Article
the average expression of the two gene sets using the CellCycleScoring
function using Seurat R package. These gene sets should be anticor-
related in their expression levels, and cells expressing neither are likely
to be in the G1 phase (not cycling).
WGCNA analysis in categorizing genes
WGCNA analysis was performed by R package ‘‘WGCNA’’^36 ,^37 (R version
3.4.3, https://cran.r-project.org/src/contrib/Archive/WGCNA; package
version 1.6.6). The WGCNA soft power value was determined by navigat-
ing the soft-threshold-mean-connectivity curve. Modules with <0.25
similarity were merged. Modules correlated with a specific cell subtype
were considered as standard modules for categorizing genes into cer-
tain cell subtypes. Seven modules were selected for neuron subtypes.
ATAC library preparation for high-throughput sequencing
ATAC-seq was performed as described previously^38 ,^39. In brief, a total of
50,000 cells were washed twice with 50 μl of cold PBS and resuspended
in 50 μl lysis buffer (10 mM Tris-HCl pH 7.4, 10 mM NaCl, 3 mM MgCl2,
0.1% (v/v) Nonidet P40 Substitute). The suspension of nuclei was then
centrifuged for 10 min at 500g at 4 °C, followed by the addition of 50
μl transposition reaction mix (10 μl 5 × TTBL buffer, 4 μl TTE mix and
36 μl nuclease-free H 2 O) from the TruePrep DNA Library Prep Kit V2
for Illumina (Vazyme Biotech). Samples were then incubated at 37 °C
for 30 min. DNA was isolated using a QIAquick PCR Purification Kit
(QIAGEN). ATAC-seq libraries were first subjected to five cycles of pre-
amplification. To determine the suitable number of cycles required for
the second round of PCR, the library was assessed by quantitative PCR
as described previously^38 and then PCR amplified for the appropriate
number of cycles. Libraries were purified with a QIAquick PCR Purifica-
tion Kit (QIAGEN). Library quality was checked using a High Sensitivity
DNA Analysis Kit (Agilent). Finally, 2 × 150 paired-end sequencing was
performed on an Illumina HiSeq X-10.
ATAC-seq data analysis
In simple terms, we removed adaptor sequences and then mapped reads
to the hg19 reference genome with the parameters: -t -q -N 1 -L 25 -X 2000
using Bowtie2 (version 2.3.4.3). All unmapped reads, non-uniquely
mapped reads and PCR duplicates were removed. The uniquely mapped
reads were shifted by +4 or −5 bp according to the strand of the read.
To visualize the ATAC-seq signal, we extended each read by 50 bp and
counted the coverage for each base. All the ATAC-seq peaks were called
by MACS2 v2.1.1 with the parameters –nolambda.
ATAC-seq data quality control
ATAC-seq data quality was evaluated for several parameters, including
the number of raw reads, alignment rate, percentage of reads mapped to
chromosome M, percentage of reads mapped to repeat regions (black
list), percentage of reads that passed MAPQ score filter, percentage
of total signal within known artefact regions and correlation between
replications.
Connecting transcription factors to target genes
To find the potential transcription factors that bind the PROX1 regula-
tory sequence (TSS ± 2k), FIMO from MEME Suite (version 5.0.4) was
used for motif enrichment analysis. To investigate the genes that are
regulates by PROX1, the PROX1 motif profile was downloaded from the
Jaspar database (http://jaspar.genereg.net/), and we used FIMO from
the MEME suite for enrichment analysis of our peaks.
Immunofluorescent staining
Tissue samples were fixed overnight in 4% paraformaldehyde, cryo-
protected in 30% sucrose, and embedded in optimal cutting tempera-
ture (Thermo Scientific). Thin 40-μm cryosections were collected on
superfrost slides (VWR) using a Leica CM3050S cryostat. For immuno-
histochemistry, heat-induced antigen retrieval was performed in 10
mM sodium citrate buffer, pH 6. Primary antibodies: mouse anti-CD45
(1:100, Abcam ab8216), goat anti-SOX2 (1:250, Santa Cruz sc-17320), rab-
bit anti-PAX6 (1:500, BioLegend 901301), rabbit anti-NEUROD2 (1:500,
Abcam ab104430), mouse anti-NEUROD1 (1:100, Abcam ab60704), rab-
bit anti-HOPX (1:1,000, Santa Cruz sc-30216), mouse anti-Ki67 (1:100,
BD 550609), mouse anti-SATB2 (1:250, Abcam ab51502), mouse anti-
MEIS2 (1:200, Santa cruz sc-81986), rabbit anti-PROX1 (1:500, Abcam
ab199359), rabbit anti-OLIG2 (1:500, Millipore AB9610), human anti-
MBP (1:1,000, Abcam ab209328), mouse anti-GFAP (1:200, CST 3670S)
diluted in blocking buffer containing 10% donkey serum, 0.5% Triton-
X100 and 0.2% gelatin diluted in PBS at pH 7.4. Binding was revealed
using an appropriate Alexa Fluor 488, Alexa Fluor 594, or Alexa Fluor
647 fluorophore-conjugated secondary antibody (Life Technologies).
Cell nuclei were counterstained using DAPI (Life Technologies). Images
were collected using an Olympus FV1000 confocal microscope.In situ hybridization
The in situ hybridization protocol has been described previously^40.
In brief, probes complementary to target human mRNA used for RNA
in situ hybridization were cloned from primary human fetal cortical
cDNA samples and reverse-transcribed using PrimeScript II 1st Strand
cDNA Synthesis Kit (Takara) with oligo dT primers. Total RNA was iso-
lated from GW27 human hippocampus using SV Total RNA Isolation
System (Promega). Specific genes were amplified using the following
primers: SEMA5A forward AGC TCG CTT GGC TTT AGT CTT A, reverse
CAA AAT AGG CTT TGA CTC CCA C; PID1 forward TGG GAT CTC TAG
TGG GGT GG, reverse TAA GGC TTC TTA GGT GCC GC; SULF2 forward
GTT TGA CAT CAG GGT CCC GT, reverse CTT TAA TGG GGT TGG CGG
CT; NRIP3 forward AGC TGT GGT TGA TGA CAA TGA G, reverse CTG
TAA TGG ATA ATG TCC CTG G; STX10 forward GGG GAA GGG ACT GAC
ATG TC, reverse GGA GGG CTG GGG TCA GAG AG; CHMP4A forward
GAT TGG GCA AGG CTG GTC CC, reverse TTG GGA GCT GGC CCT GCC
GG; BEX5 forward TCA ACA TGG AAA ATG TCC CC, reverse AGA CTG
CTT TTA AAT TGC TT; NBPF1 forward GGG TGC ACC AAG AGC AGC
CT, reverse CCT CAG CAT AAA TTT TAT GA; CASC15 forward CAA GCA
TGT AGC CCT GCC CG, reverse CTC TGT TTC TGT CAT CTC TC; primers
specific to target genes of interest were designed using Primer3 and
amplified by PCR using Q5 High-Fidelity DNA Polymerase (NEB). PCR
products of predicted band size were gel extracted and ligated into the
Hieff Clone Plus One Step Cloning Kit (Yeason). Ligation products were
transfected into Trans5α Chemically Competent E. coli (Transgene).
Cloned sequences were confirmed by sequencing. Digoxigenin-labelled
RNA probes for in situ hybridization were generated by linearizing
the pSPT18 Vector and in vitro transcribing the probe using T7 or SP6
RNA Polymerase (Roche) in the presence of DIG-RNA Labelling Mix
(Roche). Fetal brain sections of 30μm thickness were hybridized with
RNA probes at a final concentration of 500 ng/ml overnight at 64.5 °C
in hybridization solution (50% formamide, 10% dextran sulfate, 0.2%
tRNA (Invitrogen), 1 × Denhardt’s solution (Sigma) and 1 × salt solution
(containing 0.2 M NaCl, 0.01 M Tris, 5 mM NaH 2 PO 4 , 5 mM Na 2 HPO 4 ,
5 mM EDTA pH 7.5)) overnight. After the sections were washed, alkaline
phosphatase-coupled anti-digoxigenin Fab fragments (Roche) were
applied. For visualization of the labelled cRNAs, the sections were incu-
bated in the dark in NBT/BCIP solution (Roche). Images were taken
using a Leica SCN400 (Leica Microsystems).Plasmids and in utero electroporation
NBPF1 genes were cloned into a pEGFP-C1 vector. Electroporation was
performed as previously described^41. In brief, timed pregnant CD-1 mice
(E13.5) were deeply anaesthetized with isoflurane, and the uterine horns
were exposed through a midline incision. 1 μl of plasmid DNA (1–2 μg/
μl) mixed with Fast Green (Sigma) was manually microinjected into
the fetal brain lateral ventricle through the uterus, using a bevelled
and calibrated glass micropipette (Drummond Scientific) followed
by five 50-ms pulses of 50 mV with a 1 s interval delivered across the