From Bioreactors to Organoids: Tools for Culturing and Understanding Microbiota

Abstract

Emerging evidence on the role of gut microbiota on human health necessitated the investigations on gut microbial composition and functions. The diet, drug and disease are the major factors that influence the gut microbial composition and subsequent changes in gut physiology. The consequences of external factors on gut microbiota leads to immunological and neurological disorders. The in vivo studies and animal models are associated with difficulties such as surgical procedures, differences in animal and human tissue responses and ethical issues. Microbiota culture outside the in vivo systems provides mechanistic insights on the effect of external factors on gut microbiota function. In order to provide more information on gut microbiota, researchers have developed intestinal gut models that allow the culture of microbiota under controlled conditions. The multistage, dynamic in vitro fermenters have been developed to simulate the stomach, intestine and colon conditions to culture microbiota. The co-culture of host and microbiome is difficult in in vitro models due to differences in the culture media and oxygen requirements for both the cultures. Microfluidics based gut on chip models demonstrated co-culture of host and microbiota in separate channels connected by semipermeable membranes. Host–microbiota interactions play an important role in deciphering the mechanisms of microbiota related human diseases. The ex-vivo systems show more resemblance to the host signaling as these systems use the individual’s explants or host tissues. This review discusses the design considerations, pros and cons of the existing in vitro, microfluidic and ex-vivo intestinal microbiota culture models. The collective information provided in this review would be helpful to design novel in vitro microbiota culture models or methods.

Introduction

Gut microbial composition plays a pivotal role in human health and disease [91]. Microbiota composition shows similar microbial signatures within human individual microbiota across the globe [4, 6]. Microbiota composition diversity among the individuals has been observed due to variations in age, diet, drugs, and disease, environmental and geographical locations [26, 63, 115]. Understanding microbiota composition and functions are essential for the management of microbiota associated health complications [93]. Gut microbiota composition, density and diversity changes are observed along the gastrointestinal tract (GIT) due to variations in GIT structure, PH, redox potential, oxygen levels and mucus level form proximal to distal end [67]. Small intestine consists of a thin layer of mucus whereas the colon consists of a thick layer of mucus layer [89]. Mucous associated microbial communities adhere to the mucus layer and feces harbours wide variety microbial communities [27]. Majority of the gut microbiota research studies focused on fecal microbial communities due to ease of sample collections. Mucous associated microbial samples from individuals can only be obtained by biopsy samples [110]. Understanding mucus adhesive microbial communities is critical for the deciphering host–microbiota interaction as this microbiota is in close contact with gut epithelial cells [89]. Microbiota characterization without in vitro culture is limited only to abundant microbes. The less abundant and more significant bacteria might be ignored in this approach. In vitro microbiota culture allows to study the effects of various factors such as diet, drugs and environmental factors on microbial communities. Simulation of GIT conditions in in vitro experiments is challenging as GIT is a dynamic system with varied anatomy and physiology [69]. The microbial diversity is location specific along the GIT and in vitro culture requires location specific conditions [67]. For example, PH and oxygen levels vary from stomach to distal colon. The microbiota composition consists of commensal, mutualistic and pathogenic bacteria, and the dysbiosis may lead to the health disorders. The pure culture from microbiota may not provide the complete characterization but culture as a whole microbial community provides more insights. In vivo models are advantageous in terms of relevancy to biological significance [16]. Most of the in vivo models involved the end point measurements and analysis is carried out on fecal samples [14]. The gut location specific microbiota analysis needs surgical process and it involves high cost and ethical constraints [96]. In vitro models allow to create gut location specific conditions to culture and study the effect of external factors on microbiota [109]. The multistage in vitro microbiota culture models such as Simulator of the Human Intestinal Microbial Ecosystem (SHIME) and The TNO In Vitro Model of the Colon (TIM2) are advantageous in colon region specific culture of bacteria but these methods are not suitable for studying host responses. The investigations on the host microbiota interactions are a major challenge in in vitro models [78]. The host cells are aerobic and majority of the organisms of microbiota are facultative anaerobes [40]. Organ on chip models have gained attention for studying host responses to gut microbiota. The gut on chip models allows the co culture of host and microbiota in different compartments with different growth conditions and continuous flow of fluids [5]. Microfluidics chips offer the transport, absorption and host signalling studies at host microbiota interface [9]. The microfluidic gut on chip models mimics the in vivo conditions and allows the close and direct contact between host and microbiota. This allows to study the host–microbiota cross talk within the simulated gut microenvironment. Microfabricated gut on chip models allows to create gut like microenvironment such as, three dimensional (3D) like villi structures, oxygen gradient and peristalsis, but the applications are limited due to difficulties in maintaining the viability of host and microbiota cells for longer periods [103]. The microfluidic methods of co-culture utilize semipermeable artificial membranes, and it is difficult to simulate the complete in vivo para cellular membrane permeability. Mucus layer plays a critical role in the host–microbiota interactions and these gut on chip models need to be focused on simulation of the mucus layer on epithelial cells. These methods provide key mechanistic insights in host–microbiome interactions with some drawbacks. The implications of ex-vivo models, in which live and functional tissue or explants isolated and cultivated outside the source organism, provides the more mechanistic insights in host–microbiome interactions [84]. The presence of intact mucus layer, paracellular permeability and expression of transport proteins facilitate the complete resemblance to the in vivo systems. The ex-vivo models need to be explored to study the host microbiome interactions. Although ex-vivo methods are associated with some drawbacks such as changes in morphology and functionality of the tissues, these methods are more relevant to in vivo conditions [79].

This review provides the working principles, applications and challenges of in vitro, microfluidic and ex-vivo models for microbiota culture. None of the intestinal culture models completely mimics the in vivo systems and this review discusses the merits and demerits of existing intestinal models. This review will be helpful to improve the existing intestinal models and provides new clues to develop novel, cost effective and efficient intestinal models.

Bioreactor Models for Culture of Microbiota

Microbial fermentation inside the GIT can be influenced by various factors such as diet, prebiotics and drugs. The PH and oxygen levels from proximal end to distal end of the GIT varies greatly, and this variation led to the microbiota diversity in different regions of the GIT [67]. The microbial metabolites play a key role in host metabolic regulation and inflammatory responses [51]. It is difficult to study the microbial composition and microbial metabolites at specific regions of the GIT due to ethical and surgical reasons. The cultivation of microbiota in bioreactors offer controlled environment that simulates specific regions of GIT. Simple batch reactors can be used to culture the microbiota but the use of batch bioreactors have limitations due to inhibitory effects due to waste accumulation during the experiment. Continuous or semi continuous bioreactors are suitable for microbiota culture as these reactors involves replenishment of nutrients and waste removal at regular intervals [38]. Furthermore, multistage reactors with region specific conditions were first demonstrated by Macfarlane et al. [65] and this work paved the way to develop different multistage bioreactor methods. This section discusses some of the successful models to culture microbiota using multistage bioreactors.

The TNO In Vitro Model of the Colon (TIM-2)

TNO Nutrition and Food Research Institute, Netherlands, developed TIM-2 model that mimics the large intestine. This model is developed based on the TIM-1 model that mimics the small intestine [113]. The use of TIM-1 and TIM-2 in-combination provides the complete simulation of small intestine and large intestine by in vitro method. The typical design of TIM-2 model is illustrated in Fig. 1. This model consists of flexible membrane inside the four glass compartments, and this membrane mimics the luminal wall. The water of body temperature circulates between the glass wall and flexible membrane. The peristaltic waves are generated by applying pressure on water at regulated time intervals. The glass compartments are connected to ileal efflux container, pH electrode, alkali pump, N₂ gas inlet, level sensor and gas outlet. The Dialysis system with hollow fibre membrane is embedded in this model to collect accumulated microbial metabolites. The sample port is located in between the two peristaltic glass compartments and it helps to collect the luminal samples during the experiment [64]. TIM-2 offers the peristaltic movements without any stirring or shaking and microbial metabolites can be collected through dialysis system and sampling port. This model allows the mass balance without any solid –liquid separation during the experiment [113].

Fig. 1.

Schematic representation of TNO In Vitro Model of the Colon (TIM-2) model

In a study, the effect of multiple iron sources on microbiota was studied using TIM-2 model Kortman et al. [55]. This study identified the decreased abundancy of Bifido bacteriaceae and Lactobacillaceae and increased abundancy of Roseburia and Prevotella in response to iron. Furthermore, this study observed metabolome shift from sacharolytic to proteolytic profile with iron supplementation. In another study, beneficial effects of probiotic mixture (Streptococcus thermophilus, Bifidobacterium breve, Bifidobacterium longum, Bifidobacterium infantis, Lactobacillus acidophilus, Lactobacillus plantarum, Lactobacillus paracasei and Lactobacillus delbrueckii subsp. Bulgaricus) on fecal microbiota cultures was demonstrated by using TIM-2 intestine model [82]. In a recent study, the probiotic potential of mango peel was evaluated using TIM-2 intestinal model and this study reported that mango peel promotes the growth of Bifidobacterium and Lactobacillus [87]. In another study, the pectin-derived oligosaccharides extracted from the lemon peel waste shown higher abundancy of beneficial microbes such as beneficial species such as Faecalibacterium prausnitzii and rise in the levels of short chain fatty acids [71]. The xylo-oligosaccharides of sugarcane were also evaluated as a potential prebiotic using a TIM-2 adult colon model [108]. In a study, the effect of polyphenol extracts of Hibiscus sabdariffa was evaluated on microbiota of high fat diet fed mice using TIM-2 model. This study revealed that these polyphenols ameliorate the changes in microbiota induced by high fat diet [88]. Dose dependent effect of Lactulose on gut microbiota reveled that this disaccharide promotes the growth of Bifidobacterium and Lactobacillus and increase in the acetate, butyrate and lactate levels in experiments performed using TIM2 intestinal model [13]. Kovatcheva‐Datchary et al. [56] observed that Ruminococcus bromii was a primary degrader of starch in simulated proximal colon conditions. In another study, the plant sterols effects on gut microbiota of lean and obese population were studied by using TIM-2 model, and this study observed different microbial profiles in lean and obese population. Furthermore, this study identified reduced levels of cholesterol metabolites in lean populations [24]. TIM-2 model has been used to study the effect of phenolic composition of Mexican sauces [15] and functional pasta ingredients [66] on gut microbiota. The effect of high- and low-acetylated galacto-manno-oligosaccharides on elderly gut microbiota was evaluated using TIM2 model. This study suggested that low acetylated galacto gluco-manno-oligosaccharides would be beneficial substrates for elderly populations [72]. In a recent study, the effect of enzymatic feed on simulated swine gut microbiota was studied based on TIM2 model. This study observed the increased abundance of microbial fiber-degrading enzymes upon the pretreated rapeseed meal with cellulose and pectinases [62].

The microbiota preparation for TIM-2 experiments usually involves the fecal sample collection, processing and preservation [1, 55, 87]. The fresh fecal samples were mixed with dialysate solution and glycerol. Pooling of fecal samples from different individuals represents the microbiota of whole populations. Inoculum from single donor and pooled donor inoculums exhibits variations in microbial profiles and activity [1]. The Standard ileal efflux medium (SIEM) has been used in TIM-2 models and it simulates colon material. The SIEM contain various components and experiments involving the SIEM components need to be removed or lowered in concentration. For example, a standard iron source was lowered in SIEM to evaluate the effect of iron in TIM-2 model [55]. Microbiota inoculums allowed to adapt in the SIEM for 16–20 h in the TIM-2 model [24, 87, 88]. This step is followed by 4 h starvation period by disconnecting the SIEM. The sample preparations for food materials and plant materials such as, mango peel need extraction of soluble and insoluble indigestive fibers. The indigestible fiber needs to be predigested with gastric electrolyte solution that consist the pepsin and incubated with pancreatin before introducing the material into TIM-2 colon model [87]. The in vitro colon fermentation period continued up to 72 h after the starvation periods. TIM-2 model allows the controlled temperature, PH, N₂ gas levels throughout the experiment. The advantages of TIM-2 model include the maintaining the physiological water levels, peristalsis, and constant removal of wastes through the dialysis system [64]. Integration of pre digestion and analytical instruments to TIM-2 will be more advantageous to study the effect various external factors on microbiota. The TIM2 model is widely recognized for its precision in replicating the conditions of the human colon, offering unparalleled control over environmental variables such as pH and peristalsis. Its capability to continuously remove waste products enhances the reproducibility of microbial culture experiments. However, the high cost and limited scalability of this model pose significant challenges, particularly for large-scale studies. Despite these drawbacks, TIM2 remains a benchmark for studies requiring high precision.

Simulator of the Human Intestinal Microbial Ecosystem (SHIME)

The Simulator of the Human Intestinal Microbial Ecosystem (SHIME) is a dynamic model developed for the gut microbial culture in a multi compartmental bioreactors that simulates the gut conditions [59]. Typical SHIME model consists of five double-jacketed glass fermenters that are interconnected with peristaltic pumps. First two compartments mimic the stomach and intestines whereas as other three compartments mimic the ascending, transverse and descending colon (Fig. 2). This model designed to maintain the different retention times for the slurry in various compartments with different PH conditions to simulate the GIT. This model allows to operate at 37 °C and it provides the mixing of the slurry with magnetic stirrers. Anaerobic conditions maintained by supplying N₂ gas or N₂/CO₂ gas mixture in all the fermenters [107]. Twin-SHIME refers to the two simultaneous SHIME simulators to study the control and experimental studies. In-order to provide the mucosal environment in the SHIME the mucosal compartment was included in the ascending compartment of the colon. This is named as mucosal SHIME (M-SHIME) and it enriches the mucus binding microbial community in the bioreactors [107].

Fig. 2.

Schematic representation of A Simulator of the Human Intestinal Microbial Ecosystem (SHIME) B Mucosal-SHIME

SHIME offers the investigations on microbial community in a specific region of gastrointestinal tract. The effect of various treatments such as probiotic strains [120], small amounts of insecticides [49, 83], prebiotics [94, 99] and organic juices [29] on microbiota composition and metabolic activity have been studied using SHIME models. Furthermore, the effects of soil arsenic metabolism [116], probiotic-low-fat ice cream [17], dietary emulsifiers such as, polysorbate 80 and carboxymethylcellulose [18], on microbiota were also evaluated by using SHIME model. In a study, the structural integrity of food systems on microbiota was evaluated by using SHIME and this study revealed that higher amount of starch was released at distal region of the colon by intact cotyledon cells when compared with mechanically damaged cells of red kidney beans. Even though there is not much effect of structural integrity on microbial population, this study observed significant increased levels of unutilized starch in the descending colon reactor that treated with intact cotyledon cells [85]. García-Villalba et al. studied the effect of pomegranate extract and mechanism of urolithin production by microbiota using the combination of Twin SHIME and transwell co-culture system. This study observed microbiota mediated production of urolithin, and the results obtained in this method are similar to in vivo studies [37].

The effects of probiotic substrates [102] and butyrate-producing bacteria [106] on mcirobiota were well studied using M-SHIME models. In a study, M SHIME was used in combination with trans well co culture systems to evaluate the effect of a probiotic formulation on antibiotic induced gut dysbiosis. This study reported the significant reduction in the clindamycin induced gut microbiota dysbiosis by a probiotic formulation containing B. coagulans SC208 and B. subtilis HU58 [68]. In another study, propionate producing bacteria such as Lactobacillus plantarum, Bacteroides thetaiotaomicron, resorted the antibiotic induced gut dysbiosis [31]. In an interesting study, inoculation of E. coli mb6212 into the M-SHIME reactors resulted in the transfer its antibiotic resistance plasmid to the gut microbiota [57]. Identifying such kind of microbes could help to study the microbiota mediated antibiotic resistance. In another study, the spore germination, growth and metabolic activity of a probiotic was studied in a SHIME experiment using Bacillus coagulans as an example. The spore germination and metabolic activity were observed in small intestine and colon reactors, respectively [2]. In a recent study, the effect of fecal microbiota transplantation has been tested on vancomycin induced abundancy of pathogenic microbes by using SHIME models. The vancomycin induced dysbiosis of gut microbiota was successfully restored upon the Fecal microbiota transplantation (FMT) treatment and this study provided strong evidence for future implications of FMT in clinical applications [60]. [12] investigated the effect of prebiotic and probiotic combinations on young child microbiota of age between one to two years by using toddler SHIME model. This model consists of the five reactors with working volumes in the range of 28–160 ml [12]

The SHIME model is a flexible intestinal model, and it allows simulating the infant, toddler and adult microbiota conditions in multistage compartments by changing the working volumes in reactors. This model also can be operated by skipping any one or few of the compartments. For example, it can be operated by using only colon reactors by skipping the stomach and small intestine. In case of M-SHIME, either mucus associated colon reactors or lumen associated reactors can also be chosen based on the experimental requirement. Gastric emptying can be done by batch mode or continuous transfer for a specific period. Parallel experiments of control and treatment fecal samples can be cultured in SHIME models and it also allows studying the differences in the luminal and mucosal microbiome. The experimental design plays key role in the SHIME model and it depends on the specific treatments. The host microbiota interactions cannot be studied with in the reactors of the SHIME but samples from the rectors can be applied to transwell systems to study the host responses. The SHIME model excels in simulating different regions of the gastrointestinal tract, allowing for the simultaneous study of luminal and mucus-associated microbiota. This unique capability provides invaluable insights into region-specific microbial activity. Nevertheless, the complexity of its operation and the specialized skills required for its use may hinder its adoption by less experienced laboratories. Overall, SHIME’s ability to mimic gut regions makes it a robust tool for microbial research.

The SIMulator Gastrointestinal (SIMGI) Model

This in vitro model is an automated dynamic system that simulates the gastro-intestinal tract and performs the gastric digestion and colonic fermentation. This system consists of five interconnected compartments represents the stomach, small intestine and three colon regions (Fig. 3). The stomach compartment in this system consists of a reservoir with flexible silicone wall that covered by rigid methacrylate wall. The water is pumped in-between these two walls with controlled temperature and pressure to maintain the peristaltic movements within the gastric compartment. Similar to SHIME the remaining compartments consists of sampling ports, pH control, temperature control and N₂ supply. The movement of feed between compartments is automated and microbiota needs to be stabilized in colonic compartments before the experiment [8], Cueva et al. [23].

Fig. 3.

Schematic representation of SIMulator gastrointestinal (SIMGI) model

In an early study, the SIMGI model was evaluated by microbial stabilization studies upon inoculation of human fecal samples in colonic reactors. This study performed using small intestine, and three colon compartments. The variations in microbial communities observed among the colon compartments after stabilization period. This study reported the overall decrease in abundances of Bifidobacterium and Prevotella and an increase in abundance of Enterobacteriaceae in the colon compartments after stabilization period of microbiota culture. Specifically, increase in abundancy of Bacteroides in ascending colon, and higher abundancy of butyrate producer groups in transverse and descending colons were also observed in this study [8]. Cueva et al. [23] demonstrated the red wine impact on microbiota using SIMGI model. The wine feeding to the colonic bacteria caused the increased levels of microbial phenolic compounds and decreased formation of ammonium ions. This study also observed microbiota alterations in ascending colon upon wine feeding in simulated colonic conditions. In a recent study, silver nanoparticles were examined in simulated human gut conditions using SIMGI, and this study observed no significant changes in microbiota composition and microbial metabolic activity [22]. The gastric digestion and colonic fermentation effects of citrus pectin were evaluated in SIMGI model. The digestion of citrus pectin in stomach and small intestine resulted in reduced hydrolysis of pectin and colonic fermentation resulted in higher abundance of beneficial organisms such as Bifidobacterium spp., Bacteroides spp. and Faecalobacterium prausnitzii. The citrus pectin fermentation also resulted in increased levels of acetate and butyrate in colonic compartments [36] Soluble fibers, that alter the viscosity of the gastric fluids, also can be evaluated using SIMGI model. In a study, Tamargo et al. [95] evaluated the effect of chia seed mucilage on physiological condition of gastric fluids in presence of microbiota. This study observed the viscosity differences in digestion stage but it did not show any significant difference in the viscosity at colonic fermentation stage. Moreover, chia seed mucilage promoted the growth of some organism such as, Enterococcus spp. and Lactobacillus spp.

The SIMGI offers automated digestion in a simulated stomach and small intestines, and subsequent fermentation in simulated colonic regions with microbiota. Unlike TNO in vitro models (TIM1 and TIM2), the digestion and fermentation of feed integrated in a single system. Furthermore, peristalsis is simulated in the stomach compartment that is not provided in the SHIME. The SIMGI’s integration of digestion and fermentation processes in a single system is a major strength, enabling dynamic monitoring of microbial metabolism under simulated gut conditions. However, its limited ability to replicate host signaling pathways and the high resource requirements for operation are notable limitations. Despite these challenges, SIMGI offers valuable insights into microbial fermentation and metabolic processes.

Polyfermentor Intestinal Model (PolyFermS)

PolyFermS model utilizes the fecal samples that are immobilized on gel beads of 1–2 mm for colonization in the reactors. This model consists of inoculum, control and test reactors (Fig. 4). Fecal beads colonized in inoculum reactors for 48 h, and colonized medium transferred to control and test reactors by using peristaltic pumps. Test reactors can be used to study the effect of various parameters such as environment, drugs, nutrition, T in comparison with control reactor [11, 80]. This model claims that biofilm associated microbes are maintained in this reactor for longer periods of fermentation due to fecal immobilization on gel beads. It also offers the stable cultivation of intestinal microbes similar to the fecal donors.

Fig. 4.

Schematic representation of Polyfermentor intestinal model (PolyFermS)

[11] validated the polyFermS models using 6, 10 and 8-year-old male fecal samples and fermentation of microbiota. This study evaluated the metabolic activity and microbial profiles with altered pH in polyFermS model. In another study, the modulatory effects of supplementation of dietary fibers such as, β-glucan, α-galacto oligosaccharide and xylo-oligosaccharide, on proximal colon microbiota were evaluated by using polyfermS models. The typical design of this model consists of first stage inoculum reactor and second stage reactors that consists one control and test reactors [80]. In a study, two-month-old infant gut condition was simulated the by using polyFermS model. The effect of pH and retention time on lactate metabolism was studied by lactate supplementation and addition of lactate-utilizing bacteria in polyFermS reactors. This study observed that the drop in pH increases the abundance of lactate producing bacteria and increased retention time increases with abundance of lactate-utilizing bacteria [19]. In another study, infant gut microbiota was simulated by using polyFermS models to investigate the impact of nucleoside mix and yeast extract containing different levels of nucleotides of infant formula on gut microbiota. This study also evaluted the difference between the infant gut microbiota with and without pathogenic bacteria such as Salmonella enterica ssp. enterica Typhimurium N-15, Clostridium difficile DSM 1296T [28]. Fehlbaum et al. [34] demonstrated the three models of elderly polyfermS colon models. All three models consist of inoculums reactors that simulates proximal colon. The first model consists of three stage reactors represents proximal, transverse and distal colons. Second models consist of inoculums reactors that connected to second stage proximal and transverse-distal colon reactors. Third model consists of inoculums reactor that connected to second stage control and test reactors. The Elderly microbiota composition, and metabolic activity were investigated by using these models up to 80 days [34]. In another study, vegetative cells and spore forms of Clostridium difficile inoculated in the elderly polyFermS model reactors with elderly fecal microbiota. The effect of ceftriaxone induced growth and metronidalzole mediated growth inhibitory effect on Clostridium difficile were tested using polyFermS model [35]. In a swine polyFermS model, swine fecal beads were inoculated in the inoculum reactor that connected to one control reactor and four test reactors. This study simulated the swine proximal colon for 54 days and studied the composition, reproducibility and activity of microbiota in control and test reactors [97]. In a similar design of swine polyFermS model synergistic effect of prebiotics and probiotics on salmonella colonization was studied on immobilized fecal pig microbiota. This study observed synergistic inhibitory effect on salmonella colonization in polyFermS reactors upon treatment in combination with prebiotics (galacto and mannan oligosaccharides) and Bifidobacterium thermophilum RBL67 [98]

The fecal sample preparation involves the immobilization of fecal sample on gellan-xanthan gel beads, and this ensures the increased cell density of the biofilm associated microbiota in the polyFermS reactors. These methods avoid the washout of less competitive microbiota and provides the long-term stability and diversity of microbiota up to 80 days. The main advantage of this method is that it can provide the same microbiota for simultaneous test and control reactors. PolyFermS is distinguished by its ability to maintain long-term stability and microbial diversity, particularly for biofilm-associated microbiota. This makes it a powerful tool for studying dietary and environmental interventions. However, the labor-intensive process of fecal sample preparation and the need for specialized expertise in maintaining anaerobic conditions limit its accessibility. Despite these hurdles, PolyFermS is highly effective for studies requiring stable and reproducible microbial cultures.

Small Volume Bioreactors

The reactors with large volumes are associated with drawbacks in terms of large stabilization periods, complexity in maintaining the reactors such as cleaning, sterilization. Furthermore, large volume reactor requires large volumes of fecal samples or micro biota samples. Small working volumes can overcome these problems and also advantageous in studying the effect of high-cost materials such as drugs, biomolecules, on microbiota. Furthermore, small volumes can be used to experiments with increased throughput [76]. Recently, Cieplak et al. developed a ‘The smallest intestine in vitro model (TSI)’ for in vitro culture of microbiota with less working volumes in short time. This model consists of five reactors to simulate the five samples simultaneously. Each reactor working volume was set to 12 ml and each reactor connected to base vessel, bile vessel, and pancreatic juice vessel and dialysis cassette (Fig. 5). This model mimics the passage of feed through duodenum for 2 h, jejunum for 4 h and ileum for 2 h. The jejunum passage stage involves the pumping of the chyme to dialysis cassette where absorption of nutrients and bile was simulated. The ileum stage initiated by addition of 1 ml inoculum in the reactor. The temperature was maintained at 37 °C throughout the manuscript and varied pH ranges were maintained in duodenum, jejunum and ileum simulation stages. This model is advantageous in studying the effect of ileum microbiota on probiotic survival. [21]. In another study, The Copenhagen MiniGut (CoMiniGut) model and micro-Matrix™ Cassette Set-Up have been tested for the culture of microbiota with working volumes of 5 ml and 6 ml, respectively. The temperature, pH, anaerobic conditions are controlled and maintained throughout the experiments in both the models [111] Minibioreactors offers time saving and high throughput microbiota culture methods with high level of reproducibility. Integration of microfluidics method in mini bioreactors could revolutionize the in vitro micro biota culture methods [76].

Fig. 5.

Schematic representation of one of the reaction vessels of ‘The smallest intestine in vitro model’(TSI) model and its periphery instruments

The specific features and key design considerations of different bioreactor-based culture models are summarized in Table 1. The best bioreactor model should be more relevant to in vivo GIT conditions. Specifically, stomach and small intestine regions should involve in digestion of dietary components and fermentation should be carried out in ascending transverse and distal colon regions. TIM2 and SIMGI model offers the peristalsis to aid the digestion of dietary substances. TIM2 and TSI models are facilitated with dialysis process to collect the metabolites. SHIME model is advantageous for culturing mucus associated and luminal associated microbiota separately. The PolyfermS model is suitable to study the biofilm associated microbiota and this system allows long term stability of culture with enhanced diversity of microbiota. Small volume bioreactors are cost effective and less laborious than other in vitro culture models. Small volume bioreactors provide a cost-effective and less labor-intensive alternative for high-throughput microbiota culture studies. Their reduced material and operational costs make them ideal for screening interventions, especially when resources are limited. However, their simplified design may fail to capture the complexity of gut microbial ecosystems and is less suitable for long-term studies. Despite these limitations, small volume bioreactors are an efficient option for exploratory research.

Table 1.

The summary of bioreactor-based microbiota culture models

	Bio reactor models	Key design considerations	Specific features	Advantages	Disadvantages	References
1	The TNO in vitro model of the colon (TIM2)	Four glass compartments, flexible membrane	Peristalsis, dialysis, Maintains the physiological water levels	1. Precise control of environmental conditions 2. Continuous removal of waste products	1. High cost of setup and maintenance 2. Limited scalability for large-scale studies	[64, 74]
2	The simulator of the human intestinal microbial ecosystem (SHIME)	Five double jacketed glass fermenters	Culture luminal and mucus associated bacteria separately,	1. Allows simultaneous study of luminal and mucus-associated bacteria 2. Enables region-specific microbial activity analysis	1. Requires complex operational skills 2. Limited capacity for studying host–microbiota interactions	[59, 107]
3	The SIMulatorgastro-intestinal (SIMGI)	Five compartments, Stomach compartment consists flexible silicone wall	Digestion of dietary substances, peristalsis	1. Integrated digestion and fermentation processes 2. Allows dynamic monitoring of microbial metabolism	1. Limited ability to mimic host signaling pathways 2. High resource requirements for operation	[8, 23]
4	Polyfermentor intestinal model (PolyFermS)	First stage inoculum reactor and second stage control and test reactors	Long term stability and enhanced diversity of microbiota, Culture of biofilm associated bacteria,	1. Enables biofilm-associated microbiota culture 2. Suitable for studying dietary or environmental interventions	1. Complex fecal sample preparation and immobilization process 2. Requires specialized expertise to maintain anaerobic conditions	[11, 80]
5	Small volume bioreactors	Reactors with less working volumes	Less laborious and cost effective	1. Reduced material and operational costs 2. Ideal for high-throughput screening of interventions	1. Limited applicability for long-term microbial studies 2. May not capture the full complexity of gut microbial ecosystems	[21, 76]

In Vitro Host–Microbiota Culture Systems

Although microbiota culture in bioreactors offers above mentioned advantages, the bioreactors are unable to accommodate the host cells or tissues to study the host microbiota interactions. Specifically, most of the gut microorganisms are facultative anaerobes and it is difficult to co-culture with aerobic host cells or tissues. The microbiota either dies or overgrows while co-culturing with host cells in bioreactors Furthermore, co-culturing requires anaerobic conditions for microbiota and aerobic conditions for host cells and they need to different nutrient media supplementation during culture [109]. Even though, it is difficult to completely recapitulate the lumen environment in the in vitro conditions, substantial progress has been achieved in in vitro culture methods. This section discusses the existing in vitro methods to co culture the host and microbiota cells.

A typical design of transwell system was shown in Fig. 6A. This model consists of a transwell chamber with a semipermeable membrane at the bottom that allows to grow epithelial cells. This chamber is fitted in the culture well that is filled with aerobic cell culture medium. The transwell chamber allows the basal side of the epithelial cells to contact with aerobic cell culture medium and apical side the epithelial cells are exposed to anaerobic compartment. The anaerobic compartment filled with anaerobic culture media and it allows to grow obligate anaerobic microbiota on epical side of this epithelial cells. Barrier function of the epithelial monolayer can be evaluated by the transepithelial electrical resistance (TEER) electrodes fixed at aerobic and anaerobic compartments of this model system. Ulluwishewa et al. [104] demonstrated the co-culturing of obligate anaerobic microbiota with aerobic intestinal epithelial cells using apical anaerobic model of the intestinal barrier. This study validated the transwell system with an obligate anerobic bacteria, Faecalibacterium prausnitzii, grown on caco-2 cells and studied the inflammatory responses. This model has been shown to be promising for studying effects of viable anaerobic microbiota on viable epithelial cells [104]. In an experimental validation of transwell system, leucocyte–caco-2 cell interactions were studied under the presence and absence of Escherichia coli. This study designed four experiments to validate the transwell system. First experiment was conducted with basal compartment without any apical compartment to study the leucocyte and E. coli interactions. Second experiment designed with apical compartment that consists caco-2 cells but E. coli cells are directly interacted with leucocytes in the basal compartment. Third experiment was designed to evaluate the caco-2 cell effect on leucocytes-cell interaction. Third design evaluates the effect of caco-2- and E. coli cells interaction on leukocytes which is actually simulates the in vivo condition. Fourth experiment was designed to study the production cytokines by caco-2 cells in the presence of E. coli and absence of caco-2 cells. This study elucidated the E. coli mediated caco-2 and leucocyte cell cross talk by using transwell model [77]. In a previous study, inflammatory response of pathogenic and non-pathogenic bacteria was evaluated on the caco-2 leucocyte trans well culture system [39]. In a study, the murine intestinal epithelial cell lines and dendritic cells were seeded in the apical compartment and basal compartments the trans well system, respectively. This study elucidated the effect of different bacteria on induction of chemokines by epithelial cells. Furthermore, this study demonstrated the effect of bacterium on maturation of dendritic cells. In another similar kind of study, caco-2 cells and dendritic cells were seeded in apical and basal compartments of a trans well system to study the inflammatory effects of L. paracasei CNCM I-4034 and Salmonella. Both the studies demonstrated the inflammatory effects of dendritic cells in presence and absence of epithelial cells [10]. Zhang et al. demonstrated a gut microbiome (GuMI) physiome platform for co culture of bacteria and epithelial cells. This model consists of an apical unit, trans well system and basolateral units. This model is an advanced transwell system, and it controls the fluid flow and oxygen gradient in the transwell systems. This study elucidated the hypoxia responsiveness, anti-inflammatory effects of F. Prausnitzii on stable colon epithelial cells [118].

Fig. 6.

Schematic representation of in vitro host–microbiota systems A transwell model B HoxBan Model C Hanging basket model D 3D intestine scaffold intestine model E Rotating wall vessel model

The Human Oxygen-Bacteria anaerobic (HoxBan) system is a laboratory in vitro model that allows to co-culture of human epithelial cells and mutualistic bacteria in a 50 ml Falcon tubes (Fig. 6B). This system was developed by Sadabad et al. [86] to co culture aerobic caco-2 cells and an anaerobic gut bacterium, Faecalibacterium prausnitzii. The 40 ml of F. prausnitzii culture in yeast extract, casitone, fatty acid and glucose (YCFAG)-agar medium was solidified in a 50 ml Falcon tubes. The cover slip containing pre cultured caco-2 cells was placed on YCFAG-agar medium and 10 ml of Gibco Dulbecco's Modified Eagle Medium (DMEM) medium added to the falcon tube. The falcon tube lid was slightly opened to co culture aerobic cells in top DMEM medium and anaerobic cells in bottom YCFAG-agar medium. This study demonstrated the mutualistic effects that promoted the growth of F. prausnitzii cells and proinflammatory effects in caco-2 cells in co culture. In a previous study, hanging basket co culture model was used to study the interactions of microbes of bioflim that was hanged on the culture well. In this study, the multi bacterial species (Porphyromonas gingivalis, Fusobacterium nucleatum, Aggregatibacter actinomycetemcomitans and Streptococcus mitis) bioflim was developed on a coverslip. A mono layer of epithelial cells (OKF6-TERT2) grown on culture wells and coverslips containing bioflims hanged in the wells with the help of cell culture inserts. The bio film and epithelial cells maintain a distance of 0.5 mm in the wells. This study evaluated the effect of inflammatory effect of resveratrol on epithelial cells and bactericidal effect of chlorhexidine in co-culture system [73]. A typical design of hanging basket design was illustrated in Fig. 6C.

Chen et al. demonstrated 3D intestine model using scaffolding proteins to mimic the structure and function of human intestine. This model consists of hallow lumen surrounded by pours scaffold bulk (Fig. 6D). The hallow channel was fabricated by using polydimethylsiloxane (PDMS) molds and 3D scaffold was fabricated by using silk fibroin protein. Th PDMS molds generates the patterns within the lumen to mimic villi like structures. The caco-2 and HT29-MTX cells were seeded in the lumen and human intestinal myofibroblasts were seeded in the porous scaffold bulk. Bacterial binding experiments and co-culture of Y. Pseudotuberculosis and Lactobacillus rhamnosus were performed by using 3D engineered system. This model exhibited the continuous mucous secretion, low oxygen levels in the lumen and binding to the bacterial strains. This model lacks the continuous flow of fluids and nutrients and may not offer long term survival of tissues [20]. In a study, Rotating Wall Vessel (RWV) bioreactor was implicated for studying the bacterial invasion into intestinal epithelium. RWV consists of A slow turning lateral vessel with central core (gas exchange membrane), sampling ports and a filling port (Fig. 6E). The extracellular matrix coated porous microbeads were introduced in the vessel along with epithelial cells. These epithelial cells grow on rotating beads to attain 3D growth within the slow turning lateral vessel. This study reported that salmonella can invade the epithelial cells without effector proteins such as,salmonella pathogenicity island-1, salmonella pathogenicity island-2 and type three secretion system [81].

In vitro methods offer to study the host microbiota interactions that are impossible or difficult to study in in vivo systems. Most of these in vitro methods include the cancerous cells to from mono layers. Cancerous cells may not recapitulate actual epithelial cell signaling. These cell lines may lack the some of the in vivo epithelial cell characteristics like, mucus production, improper tight junctions. Moreover, these mono layers contain single type of cells whereas other intestinal cells such as goblet cells, endocrine cells are ignored in these experiments. The methods mentioned in this section lack the secondary epithelial structures like villi and crypts. Despite some limitations, these methods have been provided valuable insights on inflammatory mechanisms of host microbiota interactions.

Microfluidic Devices for Host–Microbiota Culture

The use of conventional in vitro methods is not suitable for host–microbiota interaction studies such as bacterial cell adhesion to host cells, host cell response, continuous replenishment of media and continuous removal of metabolites. The main challenge in co-culturing is maintaining the viability of all types of cells. The gut microbiota competes to grow with human epithelial cells and some organisms of microbiota are cytotoxic to human cells in in vitro conditions. The animal models for host–microbiome interactions do not completely resemble the human cell responses. Multi layered microfluidic devices facilitate the co-culture of human and bacterial cells by the use of semipermeable barriers or compartment models. Applications of organ on-a-chip microfluidic devices for host–microbiota interaction is at initial stage of research and gut on a-chip could be a potential alternative to animal models. The different culture conditions for microbiota and human cells can be created in a single, miniaturized microfluidic device. The oxygen gradient is important factor in design of gut on a-chip models for co-culturing human cells and microbiota. Mimicking peristalsis like motion and luminal fluid flow are critical for gut on a-chip models to simulate human gut conditions. Recently, microfluidic research gained the momentum for host–microbiota interaction studies and four types of microfluidic devices have been developed so far. Table 2 provides a summary of microfluidics-based host–microbiota culture systems, highlighting variations in fabrication materials, porous membranes, microbiota, host cells, co-culturing times, and oxygen sensing capabilities across different systems.

Table 2.

Summary of microfluidics-based host–microbiota culture systems

	HMI module	HuMiX device	The human Gut-on-a-Chip microphysiological system	Pneumatically-controlled island trapping system
Fabrication material	Data Not available	Super-soft silicone sheets	Poly (dimethyl siloxane) (PDMS) polymer	PDMS
Porous Membrane	Polyamide membrane	Polycarbonate membrane	PDMS membrane	–
Microbiota	Lactobacillus rhamnosus GG	Lactobacillus rhamnosus GG, Bacteroides caccae	Lactobacillus rhamnosus GG, B. fragilis	E. coli
Host cells	Caco-2 cells	Caco-2 cells	Caco-2 cells,human intestinal microvascular endothelial cells (HIMECs)	HeLa cells
Co culturing time	Up to 48 h	Up to 24 h	Upto 72 h	Up to 54 h
Oxygen sensing	–	Integrated optical sensors (optodes)	Oxygen sensor spots	–

Marzorati et al. [70] developed a Host–Microbiota Interaction (HMI) module to study the effect of specific component on luminal microbial community and host signaling. This module consists of two compartments separated by polyamide membrane that has a pore size of 0.2 µm (Fig. 7A). A mucus layer was mounted on the membrane to colonize the gut bacteria. The polyamide membrane and mucus layer allow the exchange of oxygen and metabolites between the upper and lower compartments. The upper compartment mimics the luminal side of the gut and lower compartment mimics the host. This Study demonstrated the effect of dried modified Saccharomyces cerevisiae fermentation products on luminal microbial community and host signaling. The SHIME is combined with HMI to provide the continuous supply of microbiota culture into microbial chamber of HMI device. The SHIME consists of the three reactors that mimics the stomach, small intestine and ascending colon. The culture medium was supplemented with yeast fermentate and the fecal sample was inoculated in the last reactor of SHIME that mimics the ascending colon. The microbial culture from this reactor was pumped to HMI and the host response was evaluated in the enterocytes that were present in the lower compartment of HMI module. This study also observed the reduced IL-8 production in enterocyte cells upon addition of dried yeast fermentate to culture medium. In another study, Shah et al. [92] developed a microfluidics-based Human–microbial crosstalk (HuMiX) device to evaluate the host microbiome interactions in a simulated gastrointestinal tract. This device allows the co-culture of human and microbial cells in spiral shaped microchambers. This device consists of three chambers, namely, microbial micro chamber, epithelial cell microchamber, perfusion chamber (Fig. 7B). Spiral shaped microchannels were fabricated on super-soft silicone sheets and it forms gasket. Perfusion chamber, epithelial cell microchamber were attached with semipermeable–porous polycarbonate membranes. Finally, all the three gaskets and two polycarbonate membranes were enclosed in two polycarbonate enclosures. This device was associated with the integrated optical sensors and electrodes to measure the oxygen levels and transepithelial electrical resistance, respectively. This study demonstrated the effect of culturing the commensal bacteria on human epithelial cells under aerobic and anaerobic conditions. Molecular analysis of host microbiome interactions revealed that HuMiX device is able to mimic the host–microbe interface. In another study, Kim et al. [52] developed a Gut-on-a-Chip micro physiological system that mimics the human gut. This device offers the co culturing of intestinal flora on human epithelial cells with peristaltic like movements and continuous flow of fluids. This device consists of top and bottom microchannels separated by a porous membrane (Fig. 7C). Human epithelial cells cultured on the porous membrane so that the top channel mimics the intestinal lumen-capillary tissue interface and bottom channel designed to fluid flow. Gut-on-a-Chip micro physiological system was fabricated with three polydimethylsiloxane (PDMS) layers, namely upper layer, lower layer and a middle porous membrane layer. Porous membrane was cured on silicon wafer containing post arrays with circular pillars of 10 μm in diameter, and 25 μm in height. The upper and lower layers are fabricated with three parallel chambers, and middle chambers of both the layers were separated by a porous membrane. Bonding of the three layers forms the top and bottom channels with two parallel vacuum chambers. The change in suction pressure in the vacuum chambers lead to the stretching and relaxation of porous membrane [45]. Interestingly, caco-2 cell monolayer on a porous membrane exhibited villi like structures upon continuous flow of media in top and bottom channels [52]. In recent study, oxygen sensor discs were embodied in the inlet, middle and outlet channels of the device. The oxygen levels were determined by the measurements of fluorescence emitted from optical sensitive oxygen sensor discs [48].

Fig. 7.

Schematic representation and working principles of microfluidics-based host–microbiota systems A HMI module B HuMiX device C The human Gut-on-a-Chip micro physiological system D Pneumatically controlled island trapping system

Kim et al. [53] demonstrated a microfluidic device that consists of epithelial cell zone and bacterial island region separated by pneumatic channel. Pneumatic trapping was controlled by a PDMS wall that can be lowered and raised to close and open the bacterial island form epithelial cell zone (Fig. 7D). This device allows the co culture of bacterial cells and epithelial cells. The problem of overgrowing the bacterial culture on epithelial cells can be overcome by simultaneously co-culturing in different compartments of the device, and it also allows to provide the different media for bacterial culture and epithelial cell culture. This study demonstrated the introduction of pathogenic bacteria to commensal islands and subsequent infection to epithelial cells upon the opening the pneumatic channel. This study shown the effect of commensal bacteria on pathogenic bacteria infectivity by co culturing bacterial and epithelial cells.

Although, microfluidic devices have been shown substantial success in mimicking in vivo like gut conditions, still it has long way to go. Co-culturing microbial and mammalian cells in microenvironments is still challenging due to species specific growth conditions. The co-culture of microbial cells in direct contact with mammalian cells is critical in designing the gut on chip models. Growing the cells in different layers provide the in-direct contact of the bacterial cells to mammalian cells, and molecular signal exchange through the semipermeable membranes. Compartment microfluidics models need to be developed to co culture the pathogenic organisms with commensal microbes and mammalian cells. The oxygen gradient is achieved some extent in micro chambers, and it is essential to co-culture the cells of different requirements of oxygen levels. The temperature and pH control also need to be monitored in the micro chambers to maintain the optimum growth conditions of various types of cells. Incorporation of temperature and pH sensors would be advantageous to mimic the in vivo like conditions in microchambers. The development of three dimensional organotypic models would help to study more insights of host microbiome interactions. Integration of miniaturized devices to molecular analysis instruments would be added advantage in studying the host microbiome interactions at molecular level. Multipurpose microfluidics chips need to be developed for host microbiome interactions such as, drug–microbiome interactions, microbiome related therapeutics, gut infectivity and inflammatory responses.

Ex-Vivo Models for Host–Microbiota Culture

In vitro systems consist the monolayers of only single type of cells and it does not recapitulate the in vivo like cell diversity in the intestinal barrier. Cell signaling mechanisms elucidated from in vitro methods may not be completely relevant to the in vivo mechanisms [41]. The diffusion chambers with an intestinal segment allows to study the translocation and mucosal invasion of the of microorganisms across the membranes. The explants form animal intestinal barriers offer cell cross talk as it contains different cell types. Membrane permeability and mucus uptake assays by ex-vivo methods show high relevance to the in vivo data [114]. Organ culture is an alternative to the explants and it allows to culture the intestinal segments form the stem cells of the biopsy samples. Even though, there are some difficulties in studying the host microbe interactions, organ cultures provide mechanistic insights that are highly relevant to the sample donors [84]. Microbiota culture by using ex vivo systems has started in the recent past and more advances in these methods are expected in coming years.

Ussing Chamber

Ussing chamber was developed to study the epithelial transport and intestinal permeability of drugs. The excised animal intestinal tissue segment was placed in a chamber filled with physiological buffer. This tissue segment divides (horizontally) the chamber into mucosal side and serosal side. The chamber is pumped with carbogen (O₂–CO₂ gas mixture) that supplies the oxygen to cells and circulates the fluids in both sides of tissue segment. Viability of intestinal barriers was monitored by measurements of the TEER, the potential difference, and the short-circuit current using two electrodes placed inside the chambers [105]. The typical design of Ussing chamber was illustrated in Fig. 8.

Fig. 8.

Schematic representation of Ussing chamber

The TEER is decreased when the tight junctions between the membrane cells loosens across the membrane, and the measurement of TEER provides the tight junction functional evaluation of membranes in using chamber experiments [54]. Paracellular permeability is evaluated by using non-ionic materials such as, mannitol [114]. Majority of the Ussing chamber applications involved the evaluation of treated intestinal segments or explants obtained from in vivo experiments [61]. Direct microbial treatments in mucosal chambers of the Ussing chamber with intestinal segments provided the interesting insights in the translocation of microbes. In a study, colonic mucosal invasion of E. faecalis OG1X:pAM721 and E. faecalis OG1X were evaluated in an Ussing chamber that was mounted with colonic mucosa of Wistar rats. This study observed the colonic invasion of E. faecalis OG1X:pAM72 was higher than E. faecalis OG1X due to the presence of a plasmid encoded adhesin, aggregation substance, in former E. coli strain [46]. Jutfelt et al. proved the Translocation of pathogenic and live bacterium, Aeromonas salmonicida, across the fish intestine explant in an using chamber system [50]. The cytokine response to C. difficile strains was evaluated using an Ussing chamber that mounted with T84 cell monolayer [47]. The effects of addition of bacteria, potassium cyanide and oxygen deprivation on potential difference of animal intestine membrane were evaluated using an Ussing chamber system. Furthermore, this study also showed the transmucosal passage of bacteria across the intestine segment [7]. Based on the Using chamber desgin, a Netherlands based company developed an InTESTine^Tm ex-vivo system. This system consists of apical compartment and basal compartments separated horizontally placed ex-vivo porcine intestinal tissue. Thea apical side of the tissue exposed in apical compartment and basal side of the tissue exposed to basal compartment.

Intestinal Enteroids and Organoids

Intestinal enteroids are the three dimensional, multilobulated and single lumen cell structures that exhibits the similar composition and functions of intestinal epithelial cells [79]. Intestinal enteroids cultured from isolated epithelial crypt fragments that consists of intestinal stem cells. The stemness of the intestinal entreoids controlled by stemness factors such as, wnt, notch, epidermal growth factors, and the bone morphogenetic proteins, noggin, R-spondin-1. Among these stemness factors. Noggin is responsible for the crypt like structure formation. The enteroids consists of all kind of intestine epithelial cells such as paneth cells, goblet cells entero cell, entero R endocrine cells, these enteroids also includes intestinal stem cells [100]. Similar to Intestinal enteroids, the intestinal organoids are three-dimensional epithelial cell structures supported by mesenchyme cell layer. The enteroids are structurally differ from organoids by the absence of mesenchyme cell layer. The organoids are derived from pluripotent stem cell culture by using the stemness factors such as actin-A, fibroblast growth factor, bone morphogenetic protein-4, LY294002, and CHIR99021. The culture of organoids takes more than a month, whereas enteroids can be cultured within a week [117]. The enteorids may be associated with the cells of disease condition of individuals and this may be avoided with organoids cells [79].

In a study, mouse intestinal organoids co-cultured with with lamina propria lymphocytes to evaluate the beneficial effects of Lactobacillus reuteri D8 on epithelial cell barrier. This study observed the recovery of damaged epithelial cells by stimulating regeneration intestinal stem cells in the presence of Lactobacillus reuteri D8 [43]. In another study, Hill et al. observed the epithelial barrier function and integrity upon the microinjection of Escherichia coli strain ECOR2 in to a matured human intestinal organoid. The maturation of human organoid was performed by placing it in the kidney capsule of a mouse for 10 weeks [42]. In a study, the micro injection of Clostridium difficile and stool samples from infected person into a human intestinal organoid shown inhibitory effect on Na/H exchanger 3 (NHE3) and that resulted in the increased Na+ ions and alkalinity of the lumen [32]. In another study, Clostridium difficile microinjection in to a human intestinal organoid resulted in disruption of the cells.

Organoid lumen is surrounded by epithelial cell barrier and introduction of microbiota into the organoid lumen can be performed in two ways. The microinjection is the best way to delivery of microbiota, nutrients and other materials into the lumen [30]. In a recent study, Williamson et al. developed automated, high through put microinjection platform to introduce the micro biota into the lumen of organoid. This platform is promising for the microbial quantification and monitoring the lumen physiology [112]. The alternative way of introduction of microbiota can be performed by incubating the microbes with fragments of organoid and subsequently allowing to grow to reform organoid [25]. Microbiota interaction with organoid cells also can be studied in 2D culture systems but this method does not mimic in vivo like 3D structure [33]. Figure 9 shows above mentioned methods of organoid-microbiota culture.

Fig. 9.

Organoid based host microbiota cultures

The Ussing chamber is promising for studying the translocation of microbiota into the epithelial cells but these intestinal segments cannot live more than 5 h in the Ussing chambers. Moreover, the current applied to monitor the membrane integrity in using chamber may affect the microorganism–host interactions. Even though, organoid or enteroid cultures have exhibited better in vitro and in vivo correlation, the use of these methods limited due to difficulty in introducing the microorganisms in to the lumen of organoid or enteroid. Creating anaerobic environment is difficult but recent research showed that microbial growth within the lumen of organoid generates the hypoxia in the lumen [112]. Integration of advanced technologies such as, automated microinjection, in organoid culture can enhance the utility of organoids for studying the host microbiota interactions.

Integration of In Silico Tools in Microbiota Culture Systems

The field of microbiome research has experienced a transformative shift with the integration of in silico tools, which have significantly enhanced our understanding of the intricate interactions between the gut microbiota and its host. These computational approaches have ushered in novel insights into the functional dynamics, metabolic activities, and ecological relationships within microbial communities. By enabling the development of predictive models and simulations, these tools have become indispensable for guiding experimental designs and optimizing microbiota culture systems [75, 101]. A major breakthrough in this domain is the convergence of meta-omics technologies—such as metagenomics, metatranscriptomics, and metaproteomics—with computational biology. High-throughput meta-omics methodologies, coupled with robust bioinformatics workflows, allow researchers to comprehensively characterize the taxonomic composition and functional potential of microbial communities. This integration has illuminated the complex biochemical pathways and interspecies interactions that underpin host–microbiota dynamics [44, 90]. The advent of machine learning (ML) and other artificial intelligence (AI) algorithms has further elevated the analytical power of microbiome research. These technologies enable the processing of vast, multidimensional datasets generated by microbiome studies, uncovering patterns and relationships that would otherwise remain obscure. Predictive models derived from these approaches can simulate microbial community dynamics, offering foresight into how environmental factors, dietary interventions, or pharmaceutical compounds may influence the gut microbiome [3]. In silico tools have also been pivotal in advancing the design and optimization of microbiota culture systems, such as gut-on-chip platforms and bioreactors. Computational simulations of gut environments, including pH gradients, nutrient flux, and oxygen levels, inform the configuration of culture systems to better mimic in vivo conditions. For instance, the use of in silico models helps to determine ideal operating parameters for bioreactors or to design gut-on-chip systems that replicate physiological features such as peristalsis and mucus secretion.

Despite the significant strides made, challenges remain in bridging the gap between computational predictions and experimental outcomes. Advancements in computing power, algorithm development, and data integration are crucial for unlocking the full potential of in silico tools in microbiota research. As these technologies continue to evolve, they are poised to transform our approach to understanding and manipulating the gut microbiome.

Conclusions and Future Outlook

Microbiota culture methods have emerged as indispensable tools for understanding microbial composition, metabolite production, and host–microbiota interactions. Advanced in vitro models, such as SHIME and M-SHIME, enable colon-region-specific studies, providing insights into gut microbial alterations under the influence of drugs and nutrients. TIM-2 and SIMGI models have demonstrated the digestion and fermentation of indigestible food in multistage gut compartments, with TIM-2 offering peristalsis simulation and SIMGI excelling in operational parameter control. PolyfermS, with its stable and reproducible microbial cultivation, is ideal for dietary and environmental intervention studies. Furthermore, small-volume bioreactors and models like TSI offer cost-effective and high-throughput solutions for exploratory research, albeit with some limitations in mimicking the complexity of gut ecosystems.

Microfluidic-based gut-on-chip models represent a transformative step forward by providing miniaturized environments with oxygen gradients, waste removal, and nutrient replenishment, which mimic in vivo gut conditions. These models facilitate detailed studies of host–microbiota interactions, host signaling pathways, and microbial dynamics in ways that traditional systems cannot. Ex vivo methods, employing host tissues or explants, provide the closest approximation to in vivo systems, enabling the study of personalized host responses and advancing the field of precision medicine. The next phase of microbiota research should focus on integrating advanced technologies, such as 3D printing and tissue engineering, with microfluidic platforms to develop more sophisticated and physiologically relevant models. Personalized medicine applications using patient-specific microbiota and tissues will enhance the precision of therapeutic interventions. Efforts to improve co-culture systems and high-throughput screening capabilities will accelerate drug and dietary testing. Additionally, combining experimental systems with computational in silico models will enable more predictive and mechanistic insights. These advancements will collectively drive the creation of efficient, scalable, and accessible microbiota culture systems.

Authors Contributions

All aspects of this review paper, including study conception and design, literature review, analysis, and manuscript preparation, were carried out by VKY.

Funding

No funding was received to assist with the preparation of this manuscript.

Data Availability

All data generated or analyzed during this study are included in this article.

Declarations

Conflict of interest

The author declares that there are no conflict of interest.