SUPPLEMENTARY MATERIAL for Manuscript

SUPPLEMENTARY MATERIAL for manuscript

Antibiotic-driven dysbiosis mediates intraluminal agglutination and alternative segregation of Enterococcus faecium from the intestinal epithelium

Antoni P. A. Hendrickx1,3, Janetta Top1, Jumamurat R. Bayjanov1, Hans Kemperman2, Malbert R. C. Rogers1, Fernanda L. Paganelli1, Marc J.M. Bonten1 and Rob J.L. Willems1

1Department of Medical Microbiology and 2Clinical Chemistry and Haematology, University Medical Center Utrecht, Utrecht, The Netherlands. 3To whom correspondence should be sent: Dr. Antoni P.A. Hendrickx, Heidelberglaan 100, 3584CX Utrecht, Room G04.614, The Netherlands, Phone: +31887557627, E-mail:

Supplementary Methods

Bacterial strains. A commensal (E980) and a clinical hospital associated (E1162) E. faecium isolate were used in this study. Bacterial strains were grown aerobically at 37°C on Trypticase soy agar II (TSA) plates supplemented with 5% sheep blood. For selective enumeration of E. faecium E980 and E1162 from mouse faeces and/or intestines Slanetz-Bartley (Tritium Microbiologie B.V., Netherlands) agar plates were used. The MICs of ceftriaxone and cefoxitin for E980 and E1162 are 8 and >512 µg/ml and 32 and 256 µg/ml, respectively.

Mice. Thirtyspecific pathogen-free 10 week old female wild-type Balb/c mice were purchased from Charles River Laboratories (USA). All mice were housed in the animal facility of the Academic Medical Center (Amsterdam, The Netherlands) in rooms with a controlled temperature and a 12-hour light-dark cycle. Animals were acclimated for 1 week prior the experiment and received standard rodent chow and water ad libitum.

Mouse intestinal colonization model. Intestinal colonization by E980 and E1162 was performed as previously described(1, 2). Eight mice were two days prior inoculation of bacteria (on day -2, -1 and 0) injected subcutaneously with ceftriaxone (Roche, The Netherlands; 100 µl per injection, 12 mg/ml; 5 injections total) two times daily and one time at the day of inoculation (on day 0) and left on ad libitum cefoxitin (0.125 g/l) supplemented sterile drinking water. In addition, 8 mice received 0.9% NaCl injections two days prior inoculation of bacteria (on day -2, -1 and 0; 5 injections total) and were left on normal sterile drinking water without antibiotics. Of the 8 antibiotics treated or 8 0.9% NaCl treated mice, 4 mice were inoculated on day 0 with E980 and 4 mice with E1162. The inoculum of 2 × 109CFU of E980 or E1162was prepared as described previously(2, 3). As a control, 3 mice were treated with ceftriaxone/cefoxitin antibiotics alone as described above, but did not receive E. faecium. As another control group, 3 mice were left untreated. Collection of faecal samples (2-3 faecal pellets/ml 0.9% NaCl) was done on days -2, 0, 1, 3, 6 and 10. Determination of E. faecium outgrowth was performed by enumeration of CFU´s from Slanetz-Bartley agar. After 10 days of colonization mice were anesthetized with ketamine (75 mg/kg) and medetomidine (1 mg/kg) followed by cervical dyslocation. Small intestines, cecum and colon were removed by necropsy, weighed, photographed, sectioned in parts and fixed in 2% glutaraldehyde for SEM and 4% formalin for histopathology. In addition, thismouse intestinal colonization experiment was repeated for E. faecium E1162 to confirm the reduction of the microbiota layer, mucus layer and colon wall. The experimental set-up was identical as above with 2 animals/group for untreated mice, untreated mice inoculated with E. faecium E1162, mice treated with antibiotics only, mice treated with antibiotics and inoculated with E. faecium E1162 (thus, 8 additional mice). These animals were sacrificed on day 1 and intestines were fixed in Carnoy's fixative.In total, in the two experiments, 30 mice were included, 4 mice were treated with antibiotics and inoculated with E. faecium E980 and 6 with E. faecium E1162. In addition, 4 mice were untreated (injected with 0.9% NaCl) and inoculated with E. faecium E980 and 6 with E. faecium E1162.Five mice were treated with antibiotics only and 5 mice were left untreated (injected with 0.9% NaCl).

Ethics statement. The Animal Care and Use Committee (DEC) of the University of Amsterdam, The Netherlands reviewed and approved the mouse intestinal colonization experiment (number 102902). The experiment was conducted following the Dutch law on 'Experiments on Animals' (Wet op de Dierproeven, WOD), in which the European Union guideline 2010/63/EU is implemented per 18/12/2014.

Antibodies and recombinant proteins. Polyclonal Goat anti-rabbit IgG antibody directed against mouse E-cadherin, Polyclonal Goat-anti human E-cadherin IgG antibody and Rabbit anti-mouse pIgR IgG were purchased from R&D Systems (France). Goat anti-mouse IgA and Donkey anti-Goat IgG-HRP were purchased from Southern Biotech (USA). Rabbit polyclonal anti-Muc-2 IgG antibody was purchased from Abcam/ITK Diagnostics (The Netherlands). Alexafluor 488 Goat-anti Rabbit IgG (H+L), Alexafluor 488 donkey anti-Goat IgG (H+L) and TO-PRO-3 iodide 642/661 were all purchased from Molecular Probes/Life technologies (The Netherlands). For immuno fluoresence experiments primary antibodies were used at a 1:100 dilution and the secondary antibodies at 1:500 dilution, while for Western blots dilutions were 1:2500 and 1:5000, respectively. Recombinant mouse E-cadherin (CDH1-Met 1- Val 709) and recombinant mouse pIgR (Met 1- Lys 645) were from Sino Biological/Life technologies (China) and human natural IgA protein was from Abcam. Bovine serum albumin (BSA) was purchased from Serva. All antibodies, conjugates and proteins were reconstituted by the manufacturer's instructions if necessary.

16S rRNA gene sequencing. Multiplexed 16S rRNA gene sequencing of DNA of bacteria isolated from faeces of mice obtained from day -2, 0, 1 and 10 during the course of the experiment was performed using the Illumina MiSeq platform method as described by Fadrosh et al. using an improved dual indexing approach(4). First, DNA was extracted from fecal pellets (0.01-0.1 gram) as described by Godon et al. with minor modifications(5). Briefly, the homogenized faecal pellets used for E. faecium enumeration were thawed on ice and pelleted by centrifugation for 15 min. at 14,000 x g. The pellet was resuspended in 250 µl 4M guanidine thiocyanate–0.1 M Tris (pH 7.5) and 40 µl of 10% N-lauroyl sarcosine as described previously(5). After DNA isolation, 5 µl of the sample was loaded on an agarose gel to check the DNA and 2 µl were used in a PCR using E. faecium specific primers based on the housekeeping gene ddl (D-Ala-D-Ala ligase) as described previously(6). DNA concentrations were determined using the Qubit (Life Technologies) according to manufacturer’s instructions and diluted to 5 ng/µl for 16S rRNA gene PCR. PCR products were only obtained after a 1:10 dilution of the 5 ng/µl diluted sample. PCR and normalization was performed as described and samples were pooled(4). After pooling the sample was purified using Agencourt AMPure XP beads (Beckman Coulter Genomics), i.e. beads were added to the sample in a 1:0.9 ratio, mixed and incubated for 5 min at RT. Beads were pelleted using a magnetic stand and the DNA was eluted in a total volume of 52.5 µl. The concentration was determined using the Qubit. The sample and PhiX Control (5%) were prepared according to the Illumina protocol (15039740-d) for v3 chemistry and low diversity libraries and were sequenced on 300PE MiSeq run.

16S rRNA sequence data analysis.Untrimmed paired-end reads were assembled using the FLASH assembler, which performs error correction during the assembly process(7). Pooled sequence data were de-multiplexed using split_libraries_fastq.py script of Qiime microbial community analysis pipeline (version 1.8.0)(8). There were 8 samples for which no or only a few sequences could be assigned, which were not used in further analysis. For remaining samples, a median number of sequences per sample was 61103. After removal of the barcodes, heterogeneity spacers, and primer sequences there were nearly 4.9 million sequences with a median length of 424 bases. Sequences with a minimum of 97% similarity were assigned into operational taxonomic units (OTUs) using Qiime's open-reference OTU calling workflow. This workflow was used with the “-m usearch61” option, which uses the USEARCH algorithm for OTU picking and UCHIME for chimeric sequence detection(9, 10). Taxonomic ranks for OTUs were assigned using the Greengenes database (version 13.8) with the default parameters of the script pick_open_reference_otus.py(11). Representative OTU sequences were aligned to the Greengenes core reference database using the PyNAST aligner (version 1.2.2)(12, 13). Highly variable parts of alignments were removed using the filter_alignment.py script, which is part of the pick_open_reference_otus.py workflow. Subsequently, filtered alignment results were used to create an approximate maximum-likelihood phylogenetic tree using FastTree (version 2.1.3)(14). For more accurate taxa diversity distribution, OTUs with a number of sequences less than 0.005% of total number of sequences were discarded using the filter_otus_from_otu_table.py script with the parameter “--min_count_fraction 0.00005”(15). The filtered OTU table and generated phylogenetic tree were used to assess within-sample (alpha) and between sample (beta) diversities using Qiime's core_diversity_analyses.py workflow. For rarefaction, the subsampling depth threshold of 4605 was used. The UniFrac distance was used to calculate beta-diversity of samples(16). In addition to alpha- and beta-diversity analysis and visualizations, this workflow also incorporates principal coordinates analysis and visualization of sample compositions using Emperor(17).

Histopathology. Overnight formalin fixed or Carnoy's fixed cecum and colon tissue was embedded in paraffin and sectioned and sectionswere prepared for immune fluorescence as described elsewhere(18). Cecum and colon sections were either stained with Hematoxylin and Eosin (H&E), Gram or Periodic acid-Schiff (PAS) stain according to standard procedures of the Department of Pathology of the University Medical Center Utrecht. Sections were analyzed by light microscopy after training and instruction by a pathologist. From each of the 6 groups of the mouse experiment, H&E-, Gram and PAS-stained thin sections of at least 2 animals were analyzed.Microscopy data was confirmed by repeating the animal experiment (except for E. faecium E980) and analysis of an additional 2 animals/group (thus 4 mice/group) of which the intestines were fixed in Carnoy's fixative. Integrity of the epithelium of the cecum and colon was determined from at least 50 randomly chosen sites. Microbiota diameters were measured in Gram-stained thin sections as indicated in supplemental figure 2. Mucus-layer diameters were measured in PAS-stained thin sections and colon barrier (comprising the crypts of Lieberkühn and underlying muscularis externae and serosa) diameters from H&E stained thin sections. Diameters were randomly measured along the microbiota layer/mucus layer/colon wall using ImageJ and plotted using Graphpad Prism 6.0 software and relevant groups were compared.

Immunohistochemistry. Cecum and colon sectionswere prepared for immune fluorescence as described elsewhere(18).In brief, 4 µm cecum and/or colon thin sections on glass slides were deparaffinized using 100% xylene and rehydrated by consecutive incubations in 100% ethanol, 90% ethanol, 70% ethanol and water. Antigen was retrieved by boiling slides at 95°C for 20 minutes in 0.01 M Na-citrate buffer, pH = 6) and a cooling down in the same buffer for 20 minutes at room temperature, followed by rinsing with water. Immuno staining was performed by incubating slides with 150 µl of 1:100 primary Goat anti-rabbit polyclonal IgG antibodies directed against mouse E-cadherin/pIgR/IgA for 1 hr in Hank's Balanced salt solution containing 1% bovine serum albumin (HBSS-BSA). After incubation, slides were washed with 50 ml of HBSS-BSA and subsequently incubated for 1 hr in 200µl 1:100 diluted Donkey anti-Goat IgG-Alexa Fluor 488 and 1:1000 diluted TOPRO3 nuclei stain in the dark, followed by another 50 ml wash with HBSS-BSA. Finally, bacteria were stained with 1µg/ml FM5-95 in H2O for 30 seconds and glass coverslips mounted using Prolong Gold anti-fade reagent (Invitrogen). Tissue sections of untreated mice (n = 4), untreated mice inoculated with E. faecium E1162 (n = 4) or E. faecium E980 (n = 2), mice treated with antibiotics only (n = 4), mice treated with antibiotics and inoculated with E. faecium E1162 (n = 4), mice treated with antibiotics and inoculated with E. faecium E980 (n = 2) were analyzed. Negative control sections for all antibodies were processed in an identical manner after omitting the primary antibody and showed no staining. Slides were analyzed by confocal microscopy as described below. Labeled samples were stored at 4°C to allow reanalysis.

HT-29 cell culture and adherence assay. Human colorectal adenocarcinoma HT-29 cells (obtained from the collection of Dr. S. van Mil, Wilhelmina Children's Hospital Utrecht) were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS) and 1% penicillin/streptomycin at 37°C in a 5% CO2/air atmosphere. Monolayers grown to 100% confluence were maintained for 2 weeks and culture medium was refreshed daily. For E. faecium recovering assays, HT-29 monolayers were washed twice with DMEM + FCS and incubated for 1 hr with DMEM + FCS. Subsequently, monolayers were washed twice with DMEM and incubated for another 1 hr in DMEM. E. faecium E1162 was added to the monolayers (MOI of 50 bacteria per cell) with or without varying concentrations of (0, 5, 10, 20 mM) of CaCl2, MgCl2, or KCl in a final volume of 1 mL DMEM. Monolayers with bacteria were spun down for 5 minutes at 1500 rpm to allow E. faecium to bind and were incubated for 1 hr at 37°C in a 5% CO2/air atmosphere. Media and unbound bacteria were removed and monolayers washed 3x with phosphate buffered saline (PBS), prior trypsinizing and homogenizing the HT-29 cells. Bound E. faecium was enumerated by plating CFU's from serial dilutions onto TSA agar. Recovering of E. faecium from HT-29 cells was plotted as % recovered (CFU recovered/CFU inoculum x 100%). Experiments were repeated independently for at least 3 times.

HT-29 protein extraction. To analyze E-cadherin cleavage in the presence of cations, HT-29 cells were grown to 100% confluence and monolayers were maintained for 2 weeks. Monolayers were washed as described above and incubated with varying concentrations of (0, 5, 10, 20 mM) of CaCl2, MgCl2, or KCl in a final volume of 1 mL DMEM for 1 hr at 37°C in a 5% CO2/air atmosphere. After 1 hr incubation, monolayers were washed and HT-29 cells lysed in lysis buffer (1% Triton X-100 in PBS supplemented with Complete mini protease inhibitor cocktail [Roche]) for 10 minutes at 4°C. Cell debris was removed by centrifugation for 1 min at 14,000 rpm and cell lysate was boiled for 5 minutes in sample buffer and analyzed by Western blotting as described below.

Cation determination assays. Faecal extracts of the antibiotics treated and E980 or E1162 inoculated groups and untreated control mice obtained during the course of the experiment were spun down for 10 minutes at 14,000 rpm, supernatant was transferred to a new tube and spun down again to remove debris and bacteria. Calcium, Magnesium and Potassium were measured on an AU 5811 routine chemistry analyzer (Beckman Coulter, Brea, USA). Calcium and Magnesium concentrations were determined with colorimetric assays and Potassium with an ion-selective electrode. Data obtained in mmol/l were corrected for faecal weight and plotted in mM using Graphpad Prism 6.0 software and relevant groups were compared.

Immuno fluorescence and confocal microscopy. From each of the 6 groups of the mouse experiment, cecum contents of 2 animals were analyzed. Mouse cecum contents was harvested by puncturing the cecum and contents was post fixed for 15 min onto poly-L-lysine covered glass slides and air dried. HT-29 monolayers grown on poly-l-lysine covered glass slides were incubated with varying concentrations of cations and E1162 as described above and fixed in formalin. Samples were washed with 15 mL PBS and incubated with primary Goat anti-Rabbit IgG antibody (1:100 dilution) in HBSS for 30 min followed by another 15 mL PBS wash. Subsequently, samples were incubated for 30 min with 1:500 diluted Goat anti-rabbit Alexa fluor 488 conjugate in the dark. Bacteria were stained with 1µg/ml FM5-95 in H2O for 30 seconds. After incubation, the stain was removed and coverslips were transferred to glass microscope slides with Prolong Gold anti fade reagent (Invitrogen) and analyzed by a confocal laser scanning microscope (CLSM, Leica SP5), equipped with an oil plan-Neofluar ×63/1.4 objective. AF488 was excited at 488 nm, and FM5-95 at 633nm. Images were analyzed using the LAS AF software (Leica).

Scanning electron microscopy. E. faecium cells were fixed for 15 min with 1% (v/v) glutaraldehyde (Sigma, The Netherlands) in PBS at room temperature onto poly-L-lysine covered glass slides. HT-29 monolayers on poly-l-lysine glass slides were fixed in 4% formalin. Mouse cecum and colon tissue was fixed in 2% glutaraldehyde directly after dissection and stored at 4°C. Bacterial, cell line or tissue samples were washed twice with PBS to remove excess fixative and were subsequently serially dehydrated by consecutive incubations in 1 to 5 ml of 25% (v/v) and 50% (v/v) ethanol-PBS, 75% (v/v) and 90% (v/v) ethanol-H2O, and 100% ethanol (2x), followed by 50% ethanol-hexamethyldisilazane (HMDS) and 100% HMDS (Sigma, The Netherlands). The glass slides or tissue parts were removed from the 100% HMDS and air-dried overnight at room temperature. After overnight evaporation of HMDS, samples were mounted onto 12 mm specimen stubs (Agar Scientific, The Netherlands) and bacteria coated with gold to 1 nm and tissue to 4 nm using a Quorum Q150R sputter coater at 20 mA prior to examination with a Phenom PRO Table-top scanning electron microscope (PhenomWorld, The Netherlands).

In vitro agglutination. Agglutination of E. faecium E1162 was achieved by two distinct ways. Firstly, 20 µl of cleared faecal extracts of untreated mice or mice treated with antibiotics were inoculated with 1x105 CFU/ml E. faecium E1162 and incubated for 30 min at 37°C. Secondly, 10 µg/ml of recombinant E-cadherin, pIgR, IgA or BSA was added to PBS inoculated with E. faecium and incubated for 30 minutes at 37°C. Then, bacteria were transferred to poly-l-lysine covered glass slides and stained with either FM5-95 or subjected to immunofluorescence labeling using specific antibodies and analyzed by confocal microscopy.

SDS-PAGE and Western blot analysis. Cleared faecal protein samples, HT-29 cell extracts or recombinant proteins in sample buffer (50 mM Tris-HCl pH 6.8, 12.5% glycerol, 1% sodium dodecylsulfate, 0.01% bromophenol blue) were separated through a 4-20% gradiënt SDS-polyacrylamide gel (Biorad). Prior to electroblotting, the membranes were incubated for 10 min in blot buffer (20 mM Tris, 150 mM glycine, and 20% methanol, pH 8.3). Samples were electroblotted using a Trans-Blot cell tank transfer unit at 100V onto PVDF membranes (Bio-Rad Laboratories Inc., The Netherlands). The membranes were blocked with 4% skim milk (ELK, Campina Holland, The Netherlands) in PBS-0.1% Tween 20 for 1 hr at 20°C. Incubation with primary antibodies (all at a 1:2,500 dilution) was carried out for 1 hour in 1% ELK in PBS-1% Tween 20 at 37°C, followed by a wash of 30 min in PBS-0.1% Tween 20 at 20°C. Subsequently, membranes were incubated for 1 hr with anti-Goat IgG (heavy plus light chains)-horseradish peroxidase (1: 5000 dilution, Southern Biotech) in 1% BSA in PBS (PBSb)-1% Tween 20 at 20°C. Membranes were washed twice with PBS-0.1%, Tween 20, and proteins were visualized using the ECL plus Western blotting detection system (GE Healthcare) and exposed using a LAS AF imager.