| Series | GSE86977 | 
| Title | REGION-SPECIFIC NEURAL STEM CELL LINEAGES REVEALED BY SINGLE-CELL RNA-SEQ FROM HUMAN EMBRYONIC STEM CELLS [Cel-seq] | 
| Year | 2016 | 
| Country | USA | 
| Article | Not set | 
| PMID | NA | 
| Bio Project | BioProject: https://www.ncbi.nlm.nih.gov/bioproject/PRJNA343284 | 
| Sra | SRA: https://www.ncbi.nlm.nih.gov/sra?term=SRP090072 | 
| Overall Desgin | The transcriptomes of 2684 single cells were profiled by CelSeq at different timepoints throughout a 54-day differentiation protocol that converted H1 human embryonic stem cells to a variety of brain cell types. | 
| Summary | During development of the human brain, multiple cell types with diverse regional identities are generated. Here we report a system to generate early human brain forebrain and mid/hindbrain cell types from human embryonic stem cells (hESCs), and infer and experimentally confirm a lineage tree for the generation of these types based on single-cell RNA-Seq analysis. We engineered SOX2Cit/+ and DCXCit/Y hESC lines to target progenitors and neurons throughout neural differentiation for single-cell transcriptomic profiling, then identified discrete cell types consisting of both rostral (cortical) and caudal (mid/hindbrain) identities. Direct comparison of the cell types were made to primary tissues using gene expression atlases and fetal human brain single-cell gene expression data, and this established that the cell types resembled early human brain cell types, including preplate cells. From the single-cell transcriptomic data a Bayesian algorithm generated a unified lineage tree, and predicted novel regulatory transcription factors. The lineage tree highlighted a prominent bifurcation between cortical and mid/hindbrain cell types, confirmed by clonal analysis experiments. We demonstrated that cell types from either branch could preferentially generated by manipulation of the canonical Wnt/beta-catenin pathway. In summary, we present an experimentally validated lineage tree that encompasses multiple brain regions, and our work sheds light on the molecular regulation of region-specific neural lineages during human brain development. | 
| Experimental Protocol | To generate single cell suspensions, hESC-derived cultures were dissociated from plates using Accutase  (ThermoFisher)  at  37°C.    Light  trituration  using  a P1000  pipette  was  done  every  5  min  until  nearly  all clumps  had  been  dissociated  (up  to  1  h).    Cell  suspension  was  washed  and  filtered  through  a  40  μm  cell  strainer.  Cells were washed in PBS with 1% FBS and stained with 0.5-1 μg/mL DAPI.  Single-cell suspensions were loaded onto  a  FACSAria  II  SORP  (Becton  Dickinson)  and  sorted  directly  into  PCR  strip  tubes  or  plates  held  in  chilled aluminum  blocks.    Doublets  and  dead  cells  were  excluded  based  on  forward  scatter,  side  scatter  and  DAPI fluorescence. Sorting was done using the 130  μm nozzle with the sort mode set to single cell. Accuracy of single-cell  sorts  was  confirmed  by  sorting  DAPI-stained  fixed  cells  onto  a  dry  well  of  a  96-well  plate  and  analyzing  by fluorescence microscopy.; CelSeq protocol with a few modifications. Single cells were sorted with a FACSAria (BD) into 96-well plates containing 1.2 μL 2× CellsDirect Buffer (Thermo Fisher) with 0.1 μL of External RNA Controls Consortium (ERCC) control RNAs diluted to 1 × 10-6 molecules (Thermo Fisher). After sorting, plates were then frozen and stored at -80C. For library preparation, plates were thawed on ice. mRNA was reverse transcribed using 1.25 pmol or 0.15625 pmol of oligoT  primer  carrying  a  cell-specific  8  NT  barcode  and  a  5  NT  unique  molecular  identifier  (UMI)  (Islam  et  al., 2014) (see Table S4). Barcode design ensured at least three nucleotide differences from any other barcode. Samples were  incubated  in a  PCR  machine  (Tetrad,  BioRad)  at  70  °C  with a  70 °C  heated  lid  for  3  min,  spun,  and heated again for two more minutes. Samples were reverse transcribed using Superscript III (Thermo Fisher) for two hours at 50 °C with a 52 °C lid and subsequently treated with 1 μL of ExoSAP-IT (Affymetrix). Samples were cooled on ice  for  second  strand  synthesis,  where  Second  Strand  Synthesis  Buffer,  dNTPs,  DNA  Polymerase,  and  RNAse  H (NEB) were added to the samples for a 10 μL total volume and incubated at 16 °C for 2 h. Single cells were pooled by 24 wells per library, with each library containing a water-only well and an ERCC-only well. Single cell pools or population  RNA  libraries  were  purified  with  an  equal  volume  of RNA  Clean  Beads  (Beckman  Coulter),  linearly amplified at 37 C for 15 h using the HiScribe T7 High Yield RNA Synthesis kit (NEB), and treated with DNAse I (Thermo Fisher). RNA  was fragmented using the NEBNext RNA Fragmentation Module (NEB), purified using an equal volume of RNA Clean Beads, and visualized (RNA Pico Kit, Bioanalyzer 2100, Agilent). The RNA fragments were repaired by treating with Antarctic Phosphatase and Polynucleotide Kinase (NEB)  and purified with an equal volume of  RNA Clean Beads. cDNA libraries were made using the NEBNext Small Library Prep Kit according to the  manufacturer’s  protocol,  except  Superscript  III  was  used  for  the  RT  step. Index  primers  were  used  in  PCR amplification.  Libraries  were  purified  using  an  equal  volume  of  RNA  Clean  Beads  and  were  quantified  on  the Bioanalyzer using the DNA High Sensitivity Kit (Agilent). Approximately 160-200 nmol of a pool of libraries were size  selected  to  exclude  species  <180  bp  on  a  2%  Dye-Free  cassette  on  the  Pippin  Prep  (Sage)  and  Speed  Vac concentrated    to    approximately    14    μL.    Libraries    were    then quantified    by    qRT-PCR    using    p5    (5’-AATGATACGGCGACCACCGAGA-3’) and  p7   (5’-CAAGCAGAAGACGGCATACGAGAT-3’)   primers   and visualized (DNA High Sensitivity Kit, Bioanalyzer 2100). Library pools were then sequenced on an Illumina HiSeq  using a custom read1 primer (5’-TCTACACGTTCAGAGTTCTACAGTCCGACGATC-3’) and the Illumina primer HP10. Standard Illumina primers HP12 and HP11 were used for the index read and the transcript read, respectively. PE50 kits (Illumina) were used for sequencing with read lengths of 25 nt, 6 nt, and 47 nt for read1, index, and read2, respectively.
  | 
| Data processing | Reads  were  de-multiplexed  by  CelSeq  index,  allowing  for  one  sequence mismatch.; The  transcript  reads  for  each  cell  were  aligned  to  the  RefSeq  transcriptome  (downloaded  March  2013) using Tophat with default parameters; Unaligned reads were then aligned to the genome using Bowtie (Langmead et al., 2009), followed by alignment to the ERCC spike-in controls.; Remaining unaligned  reads  were mapped  to  the  genome  again  using  GSNAP; Reads mapping  to  exons of  the  same  gene  were  collapsed  by  their  Unique  Molecular Identifier  (UMI); Reads mapped within 1kb of the 3′ end of a gene in the proper orientation are ascribed to that gene; Cells with fewer than 20,000 total cellular UMIs were discarded, and data for all remaining cells was randomly subsampled to 20,000 total cellular UMIs.; Genome_build: hg19; Supplementary_files_format_and_content: UMI_20K.2684.csv: UMI gene counts (subsampled to 20K); Supplementary_files_format_and_content: UMI.2684.csv: UMI gene counts
  | 
| Platform | GPL16791 | 
| Public On | Public on Sep 30 2016 | 
Differential expression gene List between two groups.