Targeting zonulin and intestinal epithelial barrier function to prevent onset of arthritis
Nature Communications volume 11, Article number: 1995 (2020) Cite this article
Gut microbial dysbiosis is associated with the development of autoimmune disease, but the mechanisms by which microbial dysbiosis affects the transition from asymptomatic autoimmunity to inflammatory disease are incompletely characterized. Here, we identify intestinal barrier integrity as an important checkpoint in translating autoimmunity to inflammation. Zonulin family peptide (zonulin), a potent regulator for intestinal tight junctions, is highly expressed in autoimmune mice and humans and can be used to predict transition from autoimmunity to inflammatory arthritis. Increased serum zonulin levels are accompanied by a leaky intestinal barrier, dysbiosis and inflammation. Restoration of the intestinal barrier in the pre-phase of arthritis using butyrate or a cannabinoid type 1 receptor agonist inhibits the development of arthritis. Moreover, treatment with the zonulin antagonist larazotide acetate, which specifically increases intestinal barrier integrity, effectively reduces arthritis onset. These data identify a preventive approach for the onset of autoimmune disease by specifically targeting impaired intestinal barrier function.
Altered intestinal microbiota composition, termed “dysbiosis”, is associated with autoimmune diseases, particularly type 1 diabetes 1 , multiple sclerosis 2 , systemic lupus erythematosus 3 and rheumatoid arthritis (RA) 4 , 5 . Several findings support the hypothesis that the onset of RA is linked to the intestinal microbiota, for example that cell wall fragments from various intestinal bacteria have been found to be arthritogenic 6 , 7 , some drugs used to treat arthritis have antimicrobial effects (chloroquine, sulfasalazine and minocycline) 8 , 9 , 10 and the restoration of eubiosis in arthritis patients showing clinical improvement 7 , 11 . Moreover, diet as the main microbiota influencing factor has been shown to influence arthritis 12 , 13 . However, direct mechanistic links between the gut microbiota and onset of autoimmune diseases are unclear.
The permeability against small substances of the intestinal epithelium depends on the regulation of intercellular tight junctions (TJ). Zonula occludens toxin (Zot, or zonulin), an enterotoxin secreted by intestinal epithelial cells following stimuli from diets or microbiota, was shown to be a potent regulator of TJ competency and intestinal barrier function. Importantly, breakdown of the intestinal barrier, e.g. by apoptosis of intestinal epithelial cells in response to microbial infection, enables a proinflammatory environment including differentiation of autoreactive Th17 cells and other T helper cells 14 . Our data show that in mice intestinal barrier function is impaired before the clinical onset of arthritis, while in humans serum markers associated with impaired intestinal barrier function are also increased before the onset of RA and associated with a higher risk to develop RA later on. We identify zonulin family peptide (zonulin) as molecular factor that triggers the onset of arthritis by regulating intestinal barrier function. Zonulin reduces the expression of intestinal TJ proteins, induces T-cell-mediated mucosal inflammation and controls the transmigration of immune cells from the gut into the joints. Moreover, therapeutic restoration of intestinal barrier function, e.g. by specific targeting of zonulin, not only prevents the transmigration of immune cells from the gut to the joint but also partially protects from the onset of arthritis. These data unravel a mechanism that explains the transition from asymptomatic autoimmunity to inflammatory disease. Since early treatment of autoimmune diseases such as RA leads to better long-term outcomes and higher chances of drug-free remission 15 , the direct targeting of intestinal barrier function at onset of arthritis provides an opportunity to modulate the development of autoimmune diseases, i.e. the onset of inflammation in RA.
Zonulin expression correlates with onset of rheumatoid arthritis
The permeability and barrier function of the intestinal epithelium depends on the regulation of intercellular TJ. Among other factors, intestinal permeability is regulated by zonula occludens toxin (Zot, or zonulin)-mediated disengagement of the protein ZO-1 from the TJ protein complex 16 . Zonulin is secreted by intestinal epithelial cells following stimuli from diet or microbiota. We first assessed serum zonulin levels in two independent cohorts of established RA and found elevated zonulin levels compared to healthy controls (Fig. 1a , Supplementary Fig. 1a ). Patients with chronic infections (hepatitis C infection; HCV) or cancer showed no elevation of zonulin (Fig. 1a ). Furthermore, zonulin was already increased in two independent cohorts in a subset of patients with RA-specific autoimmunity that had not yet developed the disease (“pre-RA”) (Fig. 1b , Supplementary Fig. 1a ). To add more evidence on intestinal barrier dysfunction, we also analyzed serum sCD14 levels, which has been described as marker of microbial translocation. Soluble CD14 levels were correlated with zonulin levels (Supplementary Fig. 1b ), while no such correlation was found between the C-reactive protein (CRP) or the erythrocyte sedimentation rate (ESR) (Supplementary Fig. 1c ). Longitudinal analysis revealed that pre-RA individuals with elevated zonulin levels (>10 ng/ml) had a high risk to develop RA within 1 year (Fig. 1c ).
Fig. 1: Zonulin family peptide levels and intestinal barrier dysfunction in human rheumatoid arthritis.
a Serum zonulin family peptide (zonulin) levels in healthy controls (controls) (n = 41), cancer (n = 19), hepatitis C virus (HCV) (n = 21), and rheumatoid arthritis (RA) (n = 40) patients. b Serum zonulin levels in healthy controls (n = 41) and pre-RA patients (n = 32). c Serum zonulin levels in pre-RA patients with (n = 12) or without (n = 53) later development of RA (left panel); Kaplan–Meier graph showing loss of healthy state and progression to RA patients with low ( 3.2) established RA were analyzed. These procedures were approved by the institutional review board of the Azienda Ospedaliera Universitaria Paolo Giaccone of the University of Palermo and patients provided written informed consents. Patients were balanced for age, sex and body mass index. RA patients fulfilled the ACR-EULAR 2010 criteria (38) and had to be positive for anti-CCP2 antibodies. Also, none of these patients had present or past symptoms of a chronic inflammatory bowel disease or celiac disease. Patients had to be negative for HLA27 and clinical signs of spondyloarthritis or psoriasis. Paired paraffin-embedded tissue and RNA samples were prepared from all patients to allow cross-referencing between histological assessments and qPCR gene expression analysis. Baseline characteristics of the patients are shown in Supplementary Table 1 .
Histology and immunohistochemistry in humans
Paraffin-embedded ileal samples were cut into 4-μm sections and stained with hematoxylin and eosin (H&E). Immunohistochemical analysis for CD3, CD19, CD68, occludin and claudin-1 was performed on 5-µm-thick paraffin-embedded sections from these biopsies and from tonsils (used as positive controls) as previously described 46 . Briefly, following rehydration, antigen was unmasked for 45 min at 95 °C using Dako Target retrieval solution (pH 6; Dako, Carpinteria, CA). Endogenous peroxidase was blocked for 10 min with Dako peroxidase blocking reagent, and nonspecific binding was blocked for 20 min with Dako protein block. The primary antibodies, mouse monoclonal anti-human CD3 (Leica Biosystems, dilution 1:100), CD19 (Dako, dilution 1:50), CD68 (Dako, dilution 1:100), occludin (Santa Cruz Biotechnology, dilution 1:100) and claudin-1 (Santa Cruz Biotechnology, dilution 1:100), were added and incubated for 1 h at room temperature. An isotype-matched irrelevant antibody was used as a negative control. Following three washes with Tris-buffered saline, slides were incubated for 30 min with peroxidase-conjugated Dako EnVision polymer. After three further washes, peroxidase activity was visualized using diaminobenzidine chromogen (Dako), and slides were lightly counterstained with hematoxylin before dehydration and mounting in DePex (VWR International, Oslo, Norway). The number of immune-reactive cells was determined by counting positively stained cells on photomicrographs obtained from three random high-power microscopic fields (×400 magnification) under a Leica DM2000 optical microscope, using a Leica DFC320 digital camera (Leica, Rijswijk, the Netherlands). AR and FC evaluated the ileal samples with no access to clinical data. The intra-rater agreement and the inter-rater agreement calculated by the Cohen’s K coefficient for the two observers were 0.88 and 0.78, respectively.
RT-PCR in humans
Soon after removal, ileal biopsies were stored in RNAlater® solution (Applied Biosystems, Foster City, CA, USA). RT-PCR was performed as previously described 46 . Master mix and Taqman® gene expression assays for glyceraldehyde 3-phosphate dehydrogenase (GAPDH control) and target genes occludin (Hs05465837_g1) and claudin-1(Hs00221623_m1) were obtained from Applied Biosystems (Foster City, CA, USA). Data were quantified using sds 2.1 software and normalized using GAPDH as endogenous control. Relative changes in gene expression between HCs and RA samples were determined using the ΔΔCt method. Levels of the target transcript were normalized to a GAPDH endogenous control, constantly expressed in both groups (ΔCt). For ΔΔCt values, additional subtractions were performed between RA (n = 10) and HCs (n = 10) ΔCt values. Final values were expressed as fold of induction.
RT-PCR in mice
Tissues were stored in RnaLater (Ambion) or directly transferred to TRIzol (Invitrogen). RNA was extracted according to the manufacturer’s instructions. Gene expression results are expressed as arbitrary units relative to expression of the house keeping gene GAPDH. Primer sequences are as follows: GAPDH: 5′-GGG TGT GAA CCA CGA GAA AT-3′ and 5′-CCT TCC ACA ATG CCA AAG TT-3′; ZO-1: 5′-CCA CCT CTG TCC AGC TCT TC-3′ and 5′-CAC CGG AGT GAT GGT TTT CT-3′; occludin: 5′-CCT CCA ATG GCA AAG TGA AT-3′ and 5′-CTC CCC ACC TGT CGT GTA GT-3′; claudin-1: 5′-TCC TTG CTG AAT CTG AAC A-3′ and 5′-AGC CAT CCA CAT CTT CTG-3′; claudin-2: 5′-TAT GTT GGT GCC AGC ATT GT-3′ and 5′-TCA TGC CCA CCA CAG AGA TA-3′; claudin-15: 5′-GCT TCT TCA TGT CAG CCC TG-3′ and 5′-TTC TTG GAG AGA TCC ATG TTG C-3′.
Lactulose to mannitol ratio in humans
The ratio of urinary excretion of lactulose to mannitol (LA/MA) was used to measure the intestinal mucosal permeability in ten RA patients and ten healthy controls who underwent colonoscopy, with higher ratios indicative of increased intestinal permeability as previously described 33 .
All mice were maintained under specific pathogen-free conditions at the Präklinisches Experimentelles Tierzentrum (PETZ), Erlangen, Germany and approved by the local ethics authorities of the Regierung of Unterfranken (#55.2-2532-2-424 and #55.2-2532-2-630). For all experiments, if not otherwise stated, DBA1/J female mice (8 weeks old) were used and purchased from Janvier Labs. Kaede mice were kindly provided by M. Tomura from the RIKEN Institute, Tokyo, Japan. Germ-free (GF) mice were kindly provided by M. Basic from the Hannover Medical School, Hannover, Germany. The animals were kept in the Franz-Penzoldt-Zentrum (FPZ) of the University Hospital Erlangen or the animal laboratory of Med3 under standardized husbandry conditions and hygiene management in accordance with Federation of European Laboratory Animal Science Associations (FELASA). The animals received water and feed ad libitum. The keeping rooms had a temperature of 22–23 °C and a humidity of 50–60%. There was also a 12-h light–dark rhythm in the holding rooms. Animals were kept in type II long cages, with maximum five animals.
For all experiments, mice were acclimated for 1 week, followed by a 2-week co-housing period before the experiments started. Supplementation of butyrate (all Sigma-Aldrich, Germany) was done in the drinking water at a final concentration of 150 mM and changed every 3 days and control mice received pH and sodium-matched water. Zonulin antagonist (known as Larazotide acetate, or AT-1001, or INN-202) was purchased from BOC Sciences, NY, USA or later synthesized by GENAXXON biosciences, Ulm, Germany and effectiveness was compared in one side-by-side experiment using both purchased zonulin antagonist sources (BOC Sciences, NY, USA vs. GENAXXON biosciences) to confirm their biological function. Mice received 0.15 mg/ml zonulin antagonist in the drinking water and changed every day for 10 consecutive days between 17 and 27 dpi in the CIA model. For CAIA mouse model, C57BL/6 mice received 50 µg/mouse i.v. zonulin antagonist or vehicle (3% DMSO in phosphate‐buffered saline (PBS)) 1 day after CAIA induction. CB1 receptor agonist (ACEA; arachidonyl-2-chloroethylamide, Sigma-Aldrich, Germany) treatment was i.p. injections daily with 250 µg/mouse between 17 and 27 days post CII immunization. CIA was induced in 8-week-old female Kaede (C57BL/6J) or wild-type DBA/1J mice by subcutaneous injection at the base of the tail with 100 μl with 0.25 mg chicken type II collagen (Chondrex, Redmond, WA) in complete Freund adjuvant (Difco Laboratory, Detroit, MI), containing 5 mg/ml killed Mycobacterium tuberculosis (H37Ra). Mice were challenged after 21 days by intradermal immunization in the base of the tail with this emulsion. The paws were evaluated for joint swelling and grip strength three times per week. Each paw was individually scored using a 4-point scale: 0, normal paw; 1, minimal swelling or redness; 2, redness and swelling involving the entire forepaw; 3, redness and swelling involving the entire limp; 4, joint deformity or ankylosis or both. In addition, grip strength of each paw was analyzed on a wire 3 mm in diameter, using a score from 0 to −4 (0, normal grip strength; −1, mildly reduced grip strength; −2, moderately reduced grip strength; −3, severely reduced grip strength; −4, no grip strength at all). For photoconversion of Kaede-transgenic mice, the small intestine of anesthetized Kaede-transgenic mice was subjected to lighting using a BlueWave LED Prime UVA (Dymax).
In vivo intestinal permeability measurements
Once every week, mice were housed in metabolic cages after a 4 h fasting of food and water and immediately after a gavage of 0.2 ml of a 10 ml sugar probe containing 100 mg of sucrose, 12 mg of lactulose, 8 mg of mannitol and 6 mg of sucralose. After the collection of urine the animals were placed in their respective cages, and provided with food and water. For the FITC-Dextran measurements, after 4 h fasting of food and water, mice were immediately orally gavaged with 200 μl of FITC-dextran (FD4) (440 mg/kg body weight), and blood was collected 4 h later. The concentration of the FITC-dextran was determined using a fluorimeter with an excitation wavelength at 490 nm and an emission wavelength of 530 nm. Serially diluted FITC-dextran in serum was.used to establish a standard curve.
For HPLC analysis of lactulose and mannitol, urine samples were thawed, mixed thoroughly and centrifuged for 5 min at 20,000 × g and 4 °C. Fifty microliters of supernatant was mixed with 175 µl of HPLC-grade water and 25 µl of 20% sulfosalicylic acid fresh solution, followed by incubation for 15 min at room temperature. After storage at −20 °C overnight, diluted samples were thawed again, centrifuged for 10 min at 20,000 × g and 4 °C. An amount of 200 mg ion-exchange resins (TMD-8 hydrogen and hydroxide form, Sigma-Aldrich Chemie GmbH, Taufkirch, Germany) was added, mixed well and incubated for 30 min at room temperature. After centrifugation for 10 min at 20,000 × g and 4 °C, supernatants were subjected for HPLC analysis using a DIONEX UltiMate® 3000 system with Corona Veo SD Charged-Aerosol-Detector (Thermo Scientific GmbH, Dreieich, Germany). For chromatographic separation, a hydrophilic interaction liquid chromatography (HILIC) column (Acclaim™ Trinity™ P2, 3 µm, 3.0 × 100 mm, Thermo Scientific) was used. The eluent was a mixture of acetonitrile and 100 mM ammoniumformate (pH 3.65) with an isocratic ratio of 82:18 [vol/vol] at 0.6 ml/min and 60 °C column oven temperature. For determination of sucralose, urine samples were thawed, mixed thoroughly and centrifuged for 10 min at 20,000 × g and 4 °C. Fifty microliters of supernatant was mixed with 200 µl of HPLC-grade water, mixed thoroughly and centrifuged again for 10 min at 20,000 × g and 4 °C. Supernatants were analyzed by using a DIONEX UltiMate® 3000 system with Corona Veo SD Charged-Aerosol-Detector (Thermo Scientific GmbH, Dreieich, Germany). For chromatographic separation a Synergi 4µ Polar-RP 80A column was used (250 ×1.6 mm (Phenomenex, Aschaffenburg, Germany)). The eluent was a mixture of methanol and HPLC-grade water with an isocratic ratio of 30:70 [vol/vol] at 1.0 ml/min and 50 °C column oven temperature.
Histological grading of intestinal inflammation in mice
Upon sacrifice, proximal duodenal, jejunal, ileal and colon tissue were harvested and fixed in 10% phosphate-buffered formalin. These samples were embedded in paraffin, sectioned at 4 μm, and stained with H&E or PAS for light microscopy examination. The slides were reviewed in a blinded fashion by a pathologist and were assigned a histological score for intestinal inflammation.
Tibial bones and inflamed paws with tarsal joints were fixed in 4% formalin for 24 h and decalcified in Ethylenediaminetetraacetic acid (EDTA) (Sigma-Aldrich). Serial paraffin sections (2 μm) were stained for tartrate-resistant acid phosphatase (TRAP) using a Leukocyte Acid Phosphatase Kit (Sigma) according to the manufacturer’s instructions. Osteoclast numbers were quantified using a microscope (Carl Zeiss) equipped with a digital camera and an image analysis system for performing histomorphometry (Osteomeasure; OsteoMetrics).
Paraffin-embedded tissue sections were deparaffinized with xylene and rehydrated through graded alcohol solutions. Heat-induced epitope retrieval was performed by placing the sections in Retrieval Solution pH 9, diluted 1:20 (cat. EB-DEPP-9; eubio) at 85 °C for 20 min, washed three times with PBS and blocked with PBS containing 0.2% bovine serum albumin (BSA) and 0.1% saponin for 1 h at room temperature. The sections were then incubated with primary antibodies against occludin (Abcam, dilution 1:100) and ZO-1 (ThermoFisher, dilution 1:100), diluted in PBS containing 0.2% BSA and 0.1% saponin at 4 °C overnight, washed, and incubated for 1 h with Alexa Fluor 647 donkey anti-rabbit IgG secondary antibody (Invitrogen, dilution 1:400). Some sections were incubated with PBS containing 0.2% BSA and 0.1% saponin alone in place of primary antibody as a negative control. After having been washed three times with PBS, the sections were mounted and visualized. Imaging was performed by using the Leica DMi8 TIRF Widefield Fluorescence of ×10 objective lens (Leica, Germany) and Zeiss Spinning Disc Axio Observer Z1 of ×63 objective lens (Carl Zeiss) microscopes.
To compare ZO-1 and occludin levels, we considered three samples for each of the consecutive time points (day 0, 25, 30, 35, 40, 45 and 50) in ileum and colon for each of the proteins with three images from different parts of each section. Microscopy was performed and fluorescence images were visualized and captured in DAPI and Cy5 channels. Quantification of the images was performed using ImageJ software based on the mean fluorescence intensity. We calculated for each of the samples and condition the mean intensity across the entire image and divided it by the mean intensity of the DAPI channel to account for varying tissue sizes.
Fecal supernatants preparation
Intestinal stool samples were obtained from healthy and CIA mice, at different time points of disease. Samples were suspended at 10% (wt/vol) in Dulbecco’s PBS, homogenized, and centrifuged at 4500 rpm for 10 min at 4 °C. Supernatants were passed through 1.2 µm membrane filters.
Caco-2 cells, a human colorectal carcinoma-derived epithelial cell line, were cultivated in Dulbecco’s modified Eagle’s medium (DMEM)/F12 (Gibco) containing 10% fetal bovine serum (FBS) in a humidified incubator at 37 °C, 5% CO2. Cells were maintained for 14 days post confluency, serum-starved overnight and stimulated with FSN for 24 h or PBS as a control.
Transepithelial resistance measurements
Caco-2 cells were seeded onto 12 mm polycarbonate membrane inserts (PCF membrane inserts, Millipore, Germany) and bathed in DMEM/F12 medium which contained phenol red, 10% FBS, penicillin (100 units/ml) and streptomycin (10 µg/ml). Cells were continuously maintained in a humidified incubator at 37 °C, 5% CO2. To assess that these cells form monolayer on the membrane inserts, transepithelial resistance was routinely monitored using an epithelial volt-ohm-meter including STX-2 electrodes (EVOM, World Precision Instruments, Berlin, Germany). After 12–16 days post seeding, membrane inserts containing confluent monolayer of Caco-2 cells were transferred into Ussing chambers 47 , 48 . The basolateral bath contained 1.0 ml and the apical bath 0.5 ml of the supplemented DMEM/F12 medium. The transepithelial resistance was recorded every 10 s using a CVC6 clamp device (Fiebig, Berlin Germany) 49 by measuring voltage deflections induced by injection of 20 µA symmetrical square pulses for 0.4 s. After an equilibration period of 10–40 min, AT-1002 (5 mg/ml) was applied to the apical bath by replacing 400 µl of the bath medium with 400 µl of medium containing the AT-1002 component. In time matched controls the dissolvent of AT-1002 was added to the apical bath. The transepithelial resistance was then continuously recorded for a further 3 h.
Small intestines (all data shown for the ileum) and colons were analyzed by flow cytometry. In short, tissues were incubated with PBS containing EDTA at 37 °C for 2 × 20 min, then digested twice for 35 min with a cocktail of collagenase D (1 mg/ml, Roche), dispase I (0.1 mg/ml, Roche), and DNase I (333 mg/ml, Sigma), smashed and filtered through a 40 mm gauze (BD Biosciences). Single-cell suspensions were then stimulated with phorbol 12-myristate 13 acetate and ionomycin (Sigma-Aldrich) for 6 h; monensin (BD-Bioscience) and brefeldin A (BioLegend) were added after 2 h or directly stained with the respective directly labeled FACS antibodies. All intracellular stainings were used in combination with a FoxP3staining kit (eBioscience). Flow cytometric analysis was performed on a Gallios Flow Cytometer (Beckman Coulter) or CytoFLEX (Beckmann Coulter) and evaluated using the Kaluza and cytExpert Flow Cytometry analysis software (Beckman Coulter). Antibodies used: Alexa Fluor® 647 anti-mouse CD3 antibody (BioLegend, 1:600); PE/Cy7 anti-mouse CD4 antibody (BioLegend, 1:400); Alexa Fluor® 700 anti-mouse CD8a antibody (BioLegend, 1:600); PE Mouse anti-Mouse RORγt (BD Pharmingen, 1:100); Pacific Blue™ anti-T-bet Antibody (BioLegend, 1:100); Alexa Fluor® 647 anti-mouse FOXP3 Antibody (BioLegend, 1:100); PE anti-GATA3 Antibody (BioLegend, 1:100); Brilliant Violet 421™ anti-mouse CD3ε Antibody (BioLegend, 1:200); APC/Cy7 anti-mouse/human CD45R/B220 Antibody (BioLegend, 1:200); Alexa Fluor 700 anti-mouse/human CD11b Antibody (BioLegend, 1:200); CD11c Monoclonal Antibody (N418), APC (eBioscience™, 1:400); APC anti-mouse IFN-Antibody (BioLegend, 1:50); BV421 Rat Anti-Mouse IL-4 (BD Biosciences, 1:200); PE anti-mouse IL17A Antibody (BioLegend, 1:200).
Intestinal organoid isolation and cultivation
The small intestine of wild-type C57Bl/6J was removed and washed with ice-cold PBS. It was opened longitudinally and villi were scraped out using a coverslip and discarded. The intestine was cut into 2–3 mm pieces and collected in ice-cold PBS. After sedimentation the supernatant was discarded and fresh ice-cold PBS was added to resuspend the intestinal pieces. This procedure was repeated 15–20 times until the supernatant was clear. The intestinal pieces were then incubated at 4 °C for 20–30 min in ice-cold PBS containing 2 mM EDTA (Merck) while agitating. Afterwards the tissue was sedimented by gravity and the supernatant was replaced by ice-cold PBS. The intestinal crypts were dissociated from the tissue by vigorous shaking. The supernatant was collected and screened for crypts by microscopy. This step was repeated until no crypts were left in the supernatant. The supernatant fractions enriched with crypts were passed through a 70 µm cell strainer and centrifuged at 300 × g for 5 min at 4 °C. The pelleted crypts were resuspended in 25 µl/well growth factor reduced Matrigel (Corning, New York, USA). The cell suspension was then distributed into a 48-well cell culture plate. The Matrigel was allowed to polymerize in the incubator at 37 °C and 5% CO2 for 20 min. Afterwards 300 µl of Intestinal Organoid Growth medium (mouse) (Stemcell, Köln, Germany) was added per well. Organoids were cultured in the incubator for at least 7 days before any experiments were performed. Medium was changed every 2–3 days and organoids were splitted once per week 50 .
For imaging organoids were plated in eight-well chamber slides (Ibidi, Planegg, Germany). The organoids were imaged using a Zeiss Spinning Disc confocal microscope (Zeiss, Jena, Germany). At the beginning of the experiment 1 mM Lucifer yellow (VWR International GmbH, Darmstadt, Germany) was added to each of the chambers. Afterwards the organoids were imaged for 100 min with intervals of 5 min. To ensure the capability of the organoids to get permeable, 1 mM of EGTA (Merck) was added after the observation time. Following EGTA addition organoids were imaged for another 20 min. For each time point lucifer yellow fluorescence inside and outside the organoids was quantified using Fiji ImageJ. Only organoids that were able to take up Lucifer yellow after EGTA treatment were analyzed 51 .
Caco-2 cells were washed twice with PBS and subsequently lysed. Intestinal tissue samples were homogenized in protein lysis buffer and subsequently sonicated. Protein extracts were separated on 8% SDS-polyacrylamide or 4–12% bis−tris protein gel (Invitrogen), transferred on a nitrocellulose membrane, and stained with antibodies against ZO-1 (N1N2, GeneTex, dilution 1:3000). An antibody against β-actin (AC-15, abcam, dilution 1:25,000) was used as loading control. Supplementary Fig. 10 shows the uncropped gel scans of Western blots.
Five replicates of frozen cecal samples (100 mg) were weighed into a 2 ml polypropylene tube. The tubes were kept in a cool rack throughout the extraction. 33% HCl (50 µl for cecal contents or 5 µl for serum) was added and samples were vortexed for 1 min. One milliliter of diethyl ether was added, vortexed for 1 min and centrifuged for 3 min at 4 °C. The organic phase was transferred into a 2 ml gas chromatography (GC) vial. For the calibration curve, 100 μl of SCFA calibration standards (Sigma) were dissolved in water to concentrations of 0, 0.5, 1, 5 and 10 mM and then subjected to the same extraction procedure as the samples. For GCMS analysis 1 μl of the sample (4–5 replicates) was injected with a split ratio of 20:1 on a Famewax, 30 m x 0.25 mm iD, 0.25 μm df capillary column (Restek, Bad Homburg). The GC-MS system consisted of GCMS QP2010Plus gas chromatograph/mass spectrometer coupled with an AOC20S autosampler and an AOC20i auto injector (Shimadzu, Kyoto, Japan). Injection temperature was 240 °C with the interface set at 230 °C and the ion source at 200 °C. Helium was used as carrier gas with constant flow rate of 1 ml/min. The column temperature program started with 40 °C and was ramped to 150 °C at a rate of 7 °C/min and then to 230 °C at a rate of 9 °C/min, and finally held at 230 °C for 9 min. The total run time was 40 min. SCFA were identified based on the retention time of standard compounds and with the assistance of the NIST 08 mass spectral library. Full scan mass spectra were recorded in the 25–150 m/z range (0.5 s/scan). Quantification was done by integration of the extracted ion chromatogram peaks for the following ion species: m/z 45 for acetate eluted at 7.8 min, m/z 74 for propionate eluted at 9.6 min, and m/z 60 for butyrate eluted at 11.5 min. GCMS solution software was used for data processing.
µCT imaging was performed using the cone-beam Desktop Micro Computer Tomograph “µCT 40” by SCANCO Medical AG, Bruettisellen, Switzerland. The settings were optimized for calcified tissue visualization at 55 kVp with a current of 145 µA and 200 ms integration time for 500 projections/180°. For the segmentation of 3D-volumes, an isotropic voxel size of 8.4 µm and an evaluation script with adjusted grayscale thresholds of the operating system “Open VMS” by SCANCO Medical was used. Volume of interest tibia: The analysis of the bone structure was performed in the proximal metaphysis of the tibia, starting 0.43 mm from an anatomic landmark in the growth plate and extending 1.720 mm (200 tomograms) distally.
Specific IgG detection via ELISA
Collagen type II (CII)-specific antibodies were detected by ELISA using high-binding plates (Nunc) coated overnight with 10 μg/ml mouse or chicken CII. The detection of mouse CII-specific antibodies was carried out by Eu3+-labeled anti-mouse IgG antibody and the DELFIA system (PerkinElmer) according to the manufacturer’s recommendations.
16S rRNA microbiome analysis
For Supplementary Fig. 4a–d , genomic DNA was extracted according to Qiamp Fast DNA Stool extraction kit (Qiagen) and the V3–V4 region of the bacterial 16S rRNA gene was amplified with the NEBNext Q5 Hot Start Hifi PCR Master Mix (NEB) using a dual-index strategy. Amplified fragments were purified with AMPure XP Beads (Beckmann Coulter Genomics), combined and analyzed by “2 × 300 bp paired-end” sequencing using an Illumina MiSeq device. Quality control, OTU table generation and bioinformatics analysis was done using the Usearch10 (Edgar RC Flyvbjerg H, 2015) and the Microbiomeanalyst package 52 . The database of the “Ribosomal” database project (RDP release 16) was used for classification. For Supplementary Fig. 4e , microbial DNA extraction from fecal content was done using ZymoBIOMICS 96 Magbead DNA Kit automated on a Tecan Fluent 480 liquid handling system. Amplification of the V4 region (F515/R806) of the 16S rRNA gene was performed as in previously described protocols 53 . Samples were sequenced on an Illumina MiSeq platform (PE250). Barcode-based demultiplexing was performed using IDEMP software with default parameters ( https://github.com/yhwu/idemp ). Obtained reads were assembled, quality controlled, and clustered using Usearch8.1 software package ( http://www.drive5.com/usearch/ ). Briefly, reads were merged using -fastq_mergepairs –with fastq_maxdiffs 30 and quality filtering was done with fastq_filter (-fastq_maxee 1), minimum read length 200 bp. The OTU clusters and representative sequences were determined using the UPARSE algorithm 54 , followed by taxonomy assignment using the Silva database v128 55 and the RDP Classifier 56 with a bootstrap confidence cutoff of 80% performed by using QIIME v1.8.0 57 . OTU absolute abundance table and mapping file were used for statistical analyses and data visualization in the R statistical programming environment package phyloseq 58 or web-based tool MicrobiomeAnalyst 52 .
Data are expressed as mean ± s.d. unless otherwise indicated in the figure legends. Analysis was performed using Student’s t test, single comparison, or analysis of variance (ANOVA) test for multiple comparisons (one-way or two-way ANOVA followed by Tukey’s or Bonferroni’s multiple comparisons test, respectively). All experiments were conducted at least two times. P values of 0.05 were considered significant and are shown as *p < 0.05, **p < 0.01, or ***p < 0.001. Graph generation and statistical analyses were performed using the Prism version 8 software (GraphPad, La Jolla, CA).
All relevant data are available from the authors upon reasonable request. The source data underlying Figs. 1 – 6 and Supplementary Figs. 1 – 10 are provided as a Source data file. Sequence data are deposited in BioProject with the accession code PRJNA592061 .
Endesfelder, D., Engel, M. & Zu Castell, W. Gut immunity and type 1 diabetes: a melange of microbes, diet, and host interactions? Curr. Diabetes Rep. 16, 60 (2016).
We thank all members of our laboratories at the Medical clinic 3 for their support and helpful discussions. We thank Dr. Michio Tomura from the RIKEN Institute, Yokohama City, Japan for providing the Kaede mice. We are grateful to Prof. Uwe Sonnewald for support with the GC-MS analysis. We thank Dr. Ralf Palmisano and Dr. Philipp Tripal and Dr. Benjamin Schmid from the OICE in Erlangen for the technical advice and Daniela Weidner for µCT measurements and analysis. We are thankful to Margarethe Schimpf and Viktoria Liman for technical support and to the Med3 team of study ambulance in Erlangen for access to human samples. This study was very kindly funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) DFG-FOR2886 PANDORA Project-No.1, DFG-CRC1181-Project-No.B07. Additional funding was received by the STAEDTLER® foundation, the Johannes und Frieda Marohn-Stiftung, the Else Kröner-Fresenius Foundation, the Interdisciplinary Centre for Clinical Research, Erlangen (IZKF) and the Bundesministerium für Bildung und Forschung (BMBF) BMBF-MASCARA TP No.4, the TEAM project of the European Union and the IMI-funded project EU IMI2-RTCure.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .
About this article
Cite this article
Tajik, N., Frech, M., Schulz, O. et al. Targeting zonulin and intestinal epithelial barrier function to prevent onset of arthritis. Nat Commun 11, 1995 (2020). https://doi.org/10.1038/s41467-020-15831-7