Differences in the temporal activity of alternative TSSs correspond to distinct patterns of gene expression. (a) In situ expression patterns of genes with alternative TSSs. In situ images showing the spatiotemporal expression of the CG32473, CG10120 (men), and CG1856 (tramtrack) genes during development . (b) Correspondence between time period and developmental stage. As reference, the timing of developmental stages of the Drosophila embryo is matched to a timeline of 1-hour intervals and the Affymetrix 2-hour increment time course. (c) Utilization patterns as measured by the tiling array. The TSSs identified from the most frequent 5' EST ends are listed for each gene. The patterns of peaked promoter utilization detected on the tiling array are noted according to the 12 time points measured during embryonic development. Tiling array data showed that the peaked promoter of TSS#1 was utilized at time points 3, 5, 6 and 11 (hours 4 to 6, 8 to 12, and 20 to 22), TSS#2 at 1 to 9 (hours 0 to 18), and TSS#3 was used at time points 3 to 6 (hours 4 to 12). While the pattern of utilization of the promoter of TSS#1 flipped at time points 4 and 11, the patterns for both TSS#2 and TSS#3 were contiguous. TSS#2 is maternally inherited and the utilization of its promoter extends through early zygotic stages, while the utilization of the others starts after 4 hours and is active for a shorter time. Notably, the peaked promoter of TSS#2 was the only one without a (conserved) INR motif.
Rach et al. Genome Biology 2009 10:R73 doi:10.1186/gb-2009-10-7-r73