Culturing the human microbiota and culturomics

Article metrics


The gut microbiota has an important role in the maintenance of human health and in disease pathogenesis. This importance was realized through the advent of omics technologies and their application to improve our knowledge of the gut microbial ecosystem. In particular, the use of metagenomics has revealed the diversity of the gut microbiota, but it has also highlighted that the majority of bacteria in the gut remain uncultured. Culturomics was developed to culture and identify unknown bacteria that inhabit the human gut as a part of the rebirth of culture techniques in microbiology. Consisting of multiple culture conditions combined with the rapid identification of bacteria, the culturomic approach has enabled the culture of hundreds of new microorganisms that are associated with humans, providing exciting new perspectives on host–bacteria relationships. In this Review, we discuss why and how culturomics was developed. We describe how culturomics has extended our understanding of bacterial diversity and then explore how culturomics can be applied to the study of the human microbiota and the potential implications for human health.


The importance of the gut microbiota in human health is currently in the spotlight. The composition of the gut microbiota has been linked to metabolic disorders1 and inflammatory bowel diseases2 for over a decade; however, recent advances in the field of oncology3 and the discovery of the gut–brain axis3 have broadened our understanding of the spectrum of diseases that are associated with changes in our gut microbiota. Surprisingly, interest in commensal microorganisms that have mutualistic relationships with humans and their roles in protecting us from external pathogens or contributing to metabolic pathways was neglected for a long time, despite the pioneering work of Metchnikoff, who first described the beneficial effects of probiotics4. The development of omics technologies, beginning with metagenomics, highlighted the role of gut bacteria in several disorders. However, these genomic technologies have provided a limited perspective as they cannot easily detect minority populations5. Intriguingly, they revealed that ~80% of the bacteria inhabiting the human gut were unknown and were thus considered unculturable at that time6. Concomitantly, environmental microbiologists developed new approaches to culture bacteria that cannot be cultured using conventional techniques, thus initiating the rebirth of culture in microbiology7,8. This progress motivated the development of culturomics for discovering unknown bacteria that reside in the gut. The use of multiple culture conditions combined with prolonged incubation has permitted the isolation of hundreds of new bacterial species from the gut in less than 5 years9. In addition to increasing our knowledge of the repertoire of bacterial species, culturomics can provide bacterial strains that can be used for in vitro experiments and to confirm the role of specific bacterial species in disease pathogenesis using animal models. Furthermore, some bacterial species that are identified using culturomics may provide health benefits and therefore could represent future candidates for bacteriotherapy as probiotics. In this Review, we first discuss why and how culturomics was developed. We describe how culturomics has extended our understanding of bacterial diversity and then explore how culturomics can be applied to study the human microbiota and its potential benefits for human health.

Microbiota and metagenomics

Metagenomics was first used to comprehensively describe the composition of complex environmental microbial ecosystems, such as the pioneering work that characterized microbial populations in sea water10. After seminal studies described the diversity of the gut microbiota from stool samples or colonic biopsies from healthy individuals6,11,12, further studies of bacterial diversity within the human gut led to the realization of relationships between the gut microbiota and obesity, diabetes mellitus, inflammatory bowel diseases and colorectal cancer13. Metagenomics has been widely used to identify microorganisms that are associated with a physiological state or with disease. However, the challenge of discriminating between species because of short read lengths14 makes their precise taxonomic assignment difficult. This challenge potentially explains why the most robust associations that have been observed between bacterial species and the human gut concern species that are unique in humans among their genus. Nevertheless, recent developments in sequencing have enabled identification beyond the species level (that is, the strain level), but only for dominating populations15. Major advances in bioinformatics have also enabled the analysis of entire sequence data sets using similarity algorithms or clustering tools without taxonomic assignment14. However, further developments are still needed, including, for example, the optimization of online tools that shorten the amount of time it takes to analyse sequencing data and reduce the heterogeneity in taxonomic assignment between different laboratories16,17. Beyond taxonomic identification, shotgun sequencing of microbial DNA enables functional annotation of genes and thus informs the functional potential of the community18. However, despite these advances, some limits have arisen because of differences in protocols between laboratories, particularly concerning DNA extraction or the choice of methods for bioinformatics analyses. Moreover, high-throughput sequencing methods are particularly susceptible to depth bias owing to the difficulty in detecting microorganisms at very low levels19 (Fig. 1). Several sequencing strategies have been developed to enhance the detection of minority populations when there is sufficient read depth (also called sequencing coverage)20,21, but it is difficult to evaluate how effective these methods are, as comprehensive analyses by both culture and high-throughput sequencing methods are rarely performed. Nevertheless, two studies revealed a substantial overlap between cultured and uncultured samples22,23. Despite improvements in sequencing depths, considering that stool samples contain ~1010–1011 bacteria per gram of stool and that culture is able to detect 102 bacteria per gram of stool5, 16S ribosomal RNA (rRNA) gene sequencing should generate at least 108–109 reads to achieve a comparable sensitivity to culture, which, to the best of our knowledge, has not yet been achieved. Aside from the depth bias, many sequences remain unassigned, corresponding to the microbial dark matter24. Unlike culture techniques, metagenomics cannot determine whether a bacterium is dead or alive; metagenomics sequences only a portion of the total pool of DNA25,26. Furthermore, it is challenging to determine the activity and physiological state of a microorganism using DNA sequencing. However, substantial advances in bioinformatics have led to the recent development of an algorithm that generates a replication index using metagenome sequencing data from a single time point.

Fig. 1: The strengths and weaknesses of metagenomic studies.

High-throughput sequencing methods have enabled comparisons of large cohorts in record time, as the time it takes to generate results has substantially decreased over the past decade owing to integrative workflows. Sequencing enables uncultured bacteria to be identified and associates microbial signatures with a particular physiological state or disease as the relative abundance of each taxon can be measured. In addition, bacterial species can be grouped without taxonomic assignment. Shotgun sequencing enables the function of microbial communities to be inferred through the analysis of genomes and the coding potential. However, these methods are limited by the heterogeneity of the protocols used. Discrepant results can be obtained depending on the method used to extract DNA or the primers that are used for amplification. The variety of methodologies proposed for bioinformatics analyses (for example, operational taxonomic unit clustering, taxonomic assignment or statistical analysis) can substantially affect the results. Sequencing methods cannot discriminate between live bacteria and transient DNA, and despite recent progress, they cannot easily detect minority populations (for example, bacteria that are present at <105 cells per gram of faeces).

Some of the drawbacks associated with sequencing could be addressed by assessing coding potential alongside gene expression patterns in bacterial populations by integrating metagenome with metatranscriptome data. Similarly, metabolomic and metaproteomic data and culture5,22,23,27 can be integrated with metagenomic data to provide insight into the function of bacterial communities in the gut.


Culturomics is a culturing approach that uses multiple culture conditions, MALDI-TOF mass spectrometry and 16S rRNA sequencing for the identification of bacterial species. Before describing culturomics, we will briefly describe the contribution of environmental and clinical microbiologists, as well as the advent of identification methods such as MALDI-TOF mass spectrometry and molecular tools that led to the development of microbial culturomics.

The rebirth of culture in environmental microbiology

The first studies that initiated the rebirth of culture were performed by environmental microbiologists. For example, a 2007 study mimicked the natural environment of bacteria by incubating environmental samples in a diffusion chamber to increase the number of growing bacteria and the diversity of recovered isolates28. In a recent comprehensive review, among the 14,300 known prokaryotic species, only 2,172 prokaryotes species (~15%) have been isolated at least once from humans, and the other 12,128 (~85%) have been isolated from the environment29. The “great plate count anomaly” is a phrase that was coined by Staley and Konopka in 1985 to describe the difference between the number of cells from natural environments observed using microscopy and the number that form viable colonies on agar medium, which is significantly lower30. Environmental microbiologists have largely studied the basis of this difference to increase the proportion of microorganisms that can be cultured from environmental samples31. To detect minority populations, environmental microbiologists were the first to use dilution culture32, which was subsequently used to study the human gut microbiota23. Dilution culture consists of successive dilutions of the bacterial population to the point of no bacterial growth32. In parallel, environmental microbiologists developed culture methods and media that mimic the natural environment of microorganisms. For example, Candidatus Pelagibacter ubique (previously known as SAR11) was successfully cultured using a mixture of components from sea water33. This species was considered unculturable for a long time, despite the observation that more than one-quarter of the total number of reads were derived from its genome in ocean surface water sequencing surveys. This first culture, as well as the first culture of other members of the SAR11 clade of α-proteobacteria, has been the necessary basis for several further studies that increased our understanding of relationships between marine bacteria34. The use of diffusion chambers in which two membranes allow environmental nutrients to enter but confine bacterial cells has enabled the simulation of natural environments in the laboratory7. Their use has increased the number of cultured colonies 300-fold7. In addition, co-culture has allowed the culture of bacteria from different environments as co-culture can provide growth-promoting factors that are secreted by other prokaryotes35. Finally, the elucidation of interbacterial communication pathways enabled the identification of deficiencies in nutrients and signalling molecules in standard media that are required for bacterial growth, which could then be added to medium formulations36.

The culture of fastidious bacteria by clinical microbiologists

Some microorganisms need specific culture conditions for growth. Among them, fastidious microorganisms will grow only if specific nutrients are present. Similarly, obligate anaerobes are unable to grow in the presence of oxygen.

The revival of culture in clinical microbiology laboratories was initiated by microbiologists that specialize in the culture of intracellular bacteria. For a long time, these bacteria could be cultured only by animal inoculation or the use of embryonated eggs. The use of systems to culture intracellular bacteria that are associated with human diseases was first demonstrated for some Rickettsia species, including Rickettsia felis, and Tropheryma whipplei, the causative agents of rickettsiosis and Whipple’s disease, respectively37,38. The shell-vial technique, including a fundamental step of centrifugation of clinical samples, is an example of a procedure that is used in laboratories specialized in culturing intracellular bacteria39. Later, genome sequencing of isolates obtained in pure culture enabled the identification of metabolic deficiencies and thus the design of axenic media. The design of new axenic media formulations guided by genome sequencing has been demonstrated for Coxiella burnetii (the causative agent of Q fever), T. whipplei and specific Chlamydia species (causative agents of pneumonia and sexually transmitted infections, respectively), facilitating the comprehensive study of intracellular microorganisms40. Easy culturing of these intracellular bacteria on solid media improved our knowledge of their pathogenicity, antibiotic susceptibility and virulence and led to the design of experimental models and new diagnostic tools40.

Anaerobes are dominant members of the healthy human microbiota, especially in the human gut microbiota, but only a limited number cause human infection. Culturing obligate anaerobes is a challenge for microbiology laboratories. Indeed, culturing obligate anaerobes requires specific bacteriological equipment to provide an oxygen-free environment, as well as specialized reagents, including complex media with many supplements; anaerobic culture media must contain carbon sources, macroelements, metals and some growth factors. Animal tissues (brain, meat or heart tissues) were first used in 1890 to supplement culture media and constituted the first nonselective culture media. Egg-based culture was then developed to culture Clostridium and Fusobacterium species. Thioglycolate, a versatile liquid culture medium, has been one of the most widely used media in clinical microbiology. Selective culture conditions have included the use of agents such as antibiotics, bile and dyes (such as crystal violet) or physical processes (such as heat shock).

For growth in the absence of oxygen, a range of systems have been successively designed, such as anaerobic jars, GasPak systems and anaerobic chambers40. Hungate first revolutionized the culture of extremely oxygen-sensitive microorganisms, including sulfate-reducing bacteria and methanogenic archaea41. His method was based on using roll tubes to replace atmospheric oxygen with other gases, such as N2, CO2 or H2. Nevertheless, this technique is time consuming, requires a high level of training and is used in specialized laboratories.

The first culture studies that explored the gut microbiota were performed as early as the 1970s42 and studied how diet can influence the composition of the gut microbiota43. Interestingly, these early studies observed that some bacteria are associated with cancer44, as suggested later32. Overall, Enterobacteriaceae and Veillonellaceae families were the most common among the 400 different cultured species from the human gut microbiota during the 1970s45. At that time, the normal microbiota was thought to be dominated by non-spore-forming, anaerobic, rod-like bacteria. Interestingly, many of the dominant species that were identified in 1995 (ref.46) correspond to the same dominant species that were subsequently detected by metagenomics47.

More recently, although they usually grow in strict anaerobic conditions, Ruminococcus gnavus, Prevotella nigrescens, Fusobacterium necrophorum, Finegoldia magna, Solobacterium moorei and Atopobium vaginae were successfully cultured in an aerobic atmosphere. The authors used Schaedler agar supplemented with high doses of glutathione and ascorbic acid48. Although the mechanism is unknown, the high concentration of antioxidants compared with those used in previous studies is thought to have a role. In a larger study in which uric acid was added to culture media, 276 bacterial species, including 82 strictly anaerobic bacteria, were successfully cultured aerobically49. A 2016 study isolated different strains of methanogenic archaea that are extremely oxygen-sensitive in the presence of Bacteroides thetaiotaomicron, which was used as a hydrogen source in a double-chamber flask50.

These examples exemplify the substantial contribution of clinical microbiologists in culturing prokaryotes associated with humans.

MALDI-TOF mass spectrometry for microbial identification

Sequencing universal genes is crucial for the identification and classification of bacteria. Thus, 16S rRNA gene amplification and sequencing was proposed as a universal bacterial species identification method. Nevertheless, despite the advantages of this molecular technique, it is costly and time consuming, which limits its use in high-throughput culturomic studies. Therefore, a cost-effective and rapid screening method for the identification of bacteria was needed. In 1959, the concept that microorganisms could be identified on the basis of their chemical composition was proposed51. In 1975, specific biomarkers that are possessed by Gram-negative bacteria were identified52, and for three decades it was assumed that matching protein mass spectra against libraries of references using an algorithm would enable us to accurately identify bacteria. However, technical and time limitations hindered the development of mass spectrometry for microbial identification and its use in clinical microbiology53. In 2009, the first use of MALDI-TOF mass spectrometry in a routine clinical microbiology laboratory for accurate identification at both the genus and species levels was reported54. MALDI-TOF mass spectrometry is both time effective and cost effective, and the reproducibility of the results and the fairly low level of user training that is required have increased its use worldwide, so much so that MALDI-TOF mass spectrometry became the reference method for bacterial identification in clinical microbiology55. As the culturomic approach could not exist without rapid identification tools, this breakthrough has enabled the design of culturomic studies to explore complex ecosystems, such as the human gut microbiota.

Culturomics and the human microbiota

Culturomics is a high-throughput culture method that uses MALDI-TOF mass spectrometry to identify bacterial species. In its development, the first objective was to enable the method to provide multiple culture conditions, promoting the growth of fastidious bacteria from the human gut. This result was achieved by improving culture media using blood and rumen fluid in blood culture bottles to promote the growth of minority populations (Fig. 2). In the first culturomic study, 212 culture conditions generated more than 30,000 colonies that were subsequently analysed by MALDI-TOF. Indeed, 341 bacterial species were cultured, including 31 new bacterial species and species belonging to rare phyla, such as Synergistetes or Deinococcus-Thermus5. Interestingly, more than half of these 341 bacteria were identified from the human gut for the first time. An analysis of the results of this study enabled the identification of 70 useful culture conditions, among which 18 were considered as the best for the identification of bacteria from stool samples. The identification of these optimal culture conditions enabled a larger number of samples to be tested in later studies31. Our knowledge of the repertoire of human gut bacterial species was further increased by analysing the literature to identify the weaknesses of the first culturomic studies, identifying the bacteria previously found by other studies but not yet by culturomics. Indeed, efforts were focused on culture conditions for Proteobacteria spp., microaerophilic species and halophilic species as these species were not readily identified using culturomics9. The use of fresh stool or samples collected from different levels of the gut also enabled the culture of a large number of oxygen-sensitive bacterial species. In addition, the detection of microcolony-forming bacterial species was achieved through the use of magnifying glasses9. Overall, the culture condition that yielded the most identifications was the preincubation of samples with rumen fluid and sheep blood in anaerobic conditions9. In parallel with these culturomic studies, a number of other strategies were also developed to identify bacterial species and increase our knowledge of the repertoire of bacteria in the human gut. First, in order to separate single cells to detect minority populations, the ‘dilution to extinction’ technique was employed56. This technique can be applied in microplates for high-throughput cultivation23. This principle enabled the first isolation of two bacteria considered to have clinical importance (Faecalibacterium prausnitzii and Akkermansia muciniphila)57,58. Interestingly, some specific growth requirements have been identified for these species, including supplementation with mucin for A. muciniphila and acetate for F. prausnitzii. Recently, high-throughput sequencing methods and culture-dependent techniques were combined to develop a new technique that uses microfluidics to isolate and identify new bacterial species. Using this approach, a new genus belonging to the Ruminococcaceae family59 was identified. The roles of important and abundant gut butyrate-producing bacteria, such as Faecalibacterium and Roseburia species, have also been identified using culture techniques60,61. The characterization of xylanolytic microbial communities from human faeces has allowed for the detection of new bacterial species, such as Bacteroides intestinalis, Bacteroides dorei and Roseburia intestinalis62.

Fig. 2: The culturomic workflow.

Culturomics was originally designed to identify new bacterial species in the gut microbiota, but it has since been applied to other microbiota, such as the human vaginal and urinary microbiota. The first step of culturomics is to divide the sample and diversify the sample into different culture conditions (part a). The culture conditions were designed to suppress the culture of majority populations and to promote the growth of fastidious microorganisms at lower concentrations (part b). However, targeted culture conditions were used to recover specific taxa. An important feature of culturomics is the rapid (<1 hour) identification by MALDI-TOF mass spectrometry, which relies on the comparison of the protein mass spectra of the isolate with an upgradable database (part c). If identification fails, the isolate is subjected to 16S ribosomal RNA (rRNA) sequencing (part d). If there is < 98.65% similarity to the closest official strain, the isolate could be a new species. The discovery of new taxa is confirmed by genome sequencing (part e), and taxonogenomics is used to formally describe the bacterium. All identification results are compared with a database that contains bacterial species recovered from humans. The identification of new species by culturomics has increased the repertoire of bacterial species associated with humans (part f).

More recently, targeted phenotypic culturing was used to identify sporulated bacteria by first preincubating faecal samples with ethanol to enrich the number of spores. This preincubation enabled the culture of 137 bacterial species, including 69 new bacterial taxa22. The authors found that ~55% of the bacterial genera in the gut microbiota of healthy individuals were able to produce resilient spores. In addition, they suggested that these bacteria were adapted to host-to-host transmission. This method was an interesting approach because, in contrast to culturomics, a single culture medium was inoculated.

Overall, these recent studies have more than doubled the number of bacterial species that have been isolated from the human gut, allowing us to investigate the phenotypes and functions of these isolates. In addition, whole genome sequencing of all these bacteria will facilitate the interpretation of future metagenomic studies.

The major drawbacks of culturomics remain the major workload and the inability to test as many samples as other methods, such as metagenomics. In addition, despite the progress since its inception, culturomics cannot identify the so-called ‘not yet culturable’ microorganisms. Finally, culturomics does not directly provide data on gene expression and the function of bacterial species as genome sequencing of the newly isolated microorganisms is required to assess its functional potential.

Consequences of culturomics

Increase in the number of known bacterial species

Before the initiation of the first culturomic project, 13,410 bacterial and archaeal species were known and only 2,172 species from humans had been cultured63. The application of culturomics to various projects, and specifically the human microbiota, has identified bacteria that were previously found in humans and some that were not29,64.

Culturomics has mostly been applied to the human gut microbiota in comparison with other ecosystems. Only 688 bacterial species and 2 archaeal species had been previously identified in the human gut, whereas 1,057 prokaryotic species were cultured when culturomics was first performed on a variety of stool samples, thus expanding the repertoire of bacterial species in the human gut microbiota9 (Fig. 3). Since that time, the isolation of 73 additional species from the human gut, 13 new species from the urinary tract, 15 species from the vaginal tract, 9 species from the respiratory tract, 2 species from human skin, 1 from human colostrum and 1 from the foot of an individual with osteomyelitis have been reported (Supplementary Table 1).

Fig. 3: The history and future potential applications of culturomics in clinical microbiology.

The increase in the bacterial repertoire associated with humans was facilitated by new culture methods for detecting rare and minority populations. Genome sequencing of these novel bacteria as a part of taxonogenomics enabled the sequence databases that are used for taxonomic assignment of high-throughput metagenomic studies to be updated, enabling the illumination of the microbial dark matter (that is, unassigned sequences from metagenomic studies) through precise identification of bacteria that would otherwise have been assigned an operational taxonomic unit. Sequencing also led to the identification of associations between the presence or absence of taxa with health or disease states. In parallel, updates of the MALDI-TOF databases used for routine identification of pathogens in clinical microbiology laboratories enabled the detection of new taxa from clinical specimens, contributing to understanding their roles in health and disease. Deposition of all isolates recovered in culturomic studies in a strain collection increased the availability of strains for in vitro and in vivo experiments, potentially leading to their use as probiotics. Considering the recent discovery of novel antimicrobial agents produced by human commensals, such strain collection could facilitate the search for new antibiotics.

Overall, culturomics has markedly increased the number of species associated with humans, bringing the total to 2,671 species. Indeed, culturomics enabled ~23% of the current repertoire of the bacterial species to be cultured at least once from a human sample.

From commensals to pathogens

With very rare exceptions (for example, the causative agents of cholera, plague and tuberculosis), human bacterial pathogens are all commensals as they are able to colonize human body sites without causing any infection. This observation has led microbiologists to reconsider their view of the nature of commensals and pathogens. For example, bacteria currently considered beneficial for health65, including A. muciniphila or Christensenella minuta, were first isolated as commensals57,66 but were lately recovered from clinical specimens as disease-causing agents67,68. Of note, A. muciniphila was identified in a bacteraemia episode using culturomics. Further evidence is needed to determine the potential pathogenicity of these bacterial species. In the future, the sharing of MALDI-TOF databases between culturomic and diagnostic microbiology laboratories is expected to facilitate the identification of potential pathogens from clinical specimens. In that respect, 12 new species that were previously cultured in a gut microbiota culturomic study were later isolated from 57 clinical specimens with various clinical features. Of these 12 new species, 11 were anaerobes, reflecting that a substantial proportion of new species isolated from the human gut do not tolerate oxygen. Some of these species were also recovered from clinical specimens but were thought to be present owing to their colonization rather than causing disease. As an example, Clostridium saudiense and Clostridium massilioamazoniense (formerly deposited as Clostridium sp. ND2) were cultured from stool samples and plated onto Clostridium difficile-specific media. This highlights the need to make the mass spectra from these new taxa available to enable their identification by any clinical microbiology laboratory in order to determine whether the presence of these bacteria is transitory or of clinical significance. The recurrent isolation of species in a specific clinical context, such as Actinomyces ihumi or Peptoniphilus grossensis from abscesses, suggested that they have a major pathogenic role9. Similarly, the anaerobic Bacteroides timonensis was recovered from blood, which is a sterile site; thus, it is unlikely to be a contaminant. These examples led to the consideration that clinical microbiology laboratories are likely to identify what is already known to exist rather than discover unknown microorganisms. Overall, culturomics has led to the inclusion of new taxa in MALDI-TOF mass spectrum databases, thus increasing the number of pathogens that can be identified from clinical specimens.

Reporting new taxa

Culturomics has led not only to an increase in the number of known human-associated bacteria but also to a change in the methods that are used to describe unknown bacteria. At the first stage of reporting a new bacterial species, taxonogenomics (Box 1) was introduced to describe newly isolated bacteria in a consistent format that integrates different data types such as MALDI-TOF mass spectra and genome sequencing data. New species announcement formats have also been created that include only GenBank 16S rRNA accession numbers and a collection of strain numbers to facilitate the quick dissemination of new species discovered by culturomics (Box 2).

From operational taxonomic units to bacterial species

Operational taxonomic units (OTUs) generated by metagenomic studies consist of clusters of DNA sequences from unidentified organisms and are used to classify groups of closely related microorganisms. The impact of culturomics could be measured by the reduction in the number of these unassigned OTUs by increasing the number of cultured species. Sequencing of the hypervariable regions of the 16S rRNA gene is the gold-standard marker for taxonomic assignment. During the profiling of 16S rRNA genes, all sequences are passed through quality filters and clustered. By definition, OTUs include all clustered sequences within a percentage sequence similarity threshold (typically 97%). Although one of the main advantages of OTU clustering is the time of the analyses, the biological interpretation of these OTUs remains a challenge. Each OTU, which is representative of each cluster, constitutes a tremendous source of information that highlights our current inability to access the entire repertoire of microbial genetic diversity. Integrating OTUs in metagenomics has paved the way for deciphering the genetic diversity in the microbiome. Many bioinformatic tools have been developed to assign species and OTUs in metagenomic samples69,70,71. These tools greatly simplified the attribution of OTUs to metagenomic samples. As a result, the considerable amount of data provided by OTU assignments requires rigorous classification of each OTU and systematic comparisons of metagenomic samples using standardized analyses. A recent study enabled unidentified species to be classified by integrating metagenomics with culturomics9. Indeed, preliminary studies found that only 15% of species were found using either approach5. For example, one study demonstrated that among the 247 new species that were identified, more than 50% were assigned to specific OTUs9. Notably, analyses of 84 metagenomes from stool samples revealed many potentially new species, as reflected by the considerable number of OTUs (4,158). The unexplored microbial world — known as the microbial dark matter — will be progressively revealed using multidisciplinary approaches (Fig. 3).

Exploration of changes in the gut microbiota associated with disease

As bacterial cells are the ‘units’ of a microbiota, pure culture is a crucial step in deciphering the role of specific microorganisms and/or the microbiota in host health (Fig. 2). Although metagenomics has revolutionized the study of the human microbiota, it still does not easily differentiate between strains among species, nor does it provide biological material for further studies. Consequently, pure culture remains an indispensable step in characterizing the relative role of microorganisms in health and disease. Below we describe two recent examples of the contribution of culturomics in understanding changes in the gut microbiota that are associated with disease.

Some intestinal bacteria (Clostridium neonatale, Clostridium perfringens, Clostridium butyricum and uropathogenic Escherichia coli) were recently reported as associated with necrotizing enterocolitis (NEC) in preterm neonates. Culturomics led to the detection of toxigenic C. butyricum strains in patients in a case–control study that included 15 neonates affected by NEC and 15 controls72. Interestingly, only a retrospective analysis of metagenomic data performed on the same stool samples detected an over-representation of this bacterial species in patients, whereas routine analyses did not observe any change. In addition, toxin secretion could be confirmed only using culture. Bacteriotherapy (that is, restoring the commensal gut microbiota) could be a promising treatment for this severe disease, which affects up to 10% of neonates with a very low birthweight.

In another example, after age and geography73,74, nutrition is the main determinant of gut microbiota composition and maturation. Severe acute malnutrition was found to be associated with global gut microbiota immaturity75 combined with a reduction of the bacterial load and bacterial diversity of all bacteria74,76. Different strains of the same species can be depleted or enriched following malnutrition, and they can be beneficial or deleterious when inoculated in experimental models77. One recent hypothesis is that malnutrition during infancy may result in the irreversible loss of beneficial gut microorganisms. This hypothesis may explain the ~10% therapeutic failure and the ~4% mortality rates despite a therapeutic diet and antibiotics78,79. As a result, further therapeutic options should include supplementing the absent bacteria (that is, well-characterized strains) in addition to a therapeutic diet and targeted antibiotics76,78. Culturomics was applied to a cohort of individuals from Niger and Senegal affected by kwashiorkor who were compared with controls from the same countries. Although metagenomics was helpful for screening potential microbial signatures of kwashiorkor, it was culturomics that identified 45 absent bacteria in affected individuals compared with controls. Among these, 12 species were considered potential candidates for bacteriotherapy. Importantly, the strains are stored and available for further studies76.

Future perspectives

In addition to the contribution already made by the rebirth of culture described above, culturomics along with all of the culture-dependent approaches will open new areas of research that may have far-reaching applications.

Correction of microbiota composition

Several changes in the gut microbiota composition have been associated with disease. Culturomics may have a major role in the development of bacteriotherapies to treat alterations in the microbiota in the clinic.

C. difficile infections (CDIs) are the main cause of health-care-associated diarrhoea in elderly individuals who are receiving antibiotics. Studies exploring changes in the gut microbiota in CDI mainly found a decrease in microbial diversity80,81 and changes in the composition of different phyla and genera82,83. Faecal microbiota transplantation (FMT) can restore this diversity84,85. Some focused therapeutic approaches have attempted to replace whole FMT by the administration of several specific bacteria that are potentially protective against CDI recurrence. In the first study, two individuals were treated with a pool of intestinal bacteria86. A subsequent study reported a case series of 55 individuals who were treated with a mixture of bacteria, and more recently87, one study tested the oral delivery of SER-109, which is a capsule that contains intestinal bacteria selected by ethanol treatment88. In parallel, the administration of non-toxigenic C. difficile spores in CDI appears to be a promising therapeutic approach89. In this context, culturomic studies would enable the identification of new sporulated bacteria that could be effective against CDI. Another recent example in which culturomics has been applied to the study of CDI is the discovery of the possible co-occurrence of C. difficile and Clostridium scindens in CDI90, although C. scindens had been suggested as a candidate against CDI91. Comprehensive studies of gut microbiota composition, including culturomic studies, will enable evidence-based selection of the required bacteria to replace FMT.

Correction of the microbiota composition may also be effective in the treatment of urogenital diseases. Bladder cancers are more frequently diagnosed in men than in women92. For a long time, this difference was thought to be due to a higher percentage of male smokers. However, an increased smoking rate in women has not led to a considerable increase in bladder cancer. In parallel, urine has long been considered sterile, but recent studies have shown that this is not the case93. Moreover, the urinary microbiota appears to be different in men and women94, with a notably higher proportion of Actinomycetes and Bacteroidetes species in men94. Remarkably, injections of Mycobacterium bovis, an Actinomycete, as a treatment for bladder cancer have been shown to prevent cancer relapse. This result merits a comprehensive analysis of urine composition by both metagenomic and culturomic approaches.

Bacterial vaginosis is an imbalance of the normal vaginal microbiota. It is characterized by a shift towards higher bacterial diversity with the depletion of the normal Lactobacillus-dominant microbiota, such as Lactobacillus crispatus and Lactobacillus jensenii, and the overgrowth of commensal anaerobic bacteria, such as Gardnerella vaginalis; A. vaginae; and the genera Anaerococcus, Clostridium, Peptoniphilus and Prevotella95. Similar to FMT, vaginal microbiota transplantation could be a solution for treating the mucosal microbiota imbalance by restoring beneficial vaginal microbial communities.

Therapeutic immunomodulatory effects

The impact of commensal microorganisms — in particular those from the gut — on immunity has been suggested for a long time. Recent studies have revealed a role of the gut microbiota in shaping immunity. The absence of a microbiota was found to decrease the production of tumour necrosis factor (TNF); consequently, this decreased the efficiency of immunotherapy in an experimental model96. Using a mouse model to evaluate the role of the gut microbiota in the outcome of anti-cytotoxic T lymphocyte-associated protein 4 (CTLA4) antibody tumour immunotherapy, it was demonstrated that the absence of specific taxa (particularly orders Bacteroidales and Burkholderiales) decreased the efficacy of treatment. Oral administration of bacterial strains Bacteroides fragilis, B. thetaiotaomicron and Burkholderia cepacia restored the antitumoural effect of the anti-CTLA4 antibodies. The recovery of bacteria-specific T cells from treated individuals revealed that the efficacy of anti-CTLA4 antibodies partially relies on previous exposure to bacterial commensals. Treatment efficacy of melanoma using CTLA4 blockers was found to depend on the gut microbiota composition in individuals receiving therapy97. The gut microbiota (notably Bifidobacterium spp.) was found to be involved in the onset of melanoma in an animal model; administration of Bifidobacterium spp. enhanced anti-programmed cell death 1 ligand 1 (PD-L1) cancer immunotherapy98. In addition, the microbiota also affected the toxicity of these treatments99,100. It was demonstrated that Enterococcus hirae and Barnesiella intestinihominis enhanced the efficacy of cyclophosphamide, a common alkylating immunomodulatory agent, in individuals receiving treatment for lung or ovarian cancers101. It was recently demonstrated that antibiotic therapy before immunomodulatory treatment significantly decreased the efficiency of programmed cell death protein 1 (PD-1)-based immunotherapy treatment against advanced epithelial cancers102. Culturomics is an important application in this research field and could lead to the development of new anticancer treatments, which will probably rely on the modulation of the microbiota103.

The discovery of new antimicrobial agents

As bacteria are a major source of several classes of antimicrobial agents that are widely used to treat human infections104 the expansion of the bacterial repertoire using culturomics may lead to the discovery of new antimicrobials. In addition, bacteriocins (protein or peptide toxins that are produced by bacteria and can inhibit the growth of closely related species) can be produced by members of the human gut microbiota105,106. However, owing to their narrow activity spectrum, their applications are currently limited to food preservation107. Interest in bacteriocins has increased as human microbiome studies have found many putative genes encoding these antimicrobial agents108,109. Intriguingly, the vagina, respiratory tract and oral cavity harbour the most diverse reservoirs of bacteriocins outside of the gut microbiota109. The human microbiome as a potential source of new antimicrobials was supported by the recent discovery of lugdunin109, a peptide that is produced by a Staphylococcus lugdunensis nasal strain that inhibits colonization by Staphylococcus aureus. In addition, this peptide is active against several Gram-positive pathogens, including Listeria spp., pneumococci and enterococci. A recent study demonstrated that several strains belonging to the order Clostridiales exerted inhibitory activity against Listeria monocytogenes in vitro and in vivo110. By providing a wide array of uncharacterized taxa and multiple strains per species, the isolates cultured by culturomics potentially represent a source of novel antibiotics.


We are currently witnessing a paradigm shift in which the manipulation of the human microbiota is holding promise for the treatment of infections, such as CDI85, and in the treatment of diseases that are associated with changes in the microbiome, such as cancer97,101,103. The foundation of this approach is the concept of probiotic microorganisms (as defined by Metchnikoff4), for example, Saccharomyces boulardii, which was used to treat diarrhoea111. This conceptual revolution, however, needs to rely on a sound knowledge of the microbial community inhabiting the human gut. For this purpose, substantial efforts to discover new taxa by culture are required to uncover the remaining 80% of unknown bacteria that are considered to be unculturable. In addition, these bacteria have to be incorporated into databases commonly used in metagenomic studies. In return, sequencing studies could prompt targeted culturomic approaches to culture OTUs of interest, emphasizing the complementarity between culture-dependent and culture-independent studies. This emphasis would enable the precise identification of microorganisms that are absent in specific diseases. Although FMT has a role in the treatment of CDI and more recently, in Crohn’s disease112, it does not appear to have any other therapeutic potential in its present formulation. For ‘oncomicrobiotics’ (that is, bacterial therapies that increase the efficacy of immunomodulatory treatments for cancer)101, a cocktail of beneficial bacteria that are selected depending on the pathology and profile of an individual’s gut microbiota may be more effective than a standard formulation. In conclusion, large-scale production, including fermentation steps for these beneficial bacteria (which are often difficult to culture), in combination with good medical practices could be the key to transforming bacteriotherapy for diseases that are associated with changes in our microbiota.


  1. 1.

    Turnbaugh, P. J. et al. A core gut microbiome in obese and lean twins. Nature 457, 480–484 (2009).

  2. 2.

    Lopez-Siles, M. et al. Mucosa-associated Faecalibacterium prausnitzii phylotype richness is reduced in patients with inflammatory bowel disease. Appl. Environ. Microbiol. 81, 7582–7592 (2015).

  3. 3.

    Torres-Fuentes, C., Schellekens, H., Dinan, T. G. & Cryan, J. F. The microbiota-gut-brain axis in obesity. Lancet Gastroenterol. Hepatol. 2, 747–756 (2017).

  4. 4.

    Metchnikoff, E. Les microbes intestinaux. Bull. l’Inst. Pasteur 6, 217–228 (1903). Metchnikoff is the first to describe probiotics.

  5. 5.

    Lagier, J. C. et al. Microbial culturomics: paradigm shift in the human gut microbiome study. Clin. Microbiol. Infect. 18, 1185–1193 (2012). This is the first reported microbial culturomic study.

  6. 6.

    Eckburg, P. B. et al. Diversity of the human intestinal microbial flora. Science 308, 1635–1638 (2005).

  7. 7.

    Kaeberlein, T., Lewis, K. & Epstein, S. S. Isolating “uncultivable” microorganisms in pure culture in a simulated natural environment. Science 296, 1127–1129 (2002).

  8. 8.

    Nichols, D. et al. Use of ichip for high-throughput in situ cultivation of “uncultivable” microbial species. Appl. Environ. Microbiol. 76, 2445–2450 (2010).

  9. 9.

    Lagier, J. C., Khelaifia, S. & Tidjani-Alou, M. Culture of previously uncultured members of the human gut microbiota by culturomics. Nat. Microbiol. 1, 16203 (2016). This study drastically increases the number of known human-associated microbial species.

  10. 10.

    Venter, J. C. et al. Environmental genome shotgun sequencing of the Sargasso Sea. Science 304, 66–74 (2004).

  11. 11.

    Hold, G. L., Pryde, S. E., Russell, V. J., Furrie, E. & Flint, H. J. Assessment of microbial diversity in human colonic samples by 16S rDNA sequence analysis. FEMS Microbiol. Ecol. 39, 33–39 (2002).

  12. 12.

    Suau, A. et al. Direct analysis of genes encoding 16S rRNA from complex communities reveals many novel molecular species within the human gut. Appl. Environ. Microbiol. 65, 4799–4807 (1999).

  13. 13.

    Lagier, J. C., Million, M., Hugon, P., Armougom, F. & Raoult, D. Human gut microbiota: repertoire and variations. Front. Cell. Infect. Microbiol. 2, 136 (2012).

  14. 14.

    Jovel, J. et al. Characterization of the gut microbiome using 16S or shotgun metagenomics. Front. Microbiol. 7, 459 (2016).

  15. 15.

    Truong, D. T., Tett, A. & Pasolli, E. Microbial strain-level population structure and genetic diversity from metagenomes. Genome Res. 27, 626–638 (2017).

  16. 16.

    Dhariwal, A. et al. MicrobiomeAnalyst: a web-based tool for comprehensive statistical, visual and meta-analysis of microbiome data. Nucleic Acids Res. 45, W180–W188 (2017).

  17. 17.

    Segata, N. et al. Metagenomic biomarker discovery and explanation. Genome Biol. 12, R60 (2011).

  18. 18.

    Meyer, F. et al. The metagenomics RAST server — a public resource for the automatic phylogenetic and functional analysis of metagenomes. BMC Bioinformatics 9, 386 (2008).

  19. 19.

    Lynch, M. D. & Neufeld, J. D. Ecology and exploration of the rare biosphere. Nat. Rev. Microbiol. 13, 217–229 (2015).

  20. 20.

    Cleary, B. et al. Detection of low-abundance bacterial strains in metagenomic datasets by eigengenome partitioning. Genome Res. 33, 1053–1060 (2015).

  21. 21.

    Lanza, V. F. et al. In-depth resistome analysis by targeted metagenomics. Microbiome 6, 11 (2018).

  22. 22.

    Browne, H. P. et al. Culturing of ‘unculturable’ human microbiota reveals novel taxa and extensive sporulation. Nature 533, 543–546 (2016). This study uses ethanol pretreatment to identify bacterial spores and increase the number of known human-associated bacterial species.

  23. 23.

    Goodman, A. L. et al. Extensive personal human gut microbiota culture collections characterized and manipulated in gnotobiotic mice. Proc. Natl Acad. Sci. USA 108, 6252–6257 (2011).

  24. 24.

    Rinke, C. et al. Insights into the phylogeny and coding potential of microbial dark matter. Nature 499, 431–437 (2013). This study introduces the concept of microbial dark matter.

  25. 25.

    Cangelosi, G. A. & Meschke, J. S. Dead or alive: molecular assessment of microbial viability. Appl. Environ. Microbiol. 80, 5884–5891 (2014).

  26. 26.

    Angelakis, E. et al. Gut microbiome and dietary patterns in different Saudi populations and monkeys. Sci. Rep. 6, 32191 (2016).

  27. 27.

    Zapka, C. et al. Comparison of standard culture-based method to culture-independent method for evaluation of hygiene effects on the hand microbiome. mBio 8, e00093-17 (2017).

  28. 28.

    Bollmann, A., Lewis, K. & Epstein, S. S. Incubation of environmental samples in a diffusion chamber increases the diversity of recovered isolates. Appl. Environ. Microbiol. 73, 6386–6390 (2007).

  29. 29.

    Hugon, P. et al. A comprehensive repertoire of prokaryotic species identified in human beings. Lancet. Infect. Dis. 15, 1211–1219 (2015).

  30. 30.

    Staley, J. T. & Konopka, A. Measurement of in situ activities of nonphotosynthetic microorganisms in aquatic and terrestrial habitats. Annu. Rev. Microbiol. 39, 321–346 (1985).

  31. 31.

    Lagier, J. C. et al. The rebirth of culture in microbiology through the example of culturomics to study human gut microbiota. Clin. Microbiol. Rev. 28, 237–264 (2015).

  32. 32.

    Button, D. K., Schut, F., Quang, P., Martin, R. & Robertson, B. R. Viability and isolation of marine bacteria by dilution culture: theory, procedures, and initial results. Appl. Environ. Microbiol. 59, 881–891 (1993).

  33. 33.

    Rappe, M. S., Connon, S. A., Vergin, K. L. & Giovannoni, S. J. Cultivation of the ubiquitous SAR11 marine bacterioplankton clade. Nature 418, 630–633 (2002).

  34. 34.

    Dadon-Pilosof, A. et al. Surface properties of SAR11 bacteria facilitate grazing avoidance. Nat. Microbiol. 2, 1608–1615 (2017).

  35. 35.

    Stewart, E. J. Growing unculturable bacteria. J. Bacteriol. 194, 4151–4160 (2012).

  36. 36.

    Nichols, D. et al. Short peptide induces an “uncultivable” microorganism to grow in vitro. Appl. Environ. Microbiol. 74, 4889–4897 (2008).

  37. 37.

    Raoult, D. et al. Cultivation of the bacillus of Whipple’s disease. N. Engl. J. Med. 342, 620–625 (2000).

  38. 38.

    Raoult, D. et al. A flea-associated Rickettsia pathogenic for humans. Emerg. Infect. Dis. 7, 73–81 (2001).

  39. 39.

    Gouriet, F., Fenollar, F., Patrice, J. Y., Drancourt, M. & Raoult, D. Use of shell-vial cell culture assay for isolation of bacteria from clinical specimens: 13 years of experience. J. Clin. Microbiol. 43, 4993–5002 (2005).

  40. 40.

    Lagier, J. C. et al. Current and past strategies for bacterial culture in clinical microbiology. Clin. Microbiol. Rev. 28, 208–236 (2015).

  41. 41.

    Nottingham, P. M. & Hungate, R. E. Isolation of methanogenic bacteria from feces of man. J. Bacteriol. 96, 2178–2179 (1968).

  42. 42.

    Moore, W. E. & Holdeman, L. V. Human fecal flora: the normal flora of 20 Japanese-Hawaiians. Appl. Microbiol. 27, 961–979 (1974).

  43. 43.

    Finegold, S. M., Attebery, H. R. & Sutter, V. L. Effect of diet on human fecal flora: comparison of Japanese and American diets. Am. J. Clin. Nutr. 27, 1456–1469 (1974).

  44. 44.

    Aries, V., Crowther, J. S., Drasar, B. S., Hill, M. J. & Williams, R. E. Bacteria and the aetiology of cancer of the large bowel. Gut 10, 334–335 (1969).

  45. 45.

    Rajilic-Stojanovic, M., Smidt, H. & de Vos, W. M. Diversity of the human gastrointestinal tract microbiota revisited. Environ. Microbiol. 9, 2125–2136 (2007).

  46. 46.

    Moore, W. E. & Moore, L. H. Intestinal floras of populations that have a high risk of colon cancer. Appl. Environ. Microbiol. 61, 3202–3207 (1995).

  47. 47.

    Walker, A. W., Duncan, S. H., Louis, P. & Flint, H. J. Phylogeny, culturing, and metagenomics of the human gut microbiota. Trends Microbiol. 22, 267–274 (2014).

  48. 48.

    La, S. B., Khelaifia, S., Lagier, J. C. & Raoult, D. Aerobic culture of anaerobic bacteria using antioxidants: a preliminary report. Eur. J. Clin. Microbiol. Infect. Dis. 33, 1781–1783 (2014).

  49. 49.

    Dione, N., Khelaifia, S., La, S. B., Lagier, J. C. & Raoult, D. A quasi-universal medium to break the aerobic/anaerobic bacterial culture dichotomy in clinical microbiology. Clin. Microbiol. Infect. 22, 53–58 (2016).

  50. 50.

    Khelaifia, S. et al. Aerobic culture of methanogenic archaea without an external source of hydrogen. Eur. J. Clin. Microbiol. Infect. Dis. 35, 985–991 (2016).

  51. 51.

    Wolochow, H. Detection of Airborne Microorganisms through their Unique Compounds. Naval Biosciences Laboratory Technical Reports, Report No. 211170 (Armed Services Technical Information Agency, Arlington, VA, 1959).

  52. 52.

    Anhalt, J. & Fenselau, C. Identification of bacteria using mass spectrometry. Anal. Chem. 47, 219–225 (1975).

  53. 53.

    Fenselau, C. & Demirev, P. A. Characterization of intact microorganisms by MALDI mass spectrometry. Mass Spectrom. Rev. 20, 157–171 (2001).

  54. 54.

    Seng, P. et al. Ongoing revolution in bacteriology: routine identification of bacteria by matrix-assisted laser desorption ionization time-of-flight mass spectrometry. Clin. Infect. Dis. 49, 543–551 (2009).

  55. 55.

    Seng, P. et al. Identification of rare pathogenic bacteria in a clinical microbiology laboratory: impact of matrix-assisted laser desorption ionization-time of flight mass spectrometry. J. Clin. Microbiol. 51, 2182–2194 (2013).

  56. 56.

    Gross, A. et al. Technologies for single-cell isolation. Int. J. Mol. Sci. 16, 16897–16919 (2015).

  57. 57.

    Derrien, M., Vaughan, E. E., Plugge, C. M. & de Vos, W. M. Akkermansia muciniphila gen. nov., sp. nov., a human intestinal mucin-degrading bacterium. Int. J. Syst. Evol. Microbiol. 54, 1469–1476 (2004).

  58. 58.

    Duncan, S. H., Hold, G. L., Harmsen, H. J., Stewart, C. S. & Flint, H. J. Growth requirements and fermentation products of Fusobacterium prausnitzii, and a proposal to reclassify it as Faecalibacterium prausnitzii gen. nov., comb. nov. Int. J. Syst. Evol. Microbiol. 52, 2141–2146 (2002).

  59. 59.

    Ma, L. et al. Gene-targeted microfluidic cultivation validated by isolation of a gut bacterium listed in Human Microbiome Project’s Most Wanted taxa. Proc. Natl Acad. Sci. USA 111, 9768–9773 (2014).

  60. 60.

    Duncan, S. H. et al. Wheat bran promotes enrichment within the human colonic microbiota of butyrate-producing bacteria that release ferulic acid. Environ. Microbiol. 18, 2214–2225 (2016).

  61. 61.

    Walker, A. W., Duncan, S. H., McWilliam Leitch, E. C., Child, M. W. & Flint, H. J. pH and peptide supply can radically alter bacterial populations and short-chain fatty acid ratios within microbial communities from the human colon. Appl. Environ. Microbiol. 71, 3692–3700 (2005).

  62. 62.

    Chassard, C., Goumy, V., Leclerc, M., Del’homme, C. & Bernalier-Donadille, A. Characterization of the xylan-degrading microbial community from human faeces. FEMS Microbiol. Ecol. 61, 121–131 (2007).

  63. 63.

    Hugon, P. et al. A comprehensive repertoire of prokaryotic species identified in human beings. Lancet Infect. Dis. 15, 1211–1219 (2015).

  64. 64.

    Hill, M. J. & Drasar, B. S. The normal colonic bacterial flora. Gut 16, 318–323 (1975).

  65. 65.

    Shivaji, S. We are not alone: a case for the human microbiome in extra intestinal diseases. Gut Pathog. 9, 13 (2017).

  66. 66.

    Morotomi, M., Nagai, F. & Watanabe, Y. Description of Christensenella minuta gen. nov., sp. nov., isolated from human faeces, which forms a distinct branch in the order Clostridiales, and proposal of Christensenellaceae fam. nov. Int. J. Syst. Evol. Microbiol. 62, 144–149 (2012).

  67. 67.

    Dubourg, G. et al. First isolation of Akkermansia muciniphila in a blood-culture sample. Clin. Microbiol. Infect. 23, 682–683 (2017).

  68. 68.

    Fathi, P. & Wu, S. Isolation, detection, and characterization of enterotoxigenic Bacteroides fragilis in clinical samples. Open Microbiol. J. 10, 57–63 (2016).

  69. 69.

    Schloss, P. D. et al. Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl. Environ. Microbiol. 75, 7537–7541 (2009).

  70. 70.

    Caporaso, J. G. et al. QIIME allows analysis of high-throughput community sequencing data. Nat. Methods 7, 335–336 (2010).

  71. 71.

    Huson, D. H., Mitra, S., Ruscheweyh, H. J., Weber, N. & Schuster, S. C. Integrative analysis of environmental sequences using MEGAN4. Genome Res. 21, 1552–1560 (2011).

  72. 72.

    Cassir, N. et al. Clostridium butyricum strains and dysbiosis linked to necrotizing enterocolitis in preterm neonates. Clin. Infect. Dis. 61, 1107–1115 (2015).

  73. 73.

    Yatsunenko, T. et al. Human gut microbiome viewed across age and geography. Nature 486, 222–227 (2012).

  74. 74.

    Million, M. et al. Increased gut redox and depletion of anaerobic and methanogenic prokaryotes in severe acute malnutrition. Sci. Rep. 6, 26051 (2016).

  75. 75.

    Subramanian, S. et al. Persistent gut microbiota immaturity in malnourished Bangladeshi children. Nature 510, 417–421 (2014).

  76. 76.

    Tidjani, A. M. et al. Gut bacteria missing in severe acute malnutrition, can we identify potential probiotics by culturomics? Front. Microbiol. 8, 899 (2017).

  77. 77.

    Wagner, V. E. et al. Effects of a gut pathobiont in a gnotobiotic mouse model of childhood undernutrition. Sci. Transl Med. 8, 366ra164 (2016).

  78. 78.

    Trehan, I. et al. Antibiotics as part of the management of severe acute malnutrition. N. Engl. J. Med. 368, 425–435 (2013).

  79. 79.

    Million, M., Diallo, A. & Raoult, D. Gut microbiota and malnutrition. Microb. Pathog. 106, 127–138 (2017).

  80. 80.

    Allegretti, J. R. et al. Recurrent Clostridium difficile infection associates with distinct bile acid and microbiome profiles. Aliment. Pharmacol. Ther. 43, 1142–1153 (2016).

  81. 81.

    Chang, J. Y. et al. Decreased diversity of the fecal microbiome in recurrent Clostridium difficile-associated diarrhea. J. Infect. Dis. 197, 435–438 (2008).

  82. 82.

    Schubert, A. M. et al. Microbiome data distinguish patients with Clostridium difficile infection and non-C. difficile-associated diarrhea from healthy controls. mBio 5, e01021-14 (2014).

  83. 83.

    Theriot, C. M. & Young, V. B. Interactions between the gastrointestinal microbiome and Clostridium difficile. Annu. Rev. Microbiol. 69, 445–461 (2015).

  84. 84.

    Fuentes, S. et al. Reset of a critically disturbed microbial ecosystem: faecal transplant in recurrent Clostridium difficile infection. ISME J. 8, 1621–1633 (2014).

  85. 85.

    Seekatz, A. M. et al. Recovery of the gut microbiome following fecal microbiota transplantation. mBio 5, e00893-14 (2014).

  86. 86.

    Petrof, E. O. et al. Stool substitute transplant therapy for the eradication of Clostridium difficile infection: ‘RePOOPulating’ the gut. Microbiome 1, 3 (2013). This paper presents a seminal study on the use of a cocktail of bacteria rather than a whole stool sample to perform FMT.

  87. 87.

    Tvede, M., Tinggaard, M. & Helms, M. Rectal bacteriotherapy for recurrent Clostridium difficile-associated diarrhoea: results from a case series of 55 patients in Denmark 2000–2012. Clin. Microbiol. Infect. 21, 48–53 (2015).

  88. 88.

    Khanna, S. et al. A novel microbiome therapeutic increases gut microbial diversity and prevents recurrent Clostridium difficile infection. J. Infect. Dis. 214, 173–181 (2016).

  89. 89.

    Gerding, D. N. et al. Administration of spores of nontoxigenic Clostridium difficile strain M3 for prevention of recurrent C. difficile infection: a randomized clinical trial. JAMA 313, 1719–1727 (2015).

  90. 90.

    Amrane, S., Bachar, D., Lagier, J. C. & Raoult, D. Clostridium scindens is present in the gut microbiota during Clostridium difficile infection. A metagenomic and culturomic analysis. J. Clin. Microbiol. 56, e01663–17 (2018).

  91. 91.

    Buffie, C. G. et al. Precision microbiome reconstitution restores bile acid mediated resistance to Clostridium difficile. Nature 517, 205–208 (2015).

  92. 92.

    Raoult, D. Is there a link between urinary microbiota and bladder cancer? Eur. J. Epidemiol. 32, 255 (2017).

  93. 93.

    Hilt, E. E. et al. Urine is not sterile: use of enhanced urine culture techniques to detect resident bacterial flora in the adult female bladder. J. Clin. Microbiol. 52, 871–876 (2014).

  94. 94.

    Whiteside, S. A., Razvi, H., Dave, S., Reid, G. & Burton, J. P. The microbiome of the urinary tract — a role beyond infection. Nat. Rev. Urol. 12, 81–90 (2015).

  95. 95.

    Fenollar, F. & Raoult, D. Does bacterial vaginosis result from fecal transplantation? J. Infect. Dis. 214, 1784 (2016).

  96. 96.

    Iida, N. et al. Commensal bacteria control cancer response to therapy by modulating the tumor microenvironment. Science 342, 967–970 (2013).

  97. 97.

    Vetizou, M. et al. Anticancer immunotherapy by CTLA-4 blockade relies on the gut microbiota. Science 350, 1079–1084 (2015). This study identifies relationships between the gut microbiota and the efficacy of anticancer immunotherapy.

  98. 98.

    Sivan, A. et al. Commensal Bifidobacterium promotes antitumor immunity and facilitates anti-PD-L1 efficacy. Science 350, 1084–1089 (2015).

  99. 99.

    Dubin, K. et al. Intestinal microbiome analyses identify melanoma patients at risk for checkpoint-blockade-induced colitis. Nat. Commun. 7, 10391 (2016).

  100. 100.

    Tsilimigras, M. C., Fodor, A. & Jobin, C. Carcinogenesis and therapeutics: the microbiota perspective. Nat. Microbiol. 2, 17008 (2017).

  101. 101.

    Daillere, R. et al. Enterococcus hirae and Barnesiella intestinihominis facilitate cyclophosphamide-induced therapeutic immunomodulatory effects. Immunity 45, 931–943 (2016).

  102. 102.

    Routy, B. et al. Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors. Science 359, 91–97 (2018).

  103. 103.

    Zitvogel, L., Daillere, R., Roberti, M. P., Routy, B. & Kroemer, G. Anticancer effects of the microbiome and its products. Nat. Rev. Microbiol. 15, 465–478 (2017).

  104. 104.

    Dubourg, G., Abat, C. & Raoult, D. Why new antibiotics are not obviously useful now. Int. J. Antimicrob. Agents 49, 549–553 (2017).

  105. 105.

    Millette, M. et al. Purification and identification of the pediocin produced by Pediococcus acidilactici MM33, a new human intestinal strain. J. Appl. Microbiol. 104, 269–275 (2008).

  106. 106.

    O’Shea, E. F. et al. Production of multiple bacteriocins from a single locus by gastrointestinal strains of Lactobacillus salivarius. J. Bacteriol. 193, 6973–6982 (2011).

  107. 107.

    De, V. L. & Leroy, F. Bacteriocins from lactic acid bacteria: production, purification, and food applications. J. Mol. Microbiol. Biotechnol. 13, 194–199 (2007).

  108. 108.

    Zheng, J., Ganzle, M. G., Lin, X. B., Ruan, L. & Sun, M. Diversity and dynamics of bacteriocins from human microbiome. Environ. Microbiol. 17, 2133–2143 (2015).

  109. 109.

    Zipperer, A. et al. Human commensals producing a novel antibiotic impair pathogen colonization. Nature 535, 511–516 (2016). This sequencing study demonstrates that the human microbiome contains many putative genes that encode antimicrobial agents.

  110. 110.

    Becattini, S. et al. Commensal microbes provide first line defense against Listeria monocytogenes infection. J. Exp. Med. 214, 1973–1989 (2017).

  111. 111.

    Szajewska, H. & Kolodziej, M. Systematic review with meta-analysis: Saccharomyces boulardii in the prevention of antibiotic-associated diarrhoea. Aliment. Pharmacol. Ther. 42, 793–801 (2015).

  112. 112.

    Weingarden, A. R. & Vaughn, B. P. Intestinal microbiota, fecal microbiota transplantation, and inflammatory bowel disease. Gut Microbes 8, 238–252 (2017).

  113. 113.

    Stackebrandt, E. & Ebers, J. Taxonomic parameters revisited: tarnished gold standards. Microbiol. Today 33, 152–155 (2006).

  114. 114.

    Sentausa, E. & Fournier, P. E. Advantages and limitations of genomics in prokaryotic taxonomy. Clin. Microbiol. Infect. 19, 790–795 (2013).

  115. 115.

    Fournier, P. E., Raoult, D. & Drancourt, M. New Species Announcement: a new format to prompt the description of new human microbial species. New Microbes New Infect. 15, 136–137 (2017).

Download references


This work has benefited from French State support, managed by the ‘Agence Nationale pour la Recherche’, including the ‘Programme d’Investissement d’Avenir’ under the reference Méditerranée Infection 10-IAHU-03. This work was also funded by the Prix Louis D. and by Région Provence Alpes Côte d’Azur and European funding FEDER PRIMI.

Author information

J.-C.L., G.D., M.M., F.C., M.B., F.F., A.L. and D.R. researched the data for the article. J.-C.L., G.D., M.M., J.-M.R., P.-E.F. and D.R. substantially contributed to discussion of content. J.-C.L., G.D., M.M., F.C., M.B., F.F., A.L., P.-E.F. and D.R. wrote the article. J.-C.L., G.D., M.M., F.C., M.B., F.F., A.L., J.-M.R., P.-E.F. and D.R. reviewed and edited the manuscript before submission.

Correspondence to Didier Raoult.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information



Consortia of microorganisms living in a defined environment.


Microorganism that colonizes its host without causing disease.


Living microorganisms that can be beneficial for health.


Shotgun sequencing of DNA isolated directly from a specific environment.


Administration of live microorganisms, alone or as complete ecosystems, to improve human health. FMT is an example of bacteriotherapy.

Microbial dark matter

Unassigned sequences from metagenomic studies.

MALDI-TOF mass spectrometry

Combination of a soft ionization method (MALDI) with mass spectrometry, enabling the identification of proteins through their mass and their charge number.

Diffusion chamber

A method that allows microorganisms to grow in their natural environment, the inoculum being sandwiched between semipermeable membranes of the chamber, allowing a free exchange of chemicals with the external milieu.

Obligate anaerobes

Organisms for which atmospheric oxygen concentration is toxic.

Shell-vial technique

A centrifuge-enhanced tissue culture assay.

Axenic media

Host-cell-free growth culture media.


Organisms that do not require oxygen for growth.

Schaedler agar

Solid culture medium recommended for the isolation of anaerobic bacteria from clinical specimens.

Xylanolytic microbial communities

Communities containing prokaryotes with the ability to degrade xylan.


Modern approach to describe new taxa isolated from culturomic studies combining genome sequencing and phenotypic information.

Operational taxonomic units.

Clusters of DNA sequences from unidentified organisms organized according to their DNA sequence similarity.


All genes and genomes of a microbiota.


Clinical nutritional disease including irritability, diarrhoea with indigested food, swelling of the hands and feet (nutritional oedema), general puffiness of the face (moon face) and skin changes (depigmentation and thickened black and crumpled patches with peeling and rash).

Faecal microbiota transplantation

The transfer of faecal material from a healthy individual into an individual with a condition.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Lagier, J., Dubourg, G., Million, M. et al. Culturing the human microbiota and culturomics. Nat Rev Microbiol 16, 540–550 (2018) doi:10.1038/s41579-018-0041-0

Download citation

Further reading