Since humans first became aware of microorganisms in the gut in the late 1800s, efforts to reveal them and their ecology have been ongoing for centuries. Human gut microbiome research has mainly been conducted using metagenomics over the past 2 decades. With the development of recent technologies, cultivation-based research is regaining popularity [11]. However, most large-scale cultivation-based human gut microbiota studies have been conducted on healthy individuals [19, 88–92], and only a few studies have compared the gut microbiota of patients with disease to healthy individuals [37, 93]. In our study, we used culturomics and metagenomics to investigate bacterial species enriched or depleted in the gut of UC and CD patients. We identified 215 and 170 species from UC and CD patients, respectively. The phylogenetic diversity of these isolates was broad, allowing combined analysis with our metagenomic results. Our findings underscored the complementarity of these two methods and highlighted their synergistic relationship, which was evident when interpreting dysbiosis in IBD.
UC and CD, the two most common IBD subtypes, differ in the location, extent, and symptoms of inflammation. UC patients are more likely to have occult blood in their stool than CD patients, mainly due to ulcers in the rectum or colon lining [94]. Our study also demonstrated that the feces of patients with UC had relatively high levels of hematochezia. On the other hand, patients with CD had relatively high levels of inflammatory biomarkers, including CRP and FCP (Table 1). The subtypes were distinguished by the clinical diagnostic results, determined according to the assessed disease severity (Fig. S5 in Additional file 9). The causal relationship between pathological differences and gut microbiota dysbiosis in IBD is important. For instance, luminal heme from bleeding may play a key role in worsening IBD dysbiosis by affecting the butyrate metabolism of gut microbiota [95, 96].
A shift in gut microbiota was observed in the UC and CD groups, but the imbalance was more severe in CD patients. The distribution of UC gut microbiota had a wide range and overlapped with the CD and HC groups (Table 2, Fig. 4). These findings were consistent with reports comparing the gut microbiota of UC and CD patients [97, 98]. CD causes inflammation throughout the gastrointestinal lining, which is presumed to affect not only lesions from UC but also the gut microbiota. Although there are discrepancies in gut dysbiosis in IBD between cohorts, an increased population of Proteobacteria (particularly Enterobacteriaceae) and an overall decrease in microbial diversity are commonly reported characteristics of gut dysbiosis in IBD [99, 100] and were also observed in our study (Fig. 3a, 4a). Our results also demonstrated a decreased population of Actinobacteria in the CD group (Fig. 3a). These characteristics were also confirmed by our culturomics results (Table 2).
The oxygen hypothesis proposes that increased luminal oxygen levels are a key contributor to IBD dysbiosis [101]. The main causes are the disruption of gut permeability, the release of hemoglobin-carried oxygen from intestinal bleeding, and the reduction of ꞵ-oxidation of enterocytes [102]. This induces an abundance of facultative anaerobes relative to obligate anaerobes, and the overgrowth of Enterobacteriaceae is a representative example [102]. Our culturomics results showed a significant decrease in the diversity of anaerobic bacterial species and a trend of increased diversity of aerotolerant bacterial species in CD patients, supporting this hypothesis (Table 2). Shahir et al. proposed the importance of the aerotolerance profile of mucosal gut bacteria in CD patients [103]. Thus, the anaerobes depleted in CD patients may be considered candidates for the recovery of intestinal dysanaerobiosis.
High-throughput culture-based approaches can complement metagenomics in human gut microbiome research and offer several advantages. First, the isolated bacterial strains are host-specific culture collections with active characteristics in the intestine and can be practically used for experimental research and commercialization [9, 26]. Second, large-scale strain collections are essential for understanding intraspecies diversity and strain-specificity issues [92]. Strain-level resolution has been attempted using metagenome-assembled genomes (MAGs) based on shotgun sequencing, but the accuracy and completeness of short-read sequences need improvement [104, 105]. Adding long-read sequencing techniques with a much larger data depth will overcome some limitations, but in terms of cost and time, it is impractical and inefficient to construct and analyze large-scale MAGs [106]. Third, a high-throughput culture-dependent approach significantly expands the microbial species coverage of metagenomics [89]. In our study, we identified 14,131 strains derived from gut microbiota in IBD patients, and the number of microbial species identified by culturomics was approximately 2.5 times higher than that identified by metagenomics, excluding unassigned ASVs (Fig. 5b). Additionally, the acquisition of new microbes is common in culturomics approaches, thereby expanding the human gut bacteria repertoire [17–19]. Taken together, combining culture-based and genome-based approaches in human gut microbiome research provides additional insights compared with those revealed by either approach alone.
Although culturomics has significantly improved the efficiency of high-throughput cultivation, the process remains labor-intensive [107]. Instead of applying numerous culture conditions, we used two media types for preincubation and a single medium (mGAM) for subsequent processes. The single-media cultivation strategy achieved high work efficiency and will facilitate the validation of microbial interactions and the comparison of the strains in the future [108]. Our study had a cross-sectional design. However, additional longitudinal studies will provide more detailed information because IBD patients typically experience a repetitive pattern of remission and relapse [109]. It is also crucial to identify changes in gut microbiota that occur in response to clinical interventions, such as biologics [110]. Although we focused on fecal samples, considering the recent reports describing the importance of mucosa-adherent bacteria in IBD pathogenesis, investigation of gut dysbiosis using biopsy samples from multiple intestinal regions is also necessary [88, 103]. Since the collaboration of culturomics with other omics approaches for large-scale microbiome studies remains challenging, it is crucial to improve the efficiency of culture-dependent techniques.