Expression of Adhesive Pili and the Collagen-Binding Adhesin Ace Is Activated by ArgR Family Transcription Factors in Enterococcus faecalis

Growth conditions and chemicals.E. faecalis strains were grown in M9-YE (39) or in brain heart infusion (BHI) broth (Becton, Dickenson and Co., Franklin Lakes, NJ) under static conditions at 37°C or on M9-YE or BHI agar, unless stated otherwise. Escherichia coli strains were cultured in Luria broth (Miller [LB]; Life Technologies, Grand Island, NY) or BHI medium. The following concentrations of antibiotics were used when needed: erythromycin (Em), 100 μg/ml for E. coli and 10 μg/ml for E. faecalis; chloramphenicol (Cm), 20 μg/ml; spectinomycin (Sp), 150 μg/ml for E. coli and 500 μg/ml to 1,000 μg/ml for E. faecalis; rifampin (Rif), 200 μg/ml, and fusidic acid (Fus), 25 μg/ml for E. faecalis. All antibiotics were purchased from Sigma-Aldrich (St. Louis, MO). When required for E. faecalis, 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal; Gold Biotechnology, St. Louis, MO) was added at 150 to 200 μg/ml in M9-YE plates. For the expression of cloned genes in complementation experiments, E. faecalis cultures were induced with 10 ng/ml synthetic cCF10 pheromone (Mimotopes, Victoria, Australia) in M9-YE medium.

Strain and plasmid construction.The bacterial stains and plasmids used in this study are listed in Table S6 in the supplemental material. E. faecalis strains were all derived from strain OG1RF. To characterize the regulation of pilus expression, we initially used allelic exchange to generate strains containing markerless in-frame gene deletions of ahrC, argR2, and ebpR, either singly or in combination. The construction of individual and combined deletions of ahrC and ebpR was accomplished using pCJK47, as described previously (40); deletions of argR2, singly and in combination with the other loci, were created using allelic exchange with the temperature-sensitive plasmid pCJK218, as described previously (41). To create each deletion construct used for allelic exchange, overlap extension PCR was used. First, two ∼1.0-kb fragments were amplified from OG1RF genomic DNA by using the primers listed in Table S7. Amplicons were annealed together, and the second-step PCR was performed using the primers indicated in Table S7. The resulting products were digested with the enzymes NcoI or SphI and XbaI and ligated into pCJK47 or pCJK218 (digested with the same restriction enzymes), creating pCJK47 ΔahrC, pCJK47 ΔebpR, and pCJK218 ΔargR2. E. coli strain EC1000 was used to propagate pCJK47 derivatives and DH5α was used for pCJK218 derivatives. All deletion mutants were confirmed by PCR amplification and sequenced using the primers listed in Table S7.

We examined the effects of the above deletions using plasmid-borne lacZ transcriptional reporter fusions to either the ebpR [plasmid ebpR-lacZ(−87)] or ebpA [plasmid ebpA-lacZ(−100)] promoter (see Fig. 2 and Results for a description of how the functional 5′ regulatory regions for these promoters were determined). These constructs were derived from the pTCV-LacSpec plasmid (24) (generously provided by Lynn Hancock, University of Kansas). For each reporter construct, we used the background expression of the lacZ reporter from the empty pTCV-LacSpec as a control.

To complement the various deletion mutations, we added a second plasmid carrying the cognate regulatory genes (including their native ribosome binding site [RBS]) cloned into the pheromone-inducible expression plasmid pCIE (42); all complementation experiments used cultures containing cCF10 pheromone as described above. To complement the single mutant (ΔahrC, ΔargR2, and ΔebpR) strains, previously generated plasmids (12) containing ahrC, argR2, and ebpR genes were digested with BamHI and XhoI and ligated to pCIE digested with BamHI and SalI to generate pCIE_ahrC, pCIE_argR2, and pCIE_ebpR, respectively. To create the double complementation vector, ahrC was PCR amplified using the SphI-site-containing primers indicated in Table S7. The amplicon and pCIE_argR2 were digested with SphI and ligated to create pCIE_argR2_ahrC. The constructs were sequenced, and in the case of ahrC, Western blots were performed to determine the expression of AhrC by using an anti-AhrC antibody (17). We compared the reporter expression for each complementation construct to that of an isogenic strain containing the same reporter plasmid but carrying the empty pCIE expression vector.

For the experiment analyzing the regulation of the ace promoter (Fig. 8), we amplified and cloned a 462-bp segment of the OG1RF chromosome containing this promoter (25); the cloned segment extended from the 3′ terminal segment of OG1RF_RS04590 (the gene immediately upstream from ace) through 101 bp of intergenic DNA and the first 9 codons of ace. We ligated this DNA to the RBS/lacZ in pTCV-LacSpec to generate a transcriptional fusion, ace-lacZ. We then assayed reporter activity in complemented and uncomplemented strains carrying single deletions of either ahrC, argR2, or ebpR. The sequence described above containing the functional ace promoter and the sequences of primers used for its amplification and cloning were based on the published data from reference 25.

Table S6 lists the chromosomal genotype and plasmid content for each strain examined in this study. The figure legends show the genotypes of the strains used for genetic analyses.

RNA purification, RNA-seq, and RT-qPCR.E. faecalis biofilm and planktonic cells were obtained from CDC biofilm reactors (CBRs) that were set up as previously described (23). Briefly, OG1RF and the ΔahrC mutant were streaked on BHI plates containing rifampin and fusidic acid. For each reactor, an overnight M9-YE culture was inoculated with three colonies and grown at 37°C. A CBR containing full-strength M9-YE was inoculated with 2 ml of an overnight culture at a cell density of 1.4 × 10⁹ to 2.8 × 10⁹ and incubated with stirring at 37°C for 4 h to enable bacterial attachment. M9-YE (10%) was then pumped through the reactor for additional 2 h. Next, biofilm cells from 24 polycarbonate coupons (BioSurface Technologies Corp., Bozeman, MT) and 8 Aclar fluoropolymer strips (Electron Microscopy Sciences, Hatfield, PA) from the reactors were washed, treated with RNAprotect (Qiagen), and vortexed as described in reference 23. Planktonic cells (3 ml) were treated with RNAprotect, pelleted, and stored with the biofilm cells at −80°C. Two biological replicates were run for each strain. To recover sufficient RNA from the ΔahrC mutants, two reactors were pooled for each replicate. For one of the OG1RF replicates, two reactors were also pooled.

RNA isolation and RT-qPCR were performed as previously described (23) with minor changes. Briefly, either planktonic or biofilm cells were lysed with lysozyme and mutanolysin for 10 min at 37°C. RNA was purified from the lysate using the RNeasy minikit (Qiagen) according to the manufacturer's instructions. For each replicate, 8 μg of total RNA was treated with Turbo DNase using the rigorous method (Ambion/Thermo Fisher Scientific). Universal 16S primers were used to confirm DNA removal. A Superscript III first-strand synthesis system for RT-PCR kit (Invitrogen/Thermo Fisher Scientific) was used to make cDNA using random hexamers. The RT reactions were diluted 5-fold in sterile water and aliquoted. One microliter of cDNA was used in 25-μl iQ SYBR green Supermix reaction mixtures containing 200 nM each primer. The optimal temperature for each primer pair was determined to be 53°C except for the reference gene, which was 60°C. The following three-step thermal cycling protocol was used: 3 min at 95°C (one cycle); 10 s at 95°C, 30 s at 53°C or 60°C, and 30 s at 75°C (40 cycles); and 50°C to 95°C melt curve with 0.5°C temperature shift. The RT-qPCRs were measured on an iCycler iQ5 (Bio-Rad). The primer pairs and amplicon sizes are listed in Table S7. All experiments met the following specifications: primer efficiency, 80% to 105%; r² ≥ 0.999; and C_T for no-RT control samples of >35. The Pfaffl method was used to analyze data and calculate the relative fold change normalized to OG1RF_RS05010, which was determined in RNA-seq to be unchanged in expression in planktonic and biofilm cells (43). Each reaction was performed in triplicate from each of two biological replicates for planktonic and biofilm cultures for each strain.

RNA-seq experiments were performed according to the methods previously described (42). Briefly, total RNA from planktonic and biofilm cells was treated with Turbo DNase (Ambion/Thermo Fisher Scientific) using the rigorous method. rRNA was removed using a MICROBExpress bacterial mRNA purification kit (Ambion/Thermo Fisher Scientific) according to the manufacturer's instructions. Illumina RNA-seq library creation and sequencing, using the HiSeq 2500 (paired end, 50 bp and 125 bp) platform, were performed at the University of Minnesota Genomics Center (UMGC). The fastq CASAVA (version 1.8, Illumina) sequence files were demultiplexed and deposited at the Minnesota Supercomputing Institute (MSI; University of Minnesota). Sequence analysis on paired-end reads was carried out using the Galaxy server (44–46), maintained by the MSI. Our data analysis followed the same workflow used by Weaver et al. (42) to analyze fastq files with some minor changes. Briefly, the 3′ ends of raw sequences were trimmed using FASTQ Quality Trimmer with the sliding window (47) and Cutadapt (46, 48) to remove adapter sequences. FASTQ interlacer and FASTQ de-interlacer or resync: paired-end resynchronization (MSI program) was used on the trimmed fastqsanger files to remove any unmatched paired-end reads, followed by mapping to the OG1RF reference genome (NC_017316.1 or CP002621.1) using Bowtie for Illumina (49). The SAM files were sorted by chromosome start position using Picard Sortsam (50).

Cuffmerge (51) was used on the GTF files from Cufflinks (51) to assemble the two planktonic and two biofilm files from the ΔahrC (D983) and OG1RF strains. Cuffdiff (51) was used to determine significant differences in the transcriptional expression between planktonic and biofilm cells with a false discovery rate of 0.05. The percentage of mapped reads ranged from 95% to 98% compared to total reads ranging from 22 million to 55 million for each sample and was determined using Flagstat (50, 52). Volcano plots were generated using Microsoft Excel (2016).

Immunoblot analysis of pilin protein.Three colonies from each strain were inoculated in 2 ml M9-YE supplemented with Cm 20 μg/ml and grown at 37°C overnight. The next morning, the cultures were diluted to an optical density at 600 nm (OD₆₀₀) of 0.05 in M9-YE containing 10 ng/ml cCF10 and grown for 4 h at 37°C. Whole-cell lysates were extracted from cultures normalized to an OD₆₀₀ of 1. The cells were pelleted, washed with 10 mM potassium phosphate-buffered saline (KPBS; pH 7.4), and frozen at −80°C. The frozen pellets were resuspended in 50 μl TES (10 mM Tris [pH 8.0], 1 mM EDTA [pH 8.0], 25% sucrose) containing 10 mg/ml lysozyme (Sigma-Aldrich) and 250 U/ml mutanolysin (Sigma-Aldrich) and incubated for 30 min at 37°C. Equal volumes of XT sample buffer (Bio-Rad) containing XT Reducing Agent (Bio-Rad) were added to the lysed cells and boiled for 15 min. The resulting lysates were separated on a Criterion XT precast gel, 3% to 8% Tris-acetate (Bio-Rad), and transferred to an Immobilon-P membrane (polyvinylidene difluoride [PVDF]; Millipore) in Towbin buffer (Bio-Rad) containing 0.05% SDS and 5% methanol. The membrane was stained with Ponceau S (Sigma-Aldrich) and scanned for loading uniformity. A Millipore Rapid chemiluminescent detection method was used according to the manufacturer's protocol with minor changes. Briefly, the dried membrane was incubated with primary antibody (1:10,000 dilution of anti-EbpC IgG; kindly supplied by Barbara Murray, University of Texas, Houston) in dilution buffer (5% nonfat dry milk, 10 mM KPBS [pH 7.4], 0.01% Tween 20 [vol/vol] [KPBS-T20]) at room temperature for 1 h. The membrane was washed with KPBS-T20 prior to incubating at room temperature for 30 min with secondary antibody (1:10,000 horseradish peroxidase [HRP]-conjugated goat anti-rabbit IgG_H+L; Zymo, Invitrogen) resuspended in dilution buffer. The blot was developed after washing with KPBS-T20 using SuperSignal West Pico chemiluminescent substrate (Thermo Scientific), and the signal was captured on X-ray film (Lumi-Film chemiluminescent detection; Roche/Sigma-Aldrich) by using an automatic processor.

β-Galactosidase assays.Strains were grown overnight in M9-YE with the appropriate antibiotics. For β-galactosidase (β-Gal) spot assays, the cultures were spotted on M9-YE plates containing antibiotics, 200 μg/ml X-Gal, and 10 ng/ml cCF10 and incubated overnight and photographed. For quantitating the β-Gal activity, the cultures were diluted to an OD₆₀₀ of 0.05 in M9-YE containing 10 ng/ml cCF10 and incubated at 37°C for 3 h for Fig. 8 and S4 and 4 h for Fig. 2 to 4. The cells were removed and pelleted at 6,500 × g for 15 min. The cell pellets were frozen at −80°C. On the day of the assay, β-galactosidase was measured as described by Miller (53) with the following modifications. The cell pellets were resuspended in 1 ml of Z-buffer, and 0.01 ml to 0.2 ml of culture was used for the assays and incubated for 38 min for Fig. 8 and 45 min for Fig. 2 to 4. The samples were analyzed in duplicates, and the experiments were repeated twice. Statistical analyses were performed using the Student t test (two-tailed, unequal variances).

Microtiter plate biofilm assay.Biofilm assays in 96-well microtiter plates were based on a previously described protocol (14). Briefly, overnight M9-YE cultures with antibiotics were diluted to an OD₆₀₀ of 0.05 in M9-YE containing 10 ng/ml cCF10. For each strain, 100 μl was transferred to 7 wells of a 96-well plate (Corning Costar 3595), and sterile medium was used for 7 blank wells. The plates were incubated statically for 6 h in a humidified chamber at 37°C. The cell density was measured at an OD₆₀₀ on a Modulus microplate multimode reader (Turner Biosystems, Sunnyvale, CA) prior to washing three times with double-distilled water. The plates were air dried for at least 4 h before staining with 100 μl 0.1% safranin for 20 min. The plates were washed five times with double-distilled water and air dried. The safranin-stained wells were quantified by measuring the OD₄₅₀. Two to three biological replicates were performed for each strain. Biofilm production was calculated as an index of biomass stained with safranin (OD₄₅₀ values) normalized to the cell density (OD₆₀₀). The biofilm index values for each of the replicates were averaged and are reported as percent biofilm formation relative to that of OG1RF(pCIE), which was set to 100%.

Microscopy.E. faecalis strains were grown overnight in M9-YE with antibiotic. The next morning, we inoculated 24-well plates containing submerged 11-mm-diameter Aclar fluoropolymer coupons (Electron Microscopy Sciences, Hatfield, PA) with cultures diluted to an OD₆₀₀ of 0.025 in M9-YE supplemented with 10 ng/ml cCF10. The biofilms were grown for 6 h at 37°C with shaking (125 rpm). Aclar coupons were processed as previously indicated (23). Briefly, the coupons were washed 3× in KPBS and stained with 5 μg/ml Hoechst 33342 (Molecular Probes/Thermo Fisher Scientific). The stained coupons were mounted on slides with a cover glass spacer (SecureSeal adhesive; Electron Microscopy Sciences) using SlowFade Diamond antifade mountant (Molecular Probes/Thermo Fisher Scientific). Images were captured with a Zeiss AX10 using Zen 2 (blue edition) software as a wide-field snapshot or z-stack with ×20 0.8-numerical-aperture (NA) and ×100 1.3-NA objectives. The images presented were obtained using the Fiji ImageJ package (54).

Accession number(s).The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus database (55, 56) and are accessible at GEO series accession number GSE112936 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE112936).