Organoids as a tool to study the impact of heterogeneity in gastrointestinal epithelium on host–pathogen interactions

Summary

The epithelium of the gastrointestinal (GI) tract has been extensively characterized using advanced histological and RNA sequencing techniques, which has revealed great cellular diversity. Pathogens, such as viruses and bacteria, are highly adapted to their host and often exhibit not only species-specificity but also a preference or tropism for specific GI segments or even cell types—some of these preferences are so specific, that these pathogens still cannot be cultured invitro. Organoid technology now provides a tool to generate human cell types, which enables the study of host cell tropism. Focussing on the GI tract, we provide an overview about cellular differentiation in vivo and in organoids and how differentiation in organoids and their derived models is used to advance our understanding of viral, bacterial, and parasitic infection. We emphasize that it is central to understand the composition of the model, as the alteration of culture conditions yields different cell types which affects infection. We examine future directions for wider application of cellular heterogeneity and potential advanced model systems for GI tract infection studies.

Graphical Abstract

host–pathogen interactions, human, stem cells, virus, bacteria

Introduction

The gastrointestinal (GI) tract is a structurally and functionally diverse organ complex. It is not only responsible for the processing and absorption of nutrients, but also harbours the GI microbiome and acts as a barrier against pathogens.

Classical models for studying host–pathogen interactions and disease pathogenesis have been animal models and tumour-derived cell lines. Animal models often do not recapitulate human pathology, and especially in the field of host–pathogen interaction, this can be crucial since many pathogens are highly adapted to their hosts. Therefore, animal models do not always pose a suitable model for understanding infectious diseases in humans. Classical cancer cell lines are of human origin, yet the common human GI epithelial cell lines are usually derived from adenocarcinoma samples. These models have altered intracellular signalling pathways leading to changed proliferation and cell responses to external stimuli. In addition, one cell line represents the genetic background of one single patient and only a limited number of cell lines are available. Therefore, individual host diversity is hardly reflected in classical cell line experiments.

Advances in stem cell biology have led to the development of stem cell-derived organoids, complementing the model spectrum. Organoids are defined as stem cell-derived, multicellular cell cultures, in which the cells self-organize and mimic at least some of the organ-specific functions, such as mucus secretion by gastric organoids [1]. Organoids can be generated from two sources: tissue-resident adult stem cells (ASC) and pluripotent stem cells (PSC), including induced pluripotent and embryonic stem cells [2, 3]. Together, the two technologies cover an impressive range of organs that can be mimicked in vitro. They can recapitulate aspects from tissue development to organ-specific physiological functions and patient heterogeneity. For infection biology studies, cell type diversity, innate immune signalling pathways, and genetic host diversity are of high interest, since a better understanding of pathogen tropism is expected to lead to the development of better-targeted therapeutic strategies.

Recent advances in the establishment of human organoids have been covered in several insightful reviews [4–7]. Different methods of using organoid models for studying host–pathogen interactions have been reviewed recently [8]. In this present review, we will complement the literature focussing on the heterogeneity of host–pathogen interactions, pathogen tropism towards host cells, and how organoids can be useful to study these. Since the GI tract is one of the major sites of infection, we will focus on this organ complex.

GI tract architecture and cellular composition

The GI tract is covered by a single layer of epithelium, which is organized into invaginations, called glands in the stomach and crypts in the intestine (Fig. 1). This tissue architecture follows the same principle from the stomach to colon. The small intestine has additional finger-like protrusions called villi to enhance the surface area and promote nutrient absorption. Glands or crypts and villi are lined by a single-layered columnar epithelium.

Figure 1.

GI epithelium. The key architectural elements of the GI epithelium are invaginations called gastric glands in the stomach, crypts in the intestine, and additional extrusions called villi in the small intestine. Each GI tract segment is composed of several cellular populations, some of which have short life spans of 3–5 days like pit cells in the stomach and enterocytes in the small intestine. The integrity of the GI epithelium is maintained by stem cell populations in the glands and crypts. As stem cells proliferate, they generate progenitor cells (TA cells in the intestine and isthmus cells in the stomach). As progenitor cells move upwards the crypt villus axis, they leave the stem cell niche and differentiate into specialized GI cell types. GI cells can be separated into absorptive cells, such as enterocytes/colonocytes in intestine and colon, and secretory cells, such as mucus-producing pit/neck cells in the stomach and goblet cells in intestine/colon as well as enteroendocrine cells and tuft cells. Furthermore, parietal cells secrete the gastric acid. Small intestinal Paneth cells secrete antimicrobial peptides and stem cell niche factors. M cells regulate microbiota and immune system interactions.

Within the epithelium, different cell types fulfil specific tasks. The gastric cell populations are parietal cells, chief cells, enteroendocrine cells (EEC), pit and neck cells as well as their progenitors and stem cells. The main function of parietal cells is to secrete hydrochloric acid and intrinsic factor. Chief cells secrete enzymes for lipid and peptide digestion, while EEC release local hormones such as gastrin and ghrelin, regulating digestive functions and maintenance of gastric epithelium. Pit and neck cells secrete mucin (MUC) 5AC and MUC6, respectively, that form a protective mucus layer above the gastric epithelium. Stem cells are proposed to localize at the base of the gland and at the isthmus region of the gland [9].

Like gastric stem cells, intestinal stem cells are localized at the base of intestinal crypt. The intestinal crypt is also populated by progenitor cells and Paneth cells, the latter supporting stem cells by secreting ligands for WNT and Notch receptors. The main cell lineages composing intestinal and colon epithelium are absorptive (enterocytes and colonocytes) and secretory (mainly Paneth, goblet, EEC, and tuft cells). Enterocytes and colonocytes produce enzymes and absorb nutrients including water. Goblet cells secrete mucins, generating a soft mucus layer in the small intestine and an additional thick mucus in the colon. As in the stomach, EEC are also present at low frequency in the intestinal epithelium. Colon crypts lack Paneth cells, but harbour deep crypt secretory cells, EEC and tuft cells [10, 11].

All regions, stomach, small intestine, and colon are subject to constant tissue turnover with specific cell types having life spans from several months (gastric chief cells and intestinal Paneth cells) to only 3 days (gastric pit cells and intestinal enterocytes) [9, 12]. The constant need for new cells is fuelled by the stem cells, which divide and migrate away from the stem cell niche while differentiating (transit amplifying [TA] cells in the intestine, and isthmus cells in the stomach). In the small intestine, another initially postulated population of stem cells, the ‘+4 stem cell’, named after the position in the crypt, just above the base compartment, was revealed to be an early secretory progenitor [13, 14]. In the stomach, a second population of stem cells is postulated in the isthmus region of the gland, characterized by high cellular turnover and expression of stathmin. This population of isthmus cells is responsible for fuelling the high turnover of cells in the gastric pit region [15].

In these specific locations (base or isthmus), stroma, epithelial cells, and possibly additional cells define a microenvironment for the stem cells, called stem cell niche, by secreting molecular cues such as WNT, epithelial growth factor, bone morphogenic protein (BMP) and its inhibitor Noggin, Notch, and fibroblast growth factors [16]. Understanding stem cells, their differentiation, and pathways maintaining stem cell identity has led to the development of organoids.

Organoid generation and directed differentiation

To generate organoids, stem cells are embedded in a basement membrane matrix, where they proliferate and self-organize into 3D structures [17]. The stem cells can be adult, tissue-resident stem cells, or PSC. Organoids are cultured in a cocktail of growth factors that simulate the stem cell niche cues in order to promote stem cell proliferation and long-term expansion [17, 18]. The cells anchoring to the matrix promote the polarization of the epithelial cells with the apical surface of the cells exposed to the lumen, and the basal side of the cells oriented to the supportive extracellular matrix [2, 19]. In the absence of the matrix, organoids can reverse the polarity [20].

Since organoids originate from stem cells, they have the intrinsic capacity to differentiate into a broad range of cell types. If PSC are used as a source for the organoids, they have the capacity to generate cells from all three germ layers, thus these organoids can contain mesenchymal cells and epithelial cells. If ASC are used as a source for the organoids, they have the capacity to generate cells from the organ of origin, for example, small intestinal ASC will only generate small intestinal epithelium, with the potential of generating all epithelial cell types. A single organoid can be composed of stem cells, progenitor cells, and differentiated cells.

ASC-derived GI organoids are composed of different epithelial cell types, and it is possible to enrich specific cell types by adjusting their growth conditions [16] (Fig. 2). In general, withdrawal of WNT signalling (promoted by WNT and R-spondin in the culture media) mimics the low concentrations of WNT that progenitors experience when leaving the stem cell niche, therefore leading to cell differentiation [17]. Additionally, several cues can promote stem cell differentiation towards a specific cell type. In gastric organoids, standard growth conditions that include WNT and R-spondin promote expansion, thus the organoids are enriched in progenitor cells and especially MUC6- and PGC-expressing neck/chief progenitor cells but do not contain many differentiated MUC5AC-expressing pit cells. This can be enhanced by the addition of nicotinamide (Nic). Withdrawal of WNT (and Nic) leads to increased differentiation towards pit cells [1, 23, 24].

Figure 2.

Directed differentiation of GI organoids. Modulation of GI organoid composition to increase the presence of a specific cell type by adding or removing selected growth factors is illustrated. In the stomach, the removal of WNT directs cell differentiation towards the pit cell phenotype, while the addition of Nic induces cell differentiation towards gland mucus cells. Secretory cells, such as goblet cells, can be observed when p38i and Nic are removed from the culture medium. They will increase when Notch inhibitors such as DAPT or DBZ are added. Differentiation to EEC can be further increased by the removal of WNT and TGFβ inhibitor (TGFbi) and the addition of BMP and MEK inhibitor (MEKi). Paneth cells can be induced by a stepwise protocol first removing p38i and Nic, and then reducing Wnt and adding IL-22. Growth conditions replacing p38i and Nic with IGF-1 and FGF-1 allow long-term expansion of organoids harbouring secretory cells such as goblet cells. EECs can also be induced by overexpression of neurogenin 3 (NGN3). M-cell frequency can be increased by the addition of RankL. Enterocytes can be enriched by WNT removal and reduction in R-spondin. For specific infection models, different pathogens have been successfully cultured using alternative media composition, such as HNV [21] and SARS-CoV-2 [22].

In intestinal organoids, the expansion culture medium for mouse small intestinal organoids does not require WNT, and therefore the differentiation of the organoids is quite remarkable with all cell types present. In human intestinal organoids, WNT is necessary for expansion and similar as in the gastric organoids, constant WNT stimulation also favours the presence of many progenitors in human intestinal organoids. Removal of WNT directs intestinal stem cells to the adsorptive lineage and organoids are thus mostly composed of enterocytes [17]. Addition of interferon gamma may further increase differentiation to enterocytes [25]. Removal of p38 mitogen-activated protein kinase inhibitor (p38i) and Nic from standard growth conditions leads to differentiation to the secretory lineage cells such as goblet cells [17]. Inhibition of the Notch pathway by a gamma-secretase inhibitors (2S)-N-[2-(3,5-Difluorophenyl)acetyl]-L-alanyl-2-phenyl-glycine 1,1-dimethylethyl ester (DAPT) and dibenzazepine (DBZ) further increases numbers of goblet cells [17, 26, 27]. In modified growth conditions with insulin-like growth factor 1 (IGF-1) and fibroblast growth factor 2 (FGF-2), cultures with higher numbers of secretory cells can also expand long term [18]. Addition of IL-22 into the organoid medium may increase the presence of Paneth cell marker-expressing cells [28]. Ectopic expression of an inducible neurogenin-3 can direct differentiation to EECs, similarly as a stepwise protocol removing p38i, Nic, TGFβ− inhibitor, and WNT and adding DAPT, MEKi, and BMP [29, 30]. Rank ligand (RankL)-stimulated cultures showed higher microfold cell (M cell) marker expression [31–33], while the addition of retinoic acid and lymphotoxin could increase the M-cell profile differentiation in human organoids [34]. Thus, organoid technology has provided an unprecedented source of human primary cells for experimentation in basic sciences, including infection biology.

The importance of cellular differentiation in infection

In infection biology, cell differentiation is relevant because it affects surface marker and intracellular host factor expression, chemokine secretion, and other defence mechanisms. These factors could result in different susceptibilities of the epithelium and its abilities to interact with or react to pathogens. As a result, microorganisms often exhibit a particular tropism, which means that they infect a particular organ (organ tropism) or specific cell types (cell tropism). For example, the expression of a suitable receptor is essential for intracellular pathogens, such as viruses. Entry into host cells is essential for replication and further dissemination. Human norovirus (HNV) is a good example, as some genotypes of norovirus HNV rely on fucosyltransferase 2 (FUT2) expression for cell entry. Receptor for HNV entry in human cells is yet to be discovered, but it has been shown that infection of some of norovirus strains is reliant on host expression of the FUT2 enzyme generating fucosylated histo-blood group antigens on the cell surface, which was shown by clinical and experimental data [35, 36]. Extracellular pathogens might be less reliant on specific receptor expression but on the detection of chemoattractants to the epithelium for cell attachment and niche establishment. Helicobacter pylori (H. pylori) senses urea secreted by the gastric epithelium to locate the epithelium and attach to host cells [23, 37]. In addition, host cellular differentiation impacts infection via the innate immune response since innate immune receptor expression highly varies along the different segments of the GI tract and can also vary between different cell types or even sites of the highly polarized epithelial layer [38–40]. Since organoids can be directed to differentiate into specific cell types of the epithelium, they allow the precise study of these processes (Table 1).

Table 1.

Summary of pathogen tissue and cellular tropism observed in GI organoids

Pathogen type	Pathogen	Organoid type	Infection model	Strain/variant	GI segment tropism	Cell tropism	Reference
Virus	Rotavirus	PSC	3D	Simian SA11, human clinical isolates (G9P[type 8], G1P[type 8], G3P[type 8], G2P[type 4], and G3P strains)	Small intestine	Epithelial, mesenchymal cells	[41]
		ASC	3D	Rhesus rotavirus (G3P[3]), human rotavirus strains Ito (G3P[8]), Wa (G1P[8]) and Rotarix RV1 (G1P[8]) vaccine strain, and simian rotavirus SA11	Duodenum, jejunum, and ileum	Differentiated enterocytes and EECs	[42, 43]
	Norovirus	ASC	2D monolayers	Human norovirus GII.4 variants (GII.4/ 2006a, GII.4/2006b-1-3, GII.4/2009, and GII.4/ 2012-1) and GII.3, GII.4 Sydney 2012	Duodenum, jejunum, and ileum	Enterocytes	[21, 44, 45]
	Astrovirus	ASC	2D monolayers	Classic human astrovirus, VA or MLB strains	Duodenum, jejunum, ileum, and colon	Progenitor cells, enterocytes, and goblet cells	[46, 47]
	Echovirus 11	ASC	3D	Gregory	Small intestine	EECs, enterocytes	[26]
	Human Adenovirus serotype 5 and 41	ASC	3D and 2D monolayers	Adenoid 75	Ileum	Serotype 5- Goblet cells, serotype 41 differentiated and undifferentiated cells	[48]
	Enterovirus 71	Foetal stem cell	2D monolayers	GDV083,1095	Small intestine	Intestinal epithelial cells, Goblet cells, and enterocytes	[26, 49]
	FECV	Feline ASC	2D monolayers	Serotype I	Ileum, colon	Colon epithelial cells	[50, 51]
	NL63-CoV	ASC	2D monolayer, 3D organoids	NR-470	Duodenum	Differentiated intestinal cells	[52]
	SARS-CoV-2	ASC	3D organoids	Bav Pat1	Foetal paediatric stomach, small intestine	Enterocytes secretatory TA cells, gastric pit cells, and gastric EEC	[22, 50, 51, 53–55]
	Epstein-Barr virus	ASC	2D monolayers		Stomach	Gastric cancer organoids	[56]
Bacteria	Helicobacter pylori	ASC	3D and 2D monolayers	P12, G27	Stomach	MUC5AC pit cells, GKN1, and GKN2 high	[23]
	Salmonella Typhimurium	ASC	3D and 2D monolayers	SL1344	Small intestine, colon	Apical cell surface, M cells, and enterocytes	[32, 57, 58]
	Salmonella Typhi	ASC	2D monolayers	Ty2	Small intestine, colon	Enterocytes	[58]
	Listeria monocytogenes	ASC	3D	10403s pMP74	Small intestine ileum	Basal cell surface	[57]
	Shigella flexneri	ASC	2D monolayers, transwells	Serotype 2a strain 2457T	Duodenum, ileum, and colon	M cells, basal cell surface	[58–60]
	Yersinia pseudotuberculosis	ASC	2D traswells	JM301	Small intestine, ileum	M cells	[61]
	E. coli	ASC	2D monolayers	EAggEC (042 and UD792), EPEC (B171 and E2348/69), EAEC (serotype O44:H18 and EAEC A2A clinical isolate)	Small intestine, colon		[62, 63]
PARASITES	Cryptosporidium parvum	ASC	3D and 2D monolayers	Iowa isolate	Duodenum	Diferentiated cells, LGR5+ stem cells	[64–66]
	Giardia duodenalis	ASC	2D transwell	WB6 (ATCC 50803)	Duodenum	Apical cell surface	[67]
	Ostertagia ostertagi	Bovine ASC	3D organoids	Third-stage larvae of Moredun isolate MOo2	Stomach	Organoid lumen	[68]
	Trichuris muris	Murine ASC	2D transwell	First-stage larvae E-isolate	Colon	Ki-67+ epithelial cells	[44, 69]
	Toxoplasma gondii	Feline and murine ASC	2D monolayers	ME49 strain	Jejunum		[70]

Abbreviations: 2D: 2D monolayers; 3D: 3D organoids; ASC: adult stem cell; PSC: pluripotent stem cell.

Organoid models to study infection

Various GI organoid-derived models were used to study host–pathogen interactions. Classical organoids (which are per definition 3D objects) have been microinjected with bacteria or parasites to study infections, such as H. pylori or Cryptosporidium parvum [1, 64]. An additional method has been developed by inducing organoid polarity switch by removal of extracellular matrix (ECM). These ‘apical-out’ or ‘reverse polarity’ organoids were used to study H. pylori, Salmonella enterica serovar Typhimurium, Listeria monocytogenes, and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections [37, 53, 57]. Organoids can also be dissociated and the single cells are used to generate monolayers, which can be very practical for the addition of pathogens to the apical side of the epithelium [21, 23, 62, 71–73]. Directed differentiation to specific cell types such as enterocytes in intestinal monolayers, and pit or gland cells in gastric monolayers can also be achieved [21, 23]. Additionally, organoid-derived models can be combined with more complex technologies, such as microfluidics and bioprinting, recapitulating the architecture of GI tract and supporting long-term infection modelling [65].

Organoid heterogeneity as a tool to understand the cellular tropism of viruses

Viruses depend on the cellular components to bind, internalize, and exploit the cells’ machinery to produce new viral particles. In some cases, in which the virus has a strict host specificity, human cells are key to produce the virus in the laboratory. Furthermore, if the virus has a tropism to a particular cell type, conventional cell lines may not be able to provide this. In these cases, organoid-derived models can be particularly helpful for studying GI pathogens because directed differentiation can enrich cells that are the natural host cells in vivo. Early work using PSC-derived organoids showed productive human rotavirus (HRV) infection, although only few cells were infected [41]. The low infection rate is likely caused by the undifferentiated or foetal state of the PSC-derived organoids [74]. ASC-derived organoids, which resemble adult epithelium, were successfully infected with patient-derived HRV strains [42, 43]. This was important because cell lines did not support the growth of all patient-derived strains, thus this study already suggested that ASC-derived organoids provide host cells expressing crucial factors that were not present in cell lines before [43]. This expectation was further confirmed when cells from organoids were shown to support the replication of HNV which was not culturable before [21]. This opened the field to study target cells of viruses.

Different viruses exhibit different grades of cell type specificity from specific target cells to a broader host cell spectrum. Examples for very specified viruses are echovirus 11 (E11), which infects EEC but not goblet cells and human adenovirus-5p, which infects goblet cells [26, 48]. HNV and HRV primarily infect differentiated enterocytes. In addition, HRV can target CHGA-positive EEC [43]. SARS-CoV-2, which is best known for its respiratory symptoms, can also infect the GI tract and experiments in organoids could demonstrate a preference for intestinal cells of the adsorptive lineage [22, 54]. More specifically, sc-RNA-seq analysis of SARS-CoV-2 infected colon and ileum organoid-derived monolayers showed virus tropism to a subpopulation of immature enterocytes as primary target cells and transient amplifying cells as secondary infection targets at later time points [22]. Infection of paediatric gastric organoids with reversed polarity indicated additionally a possible infection of EEC in the stomach [53]. Examples for viruses with a very broad host cell spectrum are human astrovirus (HAV) and human adenovirus-41p, which are not dependent on cell differentiation, and multiple cell types in the intestinal epithelium are susceptible to infection [46–48].

The infection specificity towards one cell type might affect aspects of their pathogenesis and organoid heterogeneity offers the opportunity to study that. The infection of monolayers with HRV activates paracrine signals in the form of calcium waves around the infected cell, which induce fluid secretion and serotonin release [75]. Animal experiments show that cramps and diarrhoea upon RV infection are mediated by the nervous system. The combination of both observations led to the hypothesis of how infected enterocytes could stimulate neighbouring EECs and these, in turn, stimulate the nervous system. The tropism of different enteroviruses might also explain aspects of their pathogenesis. In human primary intestinal epithelial monolayers, while enterovirus 71 preferentially infects goblet cells and reduces the expression of goblet cell-derived mucins, E11 preferentially targets enterocytes and abolishes epithelial structure and barrier function. As a consequence, both enteroviruses induce type III IFNs as well as low levels of type I IFN by E11, which restrict enterovirus replication in a virus-specific manner [49].

Organoids also allow the direct comparison of susceptibility of different organs or different segments of one organ. The emerging variants of SARS-CoV-2, Delta and Omicron, have changed their tropism towards the respiratory tract and the intestine. While the Delta and Omicron variants have increased capacity to infect the respiratory tract, the Delta variant is more infectious in the intestinal epithelium, contrary to the Omicron variant, which has reduced replication capacity in the intestine [55]. HNV was found to be limited to the small intestine, as only the duodenum and ileum organoids reliably supported infection, whereas HAV-infected small intestinal and colonic organoid-derived monolayers [46, 76].

Even though organoids have provided target cells for many viruses, other viruses have such specific tropism, that the target cells and infection mechanism have not yet been identified, even though organoids were used. For example, Epstein Barr virus (EBV) can cause gastric cancer and thus must infect gastric epithelial cells at some point of the cancerous cascade. However, healthy gastric organoids are not permissive for EBV infection, while gastric cancer-derived organoids are permissive and support latent EBV infection [56]. Interestingly, the proposed epithelial entry receptor for EBV, ephrin receptor A2 (EPHA2), exhibited altered cellular localization of EPHA2 in healthy or cancer organoids [56]. Future research will have to determine, whether such different presentation of putative receptors may cause different susceptibility.

Thus, in many cases, organoids have allowed to narrow down the spectrum of target cells of viruses. This lays the foundation for the identification of entry receptors, which may then enable the development of therapies.

Organoids as a tool to study pathogenic bacteria tropism in the GI tract

Pathogenic bacteria exhibit a preference for different parts of the GI tract. Examples are enteropathogenic Escherichia coli (EPEC) which causes diarrhoea and preferentially infects the small intestine, both Salmonella enterica serovars (Typimurium and Typhi) infect the small and large intestine, Shigella flexneri (S. flexneri) leads to colonic epithelium destruction, or H. pylori only infects the stomach. The reasons for these region-specific diseases are not yet well understood and organoids may provide insights to understand the mechanism behind these tropisms. Infection of organoid-derived monolayers recapitulates the preferences in vitro for some pathogens: S. Typhimurium exhibited more severe attachment in the terminal ileum, whereas S. Typhi attached better to the caecum than ileum cells [58]. Enteroaggregative E. coli showed preference for rectal organoid-derived monolayers compared to ileum monolayers, while enteropathogenic E. coli showed a tendency of higher attachment to ileum monolayers [62]. Yet, this cannot be generalized, since S. flexneri invasion shows preference for colon in some studies and no differences with small intestine in others, suggesting that tropism of some pathogens might also be influenced by additional environmental factors not yet incorporated into organoid models [58–60].

Within their preferred GI section, also pathogenic bacteria exhibit preferences for specific cells. However, contrary to viruses, bacteria can be intra- or extracellular, and only intracellular bacteria depend on receptors for host cell entry. Therefore, it is possible that extracellular bacteria may depend less on specific cell types, than intracellular bacteria. The gastric pathogen H. pylori is such an extracellular bacterium, which exhibits a preference for, but not a dependence on a particular host cell. The cellular tropism and its underlying mechanism were identified using organoids. Several organoid models have been used to study H. pylori infection in the gastric epithelium [1, 24, 77, 78]. H. pylori adheres to organoid-derived monolayers and organoids, and induces a strong nuclear factor kappa B-mediated proinflammatory response [1, 72]. The application of directed differentiation in gastric organoid-derived monolayers showed a preference for H. pylori in highly differentiated pit cells residing at the opening of the gastric gland and this preference depends on the bacterial sensing of host cell-secreted urea [23, 37].

Several intracellular pathogens of the GI tract such as Salmonella, Listeria, and Shigella infect intestinal M cells as the major entry point for intestine invasion to invade neighbouring enterocytes from the basal side or reach the underlying Peyers patches’ and spread to systemic sites. M cells are specialized for antigen sampling and transport of antigens from the luminal space to the subepithelial space to activate immune cells. M-cell markers SPIB and GP2 are present in fresh isolated intestinal crypts and ASC-derived organoids [18, 31]. Although the presence of full functional M cells in organoid cultures is still under discussion, RankL-stimulated cultures showed higher M-cell markers expression and higher internalization of bacteria such as S. Typhimurium and Yersinia pseudotuberculosis [32, 61].

To study the specificity of bacteria to the basal side of epithelial cells, organoids with reversed polarity were used. Upon ECM removal, organoids change cell polarity exposing the apical side of the epithelium [57]. Listeria monocytogenes’ tropism to the basal side, and S. Typhimuriums’ tropism to the apical side were confirmed in organoids [57]. After invasion of the intestinal epithelium S. Typhimurium replicates intracellularly and then induces extrusion of infected cells to the lumen. This has been reproduced in organoids: S. Typhimurium has been observed in actively extruding and fully extruded epithelial cells [57, 79]. The organoid structure and microinjection have also allowed the study of the full life cycle of S. Typhimurium, including the colonization of the luminal compartment, and after that cell invasion, epithelial barrier-breaking, and expulsion of the infected cell back to the lumen [79]. A similar basal tropism preference as L. monocytogenes was shown for S. flexneri using an intestinal organoid-derived transwell system [59]. The bacteria showed limited invasion of intestinal monolayers from the apical side compared to the basal side; however, the invasion was improved by directed differentiation towards the M-cell phenotype [59, 60].

Organoids for the study of parasites

The life cycle of C. parvum, a GI parasite and a causative agent of cryptosporidiosis, was studied using several organoid-derived models such as duodenum organoids and ileal organoid-derived monolayers. The differentiation state of organoids does not limit C. parvum growth in intestinal organoids as undifferentiated and differentiated intestinal organoids support the full life cycle of the parasite, including the apical shedding of newly produced oocysts. However, C. parvum targeted enterocytes and infection was more efficient in differentiated organoids [64]. The state of differentiation affects the duration of the infection, which was demonstrated in organoid-derived monolayers. In this model, the infection lasted up to 3 days in differentiated monolayers, whereas undifferentiated intestinal monolayers supported infection for up to 21 days [66].

For the protozoan Giardia duodenalis, which is a leading major cause of GI illness worldwide, infection of duodenal organoid-derived monolayers identified a sequence of molecular events, including changes in tight junction composition, leading to intestinal barrier dysfunction upon G. duodenalis infection [67]. Another important application of organoids was the identification of Toxoplasma gondii sexual development dependence on high levels of linoleic acid present in feline but not in mouse organoids [70]. T. gondii is known to reproduce sexually only in the cat family, being the main transmission vector of toxoplasmosis to humans.

Roundworm infections of the GI tract can cause chronic disease in humans and are responsible for a major production loss in cattle. The mechanism of how their larvae invade the host GI epithelia and establish infection is not well understood due to the absence of relevant in vitro models. Bovine gastric and murine caecal organoids have been adapted to study Ostertagia ostertagi and Trichuris muris [44, 68, 69]. These models revealed host tropism and the expansion of enterocytes in infection with Tri. muris, proposing this as a defence mechanism against roundworms [69].

Diverse innate immune response in the GI tract

The expression of innate immune host factors is heterogeneous at multiple levels. Contrary to classical cell lines, organoids are untransformed and maintain their segment- and cell-specific expression signature, including innate immune receptors [17, 38, 39]. Transcriptome analysis of organoids demonstrated that each GI segment expresses its specific set of innate immune receptors and expression patterns differ strongly between mice and humans. The species-specificity of innate immune responses of the epithelium may be relevant in GI infection [38].

The application of single-cell RNA technologies allows us to further understand cell type-specific responses to infection. Differential baseline expression of innate immune factors can lead to a cell type-specific response to infection. Enterocytes, TA cells, and EEC respond to HAV infection by upregulating different interferon-stimulated genes [47]. The differentiation of organoids is important, as directed differentiation to different cell types leads to differential expression profiles of host immune factors and specific responses to infection. H. pylori infection in gastric organoids enriched for gland or pit cell types yields a specific innate immune response: gland-phenotype organoids respond with higher levels of IL-8 expression than organoids harbouring pit cells [1, 24].

In addition, different variants of the pathogen can respond differently to the host innate immune response. HNV has genotype-dependent susceptibility to the Type III interferon response in intestinal organoids [80]. Also, an innate immune response can depend on the side of infection, as was shown with mammalian reovirus in colon organoids. Infections at the basolateral side of the epithelial cells generate a stronger toll-like receptor 3 response than those from the apical side [40]. The organ segment- and cell type-specific host innate immune response coupled with the importance for the infection once again highlight the potential of organoids for infection biology but also underlines the importance of understanding the cellular differentiation of the organoids.

Current challenges and future developments

Stem cell-derived organoids have been established as a valuable tool in infection research. While stem cell-derived organoids have the intrinsic capacity to generate all cell types of the organ, it must be noted that not all these cell types are present in all organoids at the same time. Thus, understanding directed differentiation and knowledge about the cellular composition of a particular model at the time of infection is a major challenge, yet of central importance for understanding host–pathogen interactions.

Currently, much effort is made to generate even more complex models and to add mesenchymal compartments, immune cells, and vasculature. Such models are likely closer to emulate an in vivo situation than the current models. The field is progressing significantly with the incorporation of organoids into microfluidic systems and advanced microengineered scaffolds and matrices. The combination of intestinal organoids with a preshaped matrix and microfluidic system resulted in villus-crypt patterning of cells, as well as an extended culture period, and could be applied in infection settings. Intestinal epithelium was combined with vascular endothelium in a complex intestine-on-a-chip to examine NL-63 coronavirus infection [52]. Bioprinted ‘Mini-guts’ in the shape of a tube containing crypts and populated with organoid-derived cells allowed observation of the full life cycle of a parasite for 4 weeks [65]. Flow and peristaltic movements applied to organ-on-a-chip can also promote infection, since it enhances crypt-like structure formation where bacteria can be passively trapped, as seen with Shigella [45]. The physical condition of the flow also affected HNV and E. coli infection, as the pathogens were more infectious than in static conditions [63]. While fascinated by the increasing complexity, we also argue that for the mechanistic understanding of a specific question, the reductionistic epithelial model may be more suitable and the model has to be chosen according to the question to be addressed [8].

The comparison of different organs and organ segments is also of high importance in drug development. For example, PSC-derived lung and colon organoids have been used for high-throughput screening of commercially available drugs for SARS-CoV-2 [50]. Future studies using a range of organoids of different donors and different organs will allow better evaluation of drug efficacy and safety. Organoids and more complex models have already been incorporated by industry into the pipeline of drug development for many diseases and will undoubtedly be also adapted to develop new drugs to combat infections.

In this respect, it should be noted that newly emerging infectious diseases regularly originate in animals (e.g., bats or swine). When the species barrier is crossed and transmission between humans is enabled, this can lead to a new pandemic and pose a serious threat. Organoids generated from animals and the study of infections in the veterinary organoids will contribute to the preparedness for newly emerging diseases [51, 81].

We also see unprecedented potential of organoids to contribute to our understanding of childhood diseases. Several pathogens affect children differently than adults, and although the maturation of the immune system likely plays a major role in this phenomenon, a contribution of the epithelium is also hypothesized. Organoids have been generated from foetal, neonate, and paediatric tissue samples of the GI tract and they can be used for infection studies [53, 82, 83]. For example, paediatric and adult gastric organoids showed differential susceptibility to SARS-CoV-2 [53]. Given the scarcity of models for human GI maturation and childhood diseases, organoids may contribute to fill this important gap in the future.

Conclusion

GI organoids are a valuable tool for examining host–pathogen interactions and dissecting mechanisms of pathogen tropism. Especially, organoids bear the possibility to generate a wide cellular heterogeneity present in the host tissue and thus enable the study of the differential susceptibility to a variety of pathogens. To unfold the potential of these possibilities, analysis of the cell types present in the model and the understanding of their maturation is of central importance. These features facilitate a novel understanding of human GI tract interactions with pathogens and increase our knowledge of possible strategies for countering infectious agents.

Abbreviations:

Abbreviations:
 
• ASC

  adult stem cells

 
• BMP

  bone morphogenic protein

 
• DAPT

  (2S)-N-[2-(3,5-Difluorophenyl)acetyl]-l-alanyl-2-phenyl-glycine 1,1-dimethylethyl ester

 
• DBZ

  dibenzazepine

 
• E11

  echovirus 11

 
• EBV

  Epstein Barr virus

 
• EEC

  enteroendocrine cells

 
• EGF

  epithelial growth factor

 
• FGF-2

  fibroblast growth factor 2

 
• FUT2

  fucosyltransferase 2

 
• GI

  gastrointestinal

 
• HNV

  human norovirus

 
• HRV

  human rotavirus

 
• HAV

  human astrovirus

 
• IGF-1

  insulin-like growth factor 1

 
• M cell

  microfold cell

 
• p38i

  p38 mitogen-activated protein kinase inhibitor

 
• SARS-CoV-2

  severe acute respiratory syndrome coronavirus 2

 
• TA cell

  transit amplifying cell

Acknowledgments

This paper is a part of ‘Organoids to study immune cell development and function’ series of reviews.

Conflict of interests

None declared.

Funding

This work was supported by the Deutsche Forschungsgemeinschaft (DFG GRK 2157; 3D Tissue Models for Studying Microbial Infections by Human Pathogens, Project 10, to S.B.).

Author contributions

Conceptualization: S.B. Literature curation: M.P. and P.S.V. Supervision: S.B. Visualization: S.B., M.P., and M.N. Writing—original draft: M.P. Writing—review & editing: S.B., P.S.V., M.N., and M.P. All authors contributed to the review and approved the final version.