Sequencing mRNA from cryo-sliced embryos to determine genome-wide spatial expression

[rflrob]

Complex spatial and temporal patterns of gene expression underlie embryo differentiation, yet methods do not yet exist for the efficient genome-wide determination of spatial patterns of gene expression. In situ imaging of transcripts and proteins is the gold-standard, but is difficult and time consuming to apply to an entire genome, even when highly automated. Sequencing, in contrast, is fast and genome-wide, but generally applied to homogenized tissues, thereby discarding spatial information. At some point, these methods will converge, and we will be able to sequence RNAs in situ, simultaneously determining their identity and location. As a step along this path, we developed methods to cryosection individual blastoderm stage Drosophila melanogaster embryos along the anterior-posterior axis and sequence the mRNA isolated from each 60\micron{} slice. The spatial patterns of gene expression we infer closely match patterns determined by in situ hybridization and microscopy, where such data exist, and thus we conclude that we have generated the first genome-wide map of spatial patterns in the Drosophila embryo. We identify numerous genes with spatial patterns that have not yet been screened in the several ongoing systematic in situ based projects, the majority of which are localized to the posterior end of the embryo, likely in the pole cells. This simple experiment demonstrates the potential for combining careful anatomical dissection with high-throughput sequencing to obtain spatially resolved gene expression on a genome-wide scale.

We're also doing an open peer review process on this, and are taking comments at the second link.

http://arxiv.org/abs/1302.4693
http://www.michaeleisen.org/blog/?p=1304