Supplementary Material:
MATERIALS AND METHODS
Hemgn promoter isolation and sequence analysis ―A mouse BAC genomic library
CITB-CJ7-B (Research Genetics, Catalogue # 96021) was screened by PCR using
Hemgn-specific primers 5’-AAA CAC ACC TCT CTC CTA CCA C and 5’-CCT ACT TTC
TGG GCT CCT TCT G. One of the BAC clones (# 567P14) contained the full length
Hemgn gene. Using clone #567P14 as template, a Hemgn promoter spanning 3171 bp
(positions -2975 to +196) was amplified with pfu Turbo DNA polymerase (Stratagene,
CA) in a PTC-200 thermal cycler. The Hemgn promoter was confirmed by DNA
sequencing. The sequence homology comparison of human and mouse Hemgn genes
was performed using VISTA (http://dcode.org) (Fig. 1B)
11.
The sequences of mouse
and human proximal promoters and the first exons were aligned using ClustalW
(http://www2.ebi.ac.uk/clustalw).
The
MatInspector
program
(http://www.gene-
regulation.com/) was used to identify putative hematopoietic-specific transcription factor
binding sites in the sequence (Fig. 1C).
Cell culture, transfections, and reporter assays ―To determine that the Hemgn
promoter is specially activated in hematopoietic cells, we tested the Hemgn promoter
activities in several cells lines originated from different tissues.
10T1/2 (a mouse
fibroblast cell line), PAC1 (a rat smooth muscle cell line), COS-7 (a monkey kidney cell
line) and C2C12 (a mouse skeletal muscle cell line) were cultured in DMEM medium
with 10% fetal bovine serum. One g of indicated Hemgn promoter-luciferase reporter
construct was cotransfected into the cells in 6-well plates using Lipofectamine Plus
1
reagent (Invitrogen) with 0.2 g of pRL-SV40 Renilla Luciferase reporter as an internal
control. Cells were harvested 24 hours after transfection and luciferase activities were
measured.
K562 cells (a human erythroleukemia cell line) were cultured in RPMI 1640
medium with 10% fetal bovine serum (Invitrogen) in 5% CO2 incubator at 37oC. Four μg
of plasmid DNA of the Hemgn promoter-luciferase reporter constructs (see below) were
transiently transfected into K562 cells in 6-well plates using DMRIE-C liposome reagent
(Invitrogen) along with 200 ng pRL-SV40 Renilla Luciferase reporter as internal control.
Cells were harvested 2 days after transfection. Luciferase activities were measured
using Dual-LuciferaseTM reporter assay system (Promega, Madison, WI, USA).
Drosophila Mel-2 cells (D. Mel-2, a Schneider S2 insect cell line) were purchased
from Invitrogen (Carlsbad, CA) and maintained in Schneider’s insect medium
supplemented with 10% fetal bovine serum and 2 mM L-glutamine plus 100U/ml
penicillin and 100µg/ml streptomycin at 28 oC. D. Mel-2 cells were cotransfected with 1
g of the Hemgn-luciferase reporter gene construct (pHemgn-2975+196luc) and GATA1
(50 to 400 ng pPacGATA1)
12
using FugeneTM 6 reagent (Roche Diagnostics
Corporation, Indianapolis, IN). Cells were harvested after 24h for luciferase assays using
the Single Luciferase Assay System (Promega). Luciferase activities were normalized to
total cell protein, measured by the Bio-Rad protein assay system.
CMS, an acute myeloid leukemia cell line [established from a 2 year old girl with
acute megakaryocytic leukemia (AMkL)]
13
was a gift from Dr. A. Fuse (National Institute
of Infectious Diseases, Tokyo, Japan). The CMS cell line was maintained in RPMI 1640
containing 10% heat-inactivated iron-supplemented calf serum (Hyclone Labs), 2 mM L-
2
glutamine, 100 U/ml penicillin and 100 µg/ml streptomycin in a humidified atmosphere at
37oC in the presence of 5% CO2/95% air. The AMkL cell line, Meg-01, was obtained
from the American Type Culture Collection (Manassas, VA) and cultured in RPMI 1640
with 10% fetal bovine serum as instructed14. The CMS cells were stably transfected with
pcDNA3-GATA1
15
by electroporation (200 V, 1000 µF). Forty-eight hours after
transfection, the cells were plated in 96-well plates with a density of 1 cell/well containing
1mg/ml G418 (Sigma). Single colonies of stably transfected cells were selected after 3
weeks, expanded, and screened for GATA1 expression by real time RT-PCR (see
below).
Construction of Hemgn promoter deletion and mutant constructs ―Deletion
mutants of the Hemgn promoter were prepared by high-fidelity PCR using pfu Turbo
DNA polymerase (Stratagene, La Jolla, CA, USA). The primers were designed to
generate a BamH I site and an Xho I site at the 5’ and 3’ ends of the PCR products
respectively. This allowed the directional cloning of the PCR fragments into the
luciferase reporter vector pXP1
16.
The constructs were verified by sequencing. The
primers for amplifying the corresponding Hemgn promoter and its deletion mutants
were: pHemgn-2975+196: 5’-CGC GGA TCC CAC ATC AGA GAC ACC TTG CC and 5’CCG CTC GAG GGT ATT GGC TTT GAC TTC AC; pHemgn-831+196: 5’-CGC GGA TCC
TTG AAC TAG GGT GGC TCT GG and 5’-CCG CTC GAG GGT ATT GGC TTT GAC
TTC AC; pHemgn-404+196: 5’-CGC GGA TCC AAC AGC CTA CCT AGG AAG AG and 5’CCG CTC GAG GGT ATT GGC TTT GAC TTC AC; pHemgn-2975+12: 5’-CGC GGA TCC
3
CAC ATC AGA GAC ACC TTG CC and 5’-CCG CTC GAG ACA CTG CAC AGG TGT
GAG GG.
The core GATA binding sequences in the plasmid pHemgn-2975+196luc construct
were mutated to cATA, a change reported to abolish GATA binding
17,18,19,
using the
QuickchangeTM site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA). The
primer for GATAbox1 mutation was: 5’-CCT CAC ACC TGT GCA GTG TcA TAA AGA
AAG TGG. The primer for GATAbox2 mutation was: 5’-GCT GTG GTC TAA CCA cAT
AAA ACT TTT AGG CGG G. Double mutations of both boxes were also generated. The
primer for GATAbox3 mutation was: 5’-GTGTTGGTCTTaAgcTTTTCAA AAAGC. The
mutant constructs (named as
mutGATA
box1+2luc)
mutGATA
box1luc,
mutGATA
box2luc,
mutGATA
box3luc
and
were re-cloned into pXP1 vector (to eliminate possible mutation in the
vector) and were verified by sequencing.
Generation and analysis of Hemgn promoter in transgenic mice―The Hemgn
promoter (pHemgn-2975+196) was linked to a lacZ reporter. The DNA fragment containing
the pHemgn-2975+196promoter-linked lacZ was isolated and used to generate transgenic
mice (Cold Spring Harbor Transgenic Core Facility). The transgenic mice were PCRgenotyped using lacZ-specific primers (forward: 5’-GTGGTGGTTATGCCGATCGC;
reverse: 5’-TACCACAGCGGATGGTTCGG).
RNAs from embryos and tissues were isolated using Trizol (Invitrogen) and the
expression of the lacZ transgene was examined by RT-PCR using lacZ-specific primers.
The transgenic fetal and adult tissues were fixed briefly in 1xPBS solution containing 2%
formaldehyde (EMS) and 0.2% glutaraldehyde (Sigma). After washing with 1xPBS three
4
times, the embryos/tissue were stained for -galactosidase activity in staining solution
(5mM ferrocyanide, 5mM ferricyanide, 2mM MgCl2, 1 mg/ml X-gal and 1xPBS) at 37oC
for 6-24 hours. The stained tissues were embedded in frozen O.C.T. medium
(Diagnostics Division Elkhart, USA) and sectioned at 10 m using MICROM GmbH
(Walldort, Germany). After hematoxylin/eosin staining of the frozen section, the
distribution of X-gal staining in tissue sections was captured by a camera (model: RT
slider 2.3.1, Diagnostic Instruments Inc.) over the Carl Zeiss (AX10phot) microscope
(Göttingen, Germany) using the SPOT Dignostic Instruments software (version 4.5.8.).
Gel shift and ChIP assays ―The nuclear extracts (3μg) from a mouse erythroleukemia
cell line (MEL)(Active Motif, Carlsbad CA) were used for each DNA binding reaction in a
volume of 20μl. The oligonucleotides used are as follows: GATAsite1: 5’CTGTGCAGTGTGATAAAGAAAGTGG;
GACACGTCACACTATTTCTTTCA-CC-5’.
5’-CTGTGCAGTGTCTTAAAG-AAAGTGG;
mutGATAsite1:
ACACGTCACAGAATTTCTTTC-ACC-5’; GATAsite2: 5’-GGTCTAACCAGATAAAACTTTTAGG;
CCAGATTGGTCTATTTTGAAAAT-CC-5’;
GGTCTAACCACTTAAAA-CTTTTAGG;
mutGATAsite2:
5’-
CCTGATTGGTGAATTTTGAAA-ATCC-5’. The
experimental procedure was described before
20.
These oligonucleotides were end-
labeled by T4 Kinase with [γ-32P] dATP. The binding reaction buffer contained 10mM Tris
(pH7.5), 50mM NaCl, 1mM dithiothreitol, 1mM EDTA, 5% glycerol and 1g of poly (dIdC). The binding reactions were conducted at room temperature for 30 min. The
samples were electrophoresed at 200V at room temperature using a 5% polyacrylamide
in 0.5×Tris-Borate-EDTA buffer. For the competition studies, 100-fold molar excess of
5
unlabeled oligonucleotide was included in the binding reaction. For supershift analysis,
2μg anti-GATA1 or anti-GATA2 antibodies (Santa Cruz Biotechnology) were incubated
with nuclear extracts respectively at room temperature for 20 min prior to their use in the
binding reactions.
The ChIP assay was performed as described previously
14.
Purified chromatin
fragments from Meg-01 (an AMkL cell line) were incubated with anti-GATA1 antibodies
(C-20, Santa Cruz, Inc.). Standard PCR for the HEMGN promoter region was performed
using
forward
(5’-CCAGACACTTCCTGGCAGAT)
and
reverse
(5’-
CACTTGACTTCCCGCCTAAA) primers spanning positions –141 to +89. A coding
region (exon 3) of the human GATA1 gene was also amplified using forward (5’TGGAGACTTTGAAGACAGAGCGGCTGAG)
and
reverse
(5’-
GAAGCTTGGGAGAGGAATAGGCTGCTGA) primers to validate the specificity of the
ChIP assays.
Western blot analyses ―The presence of GATA1 and GATA2 in MEL nuclear extracts
was detected by Western blot assays, as described before
21.
Proteins on SDS-PAGE
gel were transferred to nitrocellulose membranes (Trans-Blot Transfer Medium, BioRad) and subsequently probed with anti-GATA1 or anti-GATA2 antibodies (Santa Cruz
Biotechnology). Primary antibodies were detected with POD conjugated anti-goat-IgG
and a BM Chemiluminescence Western Blotting kit (Roche).
Real time RT-PCR ―Total RNAs were extracted from mock-transfected, and GATA1
stably transfected CMS cells or primary myeloblasts of newly diagnosed acute myeloid
6
leukemia patients using TriReagent (Molecular Research Center, Inc.).
First strand
cDNAs were prepared from 1 µg RNA using random hexamer primers and a RT-PCR kit
(Perkin Elmer), and purified with the QIAquick PCR Purification Kit (Qiagen). GATA1,
EDAG and 18S RNA transcript levels were quantitated using a LightCycler real-time
PCR machine (Roche).
PCR reactions contained 2 µl purified cDNA or standard
plasmid, 4 mM MgCl2, 0.5 µM each of sense and antisense primers, and 2 µl FastStart
DNA Master SYBR Green I enzyme-SYBR reaction mix (Roche).
Primers were to
EDAG coding sense (5’-TGTGCCAAGAAGCTGCTGTA -3’) and antisense (5’TGGTTCTGCTGGATTTTGGT -3’) sequence, GATA1 coding sequence sense (5’TGGAGACTTTGAAGACAGAGCGGCTGAG-3’)
and
antisense
(5’-
GAAGCTTGGGAGAGGAATAGGCTGCTGA-3’) sequence, and 18S RNA sense (5’GATGCGGGGCGTTATT-3’)
and
antisense
(5’-TGAGGTTTCCCGTGTTGTCA-3’)
sequence. PCR conditions consisted of an initial denaturing step of 95 oC for 10 min,
amplification with 35-55 cycles of 95oC, 59-63oC for 10 s, and 72oC for 5 s, followed by
melting curve analysis from 40oC to 99oC, and a final cooling step to 40oC. External
standard curves for EDAG and GATA1 were constructed using serial dilutions of
linearized EDAG and GATA1 cDNA constructs. An 18S RNA plasmid construct was
prepared by cloning the 18S RNA amplicon amplified from K562 cDNA with commercial
primers (Ambion) into the pGEM T-Easy vector (Promega). This plasmid construct was
linearized with ApaI, and serial dilutions were prepared and quantitated by real-time
PCR, and the data used to prepare the 18S RNA external standard curve. EDAG and
GATA1 transcript levels were expressed relative to 18S RNA. Real-time PCR results
7
were expressed as mean values from 2-3 experiments using the same cDNA
preparation.
Clinical AML samples ―Myeloblasts from children diagnosed with AML were obtained
from the Children's Hospital of Michigan leukemia cell bank. Mononuclear cells were
isolated on Ficoll-Hypaque gradients to obtain highly purified mononuclear cell fractions
consisting mostly of leukemic blasts. Total RNAs were extracted from the samples using
TRIzol reagent (Life Technologies). The research protocol was approved by the Human
Investigation Committee of Wayne State University School of Medicine.
Statistical analysis ―The nonparametric Spearman rank correlation coefficient was
used to analyze HEMGN transcript levels and their relationship to GATA1 transcript
levels. Statistical analyses were performed with StatView (Version 4.5, for Windows).
8