Microarray general methods.
Total RNA will be extracted by column purification using the Picopure (Arcturus, CA) or Qiagen RNA purification kits, depending on the material. For samples requiring both DNA and RNA, we will use the Qiagen Allprep DNA/RNA kit. Microarray methods are well developed and will be similar to previous protocols we have routinely used (12, 15, 47). Total RNA is linearly amplified once (Ambion MessageAmp II kit; (50 ?g of amplified RNA from 1 ?g of extracted RNA (12)) and labeled using a modified version of the aminoallyl-labeling method in which reverse transcription is performed in the presence of aminoallyl-dUTP. 5 ?g of each amplified RNA sample will be labeled with Cy3 or Cy5 (Amersham). The aminoallyl-labeled cDNA is purified with a modified Qiaquick PCR Purification Kit (QIAGEN), coupled to Cy3 or Cy5 dyes, and hybridized on arrays. Images of the competitive hybridization will be obtained with an Agilent scanner at the core facility at the University of Florida ICBR. Raw data will be analyzed and log transformed using Imagene. Data will be analyzed using ANOVA (6) and BAGEL (36, 37), a versatile, easy to use method that uses Bayesian inference to analyze gene expression data, to determine if there are significant differences in gene expression levels between sample groups. A false discovery rate of 0.05 or less will be used to identify genes with significant changes in expression. We will use readily available software (JMP Genomics, GeneSpring) to perform hierarchical clustering analysis and principal component analysis. We will use online Gene Ontology databases (such as http://david.abcc.ncifcrf.gov/) to determine if there is over-representation of specific gene functional categories in our data sets. Finally, we will confirm interesting candidate genes (approximately 20 per array experiment) using quantitative real-time PCR.
Quantitative real-time PCR (qRT-PCR):
Genes showing significant differences identified by gene microarray will be verified using qRT-PCR of additional individuals from treatment and control conditions. RNA will be extracted and quantified using a ND-1000 Spectrophotometer (NanoDrop Technologies). cDNA is synthesized from 150ng of RNA and gene expression monitored using qRT-PCR with an ABI 7500 (Shoemaker lab) sequence detector and SYBR green detection as described in (15). A dissociation curve and 'no template' control will be used. Housekeeping genes that do not vary in expression levels in the array studies will be used as loading controls (9, 12-15, 47, 48). Primers will be developed using PrimerExpress software (Applied Biosystems). Quantification will be based on a standard curve created from serial dilutions of genomic fire ant DNA, normalized to an internal control. For graphical representation, data will be normalized to one sample, and fold differences in relative expression levels are depicted as mean ? se. Comparisons will be analyzed statistically with either Student's t-test or ANOVA and corrected for multiple comparisons.