HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Full Text (PDF) All Versions of this Article: 19/1/147 04-2417fjev1    most recent Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Golan-Mashiach, M. Articles by Givol, D. Search for Related Content PubMed PubMed Citation Articles by Golan-Mashiach, M. Articles by Givol, D.

Design principle of gene expression used by human stem cells: implication for pluripotency

Michal Golan-Mashiach^*,,1, Jean-Eudes Dazard^*,,1, Sharon Gerecht-Nir, Ninette Amariglio^||, Tamar Fisher^||, Jasmine Jacob-Hirsch^||, Bella Bielorai^||, Sivan Osenberg, Omer Barad, Gad Getz, Amos Toren^||, Gideon Rechavi^||, Joseph Itskovitz-Eldor, Eytan Domany and David Givol^*,2

^* Department of Molecular Cell Biology,
Department of Physics of Complex Systems, Weizmann Institute of Science, Rehovot, Israel;
Department of Epidemiology and Biostatistics, School of Medicine, Case Western Reserve University, Cleveland, Ohio, USA;
Department of Obstetrics and Gynecology and Biotechnology Interdisciplinary Unit, Rambam Medical Center, Faculty of Medicine, The Technion, Haifa, Israel; and
^|| Department of Pediatric Hematology-Oncology, Chaim Sheba Medical Center and Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, Israel

²Correspondence: Department of Molecular Cell Biology, Weizmann Institute of Science, Herzl 2 St., Rehovot, Israel. E-mail: david.givol{at}weizmann.ac.il

SPECIFIC AIMS

The primary aims of the present study were to 1) compare gene expression profiles of human embryonic stem cells (ESC) with different adult stem/progenitor cells—epidermis’ keratinocytes stem/progenitor cells (KSPC), hematopoietic stem/progenitor cells (HSPC) and their corresponding differentiated cells, epidermis differentiated cells, (KDC), and hematopoietic differentiated cells (HDC); 2) define genes responsible for stem cell properties; and 3) possibly identify design principles that help to maintain the pluripotency of stem cells.

PRINCIPAL FINDINGS

We analyzed by RNA hybridization to Affymetrix U133A GeneChip® two groups of samples (each in at least triplicate) that may represent virtual steps in the differentiation pathways from ESC to adult cells: hematopoietic (H) pathway, ESC HSPC HDC, and keratinocytic (K) pathway, ESC KSPC KDC. We refer to these as "virtual pathways" to avoid being misinterpreted because we do not imply that the adult stem/progenitor cells (ASPC) evolved from our ESC samples or that our adult differentiated cells are the result of differentiation of the ASPC.

1. ES cells express more genes than differentiated cells
We wished to know whether ES cells and tissue adult cells express similar or different numbers of genes. Expression data of the ESC and the adult differentiated cells were analyzed as follows: out of 22,215 probe sets (PS) on the chip, we kept 15,918 PS that had a "present" call (obtained from MAS 5.0) for at least one sample. The log₂-transformed expression levels for each PS were centered and normalized (over the samples of ESC and adult differentiated cells). A value that exceeds 0.3 in a particular sample was interpreted as expression level significantly above average. We found that 4550 PS satisfy this criterion in ESC; this number is significantly higher than the number of similarly identified genes that are highly expressed genes in the adult differentiated state (3000 PS). ESC express a significantly higher number of genes than adult differentiated cells.

2. Marked down-regulation of expressed genes from ESC to adult tissue
For both pathways H and K defined above the genes were filtered using ANOVA, and false discovery rate (FDR) was controlled at 0.05. This left 8290 PS (corresponding to 6293 genes) that vary significantly over the three kinds of cell states in pathway H and 5432 PS (4301 genes) for K. Figure 1 A, B shows the expression levels of the significantly varying PS. The black line indicates the expression levels in ESC of 8290 PS ordered according to their expression from high levels (left) to low (right). The green dots show the expression of the same genes in HSPC (Fig. 1A ) and KSPC (Fig. 1B ), and the red dots represent the expression of the same genes in adult cells (HDC, Fig. 1A ; KDC, Fig. 1B ). The data show that ESC (black line, Fig. 1 ) express many genes at a higher level than the other cells. Many of the transcripts exhibited marked down-regulation in HSPC (Fig. 1A , green dots) and a further downward shift in HDC (Fig. 1A , red dots); 4392 PS (3483 genes) were down-regulated in HDC compared with ESC. This was accompanied by up-regulation of a smaller group, of 2638 PS (1988 genes), which have low expression in ESC and high in the HDC. A similar pattern was seen in the keratinocytic pathway (Fig. 1B ).

View larger version (60K): [in this window] [in a new window]   Figure 1. Expression levels of probe-sets (PS) that vary significantly between ESC, adult stem cells, and differentiated cells. The PS were ordered according to their ESC expression levels, marked by black circles that form a line. The expression levels in HSPC or KSPC are indicated by green dots and in HDC, and KDC by red dots. A) Expression levels of 8290 PS that vary among ESC, HSPC, and HDC. B) Expression of 5432 PS that vary among ESC, KSPC, and KDC. Only PS with P-values that passed ANOVA at an FDR level of 0.05 were plotted. In (A) 4392 PS (3483 genes) are down-regulated and 2638 PS (1988 genes) are up-regulated with differentiation. In (B) 3417 PS (2758 genes) and 1423 PS (1115 genes) are up- and down- regulated, respectively.

3. Clustering analysis shows distinct stem cell-specific genes
We clustered the 8290 PS that varied in pathway H and the 5432 PS of K separately, to identify distinct cell-state related variations of the expression profiles, and to assign genes to clusters of similar patterns of expression. Figure 2 depicts the expression matrix after clustering of the genes in the H (Fig. 2A ) and K (Fig. 2B ) pathways. Six clusters are clearly identified: clusters H1, H2, and H3 contain ESC genes that were down-regulated with differentiation in the H pathway and K1, K2, K3 were down-regulated in the K pathway. Clusters 4 and 5 contain genes that were up-regulated along the virtual differentiation pathways (H or K). Cluster 6 contains genes expressed only in adult stem/progenitor cells, whereas cluster 3 contains genes that show expression only in ESC. Clearly, ESC and adult stem progenitor cells have different gene expression profiles. The lists of genes in each of the clusters are given in tables S2-S13 in supplementary data (http://www.weizmann.ac.il/physics/complex/compphys/downloads/michalm/).

View larger version (46K): [in this window] [in a new window]   Figure 2. Clustering analysis of PS expression levels in hematopoietic and keratinocytic pathways. The expression levels of the PS taken from Fig. 1 were centered and normalized and the PS were reordered according to the dendrogram produced by the SPC algorithm. A) Expression matrix of 8290 PS in ESC, HSPC, and HDC. B) Expression matrix of 5432 PS in ESC, KSPC, and KDC. C) Overlaps between the related 6 clusters were calculated relatively to the K1–K6 clusters.

Clusters 1 and 2 contain genes that are common to ESC and ASPC and therefore may represent the "stemness" genes as previously defined. Many genes known to be markers for undifferentiated ESC or related to ESC self-renewal (e.g., NANOG, POU5F1 (OCT4), SOX2, FOXH1, TDGF1 (Cripto), LeftyA and B, Thy1) belong to clusters H3 and K3 (Fig. 2) and thus are suppressed in ASPC. Their roles are apparently taken over in ASPC by those of clusters H6 or K6, which show expression only in ASPC, and contain genes known to be essential for the self-renewal and development of the particular tissue (e.g., TP73L (p63), ITGB4, and BNC for skin; and BMI1, CD34, TIE, KIT, TAL1 (SCL), and RUNX1 for blood).

4. A common core of candidate genes that may represent an ESC signature
The genes of clusters H3 and K3 are expressed only in ESC when clustering is performed on each of the pathways, hence some of these genes may be "the signature for embryonic stemness". To search for a common core of ESC genes we intersected the H3 and K3 gene lists, yielding 179 genes, some of which are highly enriched in ESC relative to differentiated and stem/progenitors of blood and epidermis. This list may contain genes expressed in many adult tissues other than blood and epidermis. After removal of such genes by checking with datasets of various tissues, this list was reduced to 66 genes (table S1 in supplementary data). These genes may be regarded as an embryonic stem cell signature. This list includes most of the well-known genes that were analyzed in various ESC systems and shown to be essential for pluripotency and self-renewal (i.e., NANOG and OCT4, known to be important for suppression of differentiation and maintaining of pluripotential differentiation; LIN28, an RNA-binding and a negative regulator of differentiation; LEFTB, a TGFß-related protein; and frizzled (FZD), a receptor for WNT and part of the WNT/ß-catenin pathway). The list includes the transcription factor SOX2 that regulates FGF4 expression, which was found in all human ESC and is also a marker of neural progenitors, as well as TDGF1 or cripto, an autocrine growth factor stimulating cell proliferation at the expense of differentiation. Noteworthy in this list are genes that encode for proteins involved in remodeling of chromatin. An example is SMARCA1, a homolog of the yeast general transcription activator of the SWI/SNF family members known to be involved in chromatin remodeling. This complex includes members of SWI2/SNF2 family and histone deacetylase (HDAC). This suggests that a significant part of the regulation of gene expression in ESC is by epigenetic changes that regulate chromatin alteration (table S1, supplementary data).

CONCLUSIONS AND SIGNIFICANCE

We compared the complexity of gene expression in tissue progenitor cells and adult differentiated cells from two different tissues with that of ESC. Detailed measurement of gene expression indicates that ESC express more genes than the other cell states and there is a gradual decrease in gene expression complexity along the differentiation pathway. When cell fate is determined, many genes that are significantly expressed in ESC are quenched (clusters 1–3). Concomitantly, a smaller number of genes are up-regulated. These are mainly tissue-specific genes (clusters 4–5). The promiscuous expression of large number of genes in ESC may serve as a preface for future response to cell fate determination. To maintain their pluripotency, ESC keep their options open by expressing thousands of genes and selecting only a portion for the continuous expression needed for differentiation to the target tissue. The rest will be down-regulated upon commitment to the cell fate for which they are not needed. This strategy was shown to operate in hematopoietic stem cells where lymphoid- and myeloid-specific genes are expressed prior to differentiation and our data show that it is also typical to ESC.

We looked for design principles of stem cells that accounts for pluripotency. A prime candidate for pluripotential differentiation is the parsimonious "just in time" strategy: expressing genes only when needed (i.e., at the moment of commitment to a particular differentiation path). The opposite extreme is the seemingly more wasteful "just in case" strategy, which keeps a wide repertoire of expressed genes to be present just in case a particular path is selected. We hypothesize that the massive down-regulation of expressed genes is consistent with the "just in case" design principle underlying pluripotential differentiation of ESC. This is a selection model of gene expression during differentiation where the down-regulation of genes that are not needed is required for establishing the differentiated state.

View larger version (61K): [in this window] [in a new window]   Figure 3. Schematic diagram illustrating experimental design and major findings. ESC, embryonic stem cells; HSPC, hematopoietic stem/progenitor cells; HDC, hematopoietic differentiated cells; KSPC, keratinocyte stem/progenitor cells; KDC, keratinocyte differentiated cells.

FOOTNOTES

¹ These authors contributed equally to this work.

To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.04-2417fje;

This article has been cited by other articles:

				D. Xie, J. Cai, N.-Y. Chia, H. H. Ng, and S. Zhong Cross-species de novo identification of cis-regulatory modules with GibbsModule: Application to gene regulation in embryonic stem cells Genome Res., August 1, 2008; 18(8): 1325 - 1335. [Abstract] [Full Text] [PDF]

				S. Assou, T. Le Carrour, S. Tondeur, S. Strom, A. Gabelle, S. Marty, L. Nadal, V. Pantesco, T. Reme, J.-P. Hugnot, et al. A Meta-Analysis of Human Embryonic Stem Cells Transcriptome Integrated into a Web-Based Expression Atlas Stem Cells, April 1, 2007; 25(4): 961 - 973. [Abstract] [Full Text] [PDF]

				C. Allegrucci and L.E. Young Differences between human embryonic stem cell lines Hum. Reprod. Update, March 1, 2007; 13(2): 103 - 120. [Abstract] [Full Text] [PDF]

				L. Song, N. E. Webb, Y. Song, and R. S. Tuan Identification and Functional Analysis of Candidate Genes Regulating Mesenchymal Stem Cell Self-Renewal and Multipotency Stem Cells, July 1, 2006; 24(7): 1707 - 1718. [Abstract] [Full Text] [PDF]

				D. Zipori The Stem State: Plasticity Is Essential, Whereas Self-Renewal and Hierarchy Are Optional Stem Cells, June 1, 2005; 23(6): 719 - 726. [Abstract] [Full Text] [PDF]

This Article Full Text (PDF) All Versions of this Article: 19/1/147 04-2417fjev1    most recent Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Golan-Mashiach, M. Articles by Givol, D. Search for Related Content PubMed PubMed Citation Articles by Golan-Mashiach, M. Articles by Givol, D.