Our screening samples are from Yunnan province of southwestern China (n
1,013). Replication samples included samples from the Alzheimer’s Disease Neuroimaging Initiative (ADNI; n
204), samples from the US National Institute of Health (CBDB/NIMH; n
188) and samples from the Dutch Brain Imaging Genetics study (BIG; n
Screening Samples: Chinese Samples
The detailed information of the sample screening was described in our previous study 
. Briefly, a total of 1,013 unrelated healthy individuals including 460 males and 553 females were included. The identities of the subjects were self-declared and confirmed by their written ID profiles. All the sampled individuals are from Yunnan province of southwestern China. Written informed consents for this study were obtained from all the subjects, and the research protocol was approved by the internal review board of Kunming Institute of Zoology, Chinese Academy of Sciences. The ages of the 1,013 individuals range from 19 to 28 years with 98% of them being 21–26 years old.
Replication Samples: ADNI Sample
The MRI and genotyping data in this replication sample were obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database (adni.loni.ucla.edu). One goal of ADNI has been to test whether serial magnetic resonance imaging (MRI), positron emission tomography (PET), other biological markers, and clinical and neuropsychological assessment can be combined to measure the progression of mild cognitive impairment (MCI) and early Alzheimer’s disease (AD). For up-to-date information, see www.adni-info.org
. The ADNI participants consist of patients with AD, patients with MCI, and elderly healthy individuals. They were aged 55–90 years and recruited from 59 sites across the U.S. and Canada. Written informed consent was obtained from all 822 participants and the study was conducted with prior Institutional Review Board approval. Of 822 participants, 204 unrelated non-Hispanic Caucasian healthy controls were used in this study 
Replication Samples: CBDB/NIMH Sample
All subjects are the healthy control participants of Clinical Brain Disorders Branch/National Institute of Mental Health Sibling Study, a study aimed at identifying schizophrenia susceptibility genes and related intermediate biologic phenotypes 
. Subjects with good quality of structural data and genotyping were included the study.
Replication Samples: BIG Sample
In this study, a total of 486 healthy control subjects aged 18–35 years from the Brain Imaging Genetics (BIG) study at the Donders Institute for Brain, Cognition and Behaviour of the Radboud University Nijmegen (Medical Centre) were included. The BIG study is a study of self-reported healthy individuals included into earlier imaging studies at the Donders Centre for Cognitive Neuroimaging. Subjects are of European Caucasian descent and generally highly educated 
. The study was approved by the regional medical ethics committee (CMO regio Arnhem/Nijmegen) and all participants provided written informed consent prior to participation.
Measurement of Cranial Volume: Chinese Screening Sample
The cranial volume was measured and calculated as described in our previous study 
. Three principal dimensions of the cranium were measured including 1) Maximum antero-posterior length (L, measured between glabella and the inion). 2) Maximum breadth (B, biparietal diameter; measured between two parietal eminences). 3) Cranial height (H, basi-bregmatic height, measured between the internal acoustic meatus to the highest point of the vertex). Then the cranial volumes were computed using the following formula 
: Male, 0.337 (L-1.1) (B-1.1) (H-1.1) +406.01 cc; Female, 0.400 (L-1.1) (B-1.1) (H-1.1) +206.60 cc.
Measurement of Brain Volume: Replication Samples (ANDI)
3D T1-weighted brain MRI scans were acquired using a sagittal 3D MP-RAGE sequence following the ADNI MRI protocol 
. Baseline 1.5T MRI scans from 204 participants were downloaded from the ADNI public website (http://www.loni.ucla.edu/ADNI/
) onto local servers at Indiana University School of Medicine. As detailed in previous studies 
, FreeSurfer V4 software (http://surfer.nmr.mgh.harvard.edu/
), a widely employed brain segmentation and cortical parcellation tool, was used to label cortical and subcortical tissue classes using an atlas-based Bayesian segmentation procedure and to extract the measure of brain volume.
Measurement of Brain Volume: Replication Samples (CBDB/NIMH)
All structural MRI were acquired on a 1.5 Tesla GE scanner (GE Medical Systems, Milwaukee, Wisconsin) using a T1-weighted spoiled gradient recalled (SPGR) sequence (repetition time, 24 ms; echo time, 5 ms; number of excitations, 1; flip angle, 45°; matrix size 256×256; FOV 24×24 cm2
), with 124 sagittal slices (0.94×0.94×1.5 mm3
resolution). Images were processed using the FreeSurfer 
toolbox (version 5). Total Brain Volume (TBV) and Total Grey Matter volume (TGM) measurements were calculated as previously described 
. TGM was defined as sum of tissue probabilities for the grey matter region. TBV was defined as the sum of total gray matter volume, total white matter volume, and cerebrospinal fluid.
Measurement of Brain Volume: Replication Samples (BIG Sample)
Subjects were scanned at 3 Tesla (n
486) MRI scanners and T1-weighted structural magnetic resonance imaging data (3D MPRAGE) were acquired (more information on the image acquisition can be found in our previous study 
). All scans covered the entire brain and had a voxel-size of 1×1×1 mm3
. To calculate total brain volume, raw DICOM MR imaging data were converted to NIFTI format using the conversion as implemented in SPM5 (http://www.fil.ion.ucl.ac.uk/spm/software/spm5/
). Normalizing, bias-correcting, and segmenting into gray matter, white matter, and cerebrospinal fluid was performed using the VBM toolbox (VBM5.1 Toolbox version 1.19) in SPM using priors (default settings). This method uses an optimized VBM Protocol 
as well as a model based on Hidden Markov Random Fields (HMRF) developed to increase signal-to-noise ratio 
. Total volume of gray matter, white matter, and cerebrospinal fluid was calculated by adding the resulting tissue probabilities. Total brain volume was defined as the sum of white matter and gray matter volume.
SNP Selection, SNP Tagging, Genotyping and Sequencing
SNP selection was based on the previous association studies including our recent data 
. We focused on the four genes that identified by Chen et al.
recently using systematically mapping in large independent samples. In addition, our recent data and LD in Chinese were also considered. We selected 8 SNPs for the initial screening (rs3756295, rs40396, rs1291602, rs31251, rs1355095, rs2240525, rs3914025, rs31400), additional SNPs were included according to the association significance and whether they are tagging SNPs. Totally, we selected 20 SNPs for fine mapping. The 20 SNPs were genotyped using the SNaPshot method (Applied Biosystems). We sequenced the IL3 gene (including the 500 bp upstream and downstream, respectively) in 150 randomly selected individuals through direct sequencing. The conservation analyses were performed by using UCSC genome browser 
Prediction of DNA-binding Motifs
We used Dragon ERE finder 
, a web-based program for identification and interactive analyses of estrogen response elements (EREs) to predict EREs in the upstream region of IL3. AliBaba (http://www.gene-regulation.com/pub/programs/alibaba2/index.html
) was used to predict and compare DNA-binding motifs in the promoter region with alternative alleles.
Cell Culture, Treatment, and RNA Extraction
K562 cells were routinely cultured in DMEM (Gibco) supplemented with 10% FBS (Hyclone), 100 u/ml penicillin and 100 ug/ml streptomycin. Before treatments, the cells were maintained in phenol red-free DMEM containing 10% dextran-coated charcoal-stripped fetal bovine serum (DCC-FBS) (Hyclone) for a minimum of 3 days with the media changed every day. Cells were treated with 10 nM 17-beta-estradiol (E2) (Sigma) for 2 to 24 hours. Total RNA was harvested and prepared using TRIzol (Invitrogen) following the manufacturer’s instructions.
Quantitative Real-time PCR
Reverse transcription PCR (RT-PCR) was performed using the Omniscript RT Kit (Qiagen) following the manufacturer’s instructions. We carried out real-time quantitative PCR using gene specific primers, and the fold change in expression was calculated using the ΔΔCt (threshold cycle) method. The GAPDH was used as the internal control.
EMSAs were performed with a Lightshift™ chemilumescent EMSA kit (Pierce). The single-strand oligonucleotides were biotinylated with Biotin 3′ End DNA labeling Kit (Pierce) and then annealed to form double strands. The nuclear extracts of MCF-7 and U2OS were prepared by CelLytic™ NuCLEAR™ Extraction kit (Sigma). HeLa nuclear extracts were purchased from Santa Cruz Biotech. The binding reactions were performed for 20 mins at room temperature in 10 mM Tris-HCl (PH 7.5), 1 mM MgCl2, 0.5 mM EDTA, 0.5 mM DTT, 50 mM NaCl, 50 ug/ml poly (dI-dC)(dI-dC) and 4% glycerol, 35 fmol biotin 3′-end -labeled double-stranded oligonucleotides, and purified recombinant SP1 protein (Alexis) or nuclear extracts. After incubation, samples were separated on a native 6% polyacrylamide gel and then transferred to a nylon membrane. The positions of biotin end-labeled oligonucleotides were detected by a chemilumescent reaction with streptavidin-horseradish peroxidase according to the manufacturer’s instructions and visualized by autoradiography. For competition assays, we pre-incubated 100-fold excess of unlabeled oligonucleotide probe with SP1 or nuclear extracts before adding the biotin-labeled probe. The nucleotide sequences of the double-stranded oligonucledtides with either T or C allele are:
Promoter Cloning and Reporter Gene Assays
To construct IL3 promoter, we amplified fragments encompassing nucleotides -436 to +164 (relative to the ATG start codon at +1) of IL3 by PCR from genomic DNA of two individuals homozygous with respect to the corresponding genotypes (TT and CC) for rs31480, using primers tailed with Xhol and HindIII restriction sites, and directionally subcloned them into the Xhol and HindIII sites of the pGL3-Basic expression vector (Promega). We verified all recombinant clones by bi-directional DNA sequencing. HeLa, COS-7, CHO, and SK cells were routinely cultured in DMEM supplemented with 10% FBS with antibiotics. The cells were plated at 2.5×105 cells per well in a 24-well plate the day before transfection and incubated overnight at 37°C in 5% CO2. Transient transfection assays were conducted in these cells using the Lipofectamine 2000 transfection reagent (Invitrogen), all assays were performed in at least three independent experiments with minimum of three replicates. The reporters containing either T allele or C allele were transfected into these cells together with a Renilla luciferase control vector. After 24h incubation, we collected the cells and measured luciferase activity using the Dual-Luciferase Reporter Assay System (Promega).
3ERE Cloning and IL3 Activation Assays
Three repeats of Estrogen Response Elements (EREs) tailed with Xhol and HindIII restriction sites were synthesized and annealed to form double-stranded nucleotides. The sequence is 5′-CCG CTCGAG TA GGTCA GCG TGACC TA TA GGTCA GCG TGACC TA TA GGTCA GCG TGACC TA AAGCTT GGG-3′. We directionally cloned it into the pGL3-Basic vector after restriction enzyme digestion. We confirmed the construct by sequencing. The estrogen receptor alpha and beta vectors were kindly provided by professor Sylvie Mader (Faculté de Médecine, Université de Montréal) and Leigh C. Murphy (University of Manitoba). MEG-01 and HEK293T, IL3 receptor expression positive cell lines were cultured in PRMI 1640 and DMEM respectively supplemented with 10% FBS, 2 mM L-glutamine, 1 mM Sodium pyruvate and 1% antibiotics. Co-transfection of 3ERE-pGL3 (2 ug), ERα (1 ug) or ERβ (1 ug) were performed by using a Nucleofector Device from Amaxa Biosystems (Lonza Cologne AG, Germany) in MEG01 cell line, while co-transfection in HEK293T cell line were used Lipofectamine 2000 transfection reagent (Invitrogen) as previously described, and with pRL-TK as the internal control. After six hours incubation, 17-beta-estradiol and recombinant human IL3 protein (Invitrogen) were added into the medium with a final concentration of 10 nM. Cells were harvested and their luciferase activity was measured after additional 24 h incubation. All assays were performed in at least three independent experiments with a minimum of triplications.
Hardy-Weinberg equilibrium of each SNP was assessed by using GENEPOP (v 4.0) 
. Association of single SNP with total brain volume (TBV) and their additive effects on this quantitative trait were tested by utilizing PLINK or SAS statistical software using the linear regression option, with age, sex and IQ (optional) as covariates; for the analyses on total gray/white matter, TBV was also considered as a covariate 
. To account for sex-specific effects, we used a statistical model where the mean effect of SNP dose on the phenotype was allowed to differ for the two sexes. The p-value was adjusted by the conservative Bonferroni correction according to the number of independent SNPs and the divided internal samples separated by sex. We used the Haplo Stats 
to infer the haplotype frequency and to perform the haplotype association test. We used the Haploview 
to calculate pairwise LD indices r2
and D’, to define LD blocks and to select the tag SNPs. Haplotypes were inferred with the PHASE program by the Bayesian statistical methods based on the genotype data 
. Sequence alignment and assembly were conducted by DNASTAR software package. The analysis of quantitative PCR data was based on the ΔCt
The C57BL/6J mice were used in this paper and all animal procedures described herein were approved by the University Committee of Animal Resources at the University of Rochester. For the purposes of staging embryos, noon of the day a vaginal plug was detected was taken to be embryonic day 0.5 (E0.5). Mice were deeply anesthetized, perfused transcardially with PBS, followed by 4% PFA in PBS, pH 7.3. Then the brains were dissected, postfixed in 4% PFA in PBS at 4°C overnight, washed three times in PBS, and cryoprotected in 30% sucrose in PBS before rapid freezing in OCT compound (TissueTek). For antigen retrieval, cryosections (20 µm) were heated in 10 mM citrate buffer (pH 6.0) at 95°C for 10 min. The sections were permeabilized and blocked in PBS plus 0.1% Tween-20, 5% horse serum, and incubated with primary antibody overnight at 4°C, washed in PBS three times and incubated with fluorescently labeled secondary antibody for 1 h at room temperature. The primary antibodies used were mouse monoclonal anti-IL3RA (Santa Cruz, 1
100), rabbit polyclonal anti-IL3RA (Santa Cruz, 1
100), rabbit polyclonal anti-IL3RB (Santa Cruz, 1
100), goat polyclonal anti-IL3 (Santa Cruz, 1
100), rabbit monoclonal anti-β-III tubulin (Tuj1) (Covance, 1
1000), mouse anti-NeuN (Chemicon, 1
300), goat anti-SOX2 (Santa Cruz, 1
500), rabbit anti-Nestin (Abcam, 1
300), rabbit anti-TBR2 (Millipore, 1
200), rabbit anti-pH3 (Santa Cruz, 1
200), rabbit anti-PROX1 (Covance, 1
1000), anti-estrogen receptor β (Santa Cruz, 1
100), rabbit anti-Ki67 (Novocastra, 1
1000), rabbit anti-GFAP (1
2000). Secondary antibodies used were Alexa Fluor 488 (Invitrogen, 1
500), 546 (Invitrogen, 1
500) conjugated to donkey anti-mouse, rabbit or anti-goat (Invitrogen). DNA was stained with 4′,6-diamidino-2-phenylindole (DAPI; Molecular Probes). Images were acquired with a Zeiss laser confocal microscope and analysed with LSM 510 software (Carl Zeiss).
For neural progenitor’s culture, cerebral cortices from C57BL/6 mice embryos (E12.5–14.5) were dissected in HBSS solution (Invitorgen), the meninges and other parts were removed under dissecting microscope and only the cortices were retained. After several washes with HBSS, the cortices were minced and dissociated mechanically with trituration, filtered through a 70 µm cell strainer (BD Falcon). Then, the cells were culture in neurobasal medium (Invitrogen) or DMEM/F12 (Millipore) with different supplements and growth factors according to experimental requirements. For neurons culture, the procedures are same with neural progenitors except the age of mice embryos (E14.5–17.5).
For generation of neurospheres, the dissociated cells from E14.5–17.5 mice cortices were cultured for 2–5d in the medium containing neurobasal medium, B27 (Invotrogen) and N2 (Invitrogen) supplements, 100 U/mL penicillin, 100 µg/mL streptomycin, FGF2 (10 ng/ml), and EGF (10 ng/ml). For monolayer cultures, the generated neurospheres were collected and spun down (200 g for 5 minutes), then triturated with a Pasteur pipette to obtain single cells. The single cell suspensions were replated onto poly-L-lysine (50 µg/ml)/laminin (10 µg/ml)–coated Lab-Tek Chamber Slide (Thermo Scientific). Cells were treated with different concentrations of recombinant mouse IL3 (10 and 100 ng/ml) and cultured in above medium. Untreated cells served as control. After 24 or 48 hours culture, cells were fixed with 4% PFA and subjected to mmunohistochemistry. Immunostaining and morphometry were carried out to assess the numbers of proliferating progenitors (Ki67 and pH3 positive). More than 4 random microscopic fields (20×) were analyzed and about 8000 cells were counted for each condition.
Assessment of Differentiation Markers in vitro
To investigate whether IL3 can drive neural differentiation, we used quantitative real-time PCR method described previously 
to evaluate the effects of IL3 on neural differentiation. First, neural progenitors were cultured in neurobasal medium (containing B27 supplements, N2 supplements,10 ng/ml FGF2 and 10 ng/ml EGF) for 2 days, then FGF2 and EGF was removed and replaced by neurobasal medium with 1% serum (vol/vol), B27 supplements, N2 supplements and IL3 (10, 100, and 200 ng/ml). Four days after addition of recombinant mouse IL3 (Invitrogen), cells were harvested for the RNA extraction. Untreated cells served as control. RNA was isolated using Trizol reagent (Invitrogen) according to the manufacturer’s instructions and treated with DNase I (Fermentas). cDNA was synthesized from 3 µg total RNA using oligo-dT primers and Superscript III Reverse Transcriptase (Invitrogen). Quantitative PCR was performed by using the Bio-Rad iCycler & iQ Real-Time PCR Systems. The following primer pairs were used: mouse Sox2-F, GCGGAGTGGAAACTTTTGTCC
, mouse Sox2-R, CGGGAAGCGTGTACTTATCCTT
; mouse Nestin-F, CCCCTTGCCTAATACCCTTGA
, mouse Nestin-R, GCCTCAGACATAGGTGGGATG
; mouse β-III-tub-F, TAGACCCCAGCGGCAACTAT
, mouse β-III-tub-R, GTTCCAGGTTCCAAGTCCACC
; mouse GFAP-F: CCCTGGCTCGTGTGGATTT
, mouse GFAP-R: GACCGATACCACTCCTCTGTC
; mouse ENO2-F: GTCCCTGGCCGTGTGTAAG
, mouse ENO2-R: CATCCCGAAAGCTCTCAGC
; mouse nestin-F: CCCTGAAGTCGAGGAGCTG
, mouse nestin-R: CTGCTGCACCTCTAAGCGA
; mouse PLP1-F: TGAGCGCAACGGTAACAGG
, mouse PLP1-R: CCCACAAACTTGTCGGGATG
. The iCycler PCR analysis was performed using the SYBR Green master mix, according to the manufacturer’s recommendations (BioRad). The specificity of product was ensured by melting curve analysis and agrose gel electrophoresis. cDNA content of samples was normalized to the expression of GAPDH.
Cell Viability Assays
To investigate whether IL3 has protective or trophic effects on neural progenitor’s survival, cerebral cortices from E12.5 or E13.5 mice embryos were dissociated and the isolated cells (5×104) were plated onto poly-L-lysine coated 96-well plate (Corning). The survival of neural progenitors was investigated by three culture conditions. Firstly, we studied the neurotrophic effects of IL3 on neural progenitors through culturing the progenitors in the absence of any factor (with neurobasal medium only) and in the presence of IL3 (0.02–50 ng/ml). Recombinant mouse IL3 was added into the medium after the culture initiated two hours. After 36 hours incubation, cell viability was determined by measurement of cellular ATP levels (CellTiter-Glo Luminescent Cell Viability Assay, Promega). Secondly, neural progenitors were first cultured in neurobasal medium supplemented with B27 and Glutamax (Invitrogen). On day 2 of culture, the medium was replaced with serum-free neurobasal medium containing N2 supplement and different concentrations of mouse IL3 (0.1–20 ng/ml). The cultures were maintained for 3 days and cell viability was measured. Thirdly, neural progenitors were first cultured in neurobasal medium supplemented with B27 and Glutamax. On day 2 of culture, IL3 was added into the medium and the cultures were maintained for 3 days and cell viability was measured. For studying the effects of IL3 on neurons, cerebral cortices from E16.5–E18.5 mice embryos were dissociated and cultured in DMEM/F12 medium with 5% FBS. On day 2 of culture, the medium was replaced with serum-free DMEM/F12 containing N2 supplement and different concentrations of IL3 (0.1–20 ng/ml). The cultures were maintained for 2 or 3 days and cell viability was determined as described above.
Neural progenitors from E13.5 mice were first cultured in neurobasal medium under proliferating condition (containing B27 supplement, Glutamax, 10 ng/ml FGF2 and EGF) for one week. To exclude the interference of other factors, the supplements, FGF2 and EGF were removed from the medium for about 16 hours prior to IL3 treatment (3 ng/ml). Proteins from MEG01 cells (IL3 treated) and neural progenitors were homogenized in RIPA lyses buffer (Cell signaling) containing a cocktail of protease inhibitor (Sigma Chemical, MO, USA) and phosphatase inhibitor (Cell Signaling). Proteins were quantified by BCA method (Pierce). Extracted protein (40 µg) was separated by SDS-polyacrylamide gel electrophoresis and transferred to PVDF or Nitrocellulose membrane by electrophoretic transfer. The membrane was blocked, incubated with primary antibodies for overnight at 4°C, washed three times with TBST, and then incubated with secondary antibody for 1 hour at room temperature. Antibodies used in western blot are as follows: Rabbit anti-phospho-AKT (Thr308) (Cell Signaling), Rabbit anti-AKT1 (Cell Signaling), Rabbit anti-phospho-JAK2 (Tyr 1007/1008) (Cell Signaling), Rabbit anti-JAK2 (Cell Signaling), Rabbit anti-phospho-GSK3β (Ser9) (Cell Signaling), Rabbit anti-phospho-ERK1/2 (Cell Signaling), Rabbit anti-ERK1/2 (Cell Signaling), Rabbit anti-GSK3β (BD), and Rabbit anti-actin (Abcam). Immunoreactivity was detected with an enhanced chemiluminescence system (Pierce, IL, USA) with colored markers (Fermentas) as molecular size standards.