The dominant or abundant bacteria in environments commonly exhibit high environmental adaption and cell activity (Rivett and Bell 2018). If a natural environmental sample, such seawater, is used as a growth medium, combined with diluting a sample "to extinction" (Rappe et al. 2002), abundant bacteria would theoretically be able to be isolated and cultivated (Fig. 2). In fact, approaches based on simulated environments have resulted in many abundant bacteria being isolated from culturable marine environments (Table 1) (Stewart 2012). One of the best-known cases is the isolation and culturing of the SAR11 clade (Pelagibacterales) (Rappe et al. 2002). Based on this strategy, members of the microbial majority, including SAR116, OM60/NOR5, SAR92, and Roseobacter, were isolated and cultured (Henson et al. 2016). Nonetheless, it is difficult to replicate a natural environment at an arbitrarily high level of fidelity if parameters for the growth of a given bacterial taxon are unknown. Therefore, one alternative is to take the bacteria back to the environment to grow them, often by moving a portion of the environment sample into the laboratory (Stewart 2012). Based on this theory, some in situ culture methods have been developed and have isolated and cultured many previously uncultured bacteria (Table 1) (Kaeberlein et al. 2002; Nichols et al. 2010). Another method for isolating dominant bacteria in their environment has been developed based on gel microdroplets. Indeed, encapsulated gel microdroplets can be incubated in natural environments as a growth medium (Zengler et al. 2002). By using this method, some uncultured bacteria have been isolated as microcolonies in gel microdroplets (Table 1), including lineages of Planctomycetales that were previously uncultured (Zengler et al. 2002).
Figure 2. Potential strategies for culturing different types of uncultured bacteria. Dominant bacteria and rare bacteria refer to active taxa that live well and undergo cell division in natural environments. Dormant bacteria refer to bacteria that have poor adaption to local environments with low metabolic activity. For culturing active bacteria, the core strategy is simulating natural environments. The dilution method is suitable for further isolation of dominant bacteria, and culturomics and enrichment culture are two approaches for further isolating rare active bacteria. For culturing dormant bacteria, the core strategy is "wake up", and resuscitation culture is assumed to be a feasible approach. The key to growing rare or dormant bacteria is "change": culture conditions should be changed, so as to make the "rare" be "dominant", and to convert "dormancy" to "resuscitation". Various approaches can be invoked, with important innovation
Types of uncultured Strategy Cultured case Approach References Dominant bacteria Simulate environments SAR11 Using seawater as a growth medium and dilution-to-extinction Rappe et al. (2002) SAR11, SAR116, OM60/NOR5, SAR92, Roseobacter Using artificial seawater media and dilution-to-extinction Henson et al. (2016) Cytophaga-Flexibacter-Bacterioides (CFB) group
In situ culture (take the bacteria back to the environment) Kaeberlein et al. (2002); Nichols et al. (2010) SAR11, Cytophaga–Flavobacterium–Bacteroides group, SAR116 clade, and Planctomycetales Gel micro-droplets Zengler et al. (2002) Psychrobacter Co-culture Nichols et al. (2008) Faecalibacterium, Bacteroides, Bilophila, Gordonibacter, and Sutterella Co-culture Fenn et al. (2017) Firmicutes KLE1738
In the Health's "Most Wanted" list
Co-culture Strandwitz et al. (2019) Rare bacteria Culturomics (directly isolate) 247 new prokaryote species in the gut Based on 212 different culture conditions and dilution Lagier et al. (2016) Enrichment culture (increase the abundance of rare bacteria) Bradymonadales, Marinilabiliales Enrichment culture with Low-nutrient medium containing 10 mM sodium pyruvate Mu et al. (2018) 79 strains of Planctomycetes Enrichment culture based on N-acetylglucosamine and antibiotics selection Wiegand et al. (2020) Kiritimatiellaeota Enrichment culture Spring et al. (2016); van Vliet et al. (2019) ammonia-oxidizing bacteria Continuous enrichment culture Bollmann and Laanbroek (2001) Dormant bacteria Resuscitation culture Cellulose-degrading bacterial community Rpf protein Su et al. (2018) Vibrio cholerae Autoinducers Bari et al. (2013) Marinilabiaceae, Tangfeifania, Draconibacterium, Saccharicrinis Enrichment culture Mu et al. (2018)
Table 1. The approaches and strategies for isolating different types of uncultured
Co-culture, a strategy that focuses on unknown biotic factors that may be effective in isolating uncultured organisms, has been proposed as another approach to simulating an environment. In fact, co-culture dependence has been recognized since the 1900s, the classic example that is used in microbiology classes being the dependence of Haemophilus influenzae on Staphylococcus aureus (Davis 1921). H. influenzae was found to need an exogenous source of both heme and NAD for aerobic growth. In 1919, Fleming found that the dependence of H. influenzae on S. aureus can be overcome by preheating the blood (creating the oddly named "chocolate" agar); this heat treatment both releases heme and inactivates the enzyme that breaks down NAD (Fleming 1919). Recently, the groups of Epstein and Lewis noticed that some organisms forming colonies in a diffusion chamber can grow on a Petri dish but only in the presence of other species from the same environment (Kaeberlein et al. 2002; Nichols et al. 2008). In a following study, they found that short peptides (Nichols et al. 2008), siderophores (D'Onofrio et al. 2010), quinones (Fenn et al. 2017), and γ-aminobutyric acid (Strandwitz et al. 2019) were key biotic factors for culturing "uncultivable" microorganisms (Table 1). In addition to growth factors, "helper" microbes can support other forms of biotic factors to enhance the growth of an "uncultivable" microorganism. Some "helper" microbes may help other microorganisms to remove toxic compounds, such as by secreting catalase to scavenge reactive oxygen species. For example, the ubiquitous actinobacterial acI lineage was cultured by supplying the biochemical "helper" catalase (Kim et al. 2019).
To culture these low-abundance bacteria, one strategy is to tailor different synthetic media for culturing bacteria that adapt to specific environmental niches (Fig. 2). Baas Becking's dictum that "everything is everywhere, but the environment selects" (de Wit and Bouvier 2006) also suggests that different environments can be designed to select different bacteria. Based on this strategy, culturomics was developed and it has achieved great success in culturing previously uncultured members of the human gut microbiota (Table 1) (Lagier et al. 2016). In fact, the original environment was inadvertently altered when microbes were isolated. For example, marine broth 2216 was used to simulate marine environments, however, there are large differences between sea water and marine broth 2216, such as differences in concentrations of sulfur, carbon, and nitrogen (Henson et al. 2016). This may be one reason why when using marine broth 2216, most cultured bacteria have been rare species (Mu et al. 2018).
Another strategy for the isolation of low-abundance bacteria is attempting to make them dominant (Fig. 2), whereby the above approaches for culturing dominant bacteria can be applied. Enrichment is a long-used and common practice that effectively increases the populations of target organisms. This is generally achieved by introducing nutrients or environmental conditions that only allow the growth of the organism of interest. At the end of the nineteenth century, Sergei Winogradsky manipulated the concentration of ammonium salts, pH and determined the effect of the added organic material; they obtained a medium in which nitrification proceeded rapidly, and eventually isolated nitrifying microorganisms (Dworkin 2012). This was, in effect, the invention of enrichment, a technique that has proven to be a powerful tool for the isolation of specific nutritional or physiological types of microorganisms. By employing this approach, many rare, uncultured microorganisms have been isolated or highly enriched (Table 1). Sandra Wiegand et al. (2020) isolated 79 strains of the poorly represented and slow-growing phylum Planctomycetes by employing a developed enrichment strategy based on their preference for N-acetylglucosamine and resistance to certain antibiotics. Strains of the novel phylum Kiritimatiellaeota, which is a member of the Planctomycetes-Verrucomicrobia-Chlamydiae (PVC) superphylum, were isolated from saline lakes and marine sediments by an anaerobic enrichment method (Spring et al. 2016; van Vliet et al. 2019). Also, sublineage I Nitrospira were isolated by a combined method: selective enrichment using a continuous feeding bioreactor to increase the cell abundance, then purification followed by sub-cultivation via a cell sorting system by focusing on the unique characteristics of the target bacteria (Fujitani et al. 2014). Additionally, our own examinations isolated one strain within phylum Kiritimatiellaeota (16S rRNA gene similarity of less than 85% to the type strain accession number: MN909984) by anaerobic enrichment culture. The isolation of some bacteria displays a dependence on enrichment culture. For instance, Marinilabiliales [an order within the phylum Bacteroidetes (Wu et al. 2016)] contains approximately 50 species; more than 70% of these have been isolated by enrichment treatment (Mu et al. 2018). Enrichment is also a process of selection, as anaerobic enrichment can significantly decrease the abundance of obligate aerobes and enrich facultative anaerobes and obligate anaerobes. If enriched microbes are cultured on plates under aerobic conditions, enriched obligate anaerobes cannot grow, while the enriched facultative anaerobes can be isolated. This approach is efficient for isolating facultative microbes in marine sediments or intertidal zone sediments, such as some marine bacteria in the phylum Bacteroidetes (i.e., Bacteroidetes). Most members of Marinilabiliales are facultative microbes and can be enriched and isolated by employing this approach (Table 1) (Ben Hania et al. 2017; Mu et al. 2018). In addition, continuous enrichment culture using oligotrophic medium can selectively enrich slow-growing microorganisms (Table 1), for example, the isolation of ammonia-oxidizing bacteria (Bollmann and Laanbroek 2001). The approach of long-term continuous enrichment culture of archaea may also be employed in culturing slow-growing bacteria (Imachi et al. 2020).
Although the simulated environment method has been a successful approach for isolating bacteria, especially for dominant species in an environment, it has some limitations for isolating dormant species, as natural environments do not commonly have the optimum conditions for the growth of most bacteria. In fact, many bacteria are dormant in natural environments and some strains are VBNC bacteria (Mu et al. 2018). In addition, a scout model of the microbial life cycle in dormant cells and activity cells was proposed, which postulated that dormant cells were able to stochastically wake into activity in situ (Epstein 2013). It is possible that scout cells can accrue and then induce the remaining dormant cells to activity if they detect suitable environmental factors. Otherwise, the scout may not accrue and form a population in situ. As a result, the molecules in the environment that serve as a 'wake-up call' and stimulate the growth of the scout and dormant cells should be determined.
Resuscitation-promoting factor (Rpf) was the first-reported protein that revived dormant gram-positive cells and increased the growth rate of vegetative cells (Mukamolova et al. 1998); a family of related growth factors was identified in further studies (Kell and Young 2000). Rpf was later demonstrated to have a lysozyme-like structure and muralytic activity (Cohen-Gonsaud et al. 2005). One potential resuscitation mechanism of Rpf is remodeling of the peptidoglycan in the cell wall of dormant cells, generating muropeptides that may serve as a 'wake-up call', stimulating the growth of dormant cells (Kana and Mizrahi 2010). Su et al. (2018) enhanced the cellulose-degrading capability of the bacterial community in composting by using the Rpf protein to resuscitate viable but nonculturable bacteria (Table 1). In gram-negative bacteria, gene expression regulator autoinducers (AIs), with the ability to stimulate growth rates, have also been discovered (Freestone et al. 1999). Unculturable Vibrio cholerae cells were found to be resuscitated by AI-2 in aquatic reservoirs (Bari et al. 2013). In addition to purified autoinducers, the supernatant of growing cells in the late logarithmic phase has a positive effect on bacterial resuscitation (Pinto et al. 2011). Other biotic molecules, such as siderophores (Lankford et al. 1966) and pyruvate (Morishige et al. 2013; Mu et al. 2018) also promote exit from dormancy. It has been suggested that uncultured bacteria only commit to division in a familiar environment, which they recognize by the presence of growth factors released by their neighbors (D'Onofrio et al. 2010). As a result, co-culture or mixed culture is one way to 'wake up' dormant bacteria, though the actual resuscitation factors have yet to be determined (Table 1). Compared with pure cultures, enrichment culture is a mixed culture system that involves competition, cooperation, or coordination among bacterial communities (Imachi et al. 2020; Mu et al. 2018). In addition, this approach might culture "uncultured" bacteria not only by enriching the abundance of "uncultured" strains but also through the resuscitation mechanism (Mu et al. 2018).
For dominant active bacteria
For rare active bacteria
For dormant bacteria
It is impossible to culture all microbes from marine environments. As a result, it is necessary to set specific, achievable, and relevant cultivation goals. A greater effort has been made to isolate the most-wanted or key players from the marine environment. As stated by Paul Carini (Carini 2019), key bacterial players may include those that (1) have a high relative abundance, (2) play key role in biogeochemistry or bioremediation, (3) have the potential to produce natural products, and (4) substantially diverge from cultured taxa. Following Paul Carini's suggestion, Table 2 lists some marine bacteria that might be defined as key, but uncultured players in marine environments. Some key players were found to have various functions, such as Pelagibacterales (initially, this taxon was known solely by metagenomic data and known as the SAR11 clade), a group of small, carbon-oxidizing bacteria that reach a global estimated population size of 2.4 × 1028 cells. In both the euphotic zone and the deeper ocean, Pelagibacterales cells oxidize many labile organic compounds and produce CO2 as well as other volatile organic compounds (i.e., dimethyl sulfide and methanethiol) that can enter the atmosphere, playing a key role in global biogeochemistry cycling (Giovannoni 2017).
Candidate organism or group Biotopes Phylum Reason for cultivation "CandidatusAtelocyanobacterium thalassa" (Martinez-Perez et al. 2016; Thompson et al. 2012) Surface seawater Unicellular cyanobacteria group A (UCYN-A) The most abundant of the unicellular diazotrophs in oceans and key role in biogeochemistry Methylphosphonic acid aynthesis bacteria (Metcalf et al. 2012) Seawater Assorted Key roles in methane generation in the aerobic ocean Dimethylsulfoniopropionate (DMSP) synthesis bacteria (Curson et al. 2017) Seawater Assorted Key roles in DMSP generation in the aerobic ocean SAR202 (Mehrshad et al. 2018) Seawater Chloroflexi Abundant in mesopelagic waters and dark ocean SAR86 (Dupont et al. 2012) Seawater Gammaproteobacteria Abundant in surface waters and key role in biogeochemistry SAR324 or Marine Group B (Sheik et al. 2014) Seawater Deltaproteobacteria Ubiquitous in the dark ocean "Candidatus Actinomarinidae" (Ghai et al. 2013) Seawater and sediment Actinobacteria (OM1) Streamlined genome and key role in biogeochemistry "Candidatus Marinimicrobia" (Bertagnolli et al. 2017) Seawater, oxygen minimum zones (OMZs) Candidate phylum marine group A Abundant and key role in biogeochemistry "Candidatus Atribacteria" (Nobu et al. 2016) Sediment Candidate phylum Atribacteria (OP9/JS1) Key role in biogeochemistry Woeseiaceae/JTB255 (Mussmann et al. 2017) Surface Sediment Gammaproteobacteria Abundant in surface sediments and key role in biogeochemistry "Candidatus Electrothrix" and "Candidatus Electronema" (Kjeldsen et al. 2019; Muller et al. 2020) sediment Deltaproteobacteria Cable bacteria and key role in biogeochemistry Assorted bacteria (Boetius et al. 2000) sediment Proteobacteria Anaerobic oxidation of methane by sulfate 'Candidatus Methylomirabilis oxyfera' (Ettwig et al. 2010) sediment Candidate division NC10 Anaerobic oxidation of methane by nitrite Assorted bacteria (Beal et al. 2009) sediment Assorted Anaerobic oxidation of methane by iron and manganese "Candidatus Entotheonella" (Wilson et al. 2014) sponge "CandidatusEntotheonella" With large and distinct metabolic repertoire for ecological studies and drug discovery Any representative of the candidate phylum radiation Assorted Assorted Divergent from all cultured bacteria
Table 2. Examples of marine bacteria that are most-wanted in culture
Many more key players remain uncultured and their ecological roles have only been predicted by metagenomics analysis. However, phenotypic properties cannot always be predicted from sequence information and their exact ecological functions need to be further identified by isolation. The Woeseiaceae/JTB255 group are among the most abundant bacteria at the surface of coastal, abyssal, and bathyal sediments. The global estimated population of these bacteria might reach of 5 × 1026 cells (Hoffmann et al. 2020). Woeseiaceae/JTB255 cells appear to have a higher biovolume (on average 0.13 µm3) compared with the water column clade SAR11 [0.01 µm3; (Rappe et al. 2002)] and thus may account for a considerable fraction of microbial biomass in the oceans. Metagenomics analysis has shown that some representatives of Woeseiaceae/JTB255 might be involved in sulfur oxidation, carbon fixation (Mussmann et al. 2017) and cycling of a major class of refractory sediment organic matter (Hoffmann et al. 2020). However, the solo cultured strain Woeseia oceani XK5T in this group did not exhibit sulfur oxidation or carbon fixation abilities. Simulating sediment environments might be useful for isolation of Woeseiaceae/JTB255 bacteria. Nevertheless, more strains need to be isolated to determine the key factors required for culturing this group. The unicellular cyanobacteria group A (UCYN-A) is another uncultured group that is abundant and potentially significantly contributes to N2 fixation in the surface waters of oceans (Martinez-Perez et al. 2016). Unlike the uncultured Woeseiaceae/JTB255 group, these bacteria display unprecedented genome reduction and a symbiotic association with a unicellular prymnesiophyte, which might be one reason for their unculturability. In addition, as with the hosts that live in marine environments, an increasing focus is being placed on bacteria from the human microbiome, together, the potential role of host-associated unculturable bacteria is being investigated. "Candidatus Entotheonella" is a typical host (sponge)-associated marine unculturable bacteria. These bacteria are filamentous symbionts that produce almost all known bioactive compounds derived from the Lithistida sponge Theonella swinhoei (Lackner et al. 2017; Wilson et al. 2014). Genome analysis has shown that these bacteria are auxotrophic for multiple amino acids, suggesting that providing free amino acids would be crucial for cultivating them (Liu et al. 2016).
Importantly, the 'reverse genomics' approach should be considered for targeted culturing of key players (Cross et al. 2019). With this method, antibodies against predicted membrane proteins can be used to target and culture microbial cells from a specific taxonomic group (Cross et al. 2019). However, the efficiency of methods for isolating uncultured microbes from environments that harbor more complicated microbiota should be further explored. Moreover, follow-up culturing of the targeted strains is still associated with the above difficulties (Lewis and Ettema 2019).