As a C. elegans embryo develops from one cell to 558 cells, the transcriptional profile of each cell undergoes unique and dramatic changes to drive differentiation into distinct cell-types and tissues. While gene expression has been measured in isolated cell-types (Meissner et al, 2009; Spencer et al, 2011; Burdick et al, 2016), the data was primarily generated from single time point samples. Conversely, a batch whole embryo time series (Boeck et al, 2016) provides excellent temporal data from early embryo to late, but is a blend of the gene expression from all cell types present in the embryo, thus lacking cell-type specificity. We have created a more comprehensive dataset with gene expression levels for all genes, not only in distinct isolated cell-types from synchronized populations of those cells, but at discrete time points during embryonic development using high depth RNA-seq analysis. We used FACS to isolate fluorescent and non fluorescent cells from synchronized embryos containing early and highly cell-type specific tissue/lineage markers - determined by cell lineaging using 4-D microscopy as part of the Expression Patterns in Caenorhabditis (EPiC) project
(http://epic.gs.washington.edu/Epic2) (Murray et al, 2012). To cover major lineages and cell types, we have created RNA-seq libraries with highly correlated replicates for the ABa sublineage (
tbx-37), muscle (
hlh-1), intestine (
end-1), pharynx (
pha-4, excluding
end-1), hypoderm (
nhr-25), neurons (
cnd-1), and head neurons (
ceh-32). Each 5 point time series begins at the earliest expression of the marker gene with time points every 90 min. We observe dramatic differences in gene expression both between cell-types, and over time within the same cell population including differential splice isoform usage. We have determined which genes have broad expression, which have biased expression in only one or a few cell-types, and the order of transcription factor (TF) activation in each cell-type. We then used this approach to assay expression changes after TF loss by focusing on the muscle specific and essential TF
hlh-1. We crossed a
phlh-1::mCherry marker into both WT and mutant
hlh-1(ts) strains, synchronized populations, and shifted to non-permissive temperature at the YA stage. Using FACS, we isolated a time series of only
hlh-1 expressing embryonic cells (~10% of all cells), eliminating noise from non-muscle cells (~90% of all cells) unaffected by loss of
hlh-1. In the
hlh-1 mutant hundreds of genes have significantly reduced expression, most of which normally have muscle enriched expression in WT. When the most affected genes are compared with HLH-1 targets identified by ChIP-seq (Araya et al, 2014), a majority have ChIP-seq peaks indicating direct regulatory relationships. We continue to assay additional genes in this manner in an effort to gain insight into the transcriptional regulatory relationships in the major cell lineages.