| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Departments of Biochemistry and
b Microbiology, College of Medicine, Ewha Womans University, Seoul, Korea;
c College of Medicine, Seoul National University, Seoul, Korea and Macrogen Inc., Seoul, Korea
Key Words. SAGE • Megakaryocytes • Cord blood • Thrombopoietin • ex vivo expansion • Microarray
Hyung-Lae Kim, M.D., Ph.D., Department of Biochemistry, College of Medicine, Ewha Womans University, Mok-6-dong 911-1, Yangchun-Ku, Seoul, 158-710, Korea. Telephone: 822-650-5727; Fax: 822-653-8891; e-mail: hyung{at}mm.ewha.ac.kr
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Serial analysis of gene expression (SAGE), described by Velculescu et al. [6], is based on the principle that a nucleotide sequence of 9-10 bases (a gene tag) corresponds to a unique transcript. The tag frequency directly reflects the abundance of the mRNA. It allows for the establishment of both a representative and a comprehensive different gene expression profile in various cell types and organs under physiological and pathological states. Since each template contains identifiable tags corresponding to many genes, this method allows us to have global gene expression profiles that are unknown.
In the present study, we performed SAGE to analyze the gene expression profiles of MKs (CD61+ cells) derived from human CB CD34+ cells and compared them with those of non-MKs (CD61- cells). High-density oligonucleotide microarray hybridization and reverse transcription-polymerase chain reaction (RT-PCR) were used to validate the SAGE results. The data presented herein showed that many of the genes differentially expressed in MKs and non-MKs were involved in encoding proteins related to apoptosis and intracellular signaling pathways.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Ex Vivo Expansion and Separation of MK and Non-MK Fractions
The CD34+ cells were cultured at a density of 1.0 x 105 cells/ml in serum-free essential media supplemented with BSA, insulin, and transferrin (StemCell Technologies; Vancouver, BC, Canada; http://www.stemcell.com). Cultures were stimulated with recombinant human TPO (50 ng/ml; Kirin Brewery; Maebashi, Japan) alone. After 10 days, the MK fraction was separated from the non-MK fraction using an anti-CD41 (glycoprotein IIb/IIIa [GPIIb/IIIa]) monoclonal antibody (Dako; Copenhagen, Denmark; http://www.dako.dk) and a microbead-conjugated goat anti-mouse IgG (Miltenyi Biotech) followed twice by the MiniMACS system (Miltenyi Biotech). The purity of each separated fraction was verified by flow cytometry with a different antibody reacting with MKs (FITC-conjugated anti-human CD61; BD).
RNA Preparation
Total RNA was prepared from the separated MK and non-MK fractions using the TRIZOL (GIBCO BRL; Grand Island, NY; http://www.lifetech.com) according to the manufacturer’s instructions. In order to remove DNA completely, the RNA samples were incubated with RNase-free DNase I (Takara Shoji; Tokyo, Japan; http://www.takara.co.jp). The quality of the RNA prepared was confirmed by analyzing the samples by electrophoresis on a 1.2% agarose/formaldehyde gel in MOPS (3-Morpholino propanesufonic acid) buffer. RNA samples were stored at -80°C until future use.
SAGE Protocol
SAGE was performed according to the Micro Serial Analysis of Gene Expression Detailed Protocol, version 1.0. Biotinylated oligo-dT-primer-annealed total RNA, 10 µg, was converted to cDNA with a BRL cDNA synthesis kit (GIBCO BRL) in a streptavidine-coated PCR tube (Roche; Mannheim, Germany; http://biochem.boehringer-mannheim.com). The cDNA was cleaved at Nla III, and was ligated to the oligonucleotide-containing recognition sites for BsmF I. After ligation, the bound cDNA was released from the matrix by digestion with BsmF I. SAGE tag overhangs were filled with Klenow, and tags from two pools were combined and ligated to each other. The ligation product was amplified by PCR, concatemerized, and cloned into the SphI site of pZero-1 (Invitrogen; Carlsbad, CA; http://www.invitrogen.com). Clones were sequenced with the BigDye terminator kit and analyzed using ABI 3700 automated sequencer (Perkin-Elmer; Branchberg, CT; http://www.perkinelmer.com). Sequence files were analyzed by means of SAGE analysis software, version 4.12. Statistical analysis of the data (Monte Carlo test) was performed using SAGE software, version 4.12 (courtesy of Victor Velculescu and Ken Kinzler, Johns Hopkins University School of Medicine; Baltimore, MD) [6]. The identities of the mRNAs corresponding to the SAGE tags were determined through inspection and comparison with the SAGEmap (www.ncbi.nlm.nih.gov/SAGE/SAGEtag.cgi) and UniGene (www.ncbi.nlm.nih.gov/UniGene) databases.
Microarray Protocol
Total RNA (5 µg) was converted into double-stranded cDNA using the cDNA synthesis system (Roche) with T7-(dT)24 primer. Then, each cDNA was purified using the RNeasy kit (Qiagen; Valencia, CA; http://www.qiagen.com). Each Cy3- (MK), or Cy5- (non-MK) labeled cRNA was synthesized using the Megascript T7 kit (Ambion; Austin, TX; http://www.ambion.com), with Cy3-CTP and Cy5-CTP (APB; Uppsala, Sweden; http://www.apbiotech.com). The cRNA was purified using RNeasy (Qiagen). Fifteen micrograms of each purified cRNA were mixed and fragmented in fragmentation buffer (40 mM Tris [pH 8.1], 100 mM KOAc, and 30 mM MgOAC) by heating to 94°C for 15 minutes. Fragmented cRNA was mixed with hybridization buffer, containing 100 mM MES, 1 M NaCl, 20 mM EDTA, and 0.01% Tween 20, and hybridized with MAGIC II-10 K Oligo Chip (Macrogen; Seoul, Korea; http://www.macrogen.com) for 16 hours at 42°C. All preparations met Macrogen’s recommended criteria for use on their expression arrays. Arrays were then washed and scanned with an array scanner (APB). Acquired images were processed and analyzed statistically for interpretation of spot intensity results using Imagene version 4.1 software (Roche). Nonbiological factors that might contribute to the variability of data were minimized using global normalization/scaling with data from all probe sets. Each chip contains a total of 10,368 elements, of which 10,108 are unique genes/clusters. The length of oligonucleotides was 50-mer. Subsets of genes were selected based on differential Cy3/Cy5 expression ratios that were 
|2| in response.
RT-PCR
One microgram of total RNA from each sample was used as a template for the RT reaction. The total RNA was reverse transcribed in 20 µl of 10 mM Tris-HCl (pH 8.3), 6.5 mM MgCl2, 50 mM KCl, 10 mM dithiothreitol (DTT), 1 mM of each dNTP, 0.5 µg oligo dT primer, and 50 U of Superscriptase II (GIBCO BRL) for 1 hour at 42°C. The conditions of PCR were as follows, in a 25 µl reaction, 0.4 µM of each primer (Table 1
), 125 µM of each dNTP mixture, 50 mM KCl, 10 mM Tris-HCl (pH 8.3), and 1.5 mM MgCl2 AmpliTaq (Perkin-Elmer) were used. The PCR protocol consisted of denaturing for 30 seconds at 94°C, annealing for 30 seconds at 55°C, and elongating for 1 minute at 72°C. Amplification was performed for 25 to 30 cycles. Primers used were as follows. Apoptotic protease activating factor-1 (APAF-1): sense 5'-CTT GGA TGA TGT TTG GGA CTC TTG-3', antisense 5'-GAA ACG ACT TTC CAT TCC GAT CAC-3'; CD 48: sense 5'-CCA GAA CAG TGT GCT TGA AAC CAC-3', antisense 5'-TGG TCA GCC TAT ACA GTC TCT GTC C-3'; defensin: sense 5'-TTG CTG CCA TTC TCC TGG TG -3', antisense 5'-GAG GAA AGG AAA TTG AGC AGA AGG-3'; pim-1: sense 5'-CGG ATT CTA ACC TGG AGG TCA-3', antisense 5'-CTC AGA TAA AAC CAG CAG GCT ACC-3'; P-selectin: sense 5'-AAC ACA AGC CAC AGA AGC CAG G-3', antisense 5'-TGG GTC ATT TGA GGG ACA GTG AC-3'; KIAA0614: sense 5'-CTT GAA TGG ACT TGT CAG CTA CCT C-3', antisense 5'-TGC CAG CCT TTG AAC TTG CTC-3'; ß-actin: sense 5'-AAG AGG ACC CAG ATC ATG TTT GAG-3', antisense 5'-AGG AGG AGC AAT GAT CTT GAT CTT-3'.
|
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Gene tags that had no reliable matches in UniGene clusters were excluded from the list. The most abundantly expressed genes identified in both cell fractions included hemoglobin family genes, ferritin light chain, profilin 1, and genes from the apolipoprotein family, followed by housekeeping ribosomal protein coding genes (Tables 1 and 2). Survival of motor neuron 1 (0.51%), azurocidine 1 (0.19%), and CGI 120 protein coding gene (0.15%) was highly expressed in the MK fraction. By contrast, the genes encoding eukaryotic translation elongation factor 2 (0.17%) and neurogranine (0.15%) were detected only in the non-MK fraction.
|
). Each dot in Figure 1 represents a gene expressed in these two fractions. Although the expression levels of most of the transcripts were similar in both fractions, significant differences in expression levels between the two fractions were found in many transcripts.
|
show the genes differentially expressed in MKs and non-MKs. Unidentified and multiple matched genes were excluded from the tables. Table 3 shows the top 50 transcripts that had greater expression levels in MKs, than in non-MKs. The most extreme of these was identified to be eukaryotic elongation factor 1 beta 1 (45-fold), followed by CGI-135 protein (38-fold), protein associated with PRK1 (38-fold), thioredoxin peroxidase (38-fold), and so on. The transcripts with greater expression levels in MKs could be classified into several groups according to the functional relevance: genes encoding gene-expression-related proteins, such as eukaryotic translation elongation factor, CAAT/enhancer binding protein, and cyclic AMP (cAMP) responsive element binding protein; splicing-related and RNA-binding proteins, such as RNase 6, small nuclear ribonucleoprotein D2, U2 small nuclear ribonucleoprotein auxiliary factor, RD (arginin and aspartate) RNA-binding protein, and splicing factor 3a; metabolic pathway-related proteins, such as cytochrome c synthase, cytochrome b-245, lactate dehydrogenase B, malate dehydrogenase 2, and NADH dehydrogenase 1 beta; heat shock proteins, such as Hsp 90 and Hsp 40 homologue; and surface marker proteins, such as CD48, and CD62P (P-selectin).
|
|
|
) were small G-protein-related proteins, such as ras guanyl-releasing protein 2, ral guanine nucleotide dissociation stimulator, rho/rac guanine nucleotide exchange factor 2, and rab 11A; surface marker proteins, such as CD37; proteasome subunits; and ATP synthases. Table 5 shows the differentially expressed genes grouped into several categories according to their functional relevance.
Genes in the EST database or unidentified in the GenBank database were excluded from Table 5. While those genes related to the cytoskeletal system, such as tubulin and actin-binding proteins, showed almost equal or lower expression in MKs, the genes encoding transporter and channel proteins, such as arsenite transporter and transportin-SR, were expressed more in MKs. Most of the ribosomal proteins were expressed more highly in MKs than in non-MKs. Transcription factors, such as nuclear factor (erythroid-derived 2) 45 kD, nuclear factor (erythroid-derived 2)-like 2, and cAMP response element binding factor 2 were expressed more highly in MKs. These transcription factors may be important in the maturation of MKs [7]. Genes related in apoptotic event as well as in mitogen-activated protein kinase (MAPK) pathway were expressed more highly in MKs. Genes encoding the Hsp family, such as Hsp 90, chaperone containing TCP1, DnaJ homologue, and tubulin-specific chaperone a, were also expressed more highly in MKs. Expressions of the genes for surface antigens, such as CD48 and CD62, were higher in MKs.
Validation of Genes Represented in the SAGE Analysis
To verify the validity of our SAGE data, we performed a microarray experiment. Total RNA was obtained by the same way as in the SAGE experiment. This RNA was hybridized with a 10 K oligo microarray. A total of 2,807 genes found on microarray overlapped with SAGE tags. A total of 364 genes had a greater than twofold difference in expression between MKs and non-MKs and 125 genes overlapped with SAGE tags. Due to the lower sensitivity of microarray analysis, the expression ratio of some genes was underestimated (Table 5).
We picked up eight genes, which were expressed differentially in MKs and non-MKs. Using RT-PCR, selected genes were evaluated in MKs and non-MKs derived from the CB of two other volunteers (Fig. 2). The results showed that ß-actin was expressed almost equally in all cell types (MKs, 62 tags; non-MKs, 61 tags) studied, but apoptotic protease activating factor (MKs, 38 tags; non-MKs, 2 tags), CD48 (MKs, 23 tags; non-MKs, 0 tags), P-selectin (MKs, 35 tags; non-MKs, 7 tags), pim-1 (MKs, 35 tags; non-MKs, 7 tags), and defensin (MKs, 55 tags; non-MKs, 46 tags) were expressed more highly in MKs. The expression of KIAA0614 (MKs, 0 tags; non-MKs, 24 tags) was lower in MKs. These results validate our SAGE data for MKs and establish a general expression profile of the identified genes.
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We observed many housekeeping genes, such as ribosomal protein family genes, lipoprotein family genes, thymosin, transferrin receptor, and major histocompatibility complex genes, expressed highly in both MKs and non-MKs. This result is comparable with previous results from SAGE analyses of hematopoietic cells, such as dendritic cells, monocytes, and macrophages [8–10]. However, erythroid differentiation factor, serin proteinase inhibitor, CGI-120, CGI-135, and arsenite transporter genes were expressed highly only in MKs (Tables 1 and 3). The differences in the expression levels of these genes between the two fractions suggest that these genes may play roles in the differentiation into MKs. Those genes that are already known to be involved in the differentiation and function of MKs, such as nuclear factor (erythroid derived) 45 kD, P-selectin (CD62P), platelet phosphofructokinase, coagulation factor, and megakaryocyte-associated tyrosine kinase, were expressed more highly in MKs (Tables 3 and 5).
Furthermore, nuclear factor (erythroid derived) 45 kD (4.75-fold higher expression in MKs) regulates two classes of gene expression in maturating MKs that are required to reorganize the MK cytoskeleton and initiate proplatelet formation [7]. cAMP-responsive element-binding protein-like 1, which is the binding partner of nuclear factor (erythroid derived) 45 kD, was also expressed highly in MKs (38-fold higher in MKs than in non-MKs). In particular, many immediate-early genes, such as c-fos, c-jun, and d-jun, which propagate the cellular response to growth stimuli, were expressed more highly in MKs. These transcription factors may influence MK-specific gene expression.
In contrast, the levels of expression of cytoskeletal protein-coding genes, such as actin-binding proteins and tubulin family genes, were similar in both fractions. But many of the ion-channel composing protein-coding genes, such as transportin, TAP, purinergic receptor P2X, and arcenite receptor, were expressed highly in MKs (Tables 3 and 5). Nevertheless, P-selectin, one of the adhesion molecules, had a fivefold higher expression in MKs (Table 5, Fig. 2). P-selectin could mediate megakaryocyte-fibroblast interactions. Interaction between MKs and fibroblasts regulates the production of proplatelets and their migration into the sinusoidal lumina [11, 12].
Apoptotic-related genes, such as transforming growth factor beta 1 (TGF-ß1), calpain, programmed cell death interacting protein, and APAF, were predominantly expressed in MKs. However, the expression of survivin, an inhibitor of apoptosis, was low in MKs. These results are consistent with our previous report that TPO-induced apoptosis was closely associated with megakaryocytic differentiation [5]. TGF-ß1, which was highly expressed in MKs (twofold greater than in non-MKs), plays a pivotal role in the control of differentiation, proliferation, and the state of activation of many different cell types, including immune cells [13]. The molecular mechanisms involved in these apoptotic processes seem to involve the activation of the SMAD protein family [14]. The TGF-ß1 receptor initiates intracellular signaling through the activation of SMAD proteins, and specific SMAD proteins become phosphorylated and associate with other SMAD proteins. These heteromeric SMAD complexes accumulate in the nucleus, where they modulate the expression of target genes. Only the relationship between TGF-ß1 and SMAD 4 is known in apoptosis [15]. SMAD 3 and 4 together could enhance TGF-ß1-mediated transactivation. Moreover, the (JNK) pathway was shown to be activated by SMAD proteins in TGF-ß1-induced apoptosis [16]. In Tables 3 and 5, the mad 4 homologue, SMAD- and olf-interacting protein genes, and JNK-related proteins, such as c-jun and c-fos, were more highly expressed in MKs. Calcium-dependent protease and calpain-1 and 4 which are generally associated with necrosis and some forms of TGF-ß1-induced apoptosis [17], were expressed highly in MKs (over twofold greater than in non-MKs),. Apoptosis might be controlled by TGF-ß1 and SMAD pathways in MKs.
As confirmed by RT-PCR (Fig. 2), the expression of APAF was 19-fold higher in MKs than in non-MKs (Table 5). APAF is important in cell death machinery (apoptosome) and is involved in different apoptotic pathways in different states. In apoptotic conditions, caspases are often activated by APAF-1 apoptosome, a complex formed in response to many death-inducing stimuli [18, 19]. Activation of caspases is regulated directly or indirectly by antiapoptotic members of the Bcl-2 and inhibitor of apoptosis protein (IAP) family, like survivin [20, 21]. As an antiapoptotic protein, survivin inhibits caspases 3 and 7. In the present study, the expression of survivin was lower in MKs (ninefold). By contrast, stress-induced transcripts, such as some of the Hsp, were markedly greater in MKs. It has been reported that Hsp 90 has proapoptotic and antiapoptotic dual controversial potentials [22]. Expression of the Hsp 90 gene was 20-fold higher in MKs than in non-MKs. Thus, the higher expression of APAF and Hsp 90 and lower expression of survivin may contribute to the apoptotic process of MKs.
Binding of TPO to its receptor, c-Mpl, results in the activation of a variety of signaling molecules that include components of the Ras/MAPK pathway [23–25] and Janus kinase-signal transducer and activator of transcription (JAK/STAT) pathway [26]. In the present study, the genes encoding proteins related to the Ras/MAPK pathway, such as protein kinase C (PKC)-like protein, PKC substrate 80 K-H, epidermal growth factor (EGF) response factor 2, neuroblastoma RAS viral homologue, and MAP kinase kinase kinase 2, were expressed more highly in MKs. These results are consistent with the previous report that the Ras/MAPK pathway contributes to TPO-induced differentiation. Recently, it was reported that PKC activation was essential for proplatelet formation [27]. However, the JAK and STAT families, which are related to the cytokine-signaling pathway and MK proliferation, were only detected in either MKs or non-MKs. Pim-1, another Ser/Thr kinase was expressed more highly in MKs, as much as fivefold higher than in non-MKs (Fig. 2). It has been reported that the expression of pim-1 was involved in MK development [28]. Although the function of pim-1 is obscure, it might perform as a survival factor in eosinophil apoptosis [29]. These apoptotic factors and survival factors might participate in MK development cooperatively.
Interestingly, CD 48 was predominantly expressed in MKs (23-fold greater than in non-MKs, Fig. 2). It is a high-affinity ligand of 2B4 (CD244), expressed on natural killer (NK) or T cells. The interaction between CD244 and CD48 may play an important role in the regulation of NK and T cells [30]. Antimicrobial peptide, azurocicin (expressed 1.3-fold higher in MKs) and defensin (expressed 1.5-fold higher in MKs, Fig. 2) were two of the antibacterial peptides. They have a chemotactic effect on mononuclear cells and neutrophils, and induce T-cell migration for direct bacterial activity [31]. MKs might interact with NK cells or T cells through CD48 and, through secretion of azurocidin and defensin, might play the role of activator of mononuclear cells.
In summary, we identified the genes expressed differentially in MKs and non-MKs derived from human CB CD34+ cells by ex vivo expansion using TPO. The expression levels of several genes in apoptosis, such as APAF and calpain were higher in the MK fraction. Interestingly, azurocidin and defensin, known as antibacterial peptides, were highly expressed in MK cells. In contrast, the expression of survivin, an antiapoptotic protein, was lower in the MK fraction. These gene expression data from SAGE analyses may provide useful information on MK maturation, platelet production, and the possibility of interaction with other blood cells.
|
ACKNOWLEDGMENT |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y.-L. Hu, E. Passegue, S. Fong, C. Largman, and H. J. Lawrence Evidence that the Pim1 kinase gene is a direct target of HOXA9 Blood, June 1, 2007; 109(11): 4732 - 4738. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Raslova, A. Kauffmann, D. Sekkai, H. Ripoche, F. Larbret, T. Robert, D. T. Le Roux, G. Kroemer, N. Debili, P. Dessen, et al. Interrelation between polyploidization and megakaryocyte differentiation: a gene profiling approach Blood, April 15, 2007; 109(8): 3225 - 3234. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Matsumura-Takeda, S. Sogo, Y. Isakari, Y. Harada, K. Nishioka, T. Kawakami, T. Ono, and T. Taki CD41+/CD45+ Cells Without Acetylcholinesterase Activity Are Immature and a Major Megakaryocytic Population in Murine Bone Marrow Stem Cells, April 1, 2007; 25(4): 862 - 870. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. A. Senis, M. G. Tomlinson, A. Garcia, S. Dumon, V. L. Heath, J. Herbert, S. P. Cobbold, J. C. Spalton, S. Ayman, R. Antrobus, et al. A Comprehensive Proteomics and Genomics Analysis Reveals Novel Transmembrane Proteins in Human Platelets and Mouse Megakaryocytes Including G6b-B, a Novel Immunoreceptor Tyrosine-based Inhibitory Motif Protein Mol. Cell. Proteomics, March 1, 2007; 6(3): 548 - 564. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Chen, M. Hu, and R. A. Shivdasani Expression analysis of primary mouse megakaryocyte differentiation and its application in identifying stage-specific molecular markers and a novel transcriptional target of NF-E2 Blood, February 15, 2007; 109(4): 1451 - 1459. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Komor, S. Guller, C. D. Baldus, S. de Vos, D. Hoelzer, O. G. Ottmann, and W.-K. Hofmann Transcriptional Profiling of Human Hematopoiesis During In Vitro Lineage-Specific Differentiation Stem Cells, September 1, 2005; 23(8): 1154 - 1169. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Ross, R. Mahfouz, M. Onciu, H.-C. Liu, X. Zhou, G. Song, S. A. Shurtleff, S. Pounds, C. Cheng, J. Ma, et al. Gene expression profiling of pediatric acute myelogenous leukemia Blood, December 1, 2004; 104(12): 3679 - 3687. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |
