The transcriptional network that controls growth arrest and differentiation in a human myeloid leukemia cell line

Cell culture and RNA extraction THP-1 cells are heterogeneous in their ability to differentiate and respond to stimuli such as PMA and LPS 1. To increase the signal to noise ratio in our analyses, the THP-1 cell line was sub-cloned by limit dilution and one clone (#5) was selected for ability to differentiate relatively homogeneously in response to PMA as evidenced by expression of CD14 and CSF-1R quantified by qRT-PCR (Supplementary Fig. 1). These THP-1 cells were frozen in aliquots, and used fresh for all subsequent experiments. All reference to THP-1 cells refer to the cloned line. THP-1 cells were cultured in RPMI, 10% FBS, Penicillin/Streptomycin, 10mM HEPES, 1mM Sodium Pyruvate and 50uM 2-Mercaptoethanol. THP-1 was treated with 30ng/ml PMA (Sigma) over a time-course of 96h. Total cell lysates were harvested in TRIzol reagent (Invitrogen) at each time-point. Undifferentiated cells were harvested in TRIzol reagent at the beginning of the PMA time-course. Total RNA was purified from TRIzol lysates according to manufacturer's instructions. DeepCAGE The preparation of the CAGE library from total RNA was a modification of methods described by Shiraki et al. 2 and Kodzius et al. 3 , adapted to work with the 454 Life Sciences sequencer (described in detail in Detailed Methods). Analysis of deepCAGE: Promoter Construction and Expression Analysis Deep sequencing of CAGE tags was done in triplicate at 0, 1, 4, 12, 24 and 96 hours of PMA treatment for a total of 18 samples. All CAGE tags were mapped to the human genome (hg18) using the program nexalign (T. Lassmann in preparation) by aligning perfectly matching tags first, then those tags that map with a single base pair substitution and finally tags which contain a single insertion or deletion. A filter was applied to remove rRNA-derived tags. Most tags map to a unique genomic location. For tags that map to multiple locations a probabilistic model, previously described by-2-Faulkner et al. 4 , was used to assign weights to each of the possible genomic mappings. The fraction of multi-mapping tags is approximately constant across samples and discarding multi-mapping tags does not affect the expression profiles across promoters (Fig. SM-1).