Bacterial strains and suspended mixtures
Overnight cultures of E. coli K-12 PHL644/pMP4655 (carrying a gfp gene under the control of a constitutive promoter) and S. maltophilia/pBPF-mCherry were grown in Luria-Bertani (LB) broth supplemented with tetracycline (80 μg/ml) or gentamicin (20 μg/ml) at 34°C with continuous shaking (200 rpm). Cells were pelleted by centrifugation (3,300 × g, 4°C, 3 min), re-suspended, and diluted in 1× phosphate buffered saline (PBS, pH 7.4) supplied with 0.5% bovine serum albumin (BSA) (Pierce, Rockford, IL). A series of artificial mixtures of E. coli and S. maltophilia were prepared by mixing the PBS re-suspended and diluted E. coli and S. maltophilia cells at different ratios.
Biofilms were cultivated on the inner surface of silicon tubing (Cole-Parmer, Vernon Hills, IL) in flow cell systems as described previously . Briefly, a flow cell system was assembled, sterilized, and conditioned by running 0.1× LB broth (10-fold diluted LB broth, 1 ml/min) at room temperature (20-25°C). Operation was paused for one hour to allow inoculation with S. maltophilia and E. coli mixed at a ratio of 1:1. After three days of growth, biofilms were scraped into 1× PBS and pre-homogenized on ice using a homogenizer (OMNI TH, Marietta, GA) set at the lowest speed for 30 seconds. Biofilms were further dispersed into single cells using the same homogenizer set at the maximum speed for two minutes. Over 99% of bacterial cells in the biofilm matrix were dispersed into single cells. The dispersed biofilm cells were then diluted in 1× PBS (with 0.5% BSA) for IMS.
One milliliter of samples was incubated with 10 μl anti-E. coli antibody (ViroStat, Portland, ME) for 10 min with gentle shaking. Bacterial cells were pelleted by centrifugation (3,300 × g, 4°C, 3 min) and re-suspended in 100 μl separating buffer (1× PBS, 0.5% BSA, 2 mM EDTA, pH 7.4) (EDTA: ethylenediaminetetraacetic acid). 10 μl streptavidin microbeads (Miltenyi Biotec, Auburn, CA) were added and incubated at 4°C in the dark for 10 min. Separation of E. coli cells was performed in LS columns and a midi MACS® separator (Miltenyi Biotech, Auburn, CA) following the protocol provided by the manufacturer, except that one more washing step was added to remove more S. maltophilia cells. In a two-step IMS, enriched cells from the first step IMS were directly transferred into a new LS column for the second separation. Densities of E. coli and S. maltophilia cells in samples and IMS enriched collections were measured using a plate-counting method with selective agar. Cell densities were used to calculate recovery and purity of E. coli after IMS.
The protocol was amended with the use of RNAlater when enriched cells were used for microarray study. Bacterial cells were re-suspended in RNAlater rather than PBS after sample collection and kept at 4°C overnight, followed by homogenization. RNAlater was removed and cells were re-suspended in separating buffer just before IMS. During column separation, the buffer was additionally supplied with 10% (v/v) RNAlater. Enriched cells were immediately stored in RNAlater. The whole procedure was performed at 4°C. All buffers, reagents, and pipette tips were nuclease-free and pre-cooled.
Pure E. coli cultures were used to evaluate the effect of separation on the transcriptome by microarray analysis. Suspended E. coli cultures were harvested from an annular reactor (1320 LJ, BioSurface Technologies, Bozeman, MT), supplied with 0.1× LB broth (100 ml/h) for 7 days after inoculation. Aggregates were removed from broth cultures by filtration (5.0 μm Millipore, Billerica, MA). Suspended E. coli cells were immediately re-suspended in RNAlater and stored at 4°C overnight. One aliquot of RNAlater stored E. coli cells served as the control ("unsorted" cells) and was kept in RNAlater without further treatment. The other aliquot was treated to acquire "sorted" cells as described above using the amended protocol. Samples collected independently from a second annular reactor served as a biological replicate for the microarray study.
RNAlater was removed by filtration with a membrane (0.22 μm, Millipore, Billerica, MA) from E. coli cells just before RNA extraction for both "unsorted" and "sorted" cell collections. RNA extraction was based on a hot SDS/phenol protocol . A step of bead beating (BioSpec, Bartlesville, OK) for one minute was added to break cells, and all phenol/chloroform/isoamyl alcohol washes were performed in phase lock gels (5 Prime, Fisher Scientific, Pittsburgh, PA). DNA was removed from extracted RNA with Turbo DNase treatment (Ambion, Austin, TX) at 37°C for 30 min followed by purification with an RNeasy Mini Kit (Qiagen, Germantown, MD). The quality of RNA was examined by gel electrophoresis using E-gel with SYBR Safer (Invitrogen, Carlsbad, CA). High quality RNA was further re-precipitated, concentrated, and stored at -80°C.
RNA was reverse transcribed into cDNA using random hexamers (pd(N)6) (GE Healthcare, Piscataway, NJ) and labeled with Amersham CyDye Post-Labeling Reactive Dye (Amersham Biosciences, Piscataway, NJ) following the protocol provided by the Amino Allyl cDNA Labeling Kit (Ambion, Austin, TX). The quantity and labeling efficiency of cDNA was measured using a NanoDrop Spectrophotometer (ND-1000, Thermo Scientific, Wilmington, DE).
Microarray slides for E. coli were purchased from the University of Alberta (Edmonton, AB, Canada). Each slide contained three replicates of 5,978 70-mer oligonucleotides representing three E. coli strains (4,289 of them were for E. coli K-12). Sample preparation and loading, slide prehybridization, hybridization and washing were performed according to Corning protocols (GAPS II coated slides, Corning Inc., Lowell, MA). An extended 4-h prehybridization using a higher BSA concentration (1 mg/ml) was found to perform best in reducing background noise. Hybridization was in a Corning Microarray Hybridization Chamber (Corning Inc.) in 42°C water bath.
Microarray slides were scanned with a Virtek ChipReader (Virtek Vision, Waterloo, ON, Canada). Spots on scanned images were recognized and pixel intensity for each spot was quantified using the TIGR software Spotfinder (v3.1.1). Gene expression data were analyzed in the software Acuity 4.0 (Molecular Devices, Sunnyvale, CA). LOWESS normalization was performed for every microarray with three iterations using a smoothing factor of 0.4. Hybridized spots with oligonucleotides for strain E. coli K-12 having a high QC (quality control) value (> 0.1), good flag tags (A, B and C) in both Cy3/Cy5 channels were chosen for further analysis. One sample t-tests were performed across replicates. Step-down Bonferroni-Holm was used for the correction of multiple hypotheses testing. Genes with at least two-fold change in expression (p-value < 0.05) were considered to have changed expression during sample dispersion and IMS. Microarray data were deposited in NCBI Gene Expression Omnibus database (GSE22885).
Quantitative PCR (qPCR)
Primers for qPCR confirmation of the differential expression of eight identified genes in Table 1 are listed in Additional File 2: qPCR primers for nine tested genes. The primers were designed using the software Primer Premier (Palo Alto, CA) and synthesized by Integrated DNA Technologies (Coralville, IA). Annealing temperatures and qPCR efficiency were optimized with PCR products using E. coli genomic DNA as template. The 16S rRNA gene was selected as the housekeeping gene. The amplification efficiency for target genes was near 100% and within 5% of the housekeeping gene of 16S rRNA. Total RNA from sorted and unsorted E. coli cells were reverse transcribed to cDNA using a reverse transcription kit (Applied Biosystems, Carlsbad, CA). cDNA was diluted 10- and 100-fold and 1 μl was assembled for qPCR reactions using the SYBR Green PCR Master Mix (Applied Biosystems, Carlsbad, CA). Differential expression of the same gene in sorted and unsorted E. coli was calculated with the ΔΔCt method from four replicates. The PCR program included a cycle of 95°C for 10 min, 35 cycles of 30 seconds at 94°C, 30 seconds at the optimized annealing temperature for each set of specific primers and 30 seconds at 72°C, and a melting curve analysis from 60°C to 95°C at the end.