Methods

Bacterial strains and culture conditions
The bacterial strains used in this study are listed in Table 1. Lactobacillus strains were cultivated anaerobically in MRS-broth (Difco, BD, Franklin Lakes, New Jersey) at +37°C. Enterotoxigenic Escherichia coli (ETEC) was cultured in Luria-Bertani (LB) broth (Difco, BD) with agitation at +37°C in the experiments assessing the effect of Lactobacillus strains on ETEC adherence. The E. coli strains used for the cloning and expression of Slp-encoding genes were cultivated with agitation in LB broth, or in the heterologous gene expression, in M9ZB medium [30] at +37°C, with kanamycin (30 μg/ml) being added when appropriate. In the pathogen inhibition assays, all the pathogens were cultivated in tryptic soy agar (TSA) plates (Difco) and subcultured in tryptic soy broth (TSB, Difco) at +37°C with agitation.
Table 1  Strains used in this study  amolecular weight of mature protein bnot tested for adherence cnot present dculture collection of the Department of Veterinary Biosciences/Veterinary Microbiology and Epidemiology, University of Helsinki, Finland.

Cell culture
The non-transformed continuous cell line IPEC-1, isolated from the small intestine of an unsuckled, newborn piglet [32] was used as a model for porcine small intestinal epithelium. The cells were cultured in Dulbecco’s modified eagle medium/Ham’s F12 Nutrient Mixture (DMEM/Ham’s F-12 [1:1]) supplemented with 5% fetal calf serum (FCS), 1% insulin-transferrin-selenium (ITS), 16 mmol/L HEPES (all PAN-Biotech, Germany) and 5 ng/mL epidermal growth factor (EGF; BD, Franklin Lakes, New Jersey) at 39°C and 5% CO2. In the adhesion and adhesion inhibition experiments, the cells were seeded at a density of 2 x 105 /ml to a Transwell-like culture (Thincerts™, 1 μm pore size, diameter 10 mm; Greiner bio-one, Frickenhausen, Germany) and cultured for 4–5 days to allow differentiation, until the transepithelial electric resistance (TEER) value was ≥1 kΩcm2.

Detection and expression analysis of slp genes in L. amylovorus
The presence of S-layer proteins on the surface of the L. amylovorus intestinal isolates GRL 1112 – GRL 1118 has previously been described [28]. The putative slp encoding genes were identified in silico in the draft genomes of the L. amylovorus strains based on homology with the publicly available L. acidophilus slp gene sequences. The identification of the expressed slp genes was based on the observed molecular weights of the proteins, obtained by analyzing overnight cultures of the strains by standard SDS-PAGE in 12% gels, and on the amino-terminal and/or internal amino acid sequences of the Slp:s. The amino-terminal sequences were obtained by an Edman-degradation-based Procise 494 HT sequencer (Life Technologies, Carlsbad, CA), and internal peptide sequences through a peptide mapping analysis: the proteins were digested in-gel by trypsin followed by analysis with liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) carried out with an EASY-nLC liquid chromatograph (Thermo Fisher Scientific, Germany) connected to a Velos Pro-Orbitrap Elite hybrid mass spectrometer (Thermo Fisher Scientific, Germany) with a nano-electrospray ion source (Thermo Fisher Scientific, Germany). Both amino-terminal sequencing and peptide mapping were performed in the Institute of Biotechnology (University of Helsinki, Finland).

Cloning and heterologous expression of the genes encoding L. amylovorus S-layer proteins
The expressed slp genes (see above) were amplified by PCR from the chromosomal DNA of the L. amylovorus strains, cloned as NcoI-XhoI –fragments in E. coli DH5αF', sequenced to verify the correct open reading frames, and expressed in E. coli BL21 (DE3) as C-terminal hexahistidine tag-fusions, as described in the pET system manual (Merck KGaA, Darmstadt, Germany) and as previously reported [33]. Recombinant S-layer proteins were purified in the presence of 4 M guanidine hydrochloride (GuHCl) with His Trap HP columns (GE Healthcare, Little Chalfont, UK) according to the manufacturer’s instructions. The pooled protein fractions were dialysed against deionized water overnight at +4°C, centrifuged (15 000 g, 20 min, +4°C) and stored in aliquots at -80°C.

Purification of porcine intestinal mucus
The 8-week old pig used for mucus isolation was housed in a piggery of MTT Agrifood Research (Finland), treated in strict accordance with the recommendations of the Finnish Ministry of Agriculture and Forestry (Directive 2013–497) and EEC (Directive 86/609/EEC) for the care and use of animals in research, and sacrificed by bolt gun. As the pig used in this study was not specifically included in any experimental protocol on living animals before slaughtering, there was no ethical requirement for collecting mucus samples. The mucus isolation protocol was modified from [34]. Briefly, the small intestine was opened longitudinally and washed with cold phosphate-buffered saline (PBS) with 0.1 mM phenylmethylsulfonyl fluoride (PMSF) as a protease inhibitor (PBS-PMSF). Mucus was collected by gentle scraping into PBS-PMSF, centrifuged (17,000 g, 1 hour, +4°C) to remove cells and insoluble material, and the supernatant was homogenized in a domestic blender. The homogenate was concentrated in a Centricon Plus-70 filter unit (molecular weight cutoff 10,000), clarified by centrifugation (17,000 g, 30 min, +4°C), filtered twice through a glass fibre filter (GE Healthcare, Little Chalfont, UK) and once through a 0.8 μm cellulose acetate filter (Sartorius, Goettingen, Germany) and purified by gel filtration chromatography at +4°C in a Sephacryl S-200 HiPrep 16/60 column (GE Healthcare, Little Chalfont, UK) at a flow rate of 1.8 ml/min with PBS as the eluent, monitoring the A280nm values of 5 ml fractions. The protein-containing fractions were dialyzed against water and assayed for total protein by the method of Bradford (Bio-Rad Protein Assay, Bio-Rad, Hercules, CA) using bovine serum albumin (BSA) as a standard, and for glycoproteins using the Crypton™ Glycoprotein Staining Kit (Thermo Scientific, Waltham, MA), with porcine gastric mucins (Sigma), horseradish peroxidase (HRP) and soybean trypsin inhibitor (Thermo Scientific, Waltham, MA) as standards. The void volume fractions with a high glycoprotein content were pooled, lyophilized and stored at -20°C.

Adherence of L. amylovorus strains to mucus
The adherence of the L. amylovorus strains to porcine gastric mucins (type II, Sigma) or to porcine small intestinal mucus was studied essentially as described earlier [35], but by using a nucleic acid binding fluorescent stain SYTO®9 (Molecular Probes, Eugene, OR) rather than tritium for bacterial labeling, and PBS as the buffer. To label the bacterial cells in the experiments, the strains were cultivated overnight, collected, washed twice with 0.85% NaCl and suspended into the original volume of 0.85% NaCl, and then 1 μl of 5 mM SYTO®9 solution was added per 1 ml of cell suspension, followed by a 15 minute incubation in the dark with vigorous shaking, after which the cells were collected and washed twice with PBS. After the adherence assay, the input (added) and output (remaining) fluorescence values were measured in a microplate reader (Victor Multilabel Plate Reader, Perkin Elmer, Waltham, MA) and the adherence was expressed as the proportion (%) of the original fluorescence remaining, after first subtracting the background fluorescence from mucus-coated wells without bacteria (for outputs) and from wells filled with PBS (for inputs).

Adherence of L. amylovorus strains to IPEC-1 cells
L. amylovorus strains were cultivated overnight in MRS-broth containing 10 μCi 3H-thymidine/ml for metabolic labeling, collected and washed twice with PBS. IPEC-1 cells grown on Thincert™ wells were washed once with PBS, and 125 μl of the labeled bacterial suspensions in DMEM/Ham’s F-12 [1:1] medium at A600nm values of 0.25, 0.5 or 1 were added per well. The plates were incubated for one hour at +37°C, 5% CO2 followed by five washes with PBS. The cells were lysed by adding 250 μl of 1% SDS in 0.1 M NaOH per well and by incubating overnight at +37°C, and the radioactivity of the lysed samples (output) was measured by liquid scintillation counting. The input radioactivity values were determined by liquid scintillation counting of the cell suspensions in DMEM/Ham’s F-12 [1:1] (A600nm = 0.25, 0.5 or 1), first treated with an equal volume of 1% SDS in 0.1 M NaOH overnight at +37°C. The adherence was expressed as the proportion (%) of the original radioactivity remaining, after first subtracting the background radioactivity from IPEC-1 cells incubated without bacteria (for outputs) and from DMEM/Ham’s F-12 [1:1] medium (for inputs).

Inhibition of F4-fimbriated ETEC adherence to IPEC-1 cells by L. amylovorus
L. amylovorus cells were cultivated overnight, collected and washed twice with PBS. The F4+ ETEC strain was cultivated overnight in LB broth containing 10 μCi 3H-thymidine/ml for metabolic labeling. Labeled ETEC cells were collected and washed with PBS and the expression of F4 fimbriae was confirmed with the Fimbrex slide agglutination test kit (VLA Scientific, New Haw, UK). The inhibitory effects of the strains were tested in three different experimental set-ups: exclusion, competition and displacement. In each arrangement, 100 μl of L. amylovorus strains in DMEM/Ham’s F-12 [1:1] medium (A600 = 6) were added to IPEC-1 cells; in exclusion assays 1 hour before, in displacement assays 1 hour after, and in competition assays simultaneously with the addition of 100 μl of 3H –labeled ETEC in the same medium (A600 = 0.6). In the displacement assays, the unbound ETEC cells were removed by two washes with PBS before the addition of lactobacilli. The cells were further incubated for one hour at +37°C, 5% CO2, followed by five washes with PBS. Finally, the cells were lysed by adding 250 μl of 1% SDS in 0.1 M NaOH per well and by incubating overnight at +37°C, and the radioactivity of the lysed samples was measured by liquid scintillation counting. The proportion of adherent ETEC cells (%) in the presence or absence of the L. amylovorus strains was calculated as in the adhesion experiments, and the inhibition (%) was calculated according to the formula: [adherence (no La) – adherence (with La)] / adherence (no La) x 100%, where La indicates L. amylovorus.

Growth inhibition of intestinal pathogens by the culture supernatants of L. amylovorus
The supernatants collected (650 g, 20 min, +4°C) from overnight cultures of the L. amylovorus strains were filter-sterilized through 0.22 μm pore-size filters and stored at -20°C. The inhibitory effects of the supernatants were assessed by monitoring the abilities of the pathogens to grow in the presence (10% V/V) of the supernatants in a microtiter plate format as previously described [36]. Briefly, the A600nm values of the pathogen cultures were measured every 30 minutes with an automatic reader (Bioscreen C, Growth Curves Oy, Helsinki, Finland) in the presence or absence of pH-adjusted (pH 6.2) or non-adjusted culture supernatants at 36.5 +/-0.5°C, with three parallel wells for each supernatant and control. The inhibition was quantified using the area under the growth curve (AUC) obtained during the first 12 hours of growth, automatically created by the Research Express software (Transgalactic Ltd, Vantaa, Finland), and expressed as the area reduction percentage (ARP) as previously described [29]. Linear regression (SPSS) was used to estimate the relationship between the ARP values and colony forming unit (CFU) counts as previously described [36].

Isolation and generation of human monocyte-derived dendritic cells (moDCs)
Leukocyte-rich buffy coats, donated by healthy volunteers, as well as the permission to use human leukocytes, were obtained from the Finnish Red Cross Blood Service. Monocytes were purified and cultured in vitro to generate moDCs using a method described earlier [37] with minor modifications. Briefly, peripheral blood mononuclear cells were first isolated by Ficoll-Paque (GE Healthcare, Little Chalfont, UK) density gradient centrifugation using Leucosep separation tubes (Greiner Bio-One, Germany), followed by a Percoll (GE Healthcare, Little Chalfont, UK) gradient centrifugation step. After magnetic beading using anti-CD3 and anti-CD19 beads (Dynal Invitrogen, Life Technologies, Carlsbad, CA), monocytes were allowed to adhere to 24-well plates (Falcon, BD, Franklin Lakes, New Jersey) for 1 h in the presence of RPMI 1640 (Sigma) supplemented with 20 mM HEPES, penicillin and streptomycin (100 IU/ml) and 2 mM L-glutamine without serum. The adhering cells were washed twice with PBS, after which differentiation was induced by maintaining the cells in RPMI 1640 (supplemented as described above) containing 10% (v/v) FCS (Integro, Zaandam, the Netherlands), 10 ng/ml human recombinant granulocyte macrophage-colony stimulating factor (GM-CSF, Gibco Life Technologies, Carlsbad, CA), and 20 ng/ml human recombinant interleukin 4 (IL-4, Gibco Life Technologies, Carlsbad, CA). MoDCs were used on day 7 in the experiments. In each experiment, cells from four donors were used.

Stimulation of human moDCs and cytokine measurements
L. amylovorus strains were cultivated overnight, collected and washed with PBS. The A600nm values of the bacterial suspensions were normalized, and the bacterial cells were added to human moDCs at the multiplicity of infection (MOI) 1, 10, and 100 in RPMI 1640 containing FCS, HEPES, antibiotics, and glutamine. The same medium without bacteria was used as a control. After 24 h, cell culture supernatants were collected and stored at -20°C before further analyses. The supernatants were analyzed with the Bio-Rad’s Bio-Plex Pro Cytokine assay using the Bio-Plex -200 platform (Bio-Rad, Hercules, CA). Human TNF-α, IL-1β, IL-6, IL-10, and IL-12 quantification was performed for undiluted samples according to the manufacturer’s instructions. Human IP-10/CXCL10 was measured separately with the OptEIA ELISA kit (BD, Franklin Lakes, New Jersey) using samples diluted with sample matrix RPMI 1640 medium.

Purification of cell wall fragments (CWF) and coating of CWF by recombinant S-layer proteins
Cell wall fragments were purified from L. amylovorus cells as described earlier [33]. Purified cell walls were lyophilized and stored as suspensions in water at -20°C. In order to coat the cell walls, the affinity purified recombinant S-layer proteins were dissolved in 5 M GuHCl at a concentration of 30 μg/ml, dialyzed against 50 mM Tris–HCl (pH 7.0) at +4°C overnight and centrifuged (40,000 g, 30 min, +4°C) to remove large protein aggregates. The protein concentrations of the supernatants were determined by the Bradford method, immediately after which the supernatant proteins and the cell walls were combined in a ratio 1:4 (W/W) and incubated overnight at +4°C with rotation. The coated CWF were collected by centrifugation (25,000 g, 30 min, +4°C), resuspended into DMEM/Ham’s F-12 [1:1] medium and analyzed by SDS-PAGE. To verify the absence of large protein aggregates among the coated CWF, the preparations in 50 mM Tris–HCl (pH 7.0) were routinely negative stained by uranyl acetate (5 min on ice) and observed by JEOL 1200-EX II transmission electron microscope at the operating voltage of 80 kV.

Adherence of S-layer protein-coated cell wall fragments to IPEC-1 cells
CWF to be used as uncoated controls were labeled by EZ-Link Sulfo-NHS-LC-Biotin (Thermo Scientific, Waltham, MA) according to the manufacturer’s instructions by adding 0.5 mg of label per 500 μg of CWF (dry weight). IPEC-1 cells grown on Thincert™ wells were washed once with PBS, and 80 μg of Slp-coated or uncoated CWF in the total volume of 100 μl DMEM/Ham’s F-12 [1:1] medium was added per well, corresponding to approximately 8 μg (0.13-0.18 nmol) of each S-layer protein per well containing 2.5 x 105 IPEC-1 cells. The plate was incubated for two hours at +37°C and 5% CO2 followed by four washes with PBS. The cells were fixed with 4% paraformaldehyde (PFA) in PBS for 10 minutes at room temperature and washed three times with 0.1 M sodium phosphate buffer (pH 7.4). Slp-coated CWF were detected by an indirect immunofluorescence staining with Slp-specific immunoglobulins (20 μg/ml, purified by Hi Trap columns, GE Healthcare, Little Chalfont, UK) and AlexaFluor488-conjugated secondary antibodies (2 μg/ml, LifeTechnologies, Carlsbad, California), all in PBS-0.1% BSA, and uncoated cell walls were detected by staining with AlexaFluor488-conjugated streptavidin (2 μg/ml, LifeTechnologies, Carlsbad, California) in PBS-0.1% BSA. The bottoms of the Thincert™ wells were prepared for microscopy and observed in a Leica DM 4000B epifluorescence microscope (Leica Microsystems, Wetzlar, Germany). The mean number of adherent CWF was quantitated from 20 randomly selected fields of 3.5 x 104 μm2, and representative photographs were taken with the Olympus DP70 digital camera system with the cellP imaging software (Olympus Corp., Tokyo, Japan).