Evolution of biting induced Wyeomyia gene expression

This study measured the differential expression of genes from two populations of Wyeomyia smithii in the USA that differ in their propensity to bite over a geographic gradient. Population KC is a northern non-biting population from Maine; population WI is a southern low-frequency biting population from Florida. The mosquitoes were maintained at the University of Oregon under standard rearing conditions and run through two generations to minimize maternal and field effects. Selection for biting was conducted using ~14,000 individuals from the WI population. The environment and protocols used for selection were the same as for KC except biters were removed from their cage and placed into a separate “biting” cage with supplemental males from the same generation of the selected line. All hatch from the biting cage were used to generate the subsequent generations. This protocol was followed until the selected line had undergone 5 generation of selection. Through all generations of selection, including those that were not offered a host, hatch were placed on short days (LL:D=10:14) at 21°C to synchronize each generation and to miigate inadvertant direct selection on development time, generation time, or the timing of reproductive allocation. After adults of a given generation had died, their offspring were transferred to long days and reared to adulthood, as above. During experiments, all animals were in good physical condition. Animal heads were dissected then transferred into 1.5-ml microcentrifuge tubes and shock-frozen in liquid nitrogen. Samples were short-term stored at -80°C. For all experimental conditions, four independent replicates (50 heads each) were analyzed. This GEO record is for three comparisons among the three populations: (EXP1) non-biters in the base population (WI) versus biting individuals in the WI line selected for biting (WIOB) over 5 generations of selection (WI vs WIOB); (EXP2) WIOB versus northern obligate non-biters (KC); (3) WI versus KC (n=4 per treatment group for each population). We compared the gene expression patterns (EXP) for the three populations: EXP1 = WI vs WIOB. EXP2 = WIOB vs KC, EXP3 = WI versus KC using four biological replicates for each population including balanced dye-swaps. All probes including random probes were quantile-normalized across chips, subarrays, samples and replicates.