Gastrointestinal organoids in the study of viral infections

Abstract

Viruses are among the most prevalent enteric pathogens. Although virologists historically relied on cell lines and animal models, human intestinal organoids (HIOs) continue to grow in popularity. HIOs are nontransformed, stem cell-derived, ex vivo cell cultures that maintain the cell type diversity of the intestinal epithelium. They offer higher throughput than standard animal models while more accurately mimicking the native tissue of infection than transformed cell lines. Here, we review recent literature that highlights virological advances facilitated by HIOs. We discuss the variations and limitations of HIOs, how HIOs have allowed for the cultivation of previously uncultivatable viruses, and how they have offered insight into tropism, entry, replication kinetics, and host-pathogen interactions. In each case, we discuss exemplary viruses and archetypal studies. We discuss how the speed and flexibility of HIO-based studies contributed to our knowledge of SARS-CoV-2 and antiviral therapeutics. Finally, we discuss the current limitations of HIOs and future directions to overcome these.

INTRODUCTION

The human gastrointestinal epithelium is a remarkably dynamic, cellularly diverse, and histologically intricate system under tight regulation. It not only must absorb dietary nutrients to nourish every cell in the human body, but it must also form a protective barrier against the tremendous array of pathogens that make their way into the gut lumen. Viruses are among the most abundant and diverse pathogens and continue to drive the majority of gastrointestinal infections (1–3).

Historically, studies focused on the pathogenesis of enteric viruses relied on transformed cell lines that lack both the cellular diversity and coordinated function of the gastrointestinal epithelium (4). Although animal models provide a functional readout and a more holistic model of infection, they are costly, low throughput, and often not susceptible to, or exhibit different symptoms in response to, human enteric virus infection (5). Within the past decade, virologists have increasingly turned to human intestinal organoids (HIOs) as a tool to further elucidate viral-host interactions. HIOs are stem cell-derived ex vivo culture systems that are nontransformed, human origin, and recapitulate the cellular diversity of the native human intestinal epithelium (6, 7). They maintain a tight, polarized epithelial barrier and regulated secretory activity allowing for functional studies (8–10). As such, HIOs have become a valuable tool for virologists.

This review discusses recent advances in virology made possible by HIOs. We discuss how HIOs have allowed for the cultivation of previously uncultivatable viruses, facilitated studies of viral pathogenesis, and offered a fast and adaptable platform to study SARS-CoV-2. Finally, we discuss the current limitations of HIOs in virology and future directions to overcome these.

VARIATIONS AND TECHNIQUES USING HIOs

Although technological advances have allowed for the diversification of organoid models, the fundamental principle remains consistent. Broadly, organoids are derived from isolated stem cells. The term “enteroid” refers specifically to HIOs derived from multipotent stem cells isolated directly from the intestinal crypt. Although enteroids are still themselves HIOs, HIOs can also be derived from pluripotent or induced pluripotent stem cells (iPSCs; 11). HIOs derived from iPSCs maintain mesenchymal cells which are not found in enteroids (12).

It is possible to culture HIOs in a variety of platforms. Most frequently, HIOs are maintained as three-dimensional spheres in synthetic matrices (e.g., Matrigel) that support growth. Three-dimensional HIOs can be dissociated into single-cell suspensions and replated on collagen-coated culture dishes as monolayers. In addition, HIOs can be plated as monolayers on permeable membranes that separate the apical and basolateral compartments. Separation of the basolateral and apical media allows investigators to measure electrical potential across the monolayer, membrane permeability, migration, and the unique effects of compounds applied apically versus basolaterally.

Recent advances have also allowed for the enrichment of cell types that are typically rare in HIOs. Using a tetracycline-inducible promoter to overexpress neurogenin-3 allows for the enrichment of enteroendocrine cells, addition of retinoic acid and lymphotoxin promote M cell differentiation, and IL-22 increases the number of Paneth cells (13–15). Given the unique roles these cell types play in defense against pathogens, antiviral responses, and gut function, these techniques offer novel platforms to study cell type-specific contributions to pathogenesis.

HIOs FOR THE CULTIVATION OF VIRUSES

Arguably the most important advance HIOs have provided the field of virology is the ability to establish new infection models for previously uncultivatable viruses. The most notable example of this is human norovirus (HuNoV). HuNoV is a highly infectious pathogen that targets intestinal epithelial cells, causing acute gastroenteritis in immunocompetent patients and severe, chronic disease in the immunocompromised (16). Despite the significant global burden of HuNoV infection, researchers have made limited progress in the production of effective vaccines or antiviral therapies. This can be, in part, attributed to the lack of robust replication in conventional cell culture systems and animal models. Although cultured B cells were shown to support the replication of HuNoV, this system has proven difficult to consistently reproduce (17, 18). In 2016, the Estes group resolved nearly 50-years delay in establishing a robust human norovirus cultivation system by using HIOs, the first model to consistently enable replication of multiple HuNoV genotypes (Fig. 1A; 19). To date, GI.1, 11 genotypes of GII, and six variants of GII.4 HuNoV have shown productive replication in HIOs (20). Although multiple groups have attempted to use HIOs to cultivate other enteric caliciviruses, including human sapovirus and bovine norovirus, they have yet to establish optimal conditions (21).

Figure 1.

Graphic summary of virological insights from experiments using human intestinal organoids (HIOs). HIOs have allowed for the cultivation human norovirus (HuNoV; A) and determination of cell type-specific tropism for HuNoV, astrovirus, and rotavirus (RV), which infects preferentially infects differentiated enterocytes over undifferentiated stem cells (orange; B). HIOs also allowed investigators to determine that HuNoV requires both bile acids and ceramide for entry (C) and identify determinants of replication kinetics for enterovirus (D). Finally, genetically engineered HIOs allowed investigators to characterize a new role for paracrine signaling in rotavirus Ca²⁺ signaling (green) and pathogenesis (E) and understand contributors like serotonin (5HT) and chloride (Cl⁻) to luminal (red) fluid secretion and swelling (F).

Human astrovirus is another clinically relevant enteric pathogen and an important cause of acute gastroenteritis in children (22–24). Recent work identified divergent groups of astrovirus and documented their potential to cause extraintestinal infections (25–29). These studies underscore a critical need for a deeper understanding of astrovirus tropism and pathogenesis, but, similar to HuNoV, the lack of physiologically relevant cultivation system has hindered progress. In addition, although most investigators use the Caco-2 human colon adenocarcinoma cell line to propagate classical astrovirus serotypes, it does not support the replication of all nonclassical serotypes (30, 31). In 2019, Kolawole et al. (32) showed that HIOs could be used to replicate strains of human astrovirus from all three clades. This work has opened the door to study astroviruses more broadly and improve our understanding of pathogenesis.

THE BROAD CONTRIBUTIONS OF HIOs TO THE STUDY OF VIRAL PATHOGENESIS

Tropism

A key advantage of HIOs is that they recapitulate the diversity of epithelial cell types found in the gastrointestinal epithelial layer. In a proliferative state (i.e., undifferentiated), HIOs maintain a crypt-like profile of stem/progenitor and Paneth cells. Following differentiation, these cultures express mature cell types of the villus epithelium including enterocytes, goblet cells, enteroendocrine cells, and Tuft cells (7). This allows for a more precise determination of viral tropism (Fig. 1B). Both transcriptional studies and immunofluorescent microscopy (Fig. 2) have contributed to this (33–35). Using HIOs, investigators have determined that, in addition to infecting mature enterocytes, both rotavirus and norovirus are capable of infecting enteroendocrine cells (13, 34). The nonclassical human astrovirus, VA1, had previously undetermined tropism in the small intestine. Using HIOs, however, investigators were able to identify at least three distinct cell types that are infected with human astrovirus VA1: mature enterocytes, goblet cells, and intestinal progenitor cells (32). Importantly, these findings established VA1 as the first enteric virus capable of infecting progenitor cells and may provide the first hints of viral factors needed for broader tropism and extraintestinal spread. In a recent study, Triana et al. (36) used multiplexed in situ hybridization to show that astrovirus RNA colocalized with transcripts marking transiently amplifying cells and mature enterocytes, and to a lesser extent with those marking stem-like cells, goblet cells, and enteroendocrine cells.

Figure 2.

Fluorescent micrograph of multiple human intestinal organoids infected with rotavirus. The organoids were fixed and stained for E-cadherin (magenta), villin (green), nuclei (blue), and rotavirus (red). The image was acquired on a Zeiss LSM980 confocal microscope.

Beyond cellular tropism, Saxena et al. (33) also used HIOs to investigate host range restriction of different rotavirus strains, showing that HIOs are more susceptible to a human versus simian strain. In addition, using HIOs generated from the duodenum, jejunum, and ileum, they showed that HIOs generated from any segment are susceptible to rotavirus infection.

Entry

Before the establishment of a cell culture model for human norovirus, mechanistic studies relied upon replicon systems, which used transfection of self-replicating norovirus RNA (37). Although these early studies provided insight, they could not account for factors required for productive replication at all stages (binding, entry, genome replication, assembly, and release). Small animal models overcome some of these limitations, yet in these systems, it is difficult to manipulate and study host factors involved in susceptibility and permissibility using these models alone. HIOs, therefore, provided a new opportunity to robustly study host/pathogen interactions in the context of a native human norovirus infection. This breakthrough has led to the identification of components of the intestinal milieu, which are required for replication, and for the appreciation of strain-specific differences in these requirements. Ettayebi et al. (19) established that bile is required for the replication of GI.1, GII.3, and GII.17 strains of human noroviruses, and enhances the replication of GII.4 strains (Fig. 1C). Following this finding, Murakami et al. (38) went on to show that bile acids were the active component of bile that promoted GII.3 replication, and through specific interactions with the bile acid receptor, S1PR2, enhanced both entry and infection (38). These studies underlie the utility of the HIO model in understanding mechanisms of entry in physiologically relevant systems, critically broadening the potential for new antiviral therapeutic targets.

Replication Kinetics

Beyond tropism and entry, HIOs have offered valuable insight into the mechanisms underlying viral replication. This is exemplified by enteroviruses which, similar to HuNoV and astrovirus, lacked a broad physiologically relevant culture system before the advent of HIOs.

HIOs have been established as a viable platform to study multiple species of enterovirus including coxsackievirus B, echovirus 11, enterovirus A71, and poliovirus type 3 (PV-3; 39, 40). Adapting this model, Drummond et al. (39) were able to characterize inflammatory responses in HIOs that more closely mimic those seen in vivo as compared with transformed cell lines. Given that inflammatory processes often modulate viral replication, this suggests that HIOs may offer advantages when studying replication kinetics. By assessing tissue culture infectious dose (TCID50) and viral transcript levels, Tsang et al. (40) leveraged HIOs to more thoroughly characterize the kinetics of enterovirus replication (Fig. 1D). Finally, they showed a high level of expression of the receptors for EV-A71, CVB2, and PV-3 in HIOs, but low expression of the receptor (ICAM-5) for a fourth enterovirus, EV-D68, which did not replicate in HIOs. This finding is consistent with and may explain the observation that EV-D68 primarily infects respiratory epithelium in humans (41).

HOST-VIRUS INTERACTIONS AND CELL SIGNALING

When aiming to study host-pathogen interactions, HIOs offer distinct advantages over both in vitro and in vivo model systems. Multiple groups have genetically engineered HIOs to express fluorescent biosensors (42), knockout or overexpress genes of interest (20, 42, 43), or favor differentiation of certain cell populations (13, 44). These manipulations have allowed for focused investigations of specific signaling pathways involved in enteric viral pathogenesis, detailed in this section. For a broader review of strategies to genetically engineer organoids, see the review by Menche et al. (45).

Calcium Signaling

Calcium (Ca²⁺) signaling is a crucial regulator of homeostasis within the intestinal epithelium (46–48). Multiple groups had shown that rotavirus exploits Ca²⁺ signaling for replication in immortalized cell lines, but the nature of these signals and their effect on the intestinal epithelium remained unclear (48–53). Our group generated HIOs expressing the genetically encoded cytosolic calcium indicator GCaMP6s to characterize novel signals induced upon rotavirus infection (Fig. 1E; 42). Not only did we find unique Ca²⁺ signatures in infected cells, but the HIO-GCaMP6s system allowed us to characterize intercellular Ca²⁺ waves that elevate Ca²⁺ signaling in surrounding, uninfected cells. This dysregulated Ca²⁺ signaling leads to increased fluid secretory activity and serotonin release from enterochromaffin cells, contributing to rotavirus pathogenesis. Not only did this identify a novel component of virus-induced signaling, but it also offered a potential new target for antidiarrheal therapeutics. Furthermore, although intercellular Ca²⁺ waves are rare in cell lines outside of the setting of infection, this work demonstrated that HIOs display a low level of Ca²⁺ waves at baseline. This suggests a broader regulatory role of intercellular Ca²⁺ waves in the intestinal epithelium, underscoring the importance of physiologically accurate models.

Innate Immunity and Interferon Signaling

The synthesis and secretion of interferons (IFNs) are key tenants of innate immunity and early antiviral responses (54, 55). Epithelial cells are an important component of IFN production and signaling and as such, must be appropriately modeled to understand the interplay between infection and host immune responses. Studies have demonstrated that baseline interferon production and the interferon response that is mounted to an infection can be altered in senescent or immortalized cells, complicating the interpretation of immune signaling pathways in primary and transformed cell lines (56–58). In addition, transfection of viral RNA in nonsusceptible or permissive cells often induces an immune response that differs from that of native virus infection. Interferon signaling is a key element of this response. A meta-analysis of RNA sequencing on organoids infected with astrovirus, norovirus, and rotavirus demonstrated that in all cases, interferons and interferon-stimulated genes make up an important part of the host response to infection (59). Type I and type III interferon predominate in the intestine, with type I associated with controlling systemic virus spread and type III associated with localized restriction of mucosal infection (60). Cell model-dependent discrepancies in the types of interferons stimulated by infection and the susceptibility of viruses to these factors have been demonstrated for two important enteric viruses, human norovirus, and human adenovirus (HAdV).

In HuNoV studies, transfection of immortalized cell lines with RNA from clinical HuNoV isolates (35) fails to stimulate interferon production (37, 61). This is at odds with findings from other human enteric viruses and murine norovirus, suggesting a potential limitation of the transfection-based model rather than a true absence of interferon signaling (62, 63). Studies using genetically modified HIOs overcame these limitations to elucidate the role of specific interferons in HuNoV infection. By performing RNA sequencing of HuNoV-infected HIOs, Hosmillo et al. (43) demonstrated that HuNoV indeed elicits expression of interferon-stimulated genes and downstream activation of JAK/STAT signaling. Pretreatment with either IFN-β1 or IFNγ1/2 reduced HuNoV replication in HIOs, supporting an inhibitory effect of IFNs. Likewise, HIOs engineered to constitutively express interferon inhibitors offered enhanced HuNoV replication, further supporting an antagonistic role of interferons in infection. Later studies using CRISPR-Cas9 mediated knockout of IFNα, IFNγ, MAVS, STAT1, and STAT2 in HIOs allowed Lin et al. (64) to characterize strain-specific effects of IFNs on HuNoV, which may underlie broader epidemiological differences.

Studies of human adenoviruses (HAdVs) also demonstrate the advantages of HIOs over standard cell lines with regard to innate immunity. In cancer cell lines, the interferon response fails to restrict wild-type HAdV replication (65, 66). This contrasts with in vivo data, which demonstrates enhanced adenovirus load in STAT2-knockout animals compared with wild-type and in vitro data in primary human cell lines, which shows significant inhibition of Ad5 replication with exogenous IFNα and IFNγ treatment (67, 68). To resolve these differences, Holly et al. (69) cultivated clinical HAdV isolates in human ileal enteroids and transformed A549 cells. These studies recapitulated important differences in adenovirus sensitivity to IFN treatment, demonstrating that IFNβ and IFNλ3 inhibit HAdV-5p replication in ileal organoids but do not affect HAdV-5 replication in A549 cells. These studies underlie the need to carefully consider how cell culture models may confound our understanding of host responses and virus replication.

In addition, studies of astrovirus infection in HIOs revealed lineage-specific differences in expression of interferon-stimulated genes. Shortly after infection, stem and goblet cells showed higher expression of CCL2, STAT1, IFNGR1, and IRF2 relative to enterocytes. Though expression of these markers was lower in uninfected HIOs, the cell-type specific differences in interferon-stimulated gene expression persisted, suggesting that they are intrinsic. Furthermore, the antiviral response was not limited to infected cells. Bystander cells also showed an upregulation of interferon expression, albeit to a lesser extent (36).

Hormonal Signaling

Hormones produced in the gut are integral for the regulation of both the gut itself and many organ systems throughout the body (70–73). Although enteric pathogens perturb hormone signaling within the gut (74), these perturbations are unobservable in transformed cell lines that lack significant hormone production. HIOs offer a platform to study hormonal dysregulation, but the primary hormone-producing cell type, enteroendocrine cells, are typically quite rare in HIOs. This results in hormone levels that are difficult to detect through standard techniques. To overcome this, our group engineered a line of HIOs that expresses neurogenin-3 under a doxycycline-inducible promoter (13). Neurogenin-3 drives enteroendocrine cell differentiation, thus by adding doxycycline to induce neurogenin-3 overexpression, we can enrich for enteroendocrine cells. This system allowed us to detect the secretion of serotonin, monocyte chemoattractant protein-1, glucose-dependent insulinotropic peptide, peptide YY, and ghrelin upon rotavirus infection, suggesting it may be a useful model for hormone-related studies. Furthermore, this serves as a valuable proof-of-concept for those interested in studying other rare cell types in HIOs.

Functional Studies

HIOs grow as three-dimensional spheroids or as polarized monolayers, expanding their versatility for studying viral-host interactions. During viral infection, a three-dimensional HIO exhibits upregulated fluid secretion into its lumen, which increases the luminal volume (33). The change in the cross-sectional area of an infected HIO over time, measurable using standard bright-field microscopy, correlates with the degree of fluid secretion (33, 75). Given that fluid secretion is a critical driver of clinical disease, this technique, known as a swelling assay, provides means to assess factors that directly contribute to virally induced fluid loss. Using this technique, our group showed that paracrine purinergic signaling is an integral component of rotavirus-mediated fluid secretion (Fig. 1F; 42). Similarly, HIO monolayers can be plated on transwell filters, which offer independent access to both the apical and basolateral sides of a polarized monolayer without disruption of the monolayer itself (76, 77). These can be used to measure barrier integrity, solute transport, fluid secretion, and other aspects of epithelial physiology.

HIOs AS A PLATFORM FOR THE STUDY OF SARS-CoV-2

While maintaining the cellular diversity of the native intestinal epithelium, HIOs offer distinct speed and adaptability over traditional in vivo models. Establishing a novel genetic mouse model and using it to generate experimental data often takes two or more years (78). Genetic manipulation of HIOs presents challenges of its own, but an experienced user can establish a genetically engineered HIO line in as little as 2 mo (79). Furthermore, HIOs are of human origin. This is an important consideration given the host range restriction of most human viruses. SARS-CoV-2, for example, is unable to infect wild-type laboratory mice due to the incompatibility of the mouse angiotensin-converting enzyme (ACE)-2 receptor and the SARS-CoV-2 spike protein (80). As such, HIOs can expedite studies of viral pathogenesis while maintaining the benefits of nontransformed human tissue. Unsurprisingly, many groups capitalized on the speed and adaptability of HIOs to facilitate studies of SARS-CoV-2 (81–83).

Less than 2 mo after the World Health Organization declared COVID-19 a pandemic (84), three groups published work implementing HIOs to further our understanding of SARS-CoV-2 (81, 82, 85). Although clinical studies established that SARS-CoV-2 promoted gastrointestinal symptoms and was detected in the stool (86–88), studies in HIOs provided the first definitive evidence that SARS-CoV-2 could both infect and replicate in human intestinal epithelial cells (81). By using polarized HIO monolayers plated in transwells, Lamers et al. (81) also showed that SARS-CoV-2 both infected and exited enterocytes primarily at the apical membrane. Furthermore, they observed fusion of SARS-CoV-2-infected cells, suggesting that the virus is capable of inducing syncytia formation in enterocytes (82). Using HIOs from 13 different donors, Jang et al. (89) showed tremendous host-to-host variation in viral yield following SARS-CoV-2 infection. Viral titers in the culture media showed a 423-fold difference between the highest and lowest replicators. Interestingly, this variability correlated with the expression levels of the SARS-CoV-2 receptors ACE2 and TMPRSS2.

Beyond pathogenesis, HIOs were also rapidly implemented to study the activity of antiviral therapeutics against SARS-CoV-2 (90). Using HIOs in conjunction with lung organoids, Han et al. (90) showed that three different antiviral compounds inhibit replication of SARS-CoV-2 in organoids. In addition, they demonstrated that the mechanism of action of one of these compounds, Imatinib, is multifactorial, reducing receptor expression and altering expression of genes involved in lipid metabolism. In these studies, HIOs offered a physiologically relevant platform to rapidly investigate antiviral efficacy and mechanism of action at a time when speed and translatability were high priority.

CURRENT LIMITATIONS AND FUTURE DIRECTIONS OF HIOs IN VIROLOGY

Although HIOs offer many advantages, as with all model systems there are important limitations to consider. Perhaps the most salient limitation for virological studies is the lack of tissue type diversity. Although HIOs maintain the cell diversity of the intestinal epithelium, they lack the other cell types critical for gastrointestinal tract function, including immune cells, lymphatics, endothelial cells, neurons, smooth muscle, and fibroblasts, let alone integration with other tissues outside the gut. All of these tissue types play a role in host-pathogen interactions but are unobservable using organoids alone.

To overcome this, many have begun to develop systems in which HIOs are maintained in culture with other nonepithelial cell types (91). These coculture systems have allowed for broader investigation of the interactions between virally infected epithelial cells and other, nonepithelial cell types including monocytes, dendritic cells, lymphocytes, stromal cells, adipocytes, endothelial cells, and neurons. As coculture systems continue to improve, they will offer valuable platforms to study elements of pathogenesis that extend beyond the epithelium.

Furthermore, HIOs fail to maintain the intricate morphology of the intestinal epithelium. In vivo, the intestinal epithelium is neatly organized into villi and crypts with distinct cell types distributed predictably throughout (92). HIOs have crypt-like regions but generally lack villi and fail to show the same predictable distribution of cell types. Importantly, enteric viruses often show a pattern of infection that is spatially restricted along the villus-crypt axis. For example, rotavirus preferentially infects the fully differentiated enterocytes at the tips of villi (93). These morphology-dependent features of infection are difficult to assess in HIOs.

However, recent developments offer promising advancements. By using microfluidic chambers to mimic intestinal flow and peristalsis, or by inducing softening of the culture matrix in predefined patterns, investigators can push HIOs to develop the native villus-crypt architecture (94, 95). Groups have already utilized this technology for live imaging to assess epithelial responses (96). Adapting these systems for virological studies would allow for interrogation of the temporal and spatial dynamics of viral infections without the implicit difficulties of intravital imaging.

HIOs require specialized media that has yet to be universally standardized. Media supporting human intestinal organoids is typically enriched with Wnt, R-spondin, noggin, human epidermal growth factor, and serum. Although enriched media is commercially available, it is often more cost effective for investigators to create their own media using cell lines that have been engineered to express and secrete the aforementioned growth factors. Because of this, there is a high degree of variability between HIOs maintained in different laboratories (97). As HIOs increase in their popularity and accessibility, researchers must validate findings across intestinal segments, lines derived from distinct patients, and culture conditions.

To date, HIOs have facilitated many meaningful advancements in virology. As technology develops to support more versatile culture systems, HIOs will continue to play an integral role in studies ultimately aimed at lessening the burden of virological diseases worldwide.

GRANTS

This review was supported by the McNair Foundation MD/PhD Scholars Program (to J. T. Gebert); National Institute of Allergy and Infectious Diseases (NIAID), Division of Microbiology and Infectious Diseases Grant R01AI158683 (to J. M. Hyser); National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), Division of Diabetes, Endocrinology, and Metabolic Diseases (DEM) Grant R01DK115507 (to J. M. Hyser); and NIDDK Grants F31DK132942 (to F. Scribano), F32DK130288 (to K. A. Engevik), F30DK131828 (to J. T. Gebert), and NIAID Grant F31AI169983 (to J. L. Perry).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.T.G. prepared figures; J.T.G., F.S., K.A.E., and J.L.P. drafted manuscript; J.T.G., F.S., K.A.E., and J.M.H. edited and revised manuscript; J.T.G., F.S., K.A.E., J.L.P., and J.M.H. approved final version of manuscript.