Supplementary material

(1)

1 doi: 10.1136/gutjnl-2019-319919 –697. :687 70 2021; Gut , et al. Kayisoglu O

(2)

1. Supplementary figures

Supplementary figure S1 Transcriptome profiling in the human and mouse GI tract reveals segment-specific patterning of PRR signaling components. Graphs depict the normalized gene counts scaled in the heatmaps of figure 2D in the same order. (A) Mouse genes related to PRR signaling pathways. (B) Human genes related to PRR signaling pathways. Bars represent means and s.d. for 3 organoid lines.

(3)

3

Supplementary figure S2 Purity of organoid cultures and confirmation of TLR4 expression in freshly isolated epithelium. Crypts and glands were isolated from 3 mice (A;C) or 3 patients (B;D) per tissue from the segments of the GI tract. Organoids were grown from the tissue and expanded. RNA was isolated and conventional (A-B) or quantitative PCR (C-D) were performed for the listed genes. Control tissues in (A) were: mouse spleen, mouse organoid pool and mesenchyme tissue for Cd45 and Gli1; mouse corpus, duodenum and colon organoids for Cdx1, Muc2 and Muc6; liver, brain and RAW 264.7 cell line for Tlr4. Control tissues in (B) were: human mesenchyme tissue, peripheral blood mononuclear cells (PBMCs) isolated from blood and human duodenum organoids for CD45, FN1 and GAPDH; human corpus, duodenum and colon organoids for CDX1, MUC2 and MUC6; THP1 cell line, HEK293T cell line and human organoid pool for TLR4. In (C) and (D) qPCR data were normalized to Gapdh and then to expression in corpus glands. Bars represent mean with s.d. of 3 organoid lines. Results are representative of at least 3 independent experiments.

doi: 10.1136/gutjnl-2019-319919 –697. :687 70 2021; Gut , et al. Kayisoglu O

(4)

Supplementary figure S3 Polarization of human and murine gastric organoids and 2D monolayers from gastric organoids. To visualize polarization, organoids were either differentiated by withdrawing Wnt, which leads to differentiation of cells to MUC5AC-producing pit cells (A-F) or left in standard growth medium (G-H). Cells were fixed and processed for histology, then stained for mucus using Periodic Acid Schiff (PAS) staining, which shows pink mucus on the apical side (B, C, E, F). Alternatively, organoids or 2D monolayers were processed for immunofluorescent staining of occludin, which marks apical tight junctions (G and H). (A) Cartoon of 3D organoids. (B) Murine gastric 3D organoids. (C) Human gastric 3D organoids. (D) Cartoon of 2D transwell. (E) Murine 2D monolayers grown from organoid cells. (F) Human 2D monolayers grown from cells of organoids. (G) Z-projection confocal image of 3D human gastric organoids. (H) Z-projection confocal image of 2D monolayers derived from human gastric organoids. Boxed areas indicate the areas of magnification. Scale bar, 100 µm (B,C) or 10 µm (E-H).

(5)

5

Supplementary figure S4 TLR4 does not sense LPS in human gastrointestinal organoids, regardless of whether stimulation is apical or basal. (A) Human gastric organoids were seeded in two conditions: For basal stimulation, organoids were cultured as classic 3D organoids; for apical stimulation, they were seeded into 2D monolayers. Cells were treated with LPS 100 ng/mL or TNF-α 10 ng/mL in the supernatant. After 2 hours, cells were harvested, RNA was prepared, and expression of IL-8 was determined by qPCR. (B) Cells from human gastric organoids were seeded into transwells and compounds as in (A) added either to the upper compartment (apical stimulation) or the lower compartment (basal stimulation). After 2 hours, cells were processed as in (A).

(6)

Supplementary figure S5 Still frames of the supplementary movie (see online supplementary movie) depicting oscillations in p65 nuclear translocation. Organoids were generated from the GFP-p65 knock-in mouse lknock-ine and GFP-p65 translocation was monitored by live cell confocal microscopy followknock-ing stimulation with LPS. Scale bars: upper panel 100 µm, upper still frames 100 µm, lower still frames 10 µm.

(7)

7

Supplementary figure S6 (A-B) Cells from organoids were seeded in 2D (A), stimulated with 100 ng/mL LPS for the indicated time, fixed and stained for p65 and nuclei. (B) Cells with nuclear p65 were counted and are presented as percent of total cells. (C) Organoids were generated from wildtype, Myd88 -/-_{or Tlr2/4}-/-_{mice. Cells from organoids were seeded in 2D and stimulated with 100 ng/mL LPS for the} indicated time, fixed and stained for detection of nuclei and p65. Cells with nuclear localization of p65 were quantified and are presented as a percent of the total cells. All samples were compared to non-stimulated controls of the same genetic background. (D) FITC-dextran was microinjected either inside or outside of organoids and its diffusion was observed by live cell microscopy. Scale bar, 200 µm.

(8)

Supplementary figure S7 Transcriptome profiling of the adult and embryonic mouse derived gastric (Stom.) and proximal intestine (Int.) organoids. Graphs depict the normalized gene counts scaled in the heatmaps of figure 6D for the genes related to PRR signaling pathways in the same order. Bars represent means and s.d. for 3 organoid lines, except embryonic stomach with n=2.

(9)

9 doi: 10.1136/gutjnl-2019-319919 –697. :687 70 2021; Gut , et al. Kayisoglu O

(10)

3. Supplementary material and methods

Mouse strains

Wild type C57BL/6 adult mice were provided by Bernhard Nieswandt (Rudolf Virchow Zentrum, Wuerzburg, Germany). Embryos of C57BL/6 mice were provided by Manfred Gessler (Department of Developmental Biochemistry, Wuerzburg, Germany). MyD88-/- _{mice were provided by Mathias Hornef}

(Institute of Medical Microbiology, Uniklinik RWTH Aachen, Germany) [1], who originally obtained it from Jackson Laboratory (stock #009088), Tlr2/Tlr4-/-_{double knockout mice [2] by Arturo Zychlinsky} (Max Planck Institute for Infection Biology, Berlin, Germany) and GFP-p65 knockin mice [3] were provided by Manolis Pasparakis (Institute for Genetics, University of Cologne, Cologne, Germany). Organoid culturing

Human and murine organoids were generated from isolated gastric glands, small intestine and colon crypts to be maintained in culture as described previously [4–7] and passaged to a maximum of passage 5. Dissected murine stomach was separated into pylorus and corpus. Murine small intestine was cut into three equal parts as duodenum (proximal 2 cm of the proximal part), jejunum (proximal 2-4 cm of the middle part) and ileum (distal 2-4 cm of the distal part). Of the murine colon, 2 cm of proximal colon was used. From each tissue, crypts and glands were extracted using EDTA in cold chelating buffer, seeded in Matrigel (BD Biosciences, Franklin Lakes, NJ), and overlaid with medium of the corresponding tissue. Optimized EDTA concentrations and incubation times according to the organism and segment of the GI tract are provided in the table below.

Table for isolation conditions.

Tissue Type EDTA Concentration Incubation Time Temperature

M

o

u

se

Corpus 5 mM 5 min Room temp. (on bench)

Pylorus 5 mM 5 min Room temp. (on bench)

Duodenum 1-2 mM 5 min On ice

Jejunum 1-2 mM 5 min On ice

Ileum 1-2 mM 5 min On ice

Colon 2 mM 5 min On ice

H

u

m

an

Corpus 10 mM 5 min 37°C – (in hands)

Pylorus 10 mM 5 min 37°C – (in hands)

Duodenum 2 mM 5 min Room temp. (on bench)

Jejunum 2 mM 5 min Room temp. (on bench)

Ileum 2 mM 5 min Room temp. (on bench)

Colon 2 mM 5 min Room temp. (on bench)

Media were composed specifically for the species and GI segment as shown in the table below. For basal medium (AD++), Advanced Dulbecco’s modified Eagle medium (DMEM)/F12 supplemented with 10 mmol/L HEPES and GlutaMAX 1X was used. RhoK inhibitor was added only after the initial seeding and passaging of the organoids. The medium was changed every 2–3 days and mouse organoids were passaged at a ratio of 1:5 every week, and human organoids 1:5 every two weeks. For the mRNA sequencing, organoids were expanded for a maximum of 5 passages. For 2D cultures, organoids were

(11)

11

Organoid maintenance

Matrigel Corning 356231

Cell Culture

Freezing Medium Invitrogen 12648-010

Basal medium (AD++)

HEPES Invitrogen 15630-056

GlutaMAX-I Invitrogen 35050-079

Advanced

DMEM/F12 (AD) Invitrogen 12634-028

Media components Mouse organoids Human organoids

Reagent Supplier Catalog number Adult sto./ Embr. sto.

Adult duod.

jej./ Embr. int. Ile. / Col. Stomach

Sm. int. / Col.

AD++ 30% 80% 30% 30% 30%

Wnt CM Stable cell line 50% – 50% 50% 50%

R-Spondin CM Stable cell line 10% 10% 10% 10% 10%

Noggin CM Stable cell line 10% 10% 10% 10% 10%

Primocin Invivogen Ant-pm-1 100 ng/mL 100 ng/mL 100 ng/mL 100 ng/mL 100 ng/mL

B27 Invitrogen 12587-010 1x 1x 1x 1x 1x

N-Ac Sigma-Aldrich A9165-5G 1.25 mM 1.25 mM 1.25 mM 1.25 mM 1.25 mM

EGF Peprotech AF-100-15 50 ng/mL 50 ng/mL 50 ng/mL 50 ng/mL 50 ng/mL

FGF-10 Peprotech 100-26 100 ng/mL – – 100 ng/mL –

Gastrin-I Tocris 3006 10 nM – – 1 nM 10 nM

TGF-β inh. Tocris A-83-01 / 2939 – – – 2 µM 0.5 µM

p38 inh. Sigma-Aldrich SB202190 / 1264 – – – – 10 µM Nicotinamide Sigma-Aldrich N0636 – – – – 10 mM RhoK-inh. AbMole Bioscience Y-27632 / M1817 10 µM 10 µM 10 µM 10 µM 10 µM doi: 10.1136/gutjnl-2019-319919 –697. :687 70 2021; Gut , et al. Kayisoglu O

(12)

RNA sequencing and bioinformatic analysis

RNA sequencing was performed using the Illumina NextSeq 500 platform. Library preparation and sequencing were performed by the Core Unit Systems Medicine (CU SysMed), University of Wuerzburg, Germany. Libraries for RNA sequencing were prepared using the Illumina TruSeq Stranded mRNA Library Prep Kit (Cat#20020594) with 500 ng of total RNA for the construction of sequencing libraries. RNA sequencing was performed using single-end reads on the Illumina NextSeq 500 platform. Bioinformatic analysis was performed in the R software package V.3.4.4 using Bioconductor V.3.6 packages. Low quality reads and adapter sequences were removed using cutadapt (V.1.16) [9]. Reads were mapped to the mouse reference genome GRCm38 or the human reference genome GRCh38 using READemption (V.0.4.3) [10]. Raw read alignments were counted using htseq-count [11]. RUVseq batch correction was performed by calling the ruvr function [12]. Raw counts were normalized by DESeq2’s [13] median of ratios method and these values were depicted in the bar graphs. Differential gene expression analysis was performed on normalized counts using DESeq2 with a cut-off of adjusted p<0.05. The Benjamini-Hochberg method was used to adjust the p-values. For the heatmaps, the normalized counts were scaled for Z-score with row means of 0, and heatmaps were created using ComplexHeatmap [14]. Gene ontology (GO) enrichment analysis was performed using the web application of g:Profiler [15]. p<0.05 was considered as statistically significant and GO-terms were filtered to less than 3,000 genes for more definitive results.

Gene expression analysis by conventional PCR and qPCR

RNA was isolated from organoids, isolated gastric glands or intestinal crypts according to the manufacturer’s recommendations (RNeasy Mini kit; Qiagen, Venlo, Netherlands). For RNA isolation from organoids, Matrigel was first removed from the cultures by gentle disruption with cold Advanced DMEM. Complementary DNA was generated using established reverse transcriptase protocols (M-MuLV, NEB or ProtoScript® II, NEB). Quantitative reverse transcription PCR (qPCR) was performed using SYBR green (Bio-Rad) and the CFX384 Real-Time PCR Detection System (Bio-Rad). Results were calculated using the ΔΔCt method. Relative quantification was achieved by normalizing results to the values obtained for glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Conventional PCRs and qPCRs were performed using the same primers, which are listed in the table below.

Table for primers used for conventional PCRs and qPCRs of murine and human genes.

Gene Forward Reverse Product

(bp)

Reference

(if previously reported) Primer pairs for mouse genes

Gapdh 5’-GTGCCAGCCTCGTCC-3’ 5’-ACCCCATTTGATGTTAGTGG-3’ 283

Cxcl2 5’-AAGTTTGCCTTGACCCTGAA-3’ 5’-AGGCACATCAGGTACGATCC-3’ 180

Cdx2 5’-CCTAGGAAGCCAAGTGAAAA-3’ 5’-TGCGGTTCTGAAACCAAAT-3’ 185

Muc2 5’-GTGTGTTGCTCAATGAGATG-3’ 5’-TCTAGGCCATTGAAGTTTCC-3’ 211

Muc6 5’-AACCTGCAATCCTCCCCAGAA-3’ 5’-GCTGGATGCTAAAGGTGGCG-3’ 157

Cd45 5’-GACCCTATTTCTTAGGGGCA-3’ 5’-CTCTGTTGTGCTCAGTTCATC-3’ 150

Gli1 5’-AAGGGGACATGTCTAGCCCC-3’ 5’-ACAGCCTTCAAACGTGCACT-3’ 338 Farin et al. [16]

(13)

13

FN1 5’-CAGTGGGAGACCTCGAGAAG-3’ 5’-TCCCTCGGAACATCAGAAAC-3’ 168 Sudo et al.[18]

TLR4 5’-TGGATACGTTTCCTTATAAG-3’ 5’-GAAATGGAGGCACCCCTTC-3’ 507 Kumar et al.[19]

Stimulation experiments and microinjection

To investigate activation of the NF-κB pathway, 3D organoids or 2D monolayers were incubated for the indicated times with medium containing LPS from Escherichia coli (tlrl-eklps, Invivogen), flagellin from Salmonella typhimurium (tlrl-epstfla, Invivogen), Pam3CSK4 (tlrl-pms, Invivogen) or recombinant human TNF-α (BD Pharmingen, Carlsbad, CA) at the indicated final concentrations. Stimulations were performed, shortly before the day on which they were due to be passaged to allow their maturation (mouse organoids on day 5, human organoids on day 10-12). For transwell experiments, around 200,000 cells were seeded into collagen-coated (A10644-01, Gibco) transwell inserts (PIHP01250, Millipore) placed into a 24-well plate. For microinjection, 3D organoids were seeded in 50 µL of Matrigel in 4-well multidishes (Thermo Scientific, Waltham, MA), for isolation of RNA 24-well multidishes, or 8-24-well µ-slides (Ibidi, Martinsried, Germany) for microscopy. Organoids were microinjected with LPS (100 ng/mL) or fluorescein isothiocyanate (FITC)-dextran (0.5 ng/mL; sc-263323-Santa Cruz Biotechnology) on day 5 using a micromanipulator and microinjector (Eppendorf FemtoJet® 4i) together with a stereomicroscope within a sterile safety cabinet (Kojair Biowizard Silverline, Vippula, Finnland). FITC-dextran was imaged after injection by live cell microscopy using a Leica DMI6000B (Leica Microsystems) with a Leica DFC360 FX camera using a 10 X objective. For LPS stimulation, 50 organoids were injected in each well. Injections were performed in a paired manner inside and outside the organoids on the same plate. 2 h after injection of LPS, organoids were removed from Matrigel and lysed for RNA isolation. While we do not exclude that efflux from injected organoids through the injection hole is possible, live cell microscopy showed that microinjected FITC-dextran mainly remained inside the injected organoid. It quickly diffused when it was placed outside of the organoids (online supplementary figure S6D). qPCR data of LPS injected organoids showed that stimulation at the apical side (inside organoids) led to higher levels of Cxcl2 mRNA than stimulation at the basal side (outside organoids) (figure 4K). Due to the different diffusion dynamics, this does not allow comparison of TLR4 activation on the two sides in absolute quantities, but we can conclude that murine gastric TLR4 can sense both, basally and apically administered LPS.

Quantification of nuclear translocation of NF-κB

Dissociated murine gastric organoids were seeded onto Matrigel-coated glass coverslips to form 2D monolayers for immunofluorescence staining. Cells were grown for 5 days to reach approximately 80% confluency and stimulated with LPS as indicated in individual figures. Cells were fixed with 2% PFA for 20 min at room temperature, washed 3 times with PBS and permeabilized in 1X PBS supplemented with 0.3% Triton-X and 1% fresh BSA for 1 h. Immunofluorescent staining was performed with anti-p65 rabbit monoclonal antibody (8242, Cell Signaling), 1:400 in 1X PBS supplemented with 0.3% Triton-X and 1% BSA overnight at 4 °C followed by Alexa Fluor 488-conjugated anti-rabbit IgG (Cell Signaling), 1:250 in 1X PBS supplemented with 0.3% Triton-X and 1% BSA for 3 h at room

(14)

temperature. DNA was stained with Hoechst 33342. After washing 3 times with PBS, cells were mounted on glass slides in Mowiol and visualized using a fluorescence microscope (EVOS, Thermo Fischer Scientific). To quantify the percentage of cells with p65 localized in the nucleus, 9 frames per well were imaged with the EVOS fluorescence microscope and p65 nuclear translocation was manually quantified for all 9 images, resulting in a total of several hundred (usually around 1,500) cells quantified. Statistical analysis

Data represent mean±s.d. p=0.05 was taken as the maximum value for significance. The applied statistical tests and the level of significance are indicated in the figure legends respectively.

(15)

15

Corpus_org_3 Patient-32 F 47 Sleeve stomach

Pylorus_org_1 Patient-01 F 32 Sleeve stomach

Pylorus_org_2 Patient-36 M 55 Stomach carcinoma

Pylorus_org_3 Patient-42 M 58 Stomach carcinoma

Duodenum_org_1 Patient-43 F 75 Whipple

Duodenum_org_1 Patient-54 F 64 Whipple

Duodenum_org_1 Patient-55 M 60 Whipple

Jejunum_org_1 Patient-44 F 34 Stomach by-pass

Jejunum_org_2 Patient-48 M 50 Stomach by-pass

Jejunum_org_3 Patient-50 M 53 Stomach by-pass

Ileum_org_1 Patient-02 M 75 Hemicolectomy

Ileum_org_2 Patient-03 M 83 Colon carcinoma

Ileum_org_3 Patient-51 M 70 Colon carcinoma

Colon_org_1 Patient-34 M 52 Sigmoid colon carcinoma

Colon_org_2 Patient-35 F 76 Caecal carcinoma

Colon_org_3 Patient-37 F 74 Caecal carcinoma

(16)

4. Supplementary references

1 Hou B, Reizis B, DeFranco AL. Toll-like Receptors Activate Innate and Adaptive Immunity by using Dendritic Cell-Intrinsic and -Extrinsic Mechanisms. Immunity 2008;29:272–82.

doi:10.1016/j.immuni.2008.05.016

2 Weiss DS, Takeda K, Akira S, et al. MyD88, but Not Toll-Like Receptors 4 and 2, Is Required for Efficient Clearance of Brucella abortus. Infection and Immunity 2005;73:5137–43.

doi:10.1128/IAI.73.8.5137-5143.2005

3 De Lorenzi R, Gareus R, Fengler S, et al. GFP-p65 knock-in mice as a tool to study NF-κB dynamics in vivo. genesis 2009;47:323–9. doi:10.1002/dvg.20468

4 Sato T, Vries RG, Snippert HJ, et al. Single Lgr5 stem cells build crypt villus structures in vitro without a mesenchymal niche. Nature 2009;459:262–265.

5 Barker N, Huch M, Kujala P, et al. Lgr5+ve Stem Cells Drive Self-Renewal in the Stomach and Build Long-Lived Gastric Units In Vitro. Cell Stem Cell 2010;6:25–36.

doi:10.1016/j.stem.2009.11.013

6 Sato T, Stange DE, Ferrante M, et al. Long-term Expansion of Epithelial Organoids From Human Colon, Adenoma, Adenocarcinoma, and Barrett’s Epithelium. Gastroenterology 2011;141:1762– 72. doi:10.1053/j.gastro.2011.07.050

7 Bartfeld S, Bayram T, van de Wetering M, et al. In Vitro Expansion of Human Gastric Epithelial Stem Cells and Their Responses to Bacterial Infection. Gastroenterology 2015;148:126-136.e6. doi:10.1053/j.gastro.2014.09.042

8 Boccellato F, Woelffling S, Imai-Matsushima A, et al. Polarised epithelial monolayers of the gastric mucosa reveal insights into mucosal homeostasis and defence against infection. Gut 2018;:gutjnl–2017.

9 Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet.journal 2011;17:10. doi:10.14806/ej.17.1.200

10 Forstner KU, Vogel J, Sharma CM. READemption—a tool for the computational analysis of deep-sequencing–based transcriptome data. ;:3.

11 Anders S, Pyl PT, Huber W. HTSeq--a Python framework to work with high-throughput sequencing data. Bioinformatics 2015;31:166–9. doi:10.1093/bioinformatics/btu638

12 Risso D, Ngai J, Speed TP, et al. Normalization of RNA-seq data using factor analysis of control genes or samples. Nat Biotechnol 2014;32:896–902. doi:10.1038/nbt.2931

13 Klaus B. Differential expression analysis of RNA – Seq data using DESeq2. 2014;:1–24. 14 Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in

multidimensional genomic data. Bioinformatics 2016;32:2847–9. doi:10.1093/bioinformatics/btw313

15 Reimand J, Arak T, Adler P, et al. g:Profiler—a web server for functional interpretation of gene lists (2016 update). Nucleic Acids Res 2016;44:W83–9. doi:10.1093/nar/gkw199

16 Farin HF, Van Es JH, Clevers H. Redundant Sources of Wnt Regulate Intestinal Stem Cells and Promote Formation of Paneth Cells. Gastroenterology 2012;143:1518-1529.e7.

(17)

17

like receptor signaling. PLoS ONE 2012;7. doi:10.1371/journal.pone.0029830