gender: female strain: D. pseudoobscura 14011-0121.94 tissue: head developmental stage: 8 day old mated adult
Treatment protocol
Frozen whole flies were placed on 10.8 x 10.8 cm ceramic plates frozen on dry ice. Forceps were used to dissect 400-500 heads/sex for subsequent RNA extraction.
Growth protocol
We examined gene expression in male and female head samples from Drosophila melanogaster (OreR), D. pseudobscura (Drosophila Species Stock Center (DSSC) 14011-0121.94), and D. mojavensis (DSSC 15081-1352.22). Flies were grown at 22°C, 60% relative humidity in standard vials containing cornmeal media, for D. melanogaster (http://flyfood.arl.arizona.edu/cornmeal.php3) or banana-opuntia for D. pseudobscura and D. mojavensis (http://flyfood.arl.arizona.edu/opuntia.php3). All media was purchased from the Tucson Stock Center, Tucson, Arizona. Newly eclosed flies of both females and males were housed together in vials at normal density and aged for 7 days. On day 7, flies were anesthetized with CO2 and then females and males were sorted and placed into separate vials. Same sex vials were then aged an additional 24 hours at 22°C, 60% relative humidity. Whole flies were flash frozen on dry ice. Flash frozen flies were stored in -80°C before dissection and subsequent RNA extraction. Frozen fly heads were dissected using forceps from whole files on ceramic plates chilled on dry ice. All equipment used during dissections were wiped with RNAase Away (Molecular BioProducts, San Diego, California).
Extracted molecule
polyA RNA
Extraction protocol
Four biological replicate male and four biological replicate female samples were collected for each species. Each replicate consisted of a pooled sample of 400-500 fly heads. TRIzol (Invitrogen Inc, Carlsbad, California) was used to extract total RNA from heads following manufacture directions. Briefly, fly heads were immersed in 250 µl of TRIzol and thoroughly homogenized using a 1.5 ml RNAase free pestle (Kimble-Chase, Vineland, New Jersey) fitted to a Pellet Pestle Motor (Sigma-Aldrich, St. Louis, Missouri). After initial homogenization, 550 µl of TRIzol was added and the solution was further homogenized by hand. Chloroform (160 µl) was added to the homogenate and then the sample was centrifuged at 12,000 g for 15 minutes at 4°C. The aqueous layer was removed, precipitated with isopropyl alcohol and glycogen (8 µg), then centrifuged for 10 minutes at 4°C to pellet the RNA. The supernatant was removed and the RNA pellet was washed with fresh 75% ethanol. The pellet was dried briefly and then dissolved in 50 µl DEPC treated H2O. mRNA was enriched from total RNA by poly A+ selection using a Oligotex mRNA kit (QIAGEN Inc., Valencia, California) following manufacture instructions. mRNA was eluted from small spin columns twice with hot 30 µl buffer OEB. mRNA quality and quantity was determined by Bioanalyzer (Agilent, Santa Clara, California) and 260/280 ratio from Nanodrop ND-1000 UV spectrophotometer (Nanodrop, Wilmington, Delaware).
Label
Cy3
Label protocol
Arrays were used in a two-color format with mRNA labeled with Cy3 or Cy5. Four biological replicate male and four biological replicate female heads, each dye-flipped to control for potential dye artifacts from each of the three species were used in microarray experiments. For each sample of mRNA, 600 ng was simultaneously reverse-transcribed and labeled with dye along with Arabidopsis spike-in controls. Reverse-transcription/labeling reactions consisted of 15 µl 5X First Strand Buffer (Invitrogen, Carlsbad, California), 3 µl 0.1 mM DTT (Invitrogen, Carlsbad, California), 9 µl of dye-mix containing Cy3 or Cy5 labeled nonamer primer (Trilink Biotechnologies, San Diego, California), 3 µl RNase inhibitor (Ambion, Austin, Texas), 15 µl H2O plus Arabidopsis spike-in controls, and 3 µl Super-script II reverse transcriptase (Invitrogen, Carlsbad, California). Labeling reactions were incubated at 37°C for 2 hours then stopped by heat termination at 85°C for 5 min. Labeling reactions were purified using Clonetech columns (TE-10), (Clonetech, Mountain View, California) following manufacture instructions.
Hybridization protocol
Samples to be co-hybridized onto one microarray were pooled and precipitated with 120 µl DEPC treated H2O, 3 µl glycogen (1 mg/ml), 180 µl 5M ammonium acetate, and 900 µl cold 100% EtOH and by cold centrifugation at max speed for 20 min. Probe pellets were re-suspended in warm 2X hybridization buffer (250 µl formamide, 250 µl 20X SSC, 20 µl 10% SDS, 5 µl tRNA(2g/µl)) with Nimblegen feducial (Cy3 CPK6 and Cy5 CPK6) controls and heated to 95°C for 5 minutes. Nimblegen microarrays were fitted with X1 mixers according to standard protocols, warmed to 42°C on a MAUI hybridization station (BioMicro Systems, Salt Lake City, Utah), and 15 µl of hybridization cocktail was loaded onto microarrays through the X1 fill port. Vent and fill ports were sealed and arrays were loaded onto bays of the MAUI station. Mix mode B on the MAUI hybridization station was started and samples were hybridized for 18 hours at 42°C in low ozone conditions. Arrays were washed using the following modified Nimblegen protocol. An array with mixer was submerged for 10 seconds in warm (42°C) Buffer I (1000 ml H2O, 10 ml 20X SSC, 20 ml 10 % SDS, 100 µl 1 M DTT) and the mixer was removed. Arrays were placed in a fresh beaker of 42°C Buffer I and agitated for 2 minutes, submerged for 1 minute in room temperature Buffer II (1000 ml H2O, 10 ml 20X SSC, 100 µl 1 M DTT), followed by submerging for 15 sec in Buffer III (1000 ml H2O, 2.5 ml 20X SSC, 100 µl 1 M DTT). Arrays were placed in an ArrayIT Microarray Centrifuge and spun for 1 minute to dry. Dried arrays were paced in tubes and covered in aluminum foil until scanning.
Scan protocol
Arrays were scanned on a Axon GenePix® 4000B (Molecular Devices, Sunnyvale, California) scanner controlled by GenePix 6.0 software. Cy3 and Cy5 channels were balanced by visual inspection of histograms for Cy3 and Cy5 channels and by manually changing PMT settings for each laser. Each scan was saved as a TIFF file.
Data processing
Raw TIFF files for each microarray were processed using NimbleScan 2.4.27 (Roche NimbleGen, Inc. Madison, Wisconsin) software. We followed probe annotation defined in GEO platform ID GPL4629 for D. melanogaster, GPL4631 for D. pseudoobscura, and GPL4630 for D. mojavensis. This strategy created one measurement of gene expression by collapsing isoforms and should be conservative but doing so changes the number of probes assigned to gene models. The mean number of probes for D. melanogaster FBgn IDs was 27.3 (range: 20-520, N= 13446), for D. pseudobscurca the average number of probes was 25.3 per GLEANR ID (range:2 – 152 probes; N= 12901) and for D. mojavensis the average number of probes/GLEANR ID was 23.2 (range: 2-186; N=12648). All replicates for each species were hybridized together on the same day with the exception of D. mojavensis in which 3 of the 4 replicates were hybridized on one day and the fourth replicate was hybridized on a separate day. We co-hybridized male and female samples for each species on single arrays and subsequently extracted Cy3 and Cy5 channels for the replicates. Pair reports for each sample were generated using Nimblescan and individual channel pair files were loaded together and normalized using quantile normalization (Bolstad et al. 2003. Bioinformatics 19:185) and summarized using Robust Multiarray Averaging (RMA; Irizarray et al. 2003. Biostatistics 4:249) in Nimblescan software. All female and male samples were normalized together but RMA normalization was performed separately for each species.
