culturability has not been systematically explored although several recent studies have outlined strategies to address this challenge (e.g. Taguchi arrays¹⁸⁸¹, factorial design), and these will provide both new insights into the metabolic requirements and growth potential of existing isolates,¹⁷⁸-⁸⁹¹ and support the growth of fastidious gut microbes ex vivo thus providing opportunities to bring previously uncultured microbes to life.¹⁹⁰-⁹¹¹ These efforts will also be supported by the development

Figure 1 (a) With traditional microbial isolation approaches, a limited number of media and growth conditions are tested based on historical reports, empiric data and the experience of the practitioner. These typically result in limited diversity and are at best moderately selective for target microbes. Culturomics-based isolation builds on these traditional approaches but explores a broader range of media and growth conditions (n) which can significantly increase the number of novel microbes recovered. Culturomics captures greater microbial diversity although it is not typically directed towards target microbes, (b) With genome-directed isolation, genomic data from target microbes are analysed to predict their nutritional and growth preferences and then custom media are developed. These media result in low recoverable microbial diversity as they are typically permissive for growth of the target but not non-target microbes.

of small colony-picking robots that have allowed for the first time automated scalable colony picking in anaerobic chambers.