Development of a Scalable Coculture System for Gut Anaerobes and Human Colon Epithelium

Abstract

The intestinal epithelium resides at the interface of the gut microbiota and plays a pivotal role in shaping the gut ecosystem.1Maslowski K.M. Clin Exp Immunol. 2019; 197: 193-204Crossref PubMed Scopus (29) Google Scholar Owing to a lack of tractable coculture systems, the cell biological understanding of host–microbe interactions remains elusive. Organoid technology allows propagation of colonic epithelium under normoxia; however, the conflicting oxygen demands between epithelium and gut anaerobes makes their coculture difficult. We established a simple 2-chamber culture system for human colonic epithelium, termed as Intestinal Hemi-Anaerobic Coculture System (iHACS), consisting of a hypoxic apical chamber and a normoxic basal chamber. The medium in the apical chamber was equilibrated with anaerobic gas and subsequently sealed by inserting a plug made of butyl rubber (AsONE international, Santa Clara, CA). The oxygen concentration of the apical chamber was measured by a fiberoptic oxygen meter (PreSens. Regensburg, Germany). For bacterial coculture, we inoculated anaerobic bacteria (5 × 104 cells/mL in Bifidobacterium adolescentis, Bacteroides fragilis, Clostridium butyricum, and 5.0 × 105 cells/mL in Akkermansia muciniphila) in the apical chamber medium. Recently reported insulinlike growth factor 1–fibroblast growth factor 2 culture condition enabled monolayer growth of human colonic organoids within 5 days after seeding disassociated cells from 3-dimensional cultured organoids (Supplementary Figure 1A). The 2-dimensional colonic epithelial cells display an intact stem cell hierarchy and a demarcated proliferative zone, reminiscent of crypt epithelium, and transepithelial electrical resistance confirmed a highly integrated epithelial layer (Supplementary Figure 1B–E). Hypoxia is essential for the growth of obligate anaerobes, but prohibitive for the maintenance of viable epithelial layer (Supplementary Figure 1D). To surmount this tradeoff in oxygen demands, we developed iHACS, consisting of a hypoxic apical chamber and a normoxic basal chamber (Figure 1A). The plug inserts tightly, physically blocking the influx of external oxygen, which allows maintenance of hypoxia in the apical chamber, while oxygen freely perfuses the basal chamber (Figure 1B). Immunostaining of hypoxia-inducible factor 1α visualized a minimal hypoxic response in iHACS-cultured epithelium, suggesting a sufficient oxygen supply from basal membrane in this culture system (Supplementary Figure 1D). Consistent with the absence of a hypoxic response, iHACS-cultured epithelium showed intact cell polarity, stem cell hierarchy, and mucin layer (Supplementary Figure 1D, F). Of note, consistent with previous findings of oxygen consumption by mature epithelial cells,2Litvak Y. et al.Science. 2018; 362eaat9076Crossref PubMed Scopus (360) Google Scholar we observed reduced oxygen level in the apical chamber even without the plug insert, although to a lesser degree than that with the plug insert (Figure 1B). In contrast to the normoxic culture condition, iHACS successfully propagated B adolescentis, one of the major obligate anaerobe bacterial commensals, and scanning electron microscopy illustrated the attachment of B adolescentis onto the intact epithelium. (Figure 1C and D and Supplementary Figure 1G). We also confirmed colony formation of other obligate anaerobic bacteria, B fragilis, a major member of human gut microbiota, and C butyricum, a widely used anaerobic probiotic (Supplementary Figure 1H). These results indicate that the hemi-anaerobic environment is conducive to both normal human colonic epithelium and a range of anaerobic bacteria. Notably, on exposure to anaerobes, colon epithelium upregulated the expression of MUC2 (Figure 1E), which phenocopied an increased number of goblet cells in germ-free mice after the inoculation of bacterium.3Johansson M.E. et al.Cell Host Microbe. 2015; 18: 582-592Abstract Full Text Full Text PDF PubMed Scopus (310) Google Scholar We also observed significant upregulation of stem cell marker genes such as LGR5, AXIN2, and PTK7 by B adolescentis coculture (Figure 1E). Neither conditioned medium of B adolescentis nor heat-killed B adolescentis upregulated these stem cell marker genes, suggesting that B adolescentis had stem cell–promoting effects on colonic epithelium through interaction with live bacteria (Supplementary Figure 1I). The same results were confirmed with a different donor-derived organoid line (Supplementary Figure 1I and J). To demonstrate the biological effect of epithelium on gut bacterium, we used an obligate anaerobe, A muciniphila, which requires animal-derived mucin in the bacterial culture medium.4Derrien M. et al.Int J Syst Evol Microbiol. 2004; 54: 1469-1476Crossref PubMed Scopus (1376) Google Scholar Interestingly, A muciniphila degraded the mucin layer and was capable of growing without the addition of exogenous mucin in iHACS for 5 days (Figure 1F–I). To determine the essential role of goblet cell–derived mucin for the growth of A muciniphila, we cultured organoids in conditions that inhibit lineage differentiation (addition of p38 inhibitor and nicotinamide), and then cocultured with A muciniphila. This culture condition decreased MUC2 expression, while not affecting the expression of other types of mucins (Supplementary Figure 2A). Of note, the undifferentiated organoids failed to support the growth of A muciniphila (Supplementary Figure 2B). The absence of a growth-promoting effect of undifferentiated epithelium on A muciniphila was further corroborated by generating ATOH1-knockout organoids that lack goblet cells (Figure 1I and Supplementary Figure 2C). These results demonstrated that differentiated colonic epithelium can provide an essential nutrient for gut bacterium in iHACS. To determine whether the epithelium and gut microbes compete for nutrients in iHACS, we measured bacterial growth in a fresh cell culture medium or 2-day supernatant from the apical chamber of an epithelial monolayer. B adolescentis showed retarded growth in the supernatant, as compared with fresh medium (Supplementary Figure 2D). Decreased glucose level in the supernatant suggested that the epithelial glucose consumption suppressed the growth of B adolescentis (Supplementary Figure 2E). In contrast to B adolescentis, A muciniphila use mucin instead of glucose as a carbon source and grew better in the supernatant than in the bacterial medium (Supplementary Figure 2F). Interestingly, when coculturing with control organoids but not with ATOH1-organoids, A muciniphila produced acetic acid and propionic acid, indicating that A muciniphila used mucin as a sole carbon source and metabolized the substrate into short fatty acids5Ottman N. et al.Appl Environ Microbiol. 2017; 83: e01014-e01017Crossref PubMed Scopus (151) Google Scholar (Supplementary Figure 2G). These results demonstrate the use of iHACS in modeling competitive and mutually beneficial relationships between human colonic epithelium and gut microbiota in vitro. In this study, we established iHACS, a system that provides optimal growth environments for human colonic epithelium and gut anaerobes. Jalili-Firoozinezhad et al6Jalili-Firoozinezhad S. et al.Nat Biomed Eng. 2019; 3: 520-531Crossref PubMed Scopus (444) Google Scholar very recently reported similar bacterial coculture system with a colon cancer cell line and human ileal epithelium using a microfluidics device. In contrast to their system, iHACS does not need specialist devices and can be applied to a variety of experimental formats. Furthermore, iHACS is the first system to enable coculturing of anaerobic bacteria with human colonic epithelium. Human colonic epithelium is more relevant to human disease pathogenesis, and host–anaerobe interactions, than the small intestine, highlighting the broad applicability of iHACS for studying epithelial-bacterial responses in association with gastrointestinal diseases. A recent report has shown that reactive oxygen species derived from inflammatory cells modulated the luminal bioavailability of oxygen. Future study will be required to incorporate nonepithelial cells into iHACS to understand host bacterial responses in a disease setting.2Litvak Y. et al.Science. 2018; 362eaat9076Crossref PubMed Scopus (360) Google Scholar,7Campbell E.L. et al.Immunity. 2014; 40: 66-77Abstract Full Text Full Text PDF PubMed Scopus (343) Google Scholar The authors thank Kazuya Arai, Mami Matano, and Mariko Shimokawa at Sato Lab; Ryo Aoki at Ezaki Glico Co., Ltd; and Takahiro Suzuki at Miyarisan Pharmaceutical Co., Ltd. for assistance with critical experiments. Furthermore, we thank all group members of the Sato Lab for fruitful discussion. Nobuo Sasaki, PhD (Conceptualization: Lead; Data curation: Lead; Formal analysis: Lead; Funding acquisition: Lead; Investigation: Lead; Methodology: Lead; Project administration: Lead; Validation: Lead; Visualization: Lead; Writing – original draft: Lead; Writing – review & editing: Lead; Share co-corresponding author: Lead). Kentaro Miyamoto, MS (Data curation: Supporting; Formal analysis: Supporting; Investigation: Equal; Validation: Equal). Kendle M. Maskiwski, Ph.D. (Investigation: Supporting; Validation: Supporting; Writing – original draft: Supporting; Writing – review & editing: Supporting). Hiroshi Ohno, MD, PhD (Conceptualization: Supporting; Investigation: Supporting; Supervision: Supporting). Takanori Kanai, MD, PhD (Resources: Supporting). Toshiro Sato, MD, PhD (Conceptualization: Lead; Formal analysis: Equal; Funding acquisition: Lead; Resources: Lead; Supervision: Lead; Writing – original draft: Lead; Writing – review & editing: Lead). Human healthy colonic organoids were previously established.1Maslowski K.M. Clin Exp Immunol. 2019; 197: 193-204Crossref PubMed Scopus (29) Google Scholar The ethics committee at Keio University School of Medicine (Tokyo, Japan) approved the protocol (no. 20140211). Three-dimensional colonic organoids were maintained with Modified human colonic organoid (MHCO) medium,1Maslowski K.M. Clin Exp Immunol. 2019; 197: 193-204Crossref PubMed Scopus (29) Google Scholar consisting of advanced Dulbecco’s modified Eagle’s medium (DMEM)/F12 supplemented with penicillin/streptomycin, 10 or 100 mM HEPES, 2 mM Glutamax, 1 × B-27 Supplement (Thermo Fisher Scientific, Waltham, MA), 10 nM gastrin I (Sigma-Aldrich, St Louis, MO), 1 mM N-acetylcysteine (Sigma-Aldrich), 100 ng/mL recombinant mouse Noggin (PeproTech, Rocky Hill, NJ), 50 ng/mL recombinant mouse epidermal growth factor (Thermo Fisher Scientific), 100 ng/mL recombinant human insulinlike growth factor-1 (BioLegend, San Diego, CA), 50 ng/mL recombinant human fibroblast growth factor-basic (FGF-2) (PeproTech), 1 μg/mL recombinant human R-spondin1 (R&D, Minneapolis, MN), 500 nM A83–01 (Tocris, Bristol, UK), 10 μM Y-27632 (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) and 20% Afamin-Wnt-3A serum-free conditioned medium.2Litvak Y. et al.Science. 2018; 362eaat9076Crossref PubMed Scopus (360) Google Scholar Undifferentiated medium of human colonic organoid was made by additional 3 μM SB202190 (Sigma-Aldrich) and 10 mM nicotinamide (Sigma-Aldrich) into MHCO medium.3Johansson M.E. et al.Cell Host Microbe. 2015; 18: 582-592Abstract Full Text Full Text PDF PubMed Scopus (310) Google Scholar Organoids were passaged approximately every week by physical dissociation using fire-polished Pasteur pipettes. To generate monolayers derived from 3-dimensional culture of human colonic organoids, transwell culture inserts (24-well insert, 0.4 μm pore polyester membrane; Greiner bio-one, Kremsmunster, Austria) were coated with 5% Matrigel diluted with advanced DMEM/F12 medium and incubated at 37°C for 30 minutes, then Matrigel solution was removed and the membrane was dried in a tissue-culture hood for 15 minutes. Human colonic organoids were cultured for 5 to 7 days before being used to plate into monolayer culture in MHCO medium. Three-dimensional cultured organoids were treated with TrypLE Express (Thermo Fisher Scientific) to dissociate into single cells. The cells were resuspended to 1 to 2 × 106 cells/mL in MHCO medium, and 200 μL of cell suspension was added into the transwell inserts. The medium was changed every 2 days until cells were growing as confluent epithelial monolayers. Bifidobacterium adolescentis was obtained from the Japan Collection of Microorganisms (RIKEN BioResource Center, Ibaraki, Japan), Bacteroides fragilis, Clostridium butyricum, and Akkermansia muciniphila from the American Type Culture Collection (Manassas, VA). B. adolescentis, B. fragilis, C. butyricum, and A. muciniphila were anaerobically cultured in Gifu Anaerobic Broth for 1 to 5 days before starting the coculture experiment (GAM Broth, Modified; Nissui, Tokyo, Japan). The bacteria were collected by centrifugation and washed in GAM (B. adolescentis, B. fragilis, C. butyricum) or minimal mineral (MM) medium (1.9 mM NH4Cl, 810 μM MgSO4x7H2O, 680 μM CaCl2×2H2O, 290 μM K2HPO4, 20mM NaHCO3, 10 mM Thiosulfate; A. muciniphila). The bacteria were also counted under the microscope using Bacteria Counter (SLGC Japan) and were diluted to the required concentration using oxygen-depleted antibiotics free-MHCO for B adolescentis or MM medium for A muciniphila, respectively, then 200 μL of each diluted bacteria were inoculated on the confluent monolayered epithelial cells in transwell inserts. The diluted bacteria were also seeded onto GAM- or BL-based (Nissui) agar plates to count colony-forming units. To prepare heat-killed and supernatant of B adolescentis, the bacteria were anaerobically cultured in antibiotic-free MCHO, and spun down by centrifugation to separate supernatant and cells. The supernatant was filtered through a 0.22-μm pore size filter unit (Millipore, Bedford, MA), and cell pellets were incubated at 95°C for 30 minutes. For the scanning electron microscopy (SEM) analysis, monolayered epithelial cells with or without bacteria were treated with 2.5% glutaraldehyde (electron microscopy grade; Nacalai Tasque, Kyoto, Japan) overnight at 4°C and stained for 1 hour with 1% osmium tetroxide (TCI Chemicals, Tokyo, Japan) at room temperature before serial dehydration in ethanol. Samples were then coated with gold sputtering and observed under a VHX-D510 microscopy (Keyence, Osaka, Japan). Colonic epithelial cells were fixed with 4% paraformaldehyde (Wako) for immunofluorescence analysis or fixed with Carnoy’s solution for mucin layer detection, for 15 minutes at room temperature and subsequently wash by phosphate-buffered saline. After permeabilization of cells by 0.5% Triton X-100 (Sigma-Aldrich), the epithelial cells were incubated with 5% Block Ace (KAC) buffer for 1 hour at 4°C. Primary antibodies against CHGA (sc-1488, dilution 1:200; Santa Cruz, Dallas, TX), MUC2 (sc-15334, 1:100; Santa Cruz), ZO-1 (187430, 1:100; Thermo Fisher Scientific), E-cadherin (610181, 1:250; BD biosciences, San Jose, CA), or HIF1a (GTX127309, 1:300; GeneTex, Irvine, CA) were added and incubated overnight at 4°C. Cells were then washed by phosphate-buffered saline and incubated with secondary fluorescently conjugated antibodies (Thermo Fisher Scientific) for 90 minutes at room temperature. After counterstaining by Hoechst 33342 (DOJINDO, Kumamoto, Japan) or Alexa Flour 647 Phalloidin (Thermo Fisher Scientific), microscopy analysis was performed with a laser scanning confocal microscope (LSM710; Carl Zeiss, Oberkochen, Germany) and a fluorescent microscopy (Keyence BZ-X800). Total RNA was isolated from monolayered human colonic epithelium using RNeasy Mini Kit (Qiagen, Hilden, Germany), and complementary DNA was then synthesized from 350 ng total RNA by SuperScript IV Reverse Transcriptase (Thermo Fisher Scientific) according to the manufacture’s protocol. To determine gene expression, real-time quantitative polymerase chain reaction was performed using Universal Probe Library probes and FastStart Essential DNA Probes Master (Roche, Basel, Switzerland) on CFX96 Real-Time System (BIO-RAD, Hercules, CA). The level of target gene expression was normalized to GAPDH housekeeping gene expression in each sample. The barrier integrity of monolayer culture of human colonic epithelium or iHACS was measured with a transepithelial Volt/Ohm meter (Millipore MERS) following the manufacture’s protocol. The oxygen concentration was measured with potable fiber optic oxygen meter (Microx 4 trace; PreSens) that was calibrated according to the manufacturer’s instruction. Measurements were performed at a consistent position within the cell inserts and a bottom well of plates. When measuring the oxygen concentration in the iHACS, the culture plate was transferred into the anaerobic chamber, the rubber plug detached, and the oxygen meter applied as described previously. For gene knockout using CRISPR-Cas9, a human ATOH1 gene specific single-guide RNA (sgRNA) was cloned into the pX330-U6-Chimeric_BB-CBh-hSpCas9 vector (obtained from AddGene, Cambridge, MA). ATOH1 target sequences: number 1, 5ʹ- GGACGAGGTGGACGGCCGGG-3ʹ; number 2, 5ʹ- ACGGGATAACATTGCGCAGC-3ʹ. Knockout of ATOH1 gene was performed by cointroducing the CRISPR vector into organoids with a neomycin-expressing piggyBAC vector (PB533A-1; SBI, Palo Alto, CA) by electroporation. Transfected organoids were then treated with both G418 Geneticin (200 μg/mL; Thermo Fisher Scientific) and γ-secretase inhibitor DAPT (10 μM; FUJIFILM Wako Pure Chemical Corporation) to enrich organoids that had undergone successful plasmid delivery and introduce loss of function of ATOH1. The organoid electroporation protocol was previously described in detail.4Derrien M. et al.Int J Syst Evol Microbiol. 2004; 54: 1469-1476Crossref PubMed Scopus (1376) Google Scholar To confirm gene knockout of clonal organoid, each product was cloned into TOPO-TA cloning kit (Thermo Fisher Scientific) and subsequently 4 independent bacterial clones were sequenced using T7 sequencing primer. After intestinal epithelium become confluent, transwell inserts were incubated with 10 μM EdU (Alexa Fluor 594 picolyl azide; Molecular probes, Eugene, OR) for 4 days, exchanging with fresh MHCO medium containing the same dose of the EdU daily. Epithelial cells were then washed 3 times in advanced DMEM/F12 and incubated in MHCO medium for 24 hours. For confocal analysis, cells were fixed in 4% formaldehyde and following the manufacture’s instruction of Click-iT EdU (Molecular probes). Samples of fresh medium or culture supernatant of epithelium without bacteria were collected after 24- or 48-hour incubation. Samples were centrifuged at 15,350g for 10 minutes and the supernatant filtered with a 0.45-μm filter. Short-chain fatty acids were measured by a high-performance liquid chromatography following the manufacturer’s protocol (Prominence; SHIMADZU, Kyoto, japan) using a past column reaction with a detector (CDD-10A; SHIMADZU), tandemly arranged 2 columns and a guard column (Shim-pack SCR-102). The concentration of glucose was assayed by hexokinase reaction in 340-nm ultraviolet-absorption spectrophotometry following the manufacturer’s instruction. Values are presented as mean ± standard deviation from n = 3 to 12 (see figure legends). Statistical significance was determined with a 1-way analysis of variance with a Tukey’s or by were 2 conditions using Figure of gene expression with K.M. Clin Exp Immunol. 2019; 197: 193-204Crossref PubMed Scopus (29) Google Scholar in human healthy intestinal epithelium cultured with or without (MHCO) p38 inhibitor or insulinlike growth factor 1 and fibroblast growth factor of A muciniphila after in the MM medium using iHACS epithelium cultured in MHCO medium, undifferentiated medium (see Supplementary or undifferentiated medium 0.5% exogenous = 3 A of ATOH1-knockout in human colonic epithelium. in 3-dimensional organoid of in ATOH1 gene a at 104 which is the of before the the function in human colonic epithelium. goblet cells are observed in the confirmed by of MUC2 fiber and are with and Hoechst 50 A growth of B adolescentis cultured in fresh medium or epithelial cell culture supernatant The was measured at the times = 3 for each shown are with standard The concentration of glucose in fresh medium or epithelial culture supernatant measured by hexokinase to count number of A muciniphila after of culture in fresh medium, culture supernatant of or ATOH1-knockout epithelial cells = 3 for each of fatty as determined by a of supernatant of A muciniphila culture in iHACS for using normal or ATOH1-knockout epithelium. The are presented as mean ± standard from 1-way analysis of and Figure

Topics & Concepts