This protocol describes how to infect human intestinal organoids from either their apical or basolateral side to characterize host/pathogen interactions at the single-cell level using single-cell RNA sequencing (scRNAseq) technology.
Human intestinal organoids constitute the best cellular model to study pathogen infections of the gastrointestinal tract. These organoids can be derived from all sections of the GI tract (gastric, jejunum, duodenum, ileum, colon, rectum) and, upon differentiation, contain most of the cell types that are naturally found in each individual section. For example, intestinal organoids contain nutrient-absorbing enterocytes, secretory cells (Goblet, Paneth, and enteroendocrine), stem cells, as well as all lineage-specific differentiation intermediates (e.g., early or immature cell types). The greatest advantage in using gastrointestinal tract-derived organoids to study infectious diseases is the possibility of precisely identifying which cell type is targeted by the enteric pathogen and to address whether the different sections of the gastrointestinal tract and their specific cell types similarly respond to pathogen challenges. Over the past years, gastrointestinal models, as well as organoids from other tissues, have been employed to study viral tropism and the mechanisms of pathogenesis. However, utilizing all the advantages of using organoids when employing highly pathogenic viruses represents a technical challenge and requires strict biosafety considerations. Additionally, as organoids are often grown in three dimensions, the basolateral side of the cells is facing the outside of the organoid while their apical side is facing the inside (lumen) of the organoids. This organization poses a challenge for enteric pathogens as many enteric infections initiate from the apical/luminal side of the cells following ingestion. The following manuscript will provide a comprehensive protocol to prepare human intestinal organoids for infection with enteric pathogens by considering the infection side (apical vs. basolateral) to perform single-cell RNA sequencing to characterize cell-type-specific host/pathogen interactions. This method details the preparation of the organoids as well as the considerations needed to perform this work under biosafety level 3 (BSL-3) containment conditions.
Studying the cell type-specific tropism and cell type-specific immune response to human enteric viruses has been historically challenging due to the lack of primary human cellular models. This limitation has now been partially eradicated with the development of organoids1. In the case of the gastrointestinal tract, gastric and intestinal organoid models have been developed for humans and several other species (e.g., murine, bovine, feline, bat)2,3,4,5,6. Intestinal organoids reproduce the structural architecture of the human intestinal epithelium and contain crypt and villi-like structures, functional intestinal lineages and have even been used to identify previously unknown cell lineages. Two different approaches can be used to grow intestinal organoids. First, intestinal stems cells containing crypts are isolated from tissue resections or biopsies and grown under specific culture conditions (e.g., Wnt3A, R-spondin, Noggin, and EGF) to expand, and then differentiate the stem cells to most intestinal cell lineages (e.g., enterocytes, Paneth cells, Goblet cells, enteroendocrine cells)7. This method allows for the isolation of organoids from all sections of the gastrointestinal tract (e.g., stomach, duodenum, jejunum, ileum, and colon). The second method relies on human-induced pluripotent or embryonic stem cells, which are then differentiated in a stepwise process into intestinal epithelial cells8. These induced stem-cell-based organoids are often described as being more embryonic in nature as compared to patient-derived organoids. While all these organoid models have been critical to unraveling developmental cues needed to form the intestinal tract, their use in infectious disease research is still in its infancy.
Enteric virus is a broad term covering all viruses, which infect through the gastrointestinal tract, such as picornaviruses (e.g., EV-71), reoviruses (e.g., rotavirus), and caliciviruses (e.g., norovirus)9. Enteric viruses initiate their infectious life cycle through the ingestion of contaminated food and water, which leaves people in developing countries at high risk due to the discharge of untreated waste into the environment and lack of medical care after the onset of infection10. Depending on the type of pathogen, the infection can lead to gastroenteritis, vomiting, and/or watery diarrhea due to leakage of the intestinal lining. Human noroviruses are a highly prevalent and highly infectious enteric pathogen, that lead to over 600 million infections and 15 million hospitalizations worldwide11. Organoids have been key to norovirus research as they support the infection and replication of human norovirus, which was previously unable to be cultured in standard cell culture models12.
Over the past two decades, coronaviruses have emerged as key human pathogens13. This family includes the highly pathogenic MERS, SARS-CoV-1, and SARS-CoV-2, which require strict safety level containments when performing research on these viruses. Interestingly, while all three of these pathogens are mostly recognized for their induced respiratory symptoms and distress, it is now evident that these viruses do not solely infect the respiratory tract but also other organs. An important pathology induced in SARS-CoV-2 infected patients besides the respiratory distress is the presence of gastrointestinal symptoms14. A fraction of SARS-CoV-2 infected patients displays such symptoms, ranging from very mild to severe diarrhea. Additionally, SARS-CoV-2 genomes can be detected in stool and gastrointestinal tract biopsies of infected patients15. Importantly the presence of gastrointestinal symptoms is not limited to SARS-CoV-2 as they were also observed in MERS and SARS-CoV-1 infected patients. To understand how SARS-CoV-2 induces gastrointestinal distress and precisely identify the tropism of SARS-CoV-2 in the gastrointestinal tract, human intestinal organoids have been a key tool and are now exploited to unravel cell type-specific responses to this pathogen16,17.
Transcriptional profiling of a cell population (bulk RNA sequencing) has been standard practice when evaluating pathogen infections of both immortalized cell lines as well as with organoids. While this allows us to determine global changes in response to pathogens (e.g., upregulation of cytokines), bulk RNAseq does not allow us to determine why specific cells in a population are more prone to infection than others. Single-cell RNA sequencing (scRNAseq) has become a powerful tool to unravel cell lineage-specific transcriptional programs and can be used to determine how these programs support or repress virus infection18,19. The first description of scRNAseq was in 2009 and was used to evaluate the transcription profiles of the different cells found in a mouse blastomere20. These technologies have now been expanded and can be implemented through several different platforms. Early versions of this technology applied fluorescence-activated cell sorter (FACS) to separate individual cells for sequencing, which was often limited to 96- or 384-well plates, thereby giving 300 individual cells to analyze per sample21. These methods have now been advanced by the single-cell sequencing platforms, which use a microfluidic device to encapsulate single cells into individual droplets with barcode containing beads. This technology allows for up to 10,000 cells to be captured per sample condition.
Combining organoids technology with scRNAseq allows us to study how enteric pathogens impact the gastrointestinal tract in a cell type-specific manner. However, several technical and biosafety considerations need to be taken. First and foremost, classical organoids culture methods (3-dimensional (3D) organoids, embedded in an extracellular matrix (ECM)) expose the basolateral side of the epithelial cells to the outside of the organoid. As enteric pathogens initiate their infection through ingestion of contaminated food/water, the infection most often initiates from the apical side of the cells, which is not accessible in these 3D intestinal organoids. Therefore, organoids need to be prepared to make the apical side accessible for pathogen infection either through 2D seeding, thereby directly exposing the apical side of the cells, or through microinjection22,23. Second, to perform scRNAseq of infected biological samples, it is important to consider their infectious nature. While methods to fix cells and inactivate pathogens prior to single-cell isolation for subsequent RNAseq have been proposed, these methods often lead to a decrease in sequencing quality18. The protocol below will describe several approaches to infect intestinal organoids with enteric viruses considering the infection side (apical vs. basolateral infection) (Figure 1). Additionally, the protocol will include a workflow to dissociate and isolate single cells from organoids infected with highly pathogenic viruses for scRNAseq. The protocol will highlight the key steps that need to be implemented when working under biosafety level-3 (BSL-3) containment conditions to avoid the generation of aerosols and potential contamination.
Human tissue was received from colon resection or ileum biopsies from the University Hospital Heidelberg for the following protocol. This study was carried out under the recommendations of the University Hospital Heidelberg with informed written consent from all subjects in accordance with the Declaration of Helsinki. All samples were received and maintained in an anonymized manner. The protocol was approved by the Ethics commission of the University Hospital Heidelberg under protocol S-443/2017.
1. Maintenance and passaging of intestinal and colon organoids
CAUTION: Human intestinal organoids are derived from human tissue or from induced pluripotent/embryonic stem cells, and as such, ethical approval is required. Country-specific regulations need to be followed. Human material is generally not tested and is therefore often considered BSL-2 material. Proper containment conditions need to be confirmed in the country in which the experiment takes place.
2. Preparation of organoids in two-dimensions (2D) for apical infection
NOTE: The following protocol will describe how to seed intestinal organoids as a monolayer of cells in a cell culture plate to infect intestinal epithelium cells from their apical side. Use the 48-well plate for the sequencing experiments and the 8-well glass-bottom chamber slide to control infection using immunofluorescence approaches.
3. Preparation of organoids in three-dimensions (3D) for apical and basolateral infection
4. Preparation of single cell solution and preparation of Gel beads-in-emulsion (GEMs) in biosafety level 3 (BSL-3) conditions
NOTE: Completion of the following steps requires that the single cell sequencing equipment (Table of Materials) and a PCR machine capable of handling 100 μL reactions are present inside a BSL-3 facility. Organoids are grown as described in section 1 and infected as described in sections 2-3 depending on the pathogen and entry route. At a pre-determined time post-infection, organoids are harvested. Below the method used to harvest intestinal organoids is described.
Preparation of organoids for single-cell sequencing
Single-cell sequencing results are highly dependent on using good quality cells. To ensure that organoids are of good quality, they should be properly maintained and observed on a daily basis to determine when they are ready to be split (Figure 2). The timing of splitting organoids is donor-dependent; some donors grow more quickly and need to be split every 5 days, while others are slower and need to be split every 10 days. On average, organoids are split one time per week when the centers become dark (Figure 2B). If organoids are allowed to become too large and accumulate too many dead cells in the center, the organoid will die.
Organoids are maintained in a media that contains high amounts of Wnt3A. This supports the stem cell niche and promotes the organoids to continue to grow and proliferate. Under these growth conditions, the organoids contain high amounts of stem cells and transit-amplifying cells and a lower amount of differentiated cell populations such as mature enterocytes, Goblet cells, and enteroendocrine cells. However, to mimic the cellular complexity found within the human intestine, it is important to push cell differentiation and produce more of these cells. This is accomplished by changing the media conditions and removing Wnt3A, and reducing R-Spondin and Noggin (Table 1). Normally, cellular differentiation toward enterocytes, Goblet cells, and enteroendocrine cells requires 4 days of differentiation media (Figure 3). It is key to obtain a good differentiation; otherwise, evaluating pathogen tropism and cell type-specific responses will become difficult.
Confirmation of BSL-3 pathogen inactivation
Full inactivation must be confirmed with the pathogen of choice and validated that it is safe to remove the cDNA from the BSL-3. For SARS-CoV-2, full inactivation of the virus was validated by taking 100 μL of SARS-CoV-2 and incubating it in a PCR machine for 5 min at 85 °C. The virus was then added back to naïve Vero cells, and virus infection was compared to non-heat treated virus by immunofluorescence and plaque assays at 24 h, 48 h, and 72 h post-infection to ensure that all particles were no longer infectious. These results were sent to the local regulatory agency, and upon their approval the single-cell experiment and cDNA processing was performed.
Single-cell sequencing results
To evaluate how SARS-CoV-2 infects human colon and ileum organoids, single-cell sequencing was performed. Organoids were prepared as described above and infected in a 2D format to allow for apical infection by SARS-CoV-2. Infected cells were harvested at 12 h and 24 h post-infection and were processed for single-cell sequencing as described above. Analysis of the single-cell sequencing data allowed us to determine that only a subpopulation (immature enterocyte 2) of human intestinal epithelial cells supported the infection of SARS-CoV-2 (Figure 4). Additionally, as not all cells in a population were infected, both infected cells and non-infected bystander cells were analyzed (Figure 5). These results showed that SARS-CoV-2 induced a pro-inflammatory signal cascade in infected cells while non-infected bystander cells showed an interferon-mediated immune response. Additionally, scRNA-Seq showed that infected cells were unable to sense interferons due to virus-mediated blockage of the pathway (Figure 5). It was not possible to obtain this information when using bulk RNA sequencing.
Figure 1: Schematic depicting the three different methods to prepare human intestinal organoids for infection with enteric pathogens. Apical infection can be achieved through seeding intestinal organoids in 2D. An apical and basolateral infection can be performed by disrupting the 3D organoid. Finally, a basolateral only infection can be performed by infecting intact 3D intestinal organoids. Each of these methods can be used to generate samples for single cell sequencing. Please click here to view a larger version of this figure.
Figure 2: Organoid maintenance schematic and representative brightfield images. (A) Schematic for maintenance and passaging of human intestinal organoids. (B). Representative brightfield images of days 1, 3, 5, and 7 post-splitting. By day 7 the organoids become large and dark due to the accumulation of dead cells and are ready to be split. Scale bar indicates 25 μm. Please click here to view a larger version of this figure.
Figure 3: Representative qPCR of human intestinal organoids 4 days post-switching to differentiation media. Intestinal organoids were maintained in growth media or switched to differentiation media for 4 days. RNA was harvested and qPCR was performed for markers of stem cells (OLFM4), Paneth cells (LYZ), Goblet cells (MUC2), and enterocytes (SI). N = 5. Please click here to view a larger version of this figure.
Figure 4: Identification of the cell population infected by SARS-CoV-2. Human colon and ileum derived organoids were infected with SARS-CoV-2. After 12 and 24 h post-infection cells were harvested and subjected to single-cell RNA sequencing to identify which cell populations supported SARS-CoV-2 infection. Virus infection was found to increase over time and mainly infect immature enterocyte 2. This figure has been modified from Triana et al.19. Please click here to view a larger version of this figure.
Figure 5: Determination of intrinsic innate immune response. Human colon derived organoids were infected with SARS-CoV-2. After 12 and 24 h post-infection cells were harvested and subjected to single cell RNA sequencing to determine the intrinsic innate immune response in both virally infected and non-infected bystander cells. SARS-CoV-2 infected cells displayed a strong pro-inflammatory response while non-infected bystander cells displayed an interferon-mediated response. Figure modified from Triana et al.19. Please click here to view a larger version of this figure.
Growth Media | |
Compound | Final concentration |
Ad DMEM/F12 | GlutaMAX (1X) |
+GlutaMAX | HEPES 1 mM |
+HEPES | Pen 10 U/mL |
+P/S | Strep 10 µg/mL |
L-WRN | 50% by volume |
B27 | 01:50 |
N-acetyl-cysteine | 1 mM |
EGF | 50 ng/mL |
A83-01 | 500 nM |
IGF-1 | 100 ng/mL |
FGF basic | 50 ng/mL |
Gastrin | 10 mM |
Differentiation Media | |
Compound | Final concentration |
Ad DMEM/F12 | GlutaMAX (1x) |
+GlutaMAX | HEPES 1 mM |
+HEPES | Pen 10 U/mL |
+P/S | Strep 10 µg/mL |
B27 | 01:50 |
N-acetyl-cysteine | 1 mM |
R-spondin | 5% by volume |
Noggin | 50 ng/mL |
EGF | 50 ng/mL |
Gastrin | 10 mM |
A83-01 | 500 nM |
Table 1: Media composition for growth and differentiation media.
Enteric pathogens most often initiate their lifecycle by infecting intestinal epithelial cells from their apical side facing the lumen of the gut. While organoids are well recognized to be a good model to reproduce the cellular complexity and organization of the intestinal epithelium, their organization as three-dimensional, closed structures make their apical membrane inaccessible to the pathogen. This protocol described methods for infecting intestinal organoids from their apical side, their basolateral side, or both with BLS-3 pathogens. These protocols can easily be adapted to study any enteric pathogen under BSL-2 or BLS-3 containment or any other organoid model by following a few critical steps that are highlighted below. The method described above is for the isolation and preparation of single-cell droplets in accordance with the regulations in Germany. As a disclaimer, this protocol does not describe the biosafety handling measures (standard operating procedures) that need to be taken while working under BSL-3 conditions. It is also important to insist that the regulations may vary in other countries and that the local authorities must be contacted to make sure that all local regulations are respected.
One of the critical steps in seeding organoids in two dimensions for apical infection is controlling that cells will similarly differentiate compared to when grown as classical three-dimensional organoids. Depending on the enteric pathogen, tropism could be restricted to very rare cells or to cells that need to be highly differentiated. In this case, using a two-dimensional organoid that did not fully differentiate could result in the misconclusion that this enteric pathogen cannot infect intestinal organoids. It is suggested, if possible, to perform infections using the three configurations of this protocol: 2D organoid for apical infection only (section 2), cracked open 3D organoids for apical and basolateral infection (section 3), and full 3D organoids for basolateral infection only (section 3). This approach will help to discern the entry route of the pathogen (apical vs. basolateral) and will also control that a similar level of differentiation has been achieved. An alternative for 2D apical infection is microinjection, which will use a 3D organoid but deliver the pathogen directly into the apical side (see Bartfeld et al.27 for details). This method requires a skilled injector to ensure that the pathogen is properly placed, and the organoids remain intact. Microinjection is commonly used in BSL-2 containment and may not be suitable for BSL-3 containment.
An additional important consideration when performing infection experiments in 2D seeded organoids is the final cell density. As mentioned in step 2.3, 100-150 organoids will be seeded in one well of a 48-well plate or one well of an 8-well glass-bottom chamber slide. Depending on the organoid line and on the person handling the organoids, the size of these organoids can be significantly different. This could result in very different cell densities in the 48-well plate or 8-well glass-bottom chamber slide. Depending on the enteric virus, some viruses prefer more sparse cells, while others will also be able to infect confluent cells. The molecular origin of such differences in infectivity for different cell confluences is not clear; therefore, pilot experiments aiming at finding the best cell density for the enteric pathogen of choice should be performed prior to performing further downstream characterization.
Often FACS sorting is performed prior to performing the single-cell droplet emulsion. This step is often used to separate dead from live cells and single cells from doublets. When working with BSL-3 pathogens, it requires that the facility is equipped with an appropriate FACS sorter, which is not often the case. Further, not all cells in an organoid have the same size, and it is often hard to discriminate between a doublet or a larger cell, thereby causing a risk of negatively selecting against a specific cell type. Also, there is still discussion in the field whether the time needed for sorting between 5,000-10,000 for each sample could result in a significant modification of the transcript profile of the individual cells. While cell fixation methods compatible with single-cell sequencing (e.g., methanol and RNAassist) have been described, it was observed that this leads to a decrease in the quality of sequencing18. Finally, it is suspected that sorting cells using cell death markers can also lead to a bias. Given the directional proliferation and differentiation of the cells through the crypt-villi axis, the most differentiated cells, which are going to be shed and released, are located at the tip of the villi. These cells are often positive for different markers of cell death pathways (e.g., apoptosis, necrosis, and necroptosis); however, when looking at rotavirus infection of mouse intestine, the tip of the villi is the most infected area28. Thus, filtering out cells that may look positive for death markers would result in a negative selection of the infected cells that may represent the physiological infection. Currently, there is no good solution for sorting and fixing organoids prior to single-cell sequencing. Using live, unsorted cells is recommended as further studies are needed to find suitable alternative protocols.
Single-cell sequencing has revolutionized how cellular responses can be evaluated. This technique allows for the identification of cell lineage-specific responses both in basal conditions and under pathogen infections. This method has opened doors in many fields that were previously limited by bulk readouts. While this method is very powerful, it has its limitations. A key limitation is the extensive bioinformatic analysis that is required downstream of the sequencing. This is especially key when analyzing tissues and assigning cell types where there is currently no annotation. Having a skilled bioinformatician is required to support all single-cell studies.
This protocol describes how to seed and handle human intestinal organoids, infect them with enteric pathogens, and perform scRNAseq. Adapting this approach to other organs is now possible, as organoid model systems have been developed for most organs. Lung and liver organoids are similarly organized compared to intestinal organoids, and as such, using an analogous approach could be transposed to these organoids. The critical control will be to validate that when grown in two dimensions or cracked open, these organoids achieve similar differentiation as their 3D organoid counterparts. The specific features and genes that define a differentiated status are specific for each organ model. Other organoid models such as kidney and vascular organoids, large dense structures, will need additional methods to serially dissociate these structures into single cells.
The authors have nothing to disclose.
Megan Stanifer and Steeve Boulant were supported by research grants from the Deutsche Forschungsgemeinschaft (DFG): (Project number 240245660, 278001972, 415089553, and 272983813 to Steeve Boulant and 416072091 to Megan Stanifer), the state of Baden-Wuerttemberg and the Bundesministerium für Bildung und Forschung BMBF 01KI20239B to MS and 01KI20198A and (NUM-COVID 19, Organo-Strat 01KX2021) to SB. Schematics were created in BioRender.
Recombinant mouse noggin | Peprotech | Cat#250-38 | |
[Leu15]-Gastrin I | Sigma-Aldrich | Cat# G9145 | |
0.05% Trypsin-EDTA | Thermo Fischer Scientific | Cat#25300054 | |
24-well non-cell culture treated plate | Corning | Cat#3738 | |
8-well glass bottom chamber slide | iBIDI | Cart#80827 | |
A83-01 | Tocris | Cat#2939 | |
Advanced DMEM/F12 | Thermo Fischer Scientific | Cat# 12634010 | |
B27 | Thermo Fischer Scientific | Cat#17504-044 | |
Chromium Controller & Next GEM Accessory Kit | 10X Genomics | Cat#1000202 | Used in the preparation of single cell solution and preparation of Gel beads-in-emulsion (GEM) |
Chromium Next GEM Chip G Single Cell Kit | 10X Genomics | Cat #1000127 | Used in the preparation of single cell solution and preparation of Gel beads-in-emulsion (GEM) |
Chromium Next GEM Single Cell 3′ Kit v3.1 | 10X Genomics | Cat#1000268 | Used in the preparation of single cell solution and preparation of Gel beads-in-emulsion (GEM) |
Collagen from human placenta | Sigma Aldrich | Cat#C5533-5MG | |
CYP34A forward | Eurofins | GATGGCTCTCATCCCAGACTT | Primers used to check for differentiation |
CYP3A4 reverse | Eurofins | AGTCCATGTGAATGGGTTCC | Primers used to check for differentiation |
DMEM/F12 | Thermo Fischer Scientific | Cat#11320074 | |
EDTA | Sigma Aldrich | Car#E9884 | |
Fast Read 102 counting slides | Biosigma | Cat# BVS100 | |
Fetal Bovein Serum (FBS) | Capricorn | Cat#FBS-12A | |
GlutaMAX | Thermo Fischer Scientific | Cat# 35050061 | |
HEPES | Thermo Fischer Scientific | Cat3 15630080 | |
L-WRN cells | ATCC | CRL-3276 | This cell line is used to make the conditioned media containg Wnt 3A, R-Spondin and Noggin. The protocol for the production of the conditioned media can be found on the manufacterures site. |
MatriGel. GFR, LDEV free | Corning | Cat#354230 | |
MUC-2 forward | Eurofins | TGTAGGCATCGCTCTTCTCA | Primers used to check for differentiation |
MUC-2 reverse | Eurofins | GACACCATCTACCTCACCCG | Primers used to check for differentiation |
N-acetyl-cysteine | Sigma Aldrich | Cat# A9165 | |
OLMF4 forward | Eurofins | ACCTTTCCCGTGGACAGAGT | Primers used to check for differentiation |
OLMF4 reverse | Eurofins | TGGACATATTCCCTCACTTTGGA | Primers used to check for differentiation |
Penicillin/Streptomycin | Thermo Fischer Scientific | Cat#15140122 | |
Recombinant human FGF basic | Peprotech | Cat#100-18B | |
Recombinant human IGF-1 | BioLegend | Cat#590904 | |
Recombinant mouse EGF | Thermo Fischer Scientific | Cat# PMG8043 | |
SI forward | Eurofins | AATCCTTTTGGCATCCAGATT | Primers used to check for differentiation |
SI reverse | Eurofins | GCAGCCAAGAATCCCAAT | Primers used to check for differentiation |
TrypLE Express | Thermo Fischer Scientific | Cat#12605036 | |
Y-27632 | Caymann Chemicals | Cat#10005583 |