Am9930

E11 embryos were obtained from Runx1-GFP/+ mice mating and AGM cells stained with NG2 Cy3 (1:100, Millipore, ab5320c3), PDGFRβ APC (1:100, Biolegend, 136008), CD31 PECy7 (1:4000, eBioscience, 25-0311-82), CD45 PerCypCy5.5 (1:400, BD Pharmingen, 550994) and ckit BV421 (1:500, BD Horizon, 562609) for 30 min at 4 °C, washed, centrifuged, then resuspended in PBS/FCS/PS for analysis and cell sorting. Cells were purified using BD FACS ARIA Fusion cell sorter (BD Bioscience) and Sytox AAD was used to select viable cells. Cells were sorted and collected directly into 10–20 µl of lysis buffer containing Nuclease-free water (Ambion AM9930), 0.2% Triton and 1/20 RNAse inhibitor (Thermoscientific 10777019). Full-length cDNA was generated from 2.3 µl of this cell lysate using the Smarter2 procedure as described^{75 (link)}. Sequencing libraries were generated from 500 pg of cDNA with Illumina’s Nextera XT sample prep kit (Illumina Inc., U.S.A) and according to the Illumina TruSeq Rapid v2 protocol on the HiSeq2500 with a single read 51 bp and dual 9 + 9 bp index (Illumina Inc., U.S.A). Reads were aligned against the GRCm38 reference using HiSat2 (version 2.0.4). We called gene expression values using GENCODE M19 gene annotation file and the union mode in the Bioconductor Genome Alignments Package (v1.8.1).

Cells were sorted and collected directly into 20μL of lysis buffer containing Nuclease-free water (Ambion AM9930), 0.2% Triton and 1/20 RNAse inhibitor. We generated full-length cDNA from 3.4 μL of cell lysate according to the Smarter2 technology as described (Picelli et al., 2013 (link)). 500 pg of cDNA was tagmented with Illumina’s Nextera XT sample prep kit (Illumina Inc., U.S.A) and converted into sequencing libraries. After pooling, single-read 43bp sequencing was performed on Illumina HiSeq2000 using the Truseq v3 sequencing chemistry (Illumina Inc., U.S.A). TopHat2 version 2.0.10 (Kim et al., 2013 (link)) was used to align the reads to the mouse reference genome (mm10). Gene expression values for RefSeq transcripts were calculated as TPMs using Cufflinks (Trapnell et al., 2012 (link)) (2.1.1). Raw reads were counted with the summarizeOverlaps function with the GENCODE M19 gene annotation using the union mode from the Bioconductor Genomic Alignments package (Lawrence et al., 2013 (link)) (v1.18.1).

Total RNA was extracted from 200 µL of cleared CSF using PAR and the Plasma/Serum RNA Purification Kit (Norgen #55000) (NOR). The isolated RNA was eluted in 100 µL of nuclease‑free water (Ambion #AM9930). Before each RNA extraction, the samples were spiked with cel-miR-39 (2x10^-4 nmoles added) (Invitrogen) (Figure 1).