Resolution of ntla-dependent transcriptome at 9 hpf using caged molecules

Transcription factors play diverse roles during embryonic development, combinatorially controlling multiple cellular states in a spatially and temporally defined manner. Resolving the dynamic transcriptional profiles that underlie these patterning processes is essential for understanding embryogenesis at the molecular level. Here we show how temporal, tissue-specific changes in embryonic transcription factor function can be discerned by integrating caged morpholinos (cMOs) with photoactivatable fluorophores, fluorescence-activated cell sorting (FACS), and microarray technologies. As a proof of principle, we have dynamically profiled No tail-a (Ntla)-dependent genes at different stages of axial mesoderm development in zebrafish, discovering and characterizing discrete sets of transcripts that are coincident with either notochord cell fate commitment or differentiation. Our studies demonstrate how optically controlled chemical tools can be use to probe developmental processes with spatiotemporal precision and reveal the sequential activation of distinct transcriptomes within a cell lineage by a single transcriptional factor. Zebrafish zygotes were injected with a mixture of ntla caged morpholino (cMO) and caged fluorescein dextran (cFD), and a 100 µm-diameter region of the shield was UV-irradiated at 6 hours post fertilization (hpf). The UV irradiation generated an active morpholino targeting the ntla 5'UTR and simultaneously labeled the cells with green fluorescence. By 36 hpf, the irradiated, green-fluorescent cells contributed to the medial floor plate rather than the notochord, consistent with earlier proposals that Ntla acts as a transcriptional switch between these two cell fates. Combining these caged reagents with microarray analysis identified transcriptional changes coincident with loss of ntla at 9 hpf and subsequent notochord fate respecification. Zebrafish zygotes were injected with a mixture of ntla cMO/cFD, and a 100 µm-diameter region of the shield was UV-irradiated at 6 hpf. Control embryos were injected with cFD and similarly irradiated at 6 hpf. Experimental and control sets of embryos were enzymatically dissociated at 9 hpf, green-fluorescent cells were isolated by FACS, and collected in trizol. Thirty embryos were used for each experiment or controls, each yielding approximately 8,000 green fluorescent cells. Five biological replicates were performed along with five paired controls. Total RNA was isolated from each sample, reverse transcribed, and amplified using WTA2 TransPlex Complete Whole Transcriptome Amplification kit (Sigma) using a miniaturized procedure and manufacturer-recommended incubation steps. The synthesized cDNA was sent to Nimblegen for dye labeling and hybridization. Samples were hybridized to the Nimblegen 2007 (Zv 7) Danio rerio Gene Expression Array chip using one dye per chip. Raw probe data (.pair file) was subjected to RMA, normalization, background correction, and generation of gene expression summary (.calls file) by Nimblegen. Processed data was analyzed in ArrayStar software (DNAStar). Significantly affected genes (fold change > 2 and p-value < 0.10) were identified using a moderated t-test.