Batch fermentations of C. thermocellum ATCC 27405 wild-type strain were conducted in 3L BioStat B jacketed glass fermentors (Sartorius Stedim Biotech, Bohemia, NY) using a 2L working volume of MTC medium (mineral salt medium containing 1g/L yeast extract; [16]) at 58C and 300rpm, with pH controlled at 7.0 using 3N NaOH. Fermentors with medium containing only the carbon substrate, 5g/L crystalline cellulose (Avicel® PH105, FMC Biopolymer, Philadelphia, PA), were sparged with ultra-high purity nitrogen and vigorously agitated overnight, followed by addition of the remaining medium components and sparged for an additional 2-3hrs with nitrogen gas. A 10% v/v inoculum of overnight (16-20hrs) 5g/L Avicel® bottle cultures was used to inoculate the fermentors and the gas inlet/exhaust lines were clamped post inoculation.
Extracted molecule
total RNA
Extraction protocol
Fermentation samples for RNA isolation were harvested by spinning down ~30mL culture in 50mL Oak Ridge tubes at 8000 rpm and 4C for 10-15mins and the supernatant was discarded. The solid pellet fraction containing cells and any residual Avicel® was resuspended in 1mL of TRIzol (Invitrogen, Carlsbad, CA), flash frozen in liquid nitrogen and stored at -80C until further use. Total RNA was extracted from the cell pellets as follows. Briefly, the frozen cell solution in TRIzol was thawed on ice and the cell solution (~1mL) was added to a 2mL tube containing 1mL of 0.1mm glass beads (BioSpec Products, Bartlesville, OK) ashed at 250C overnight. Cells were lysed by rapid agitation of the tubes at 6500rpm for 1min in three 20s-On/20s-Off cycles using the Precellys® bead beater (Bertin Technologies, France). Subsequently, the cell lysate (~0.8mL) in TRIzol was phase separated by addition of 200uL chloroform and the RNA was precipitated by addition of 500uL 100% isopropanol. The precipitated RNA pellet was washed with 1mL of 75% ethanol and resuspended in 100uL of RNase-free water. Any contaminating DNA was digested by in-solution DNase-I (Qiagen, Valencia, CA) treatment and the RNA sample was cleaned using the RNeasy mini kit (Qiagen, Valencia, CA) as per manufacturer’s instructions.
Label
Cy3
Label protocol
The 6hr time-point total RNA sample was used as the reference and all other time-point samples (8, 10, 12, 14, 16hr) were compared to the reference in cDNA/cDNA arrays. For each time-point comparison, equal amount of the extracted total RNA samples was labeled with Cy3-dUTP/Cy5-dUTP fluorescent dyes (GE Healthcare, Piscataway, NJ), mixed and hybridized onto custom oligo-arrays in dye swap experiments as described earlier [Chhabra SR, J Bacteriol., 2006 188(5); PMID 16484192].
Hybridization protocol
For each time-point comparison, equal amount of the extracted total RNA samples was labeled with Cy3-dUTP/Cy5-dUTP fluorescent dyes (GE Healthcare, Piscataway, NJ), mixed and hybridized onto custom oligo-arrays in dye swap experiments as described earlier [Chhabra SR, J Bacteriol., 2006 188(5); PMID 16484192].
Scan protocol
Microarray slides were scanned in ScanArray Express scanner (Perkin Elmer, Waltham, MA) and signals were quantified in ImaGene version 6.0 (BioDiscovery Inc., El Segundo, CA).
Description
Microarrays containing 2980 unique and 10 group 70-mer oligonucleotide probes representing ~97% of the 3163 Open Reading Frames (ORFs) in the 2003 draft assembly of C. thermocellum ATCC 27405 were constructed as described earlier [Brown SD et al., Appl Biochem Biotechnol 2007 137-140(1-12); PMID: 18478424]. The probe sequences were later compared to the completed genome sequence using reciprocal BLAST analysis and assigned new ORF numbers. Based on the comparison, 79 probes which did not have any BLAST hits and 108 probes that only had partial hits to annotated ORFs in the closed genome were either excluded or marked-up during downstream data analysis. Cy3_13454094
Data processing
Statistical data analysis was conducted using JMP Genomics software (SAS Institute Inc., Cary, NC). The array signal intensities were background-corrected, log2-transformed and data for duplicated probes on the arrays were averaged and normalized using the Data-Standardize method. Low-quality array data were discarded based on scatter plots, correlation coefficients, principal component analysis and other quality control criteria. One-way ANOVA analysis was conducted with the 6hr samples as the control at a False Discovery Rate of 2% (P-value < 0.01) to identify differentially expressed genes of statistical significance. ANOVA analysis data are linked as a supplementary file on the Series record.
