Unique Gene Expression Signature by Human ... - Wiley Online Library

Embryonic Stem Cells–Characterization Series Unique Gene Expression Signature by Human Embryonic Stem Cells Cultured Under Serum-Free Conditions Correlates with Their Enhanced and Prolonged Growth in an Undifferentiated Stage ¨ HELI SKOTTMAN,a,b ANNE-MARIE STROMBERG ,c EIJA MATILAINEN,c JOSE INZUNZA,d OUTI HOVATTA,b,c RIITTA LAHESMAAa a

Turku Centre for Biotechnology, University of Turku and Åbo Akademi University, Turku, Finland; bREGEA, Institute for Regenerative Medicine, University of Tampere and Tampere University Hospital, Tampere, Finland; c Department of Obstetrics and Gynecology, CLINTEC, Karolinska Institutet, Karolinska University Hospital Huddinge, Stockholm, Sweden; dDepartment of Medical Nutrition, Karolinska Institute, Karolinska University Hospital Huddinge, Stockholm, Sweden

ABSTRACT Understanding the interaction between human embryonic stem cells (hESCs) and their microenvironment is crucial for the propagation and the differentiation of hESCs for therapeutic applications. hESCs maintain their characteristics both in serum-containing and serum-replacement (SR) media. In this study, the effects of the serum-containing and SR culture media on the gene expression profiles of hESCs were examined. Although the expression of many known embryonic stem cell markers was similar in cells cultured in either media, surprisingly, 1,417 genes were found to be differentially expressed when hESCs cultured in serum-containing medium were compared with those cultured in SR medium. Several genes upregulated in cells cultured in SR medium

suggested increased metabolism and proliferation rates in this medium, providing a possible explanation for the increased growth rate of nondifferentiated cells observed in SR culture conditions compared with that in serum medium. Several genes characteristic for cells with differentiated phenotype were expressed in cells cultured in serum-containing medium. Our data clearly indicate that the manipulation of hESC culture conditions causes phenotypic changes of the cells that were reflected also at the level of gene expression. Such changes may have fundamental importance for hESCs, and gene expression changes should be monitored as a part of cell culture optimization aiming at a clinical use of hESCs for cell transplantation. STEM CELLS 2006;24:151–167

INTRODUCTION

cultured in the presence of animal proteins express an immunogenic nonhuman sialic acid [3]. Therefore, nonhuman sera and feeder cells in the hESC culture systems should be replaced by alternatives. The establishment of hESC lines was first described by using fetal calf serum (FCS)-containing medium and fetal mouse fibroblasts as feeder cells [4, 5]. hESC lines have

Human embryonic stem cells (hESCs) provide new opportunities for cell transplantation in severe degenerative diseases [1, 2]. To enable the use of hESCs in cell transplantation in humans, it is essential to eliminate the risk of infection transmitted by animal pathogens. Recent data have also suggested that hESCs

Correspondence: Heli Skottman, Ph.D., REGEA Institute for Regenerative Medicine, University of Tampere and Tampere University Hospital, 33520 Tampere, Finland. Telephone: 358-3-3551-4119; Fax: 358-3-3551-8498; e-mail: [email protected] Received August 11, 2004; accepted for publication June 27, 2005; first published online in STEM CELLS EXPRESS August 11, 2005. ©AlphaMed Press 1066-5099/2006/$12.00/0 doi: 10.1634/stemcells.2004-0189

STEM CELLS 2006;24:151–167 www.StemCells.com

Molecular Signature of hESCs in Serum-Free Medium

152

recently been successfully cultured under serum-free conditions using serum replacement (SR) [6] and without feeder cells on matrigel using mouse embryonic fibroblast (MEF)-conditioned medium [7, 8]. Completely serum-free culturing conditions using SR medium and postnatal human fibroblasts as feeder cells have been described [9]. Recently, a feeder- and serum-free system for culturing hESCs on fibronectin matrix was described [10]. In the study, transforming growth factor ␤1 (TGF␤1), basic fibroblast growth factor (bFGF), and leukemia inhibitory factor (LIF) were added to the 20% SR culture medium, and the cells were successfully propagated. However, judging by morphology, differentiation of cells was seen in the colonies [10]. hESCs have previously been shown to maintain all ES cell characteristics as nondifferentiated cells when cultured either in FCS- or SR-containing medium with adequate supplements [4 – 6]. In mouse ESC cultures, the feeder layer can be replaced by addition of LIF to the growth medium [11]. However, hESCs seem to lack this response to LIF [4, 5] and therefore the use of LIF in the growth medium is probably unnecessary. The SR medium has been shown to require the supplementation of bFGF to prevent differentiation of hESCs [6]. FCS is a complex mixture containing compounds with both beneficial and adverse effects on hESCs, and FCS batches vary in capability of maintaining hESCs at an undifferentiated stage. To avoid these problems and to develop entirely animal-free culture conditions, we have optimized the SR culture conditions for our hESC lines that have been derived and growing on postnatal human foreskin fibroblasts [12]. Our hESCs maintained their pluripotency both in serum-containing and SR medium, but in SR medium they proliferated faster [13]. It is unknown why SR medium supports the growth of hESCs better than serum-containing medium. Because both culture media support the pluripotent nondifferentiated growth of hESCs, the factors responsible for the better growth rate seen in SR medium are unlikely to regulate the mechanisms responsible for the pluripotency of hESCs. In the present work, we studied the effects of two different culture media, a serum-containing and a serum-free one, on the gene expression profiles of hESCs. We used DNA microarray analysis, which allows a large-scale gene expression profiling from a limited amount of starting material. High-density oligonucleotide microarrays, containing most of the known human genes as well as thousands of unknown ESTs, are especially useful tools for studying unknown changes in hESCs caused by different culture media.

MATERIALS

AND

METHODS

hESC Lines The hESC lines HS181, HS235, and HS237 from Karolinska University Hospital Huddinge were derived and cultured in

FCS-containing medium on human foreskin fibroblast feeder cells. The derivation and characterization of the line HS181 has been previously described [12], and HS235 and HS237 lines have been characterized accordingly. All of these lines have a karyotype 46XX [14]. They express markers typical for hESCs, alkaline phosphatase, SSEA-4, TRA-1-0, TRA-1-1, and Oct-4 but are SSEA-1 negative. The pluripotency has been demonstrated by the formation of teratomas when injected to severe combined immunodeficiency mice and by in vitro differentiation of embryoid bodies expressing markers from three embryonic germ cell layers [12, 13].

hESC Cultures The human foreskin fibroblasts (CRL-2429, American Type Culture Collection, Manassas, VA, http://www.atcc.org) used as feeder cells were mitotically inactivated by irradiation (35 Gy) and cultured in Iscove’s medium supplemented with 10% FCS (Stem Cell Technologies, Vancouver, British Columbia, Canada, http://www.stemcell.com). After the formation of a confluent monolayer, feeder cells were cultured in hESC medium containing serum or SR. The hESCs were cultured on feeder cells in two different media. The serum medium consisted of Knockout Dulbecco’s modified Eagle’s medium (DMEM), 20% FCS (Stem Cell Technologies), 2 mM L-glutamine, 0.1 mM beta-mercaptoethanol, 1% penicillin streptomycin, 1% nonessential amino acid, and 1 ␮l/ml recombinant human LIF (Chemicon, Temecula, CA, http://www.chemicon.com). Originally the hESCs were cultured in serum medium but moved into SR medium for 12 weeks (⬃16 passages). The SR medium consisted of Knockout DMEM, 20% Knockout SR, 2 mM L-glutamine, 1% penicillin streptomycin, 1% nonessential amino acids, 0.5 mM beta-mercaptoethanol, 1% ITS (SigmaAldrich, St. Louis, http://www.sigmaaldrich.com), and 8 ng/ml bFGF (R&D Systems, Minneapolis, http://www.rndsystems. com). All of the chemicals were from Gibco (Grand Island, NY, http://www.invitrogen.com) unless stated otherwise.

Oligonucleotide Microarray Analysis Using the HS237 Line The total RNA was isolated from 5 to 10 HS237 hESC colonies cultured for 37 passages using an RNeasy mini kit (Qiagen Nordic, West Sussex, UK, http://www1.qiagen.com). hESC colonies were selected under a light microscope to avoid colonies that might have started to differentiate. Because the possible RNA contamination from the fibroblasts in the isolated RNA cannot entirely be avoided, the total RNA from human foreskin fibroblast was included in the study as a control sample. Human foreskin fibroblasts were plated in serum-containing medium and transferred into SR-containing medium for 1 week before the RNA isolation. From all RNA samples, 100 ng of total RNA

Skottman, Stromberg, Matilainen et al. was used as a starting material for the microarray sample preparation. The sample preparation was performed according to the Affymetrix two-cycle GeneChip eukaryotic small-sample target labeling assay version II (Affymetrix, Santa Clara, CA, http:// www.affymetrix.com). Biotin-labeled cRNA 15 ␮g was fragmented and hybridized to HG-U133A and HG-U133B arrays. Arrays were stained and scanned according to Affymetrix protocols. Microarray analyses were performed for two biological replicates of HS237 cells and fibroblasts. The gene transcript levels were determined from data images with algorithms in the GeneChip Microarray Suite software (Affymetrix MAS version 5.0), and further analysis of data was performed with Kensington software (InforSense, London, http://www.inforsense.com). At the detection level, each probe set was assigned to call of present (P), absent (A), or marginal (M). A gene with detection call “present” was considered to be expressed. The comparison level analysis of the cells cultured in serum or SR medium defined a gene as differentially expressed if change call (change p ⬍ .05) was increased (I/MI) or decreased (D/MD). A gene was defined as significantly upregulated if the signal fold-change (FC) between the target sample and the reference sample was larger than 2 and target sample was present. A gene was defined as significantly downregulated if FC was less than -2 and the reference sample was present. As recommended by Affymetrix, the probe sets were excluded if the detection call for both the target and the reference was “absent” or if the change call gave no change (NC) (change p ⬎ .05) in the comparison analysis. Only genes that fulfilled all the filtering criteria reproducibly in two biological replicates were considered significant. The signal values and detection calls for all the genes gained with algorithms in the MAS software are available upon request.

Validation of the Microarray Results To validate microarray results gained with HS237 line, the HS181 and HS235 hESC lines were cultured in similar culture conditions as the HS237 line. At the onset of RNA isolation, the hESC lines HS181 and HS235 had been cultured for 34 and 50 passages, respectively. For the validation of the microarray results with TaqMan real-time reverse transcription–polymerase chain reaction (RT-PCR) and RT-PCR, a set of 15 genes was selected (Table 1). Primers and probes were designed by Primer Express software (Applied Biosystems, Foster City, CA, http:// www.appliedbiosystems.com) and made by CyberGene AB (Huddinge, Sweden, http://www.cybergene.se) and DNA Technology (Aarhus, Denmark, http://www.dna-technology.dk). The sequences for the primers and probes and the cycle conditions for RT-PCR are listed in Table 1. Total RNA 100 ng from cells was used for cDNA synthesis using Sensiscript Reverse Transcription kit (Qiagen). The TaqMan experiments were perwww.StemCells.com

153

formed using the ABI PRISM 7700 Sequence Detection System (Applied Biosystems) as described previously [15]. The relative levels of the target mRNA expression were normalized against GAPDH expression. All measurements were performed in duplicate in two separate runs using samples derived from two biological replicates. The standard deviation of individual TaqMan measurements had to be less than 0.5.

RESULTS DNA Microarray Analyses Expression Profiling of hESCs Cultured in SerumContaining and SR Media The mRNA expression patterns of HS237 hESCs cultured in serum-containing or SR medium were compared using Affymetrix microarrays, which enable a large-scale gene expression profiling of ⬃39,000 transcripts and variants, including more than 33,000 well-substantiated human genes. Biological replicates for all of the samples showed a correlation coefficient ⱖ 0.963, indicating a high reproducibility of the data. The average signal fold-change (AverageFC) of the genes expressed in hESCs cultured in serum-containing versus SR medium was calculated from two biological replicates. To exclude redundant genes included in arrays (total number of probe sets, 44,928), Unigene IDs were used in analyses. Using microarrays, we identified 10,460 nonredundant transcripts (15,707 probe sets) expressed both in cells cultured in serum and SR medium (Fig. 1). A total of 947 of these 10,460 shared genes were upregulated (change p ⬍ .05), and 82 of these genes were more than twofold upregulated in the cells cultured in serum-containing medium compared with the cells cultured in SR medium. On the other hand, 470 of the 10,460 shared genes were upregulated (change p ⬍ .05) and 13 were more than twofold upregulated in the cells cultured in SR medium compared with the cells cultured in serum-containing medium. There were also genes that were characteristic for the cells cultured in one of the media used; 156 and 677 transcripts were expressed specifically in cells cultured in SR or serum medium, respectively (Fig. 1). Ninety-one of these medium-specific genes were expressed more than twofold in cells cultured in serum medium compared with the signal level of the cells cultured in SR medium. Thirteen of these medium-specific genes were expressed more than twofold in the cells cultured in SR medium compared with the signal level of the cells in serum-containing medium.

Expression of ESC Markers Among the genes expressed at similar levels (NC, change p ⬎ .05) in cells cultured in either medium, there were many ESC


154

Table 1. Real-time quantitative reverse transcription–polymerase chain reaction (RT-PCR) and RT-PCR primers and probes 1) 5ⴕ-Forward Primer-3ⴕ 2) 5ⴕ-Reverse Primer-3ⴕ 3) 5ⴕ-Probe-3ⴕ for TaqMan

RT-PCR Cycle Conditions Oct-4

Nanog

DNMT3B

FST

Dusp6

LIFR

gp130

GAPDH

UTF1

Gata6

Sox17

(annealing 57°C, 35 cycles)

EOMES


Sulf1


Tbx5


Serpine


markers, such as Oct-4, Nanog, Cripto, and DNMT3B (Table 2). Some of the known ES cell markers, such as LIFR, IL6ST(gp130), connexin 43, CD9, Tcf3, and galanin, were also expressed in the fibroblast control sample. The use of this group of genes as markers in expression studies is complicated because there is a possibility that some signal may come from the feeder cell. Among these genes, gp130 was upregulated more than twofold (AverageFC, 2.39) in cells cultured in serum medium compared with cells in SR medium. Also, LIFR was slightly upregulated (AverageFC, 1.19) in the cells cultured in serum medium, although its relative expression level was very low. The telomere repeat binding factor (TERF) expression was downregulated (AverageFC, ⫺1.1) in cells cultured in serum medium compared with the cells cultured in SR medium. The

1) 2) 3) 1) 2) 3) 1) 2) 3) 1) 2) 3) 1) 2) 3) 1) 2) 3) 1) 2) 3) 1) 2) 3) 1) 2) 3) 1) 2) 3) 1) 2) 1) 2) 1) 2) 1) 2) 1) 2)

TCTGCAGAAAGAACTCGAGCAA AGATGGTCGTTTGGCTGAACAC CCTCTTCTGCTTCAGGAGCTTGGCAA TGCAGTTCCAGCCAAATTCTC CCTAGTGGTCTGCTGTATTACATTAAGG TCCAAAGCAGCCTCCAAGTCACTGG CGAAAGGATGTTTGGCTTTCC GACCTTCCCAGCAGCTTCTG ACAGACGTGTCCAACATGGGCCGT GTAAGTCGGATGAGCCTGTCTGT CAGCTTCCTTCATGGCACACT CCAGTGACAATGCCACTTATGCCAGC GCTGTGGCACCGACACAGT ACTCGCCGCCCGTATTCT CTCTACGACGAGAGCAGCAGCGACTG ACTGTGGAAGATATAGCTGCAGAAGA CACTGTTGCTGTCTATGGATCTAGGA ATAAAACTGCGGGTTACAGACCTCAGGCC GCCTCAACTTGGAGCCAGATT GTTTAAGGTCTTGGACAGTGAATGAAG CTCCTGAAGACACAGCATCCACCCGA GTTCGACAGTCAGCCGCATC GGAATTTGCCATGGGTGGA ACCAGGCGCCCAATACGACCAA GGCACCTGGGCGACATC TCCACGTGCTGGTTCAAGGT AACATCCTGGGCCCGCTGCG GAGCACCAATCCCGAGAACA GCGAGACTGACGCCTATGTAGA CCCATCTTGACCCGAATACTTGAGCTCG CGCACGGAATTTGAACAGTA CACACGTCAGGATAGTTGCAG ACCCCCTTCCATCAAATCTC CCATGCCTTTTGAGGTGTCT TCTTGGGGAGCTGAATAGGA TGTAAGACCTCACCAAGTTCTGA AGCACTTCTCCGCTCACTTC CCGTGCACAGAGTGGTACTG TCCAGTTTTGTCCCAGATGA ATCGAGGTGAACGAGAGTGG

transcriptional coactivator UTF1 was expressed only in the cells cultured in SR medium (Average FC, ⫺2.0). Markers for differentiated cell phenotypes, such as GFAP, SOX-1, MYF5, and GATA-2, were not expressed in hESCs cultured in either medium. Among the early differentiation markers (keratin 8, keratin 18, beta tubulin 5, cardiac actin, and troponin T1) reported by another study [16], all genes except for beta tubulin 5 were expressed in hESCs cultured in either medium, suggesting that these genes may also be useful markers for early differentiation in our culture conditions. According to our microarray results, the differentiation markers GATA6 and SOX17 were expressed only in cells cultured in serum medium (Average FC, 3.43 and 3.1, respectively). All of these differentiation markers, except for troponin T1, beta tubulin 5, and

Skottman, Stromberg, Matilainen et al.

Figure 1. Combination Venn diagram of shared and specific genes expressed in HS237 hESCs cultured in two different culture media, serum-containing and SR medium. The region of overlap between all areas indicates the number of genes expressed in hESCs cultured in either culture medium. Regions not overlapping indicate genes expressed specifically in hESCs cultured in the indicated culture medium. Ellipses represent differentially expressed (p ⬍ .05) genes; rectangles, greater than twofold differentially expressed genes in hESCs cultured in serum or SR medium. Abbreviations: hESC, human embryonic stem cell; SR, serum replacement.

SOX17, were also expressed in the fibroblast control sample, but the expression was higher in hESCs.

Differentially Expressed Genes in hESCs Cultured in Serum-Containing and SR Media Among the 10,460 genes expressed in cells cultured in serum and SR media, we detected 1,417 differentially expressed genes, 947 upregulated in cells cultured in serum medium and 470 upregulated in cells cultured in SR medium. The full list of these genes is available on request. For the analyses we used a hierarchical clustering of log-transformed signal values. The gene expression profiles obtained using the biological replicates of each sample clustered well together. As expected, the hESCs cultured in different conditions were more closely related to each other than to fibroblast samples (Fig. 2A). On the other hand, we performed clustering analyses also for the genes whose expression showed more than twofold difference between the cells cultured in serum and SR media. In those analyses, the gene expression profiles of the cells cultured in serum medium clustered more closely together with those of fibroblasts, perhaps indicating that the cells cultured in serum medium were more differentiated than those cultured in SR medium (Fig. 2B). A total of 1,417 differentially expressed genes were further classified by biological function using NetAffex (Affymetrix) www.StemCells.com

155

database. Approximately 40% of these differentially expressed genes had no known biological function (Fig. 3). Among the 947 genes whose expression was increased in cells cultured in serum medium, there were genes mainly involved in signaling, development, protein modifications, proliferation/regulation of proliferation or cell cycle, regulation of transcription, and cell adhesion (Fig. 3A). On the other hand, among the 470 genes whose expression was increased in cells cultured in SR medium, there were genes mainly involved in signaling, development, and cell proliferation (Fig. 3B). Next, we used twofold cut-off in expression level to identify the most differentially expressed genes. Eighty-two of the shared but differentially expressed genes were upregulated more than twofold in the cells cultured in serum medium compared with the cells cultured in SR medium (Fig. 1, Table 3). This group included genes involved in the regulation of transcription (EOMES, MAFF), development (COL12A1, COL2A1, FBN1, ACTC, ACTA2, TAGLN2, CALD, VEGFC, POSTN, GREM1), the regulation of cell growth (IGFBP3, IGFBP7, CTGF), and cell adhesion (FLRT3, OSF2, TNC, TGFBI, CD44, CDH11). Among the shared but differentially expressed genes, only 13 of the genes were upregulated more than twofold in cells cultured in SR medium compared with the cells cultured in serum medium, including genes involved in metabolism (NME4, INDO) and 6 genes without a known biological function (Fig. 1, Table 4). hESCs cultured in serum or SR medium also expressed genes that were specific for a certain culture medium (Fig. 1). Ninety-one of these medium-specific genes were expressed more than twofold in cells cultured in serum medium but not in cells cultured in SR medium (Table 5). These 91 transcripts included genes involved in development (ACTA1, COMP, BDNF, NRPI, Nodal), development/regulation of transcription (HOXA1, GATA6, SOX17, TBX5, TBX3, SIX1, BHLHB2, FOXD1, INSM1, NR2F1), and cell adhesion (ALCAM, PARVA, PCDH10). The 13 transcripts, expressed only in cells cultured in SR medium, included two genes involved in ion transport (KCNK12, SCNNIG), UTF1 involved in the regulation of transcription, and seven genes without a known biological function (Table 6). Given the differences observed in the growth rate of hESCs cultured in serum and SR medium [13], we assumed that cells may display additional differences in their signaling pathways related to cell growth. Wnt signaling has been implicated in maintaining undifferentiated ES cells. There were 35 transcripts involved in the Wnt signaling pathway that were expressed in cells cultured either in serum or SR media. Six of these genes (PAFAH1B1, PLAU, LDLR, WNT5A, CCND1/Cyclin D1, GSK3B) were upregulated in the cells cultured in serum medium compared with the cells cultured in SR medium but by less


156

Table 2. A set of embryonic stem cell (ESC) markers expressed according to microarray results at similar levels in HS237 cells cultured in serum-containing or serum-replacement medium ESC Marker Nanog [37, 38] POU5F1, Oct-4 [35, 39] Cripto (TDGF1) [40] Connexin 45 [16, 41] DNMT3B [42] ACVR2B [16] ABCG2 [16] PODXL [43] REX-1 (Zfp42) [44] LIN28 [16] GDF3 [45, 46] PUM1 and PUM2 [47] Nanos 1 [48]

Unigene ID

Annotation

Hs.329296 Hs.249184 Hs.385870 Hs.532593 Hs.251673 Hs.23994 Hs.194720 Hs.16426 Hs.458361 Hs.86154 Hs.86232 Hs.153834/Hs.133543 Hs.351851

homeobox transcription factor homeobox transcription factor teratocarcinoma-derived growth factor 1 gap junction protein, alpha 7, 45 kDA DNA (cytosine-5-)-methyltransferase 3 beta activin A receptor, type II ATP-binding cassette, subfamily G, member 2 podocalyxin-like zinc finger protein 42 lin-28 homologue (C. elegans) growth-differentiation factor 3 Pumilion 1 and 2 Nanos homologue 1

than twofold. Recently, one group reported a finding that the activation of the canonical Wnt pathway by inactivation of GSK3 is sufficient to maintain the self-renewal of hESCs [17]. In our study, the level of GSK3 expression was slightly upregulated (AverageFC, 0.7) in the cells cultured in serum medium, suggesting possible early differentiation of the cells compared with the cells cultured in SR medium. The negative regulation of the TGF␤ pathway might be critical for the maintenance of the undifferentiated hESCs [18]. The hESCs that were cultured in either one of the culture media expressed 26 genes involved in the TGF␤ pathway. Seven of these genes (THBS1, TGIF, FST, SPP1, SMAD7, TGFBR1, SMAD1) were upregulated, and THBS1 and FST were upregulated more than twofold in the hESCs cultured in serum medium compared with those in SR medium. According to microarray results, one of the TGF␤ pathway target genes, Serpine, was highly expressed (AverageFC, 3.4) in cells cultured in serum medium, and these results suggested a higher activity of TGF␤ pathway in the hESCs cultured in serum medium compared with those in SR medium. Interestingly, the expression of Nodal (mesodermal and endodermal inducer) and, on the other hand, nodal signaling inhibitors CER1, LeftB, and FST was also upregulated more than twofold in cells cultured in serum medium compared with those cultured in SR medium, suggesting higher activity of Nodal signaling pathway in cells cultured in serum medium. Because the SR medium requires a supplementation with bFGF to prevent the differentiation of hESCs [6], we checked the expression of genes related to the FGF signaling pathway. Among these genes, the expression of endogenous FGF2 and FGF13 was slightly upregulated (AverageFC, 1.9 and 1.0, respectively) in cells cultured in serum medium. One of the bFGF receptors, FGFR1, was upregulated (AverageFC, ⫺1.5) in cells cultured in SR medium, possibly indicating a higher FGF receptor activity in the cells cultured in SR medium containing bFGF. The enrichment of FGFR1 expression in the undifferen-

tiated hESCs compared with differentiated cells has been published [18].

Comparison of Differentially Expressed Genes with Previously Reported Microarray Results To determine if differentially expressed genes in HS237 hESC line cultured in serum and SR media include ESC genes reported previously, we compared our data with the microarray data of hESCs cultured in various other culture conditions. A summary of the data comparison is presented in Table 7. Sato et al. [18] reported 918 genes that were upregulated in undifferentiated hESCs cultured with matrigel and MEF-conditioned medium compared with the differentiated cells. In our study, 45 of the genes upregulated in the cells cultured in serum medium compared with the cells in SR medium were found in their list of genes upregulated in undifferentiated cells. On the other hand, in our study, 46 of the genes upregulated in the cells cultured in SR medium compared with cells in serum medium were found in their list of genes upregulated in undifferentiated cells. Sperger et al. [19] compared the gene expression profiles of five hESC lines cultured with MEF feeders and 20% SR medium to somatic cell lines. By using their Unigene ID annotation, we were able to compare their data with ours: 55 of the genes upregulated in our study in the cells cultured in serum medium compared with the cells in SR medium were among the genes found to be differentially expressed between hESC and somatic cell lines. On the other hand, 28 of the genes upregulated in our study in the cells cultured in SR medium compared with the cells in serum medium were among their genes. Bhattacharya et al. [16] compared six hESC lines to universal RNA and reported 92 genes that were upregulated in all six hESCs compared with universal RNA. Their cell lines were cultured on MEF feeders in a medium containing both serum and SR. In our study, only 11 of the genes upregulated in cells cultured either in serum or SR medium were found in their data. Among these genes was early differentiation marker ␣-cardiac actin, which in


157

Figure 2. Hierarchical clustering of biological replicates. Log-transformed signals from all samples were clustered using hierarchical clustering. The colors indicate the relative expression level of each gene in all analyzed samples, with red indicating higher expression and dark color indicating lower expression. The length of the arms is proportional to the similarities of expression pattern, with shorter length representing a closer relationship. (A): The dendrogram presenting expression pattern of 1,417 differentially expressed genes in HS237 hESCs cultured in serum and SR media. (B): The dendrogram presenting expression pattern of 95 genes differentially expressed greater than twofold in HS237 hESCs compared with those cultured in serum or SR media. A full list of genes differentially expressed greater than twofold is presented in Tables 3 and 4. Samples in both dendrograms are hESCs in SR medium (1, 2), hESCs in serum medium (3, 4), and fibroblasts (5, 6). Abbreviations: hESC, human embryonic stem cell; SR, serum replacement.

www.StemCells.com

158


Figure 3. A pie chart presenting biological function of 1,417 differentially expressed genes in HS237 human embryonic stem cells when comparing cells cultured in serum or SR media. (A): Percentages of 947 genes upregulated in cells cultured in serum medium compared with cells in SR medium. (B): Percentages of 470 genes upregulated in cells cultured in SR medium compared with cells in serum medium. Annotation was made using NetAffex database (Affymetrix). Abbreviation: SR, serum replacement.

our study was only expressed in the cells cultured in serum medium but not in the cells cultured in SR medium, suggesting the presence of cells with a differentiated phenotype in the serum medium. These three comparisons suggest that some of the genes in our study, which actually had less than twofold change in expression level between cells cultured in serum and SR media, may have a significant role in hESC characteristics and early differentiation because their expression was changed when undifferentiated hESCs were compared with the differentiated cells in the data published by Sato et al. [18], Sperger et al. [19], and Bhattacharya et al. [16]. This data comparison clearly demonstrates that when the gene expression profiles of hESCs cultured in various types of culture conditions are compared, only few genes are expressed in a similar manner.

Validation of Microarray Results Using Real-Time RT-PCR and RT-PCR Selected genes were further analyzed by TaqMan real-time RT-PCR (Table 1), and the average fold-change values of the real-time RT-PCR from two separate runs are presented in Figure 4. Based on the microarray results, the expression of many known ESC markers were expressed at similar levels in HS237 cells cultured either in serum or SR medium. These findings were further studied by analyzing the kinetics of DNMT3B, Oct-4, and Nanog expression also in two other hESC lines cultured in a serum-containing and SR medium. TaqMan detection verified that the expression of Nanog was downregulated in all three hESC lines in serum medium compared with cells in SR medium, but by less than twofold. The expression of Oct-4 was downregulated in all three hESC lines cultured in serum medium, and this downregulation was more than fivefold in the HS181 line compared with the cells in SR medium. The

expression of DNMT3B was also downregulated in all three hESC lines cultured in serum medium, and this downregulation was more than 25-fold in the HS181 line compared with the cells in SR medium. According to microarray results, some of the genes were expressed more than twofold in HS237 cells cultured in serum medium compared with cells cultured in SR medium. To further study these findings, the expression of FST and DUSP6 was studied in all three hESC lines. TaqMan analysis demonstrated that the expression of these genes was increased more than twofold only in HS237 cells cultured in serum medium compared with those cultured in SR medium showing hESC line– specific expression pattern. To further study our microarray findings showing that LIF receptors LIFR and gp130 were upregulated in HS237 cells cultured in serum medium, the expression of LIFR and gp130 was studied in all three hESC lines. These results showed that gp130 expression was increased more than twofold in all three lines cultured in serum medium compared with the cells in SR medium. LIFR also showed greater than twofold increases in HS181 and HS237 cells but, on the other hand, greater than twofold decreases in HS235 cells cultured in serum medium compared with the cells cultured in SR medium. According to the microarray results, there were also genes that were characteristic for the cells cultured in one of the media used; GATA6 was expressed specifically at a greater than twofold higher level in HS237 cells cultured in serum medium and UTF1 in HS237 cells cultured in SR medium. These findings were further confirmed in HS181 and HS235 lines using TaqMan detection. RT-PCR showed the expression of Eomes in all three hESC lines cultured in both culture media, and this result was consistent with DNA microarray data from the HS237 line. Interestingly, SULF1 was not expressed in HS235 cells cultured in serum-containing medium but was expressed in both


159

Table 3. Genes expressed in HS237 hESCs cultured in either medium but upregulated more than twofold in hESCs cultured in serum medium compared with hESCs cultured in SR medium

Gene Symbol IGFBP3 EOMES POSTN ANXA1 ACTA2 SULF1 ANXA3 CTGF SPP1 KIAA1199 GREM1 COL12A1 TAGLN TGFBI URB IGFBP7 FBN2 — TNC — CDH11 CALD1 ACTC FST COL2A1 COL8A1 CD44 FER1L3 LUM TNFAIP6 ARK5 NNMT COL5A2 HIC LOX PAPPA HAK S100A10 DKFZp434L142 FLJ23091 AMIGO2 DUSP6 LEFTB SIPA1L2 CYBRD1 DDR2 IL6ST SLC40A1 FLRT3 C14orf31 MGC4677 RBM24 THBS1 PTX3 IER3 NEK7 WNT5A

Gene Description

Average FC to Cells in SR Medium

Unigene ID

insulin-like growth factor binding protein 3 eomesodermin homologue (Xenopus laevis) periostin, osteoblast-specific factor annexin A1 actin, alpha 2, smooth muscle, aorta sulfatase 1 annexin A3 connective tissue growth factor secreted phosphoprotein 1 (osteopontin) KIAA1199 gremlin 1 homologue, cysteine knot superfamily (Xenopus laevis) collagen, type XII, alpha 1 transgelin transforming growth factor, beta-induced, 68 kDa steroid-sensitive gene 1 insulin-like growth factor binding protein 7 fibrillin 2 (congenital contractural arachnodactyly) Full-length cDNA clone CS0DF022YN12 of fetal brain of Homo sapiens tenascin C (hexabrachion) full-length cDNA clone CS0DN003YO15 of adult brain of Homo sapiens cadherin 11, type 2, OB-cadherin (osteoblast) caldesmon 1 actin, alpha, cardiac muscle follistatin collagen, type II, alpha 1 collagen, type VIII, alpha 1 CD44 antigen (homing function and Indian blood group system) fer-1-like 3, myoferlin (C. elegans) lumican tumor necrosis factor, alpha-induced protein 6 AMP-activated protein kinase family member 5 nicotinamide N-methyltransferase collagen, type V, alpha 2 I-mfa domain–containing protein/I-mfa domain–containing protein lysyl oxidase pregnancy-associated plasma protein A, pappalysin 1 heart alpha-kinase S100 calcium-binding protein A10 hypothetical protein DKFZp434L142 putative NFkB-activating protein 373 amphoterin-induced gene 2 dual-specificity phosphatase 6 left-right determination, factor B signal-induced proliferation-associated 1 like 2 cytochrome b reductase 1 discoidin domain receptor family, member 2 interleukin 6 signal transducer (gp130, oncostatin M receptor) solute carrier family 40 (iron-regulated transporter), member 1 fibronectin leucine-rich transmembrane protein 3 chromosome 14 open reading frame 31 hypothetical protein MGC4677 RNA-binding motif protein 24 thrombospondin 1 pentaxin-related gene, rapidly induced by interleukin-1 beta immediate early response 3 NIMA (never in mitosis gene a)-related kinase 7 wingless-type MMTV integration site family, member 5A

4.4 4.3 4.2 4.1 3.7 3.6 3.5 3.5 3.4 3.4 3.3 3.3 3.3 3.2 3.0 2.9 2.9 2.9 2.8 2.8 2.8 2.7 2.7 2.7 2.6 2.6 2.6 2.6 2.6 2.6 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.4 2.4 2.4 2.4 2.4 2.4 2.3 2.3 2.3 2.3 2.3 2.3 2.3 2.2

Hs.450230 Hs.147279 Hs.136348 Hs.287558 Hs.208641 Hs.409602 Hs.442733 Hs.410037 Hs.313 Hs.212584 Hs.40098 Hs.101302 Hs.410977 Hs.421496 Hs.356289 Hs.435795 Hs.79432 Hs.5921 Hs.98998 Hs.371609 Hs.443435 Hs.443811 Hs.118127 Hs.9914 Hs.408182 Hs.114599 Hs.306278 Hs.362731 Hs.406475 Hs.407546 Hs.200598 Hs.364345 Hs.283393 Hs.132739 Hs.102267 Hs.440769 Hs.388674 Hs.143873 Hs.323583 Hs.297792 Hs.121520 Hs.298654 Hs.278239 Hs.406879 Hs.31297 Hs.440905 Hs.71968 Hs.409875 Hs.41296 Hs.439190 Hs.432419 Hs.201619 Hs.164226 Hs.2050 Hs.76095 Hs.24119 Hs.152213 Continued on following page

www.StemCells.com


160 Table 3. (Continued)

Gene Symbol TFPI2 TAGLN2 CAV1 NRIP1 KIAA0992 — FLJ38507 CDK6 RASGRF2 COL15A1 VEGFC — PRNP AMOTL2 SLC7A5 JAK1 MYLK CTHRC1 PDGFC NQO1 MAFF FLNC CAMK2D DKFZP434B044 CYR61

Gene Description tissue factor pathway inhibitor 2 transgelin 2 caveolin 1, caveolae protein, 22 kDa nuclear receptor–interacting protein 1 paladin MRNA; cDNA DKFZp762M127 (from clone DKFZp762M127) colon carcinoma–related protein cyclin-dependent kinase 6 Ras protein–specific guanine nucleotide-releasing factor 2 collagen, type XV, alpha 1 vascular endothelial growth factor C MRNA for hypothetical protein (ORF1), clone 00275 prion protein (p27–30) angiomotin-like 2 solute carrier family 7, member 5 Janus kinase 1 (a protein tyrosine kinase) myosin, light polypeptide kinase collagen triple helix repeat-containing 1 platelet-derived growth factor C NAD(P)H dehydrogenase, quinone 1 v-maf musculoaponeurotic fibrosarcoma oncogene homologue F (avian) filamin C, gamma (actin-binding protein 280) calcium/calmodulin-dependent protein kinase (CaM kinase) II delta hypothetical protein DKFZp434B044 cysteine-rich, angiogenic inducer, 61


Unigene ID

2.2 2.2 2.2 2.2 2.2 2.2 2.2 2.2 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0

Hs.438231 Hs.406504 Hs.74034 Hs.155017 Hs.194431 Hs.12853 Hs.435013 Hs.38481 Hs.410953 Hs.409034 Hs.79141 Hs.152129 Hs.438582 Hs.426312 Hs.184601 Hs.436004 Hs.386078 Hs.283713 Hs.43080 Hs.406515 Hs.460889 Hs.58414 Hs.111460 Hs.262958 Hs.8867

Abbreviations: FC, fold change; hESC, human embryonic stem cell; SR, serum replacement.

Table 4. Genes expressed in HS237 hESCs cultured in either medium but upregulated more than twofold in hESCs cultured in SR medium compared with hESCs in serum-containing medium Gene Symbol NME4 INDO GUCA1A EFHC2 TRO APOBEC3B NMI PGAP1 MEG3 — ARHGEF9 HLA-DQB1 PCSK9

Gene Description

Average FC to Cells in Serum Medium

Unigene ID

nonmetastatic cells 4, protein expressed in indoleamine-pyrrole 2,3 dioxygenase guanylate cyclase activator 1A (retina) EF-hand domain (C-terminal)-containing 2 trophinin/trophinin apolipoprotein B mRNA-editing enzyme, catalytic polypeptide-like 3B N-myc (and STAT) interactor GPI deacylase maternally expressed 3 MRNA; cDNA DKFZp586E2317 (from clone DKFZp586E2317) Cdc42 guanine nucleotide exchange factor (GEF) 9 major histocompatibility complex, class II, DQ beta 1 proprotein convertase subtilisin/kexin type 9

2.6 2.5 2.4 2.2 2.2 2.2 2.2 2.1 2.1 2.0 2.0 2.0 2.0

Hs.9235 Hs.840 Hs.92858 Hs.301143 Hs.434971 Hs.226307 Hs.54483 Hs.528683 Hs.418271 Hs.293563 Hs.54697 Hs.409934 Hs.18844


culture conditions in HS181 and HS237 cells. In contrast, among the genes that were according to microarray results specifically expressed in HS237 cells cultured in serum (SOX17, Serpine, TBX5), only SOX17 showed a specific expression in HS237 cells cultured in serum medium. Serpine and TBX5 expression were detected in all three hESC lines in both conditions when moresensitive RT-PCR was used for detection (Fig. 5).

DISCUSSION We have previously shown that our hESC lines maintain their characteristics as nondifferentiated pluripotent cells when cultured either in a serum-containing or SR medium and that cells proliferate faster in a SR medium [12–14]. In this study, our goal was to study whether these two culture conditions influence the gene expression profiles of hESCs. Moreover, we were


161

Table 5. Genes specifically expressed in HS237 hESCs cultured in serum-containing medium and with greater than twofold level compared with signal level in hESCs cultured in SR medium

Gene Symbol SP5 HOXA1 NR2F1 RGS5 MFAP5 COL12A1 — COMP KIAA0882 SERPINE1 PARVA URB FLJ21069 SLC22A3 — MGC10946 PAG LOC285458 NPPB GLIPR1 HRB2 PPP1R14C — TBX5 NODAL APOA1 INSM1 C9orf13 GATA6 ALDH1A1 SYTL2 BHLHB2 SDCCAG33 ALCAM ARHGAP24 — MBD2 IGFBP3 FBLN2 TBX3 AGPAT3 — SEMA3C EPHA2 B1 PCDH10 FER1L3 P4HA3 ADAMTS5 DKFZP564O0823 KIAA1245 EBF GREM2 AHNAK FN1 TSLP

Gene Description Sp5 transcription factor homeobox A1 nuclear receptor subfamily 2, group F, member 1 regulator of G-protein signaling 5 microfibrillar-associated protein 5 collagen, type XII, alpha 1 full-length cDNA clone CS0DI014YH21 cartilage oligomeric matrix protein KIAA0882 protein serine (or cysteine) proteinase inhibitor, clade E, member 1 parvin, alpha steroid-sensitive gene 1 hypothetical protein FLJ21069 solute carrier family 22 (extraneuronal monoamine transporter), member 3 CDNA FLJ38181 fis, clone FCBBF1000125 — phosphoprotein associated with glycosphingolipid-enriched microdomains hypothetical gene supported by AK096952 natriuretic peptide precursor B GLI pathogenesis-related 1 (glioma) HIV-1 rev binding protein 2 protein phosphatase 1, regulatory (inhibitor) subunit 14C CDNA FLJ44429 fis, clone UTERU2015653 T-box 5 nodal homologue (mouse) apolipoprotein A-I insulinoma-associated 1 chromosome 9 open-reading frame 13 GATA-binding protein 6 aldehyde dehydrogenase 1 family, member A1 synaptotagmin-like 2 basic helix-loop-helix domain–containing, class B, 2 serologically defined colon cancer antigen 33 activated leukocyte cell adhesion molecule Rho GTPase-activating protein 24 — methyl-CpG-binding domain protein 2 insulin-like growth factor–binding protein 3 fibulin 2 T-box 3 (ulnar mammary syndrome) 1-acylglycerol-3-phosphate O-acyltransferase 3 hypothetical gene supported by BX647608 sema domain, immunoglobulin domain EPH receptor A2 parathyroid hormone–responsive B1 gene protocadherin 10 fer-1-like 3, myoferlin (C. elegans) procollagen-proline, 2-oxoglutarate 4-dioxygenase A disintegrin-like and metalloprotease (reprolysin type) with thrombospondin type 1 motif DKFZP564O0823 protein hypothetical protein MGC8902 early B-cell factor gremlin 2 homologue, cysteine knot superfamily (Xenopus laevis) AHNAK nucleoprotein (desmoyokin) fibronectin 1 thymic stromal lymphopoietin


Unigene ID

5.6 5.3 4.6 4.5 4.5 4.4 4.3 4.3 4.3 4.1 4.1 4.1 4.1 4.0 3.9 3.8 3.8 3.7 3.7 3.7 3.7 3.6 3.6 3.5 3.5 3.5 3.5 3.5 3.4 3.4 3.4 3.3 3.3 3.3 3.2 3.2 3.2 3.1 3.1 3.0 3.0 2.9 2.9 2.8 2.8 2.8 2.7 2.7 2.7

Hs.368802 Hs.67397 Hs.361748 Hs.24950 Hs.300946 Hs.101302 Hs.23871 Hs.1584 Hs.411317 Hs.414795 Hs.44077 Hs.356289 Hs.341806 Hs.242721 Hs.143134 Hs.130692 Hs.266175 Hs.381187 Hs.219140 Hs.511765 Hs.269857 Hs.192822 Hs.86538 Hs.381715 Hs.370414 Hs.93194 Hs.89584 Hs.385887 Hs.50924 Hs.76392 Hs.390463 Hs.171825 Hs.284217 Hs.10247 Hs.442801 Hs.124776 Hs.25674 Hs.450230 Hs.198862 Hs.129895 Hs.443657 Hs.411391 Hs.171921 Hs.171596 Hs.79340 Hs.146858 Hs.362731 Hs.440644 Hs.58324

2.7 2.7 2.6 2.6 2.6 2.6 2.5

Hs.105460 Hs.445556 Hs.120785 Hs.207407 Hs.378738 Hs.418138 Hs.389874

Continued on following page

www.StemCells.com


162 Table 5. (Continued)

Gene Symbol COL4A1 ACTA1 RUNX2 FAM46A SIX1 LOC339535 RGC32 FOXD1 SIAH2 ODZ2 LTBP3 NAV3 — SEMG1 WASPIP AMOTL1 NRP1 NEXN DDX3Y ZNF25 TMEPAI DACH1 THSD2 LMO7 MAP4 — CDC42EP3 MFAP3L PDE4DIP FBLN5 FLJ23091 LAMA2 HDLBP CER1 BDNF


Gene Description collagen, type IV, alpha 1 actin, alpha 1, skeletal muscle runt-related transcription factor 2 family with sequence similarity 46, member A Sine oculis homeobox homologue 1 (Drosophila) hypothetical protein LOC339535 response gene to complement 32 forkhead box D1 seven in absentia homologue 2 (Drosophila) Odz, odd Oz/ten-m homologue 2 (Drosophila) latent transforming growth factor beta–binding protein 3 neuron navigator 3 CDNA FLJ43100 fis, clone CTONG2003100 semenogelin I Wiskott-Aldrich syndrome protein-interacting protein angiomotin-like 1 neuropilin 1 nexilin (F actin–binding protein) DEAD (Asp-Glu-Ala-Asp) box polypeptide 3, Y-linked zinc finger protein 25 (KOX 19) transmembrane, prostate androgen–induced RNA dachshund homologue 1 (Drosophila) thrombospondin, type I, domain-containing 2 LIM domain 7 microtubule-associated protein 4 CDNA: FLJ23131 fis, clone LNG08502 CDC42 effector protein (Rho GTPase–binding) 3 Microfibrillar-associated protein 3-like phosphodiesterase 4D–interacting protein (myomegalin) fibulin 5 putative NFkB-activating protein 373 laminin, alpha 2 (merosin, congenital muscular dystrophy) high-density lipoprotein–binding protein (vigilin) cerberus 1 homologue, cysteine knot superfamily (Xenopus laevis) brain-derived neurotrophic factor

2.5 2.5 2.5 2.5 2.5 2.4 2.4 2.3 2.3 2.3 2.3 2.3 2.3 2.3 2.3 2.3 2.2 2.2 2.2 2.2 2.2 2.2 2.2 2.2 2.2 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.0 2.0 2.0

Unigene ID Hs.437173 Hs.1288 Hs.122116 Hs.10784 Hs.54416 Hs.532047 Hs.76640 Hs.96028 Hs.20191 Hs.173560 Hs.289019 Hs.306322 Hs.440492 Hs.1968 Hs.401414 Hs.292781 Hs.173548 Hs.22370 Hs.99120 Hs.5856 Hs.83883 Hs.63931 Hs.135254 Hs.5978 Hs.31095 Hs.301296 Hs.352554 Hs.178121 Hs.502577 Hs.11494 Hs.297792 Hs.445120 Hs.427152 Hs.248204 Hs.439027

Abbreviations: FC, fold change; hESC, human embryonic stem cell; SR, serum replacement. Table 6. Genes specifically expressed in HS237 hESCs cultured in SR medium and with greater than twofold level compared with signal level in hESCs cultured in serum-containing medium Gene Symbol

Gene Description

Average FC to Cells in Serum Medium

MGC34827 SCNN1G — —

hypothetical protein MGC34827 sodium channel, non–voltage-gated 1, gamma Similar to 6-pyruvoyl-tetrahydropterin synthase LOC388889 /// CDNA FLJ25967 fis, clone CBR01929

4.8 4.1 3.9 3.9

WDR1 KCNK12 RPL23 — CDK11 ZNF334 C10orf110 UTF1 PIWIL2

WD repeat domain 1 potassium channel, subfamily K, member 12 ribosomal protein L23 LOC388920 cyclin-dependent kinase (CDC2-like) 11 zinc finger protein 334 chromosome 10 open reading frame 110 undifferentiated embryonic cell transcription factor 1 piwi-like 2 (Drosophila)

2.5 2.4 2.4 2.2 2.1 2.1 2.1 2.0 2.0


Unigene ID Hs.31110 Hs.145645 Hs.14204 Hs.355618 Hs.517501 Hs.85100 Hs.252617 Hs.406300 Hs.120377 Hs.129836 Hs.192662 Hs.283652 Hs.458406 Hs.528649


163

Figure 4. Verification of microarray results with TaqMan real-time quantitative RT-PCR. The expression of DNMT3B, gp130, UTF1, GATA6, FST, LIFR, DUSP6, OCT-4, and Nanog was analyzed using real-time RT-PCR. RNA isolated from three human embryonic stem cell lines cultured in serum-containing and SR medium was analyzed. All measurements were performed in duplicate in two separate runs using samples derived from two biological replicates. The relative levels of gene expression of target mRNA was normalized against GAPDH expression. The average ⌬⌬Ct values of gene expression are presented as fold change (fold change ⫽ 2⫺⌬⌬Ct) for cells cultured in serum medium as relation to expression in cells cultured in SR medium. Abbreviations: RT-PCR, reverse transcription–polymerase chain reaction; SR, serum replacement.

Figure 5. RT-PCR analysis of selected genes in three hESC lines cultured in serum-containing and SR media. A subset of genes was analyzed, including two genes expressed in HS237 cells cultured in either medium (SULF1 and Eomes) and three genes expressed only in HS237 cells cultured in certain culture medium (SOX17, Tbx5, Serpine). Only SOX17 showed medium-specific expression in HS237 cells cultured in serum medium, and other analyzed genes were expressed in all hESC lines cultured in either medium. ⫺ indicates RT-PCR–negative control. Abbreviations: hESC, human embryonic stem cell; RT-PCR, reverse transcription–polymerase chain reaction; SR, serum replacement.

interested in elucidating whether factors in SR medium support the undifferentiated growth of hESCs better than those in serumcontaining medium. According to microarray results, there were many known ES cell markers among the genes whose expression profiles were similar in the cells cultured in either medium, supporting our assumption that the factors responsible for the better growth rate seen in SR medium are unlikely to regulate functions behind the pluripotency of hESCs. Using real-time RT-PCR, a slight downregulation in the expression of DNMT3B, Oct-4, and Nanog in all three hESC lines cultured in serum medium was seen when compared with the gene expression in cells cultured in a SR medium (Fig. 4). On the other hand, according to the microarray results, the downstream genes of Oct-4, such as platelet-derived growth www.StemCells.com

factor ␣ receptor and Osteopontin [20], were upregulated in HS237 cells cultured in serum medium, showing that the slight decrease of Oct-4 expression had no effect on their expression. According to microarray results, the expression of Sox-2 was also slightly downregulated (AverageFC, ⫺1.5) in HS237 cells cultured in serum medium compared with the cells cultured in SR medium. Sox-2 cooperates with Oct-4 in stimulation of UTF1 transcription [21]. In mouse ESCs, the disappearance of UTF1 expression precedes that of Oct-4 and Sox-2 [21]. Similarly, in our culture conditions, UTF1 may be an important marker for an early differentiation because it was downregulated more than twofold in all three hESC lines cultured in serum medium compared with those cultured in SR medium (Fig. 4). Also, the downregulation of TERF in HS237 cells cultured in serum medium compared with the cells cultured in a SR medium suggested a presence of differentiated cells in serum medium. Although human ESCs seem to lack the response to LIF [4, 5], it is possible that added LIF in serum medium may induce the expression of LIF receptors (LIFR and gp130). Recent data showed that the STAT3 signaling pathway can be stimulated by LIF in hESCs but that the level of activation is much lower than in mouse ESCs [17]. In HS237 cells, STAT3 and SOCS-1, inhibitor of STAT3 signaling, were expressed in the cells cultured in either culture medium without changes, but SOCS-3 was upregulated (AverageFC 1.5) in the cells cultured in serum medium. This is in concordance with the results in mouse ESCs, which showed that the expression of SOCS-3 (but not SOCS-1) was increased in the presence of LIF [22]. It has been suggested that constitutive expression of SOCS1 and SOCS3 may inhibit LIF


164

Table 7. Comparison of our results on genes differentially expressed between HS237 hESCs cultured in serum-containing and SR medium with previously reported microarray results of other hESC lines cultured in various culture conditions 91 Genes Expressed 947 Genes Upregulated in 470 Genes Upregulated in More Than Twofold Cells Cultured in SerumCells Cultured in Only in Cells Cultured Containing Medium SR-Containing Medium in Serum Medium Sato et al. [18] 918 genesa

AHCY, CD59, EPB41L2, MYST2, FUS, NDUFB8, ACTA1, DDX3Y NASP, NID, ALDH3A2, ALDOC, BIRC5, CASP3, TGIF, GULP1, RNASEH2A, KIF5C, MT1X, FGF2, CSPG2, TRIM14, DFFA, TERF1, NAP1L3, GPC4, IMP-3, PCCA, SGNE1, FGF13, DIAPH2, PIM2, RIMS3, DBT, CALB1, EBAF, LEFTB, UGP2, MAP2K6, SEMA3A, CCND1, POLG2, RBPMS, CYB5, DUSP6, ILF3, GCDH, SUMO2, SAP18, SLC39A8, DUT, LTA4H, PRKCBP1, RANBP5, H2AFV, ABCB7, NTHL1, CDT1, RRAS2, MDN1, SILV, KCNQ2, C1orf38, PLEKHE1, TA-LRRP, LRIG1, HLA-DQB1, NEK3, ADCY2, DKC1, USP32, SLD5, OLFM1, LOC283824, SAV1, DATF1, NPTX2, SFRS7, MRS2L, NUDT15, B1, PPP2R2B, BTD, AASS, KLHL7, E2F5, BM039, ZNF451, FGFR1, PASK, TEAD4 ⫹ 2 ESTs (45) PAI-RBP1, POLR3E, PLCXD1, FLJ20171, ZNF331, MGC8407, ZNF589, ZBTB3, MLSTD1, GNB2L1 (46) Sperger et al. [19] CYBRD1, FLJ20421, LOC90799, HSMPP8 ⫹ 1 FLJ21069, PPP1R14C, Negatively significant MANEA, SLC40A1, EST SDCCAG33 913 genesb IRF2BP2, FBXL3, DKFZp761B1514, NDFIP2, ARHGAP18, C14orf32, ZAK, C6orf89, MGC46719, T2BP, RAI1, GLTP, PS1TP4, MGC8902, HECTD2, TIMP2, GOLPH4 ⫹ 2 ESTs (23) Sperger et al. [19] TMEPAI, RAMP, EGLN3, MGC2749, MTA3, BCOR, — Positively significant CXXC5, NID67, FAM44B, PHC1, ERBB3, c 1,471 genes PCDHA1, SEMA6A, LOC55971, PCSK9, LOC83690, MIG-6, NOPE, NEURL, ZCCHC3, HDGFRP3,ADRBK2, NFATC3, MGC18216, RNF149, GRTP1, JUB, NSE2, COL8A1, FAM46B, CDCA7, DKFZp762C1112, LOC152485 ⫹ 9 EST ZNRF3, DAM12, (25) SOCS3, DSC2, MAP3K3, GPR, PARD6G, DRCTNNB1A, ASAM, PAPPA, INM01 ⫹ 4 ESTs (32) Bhattachary et al. [16] KRT18, ACTC, LEFTB, RPL7, SFRP2, EIF4A1 (3) — 92 genesd KRT8, MGC4083, TUBB4, RAMP, SEMA6A (8)

16 Genes Expressed More Than Twofold Only in Cells Cultured in SR Medium

Culturing Condition

—

matrigel, MEFconditioned mediume

—

MEFs ⫹ 20% SR mediumf

EST (Hs.517501) MEFs ⫹ 20% SR mediumf

—

MEFs ⫹15% serum ⫹ 5% SR medium

a

Differentially expressed genes in hESCs compared with differentiated cells. bCorresponding genes found from HG-U133A and B arrays by Unigene annotation. Greater than threefold downregulated genes in hESCs compared with somatic cell lines. cCorresponding genes found from HG-U133A and B arrays by Unigene annotation. Greater than threefold upregulated genes in hESCs compared with somatic cell lines. d Greater than threefold differentially expressed genes in six hESC lines compared with universal RNA. eCulture conditions described elsewhere [8]. fCulture conditions described elsewhere [7]. Abbreviations: hESC, human embryonic stem cell; MEF, mouse embryonic fibroblast; SR, serum replacement.

signal transduction in embryonal carcinoma cells and that the silencing of endogenous SOCS-1 in hESCs could make the culturing of these cells more feasible [23]. In our study, in Smad7 and especially Smad1, involved in a complex formation with STAT3 [24], the expression was slightly increased in HS237 cells cultured in serum medium compared with the cells cultured in SR medium. Recently, one group reported

that four of their hESC lines cultured without LIF expressed no LIF receptors [7], supporting the possibility that LIF may increase the expression of its receptors and genes related to LIF signaling when added in a culture medium. On the other hand, our preliminary data suggest that the expression of these receptors is not induced by LIF in a dose-dependent manner (Aghajanova et al., unpublished data), and it is most

Skottman, Stromberg, Matilainen et al. likely that the expression of LIF receptors is increased during hESC differentiation. As seen in microarray data, among the genes upregulated more than twofold in HS237 cells cultured in serum medium compared with cells in SR medium, many genes known to be expressed in differentiated cells were identified. The cluster analysis of these genes also showed that hESCs cultured in a serum medium clustered more closely to fibroblasts than to the hESCs cultured in SR medium. These results suggest the presence of differentiating cells in serum medium, although these cells were classified as undifferentiated cells by morphology. Based on our data, the downregulation of UTF1 and especially the upregulation of GATA6, an inducer of endodermal differentiation [25], are good markers for early differentiation. GATA6 expression was shown to increase when the expression of Oct-4 was decreased during hESC differentiation [26]. We found 1,417 shared genes that were differentially expressed (p ⬍ .05) in HS237 cells cultured in similar conditions. The difference in the two culture media was that one contained serum and LIF and the other contained SR and bFGF. Approximately 40% of the shared but differentially expressed genes have no known biological function. Some of these genes may explain previously observed differences in the growth rate of hESCs [13], namely that the hESCs were growing faster in SR medium compared with those cultured in serum-containing medium. Several genes involved in the regulation of transcription, RNA processing, and cell proliferation were upregulated in HS237 cells cultured in SR medium, reflecting higher proliferation rates of cells in these conditions. TGF␤ signaling pathway regulates cell proliferation, differentiation, and extracellular matrix production of cells, and TGF␤ has been shown to inhibit the growth of epithelial cells through TGFB receptors [27, 28]. The expression of TGFBR1 receptor was upregulated in HS237 cells cultured in serum medium, suggesting higher activity of TGF␤ signaling pathway in cells cultured in serum medium compared with the cells cultured in SR medium. The downregulation of SULF1 expression may enhance growth signaling in cancer cells, and cells expressing SULF1 have diminished proliferation [29]. Interestingly, in our data, the expression of SULF1 was upregulated (AverageFC, 3.6) in HS237 cells cultured in serum medium, suggesting a possibility that also in hESCs, SULF1 expression may diminish cell proliferation rate. The expression of Nodal and nodal signaling inhibitors CER1, LeftB, and FST was upregulated in HS237 cells cultured in serum medium compared with those in SR medium. CER1 inhibits Nodal signaling during embryonic development in mouse, and cell proliferation is inhibited in the same regions where CER1 is expressed [30]. In our study, CER1 was expressed only in HS237 cells cultured in serum medium, suggesting similar effects in hESCs, i.e., reduced cell proliferation in serum medium. www.StemCells.com

165

Another gene, related to the inhibition of T-cell proliferation [31], GADD45A, was also slightly upregulated (AverageFC, 1.5) in HS237 cells cultured in serum-containing medium, but the functional impact of this and other genes of interest in this study on the proliferation of hESCs remains to be studied. In addition to differences in growth rate in cells cultured in serum or SR medium, we have observed that hESCs attach better to culture plates in serum-containing than in SR medium. In our data, 50 cell adhesion–related genes, such as integrins, laminin receptors, and TGFBR1, were upregulated in HS237 cells cultured in serum medium compared with the cells cultured in SR medium. TGFBR1 has been shown to induce fibronectin expression [27], and we noticed that the expression of fibronectin was upregulated (AverageFC, 1.5) in HS237 cells cultured in serum medium. Homeobox genes such as HOXA1 may control the expression of genes encoding cell adhesion molecules [32]. This gene was specifically and highly expressed in HS237 cells cultured in serum medium. ALCAM, a transmembrane cell adhesion molecule, plays an important role in cell-to-cell interaction and is expressed in human blastocysts but not in eightcell embryos or morulae [33, 34]. Our data suggest that ALCAM plays an important role in hESC attachment as well because it was specifically expressed in HS237 cells cultured in serum medium but not in the cells cultured in SR medium. Generally, gene expression studies have focused on large differences due to the assumption that larger expression changes are biologically more important. However, it has been clearly demonstrated that, for example, differences of less than twofold in the amount of Oct-4 expression have important biological effects in ES cells [35]. If we consider only changes greater than twofold as biologically significant, we may lose a lot of important data, especially if we can rule out the effect of genetic variation on gene expression when a single hESC line is studied. As our data comparison with other microarray data shows, some of the genes that actually had less than twofold changes in expression level between HS237 cells cultured in serum and SR media may have a significant role in hESC characteristics and early differentiation. The detection of small changes challenges the microarray technology, but oligonucleotide microarrays are more sensitive to detect small changes in gene expression than cDNA microarrays [36]. Although hESCs cultured in two different culture conditions have shown similar ESC characteristics, our data clearly indicate that the manipulation of hESC culture conditions results in phenotypic changes of the cells. Such changes are also reflected at the level of gene expression. Because gene expression changes may have a fundamental importance for hESCs, such changes should be monitored as a part of cell culture optimization. The results presented in this study clearly support the previous findings favoring the use of SR medium rather than

166

serum-containing medium for culturing of hESCs. The SR currently available is not entirely animal free because it contains animal proteins [10], so the development of totally animal-free culture systems aiming at clinical use of hESCs for cell transplantation in humans requires more effort.

ACKNOWLEDGMENTS O.H. and R.L. contributed equally to this study. We thank Jonna Rinne for reviewing the language of this manuscript, Miina Miller for technical advice on Affymetrix technology, Tuomas Nikula for advice on Kensington software, and the Finnish DNA

REFERENCES 1

Edwards RG. IVF and the history of stem cells. Nature 2001;413:349 – 351.

2

Keller G, Snodgrass HR. Human embryonic stem cells: the future is now. Nat Med 1999;5:151–152.

3

Martin MJ, Muotri A, Gage F et al. Human embryonic stem cells express an immunigenic nonhuman sialic acid. Nat Med 2005;11:1–5.

4

Thomson JA, Itskovitz-Eldor J, Shapiro SS et al. Embryonic stem cell lines derived from human blastocysts. Science 1998;282:1145–1147.

5

Reubinoff BE, Pera MF, Fong CY et al. Embryonic stem cell lines from human blastocysts: somatic differentiation in vitro. Nat Biotechnol 2000; 18:399 – 404.

6

Amit M, Carpenter MK, Inokuma MS et al. Clonally derived human embryonic stem cell lines maintain pluripotency and proliferative potential for prolonged periods of culture. Dev Biol 2000;227:271–278.

7

Carpenter MK, Rosler ES, Fisk GJ et al. Properties of four human embryonic stem cell lines maintained in a feeder-free culture system. Dev Dyn 2004;229:243–258.

8

Xu C, Inokuma MS, Denham J et al. Feeder-free growth of undifferentiated human embryonic stem cells. Nat Biotechnol 2001;19:971–974.

9

Amit M, Margulets V, Segev H et al. Human feeder layers for human embryonic stem cells. Biol Reprod 2003;68:2150 –2156.

10 Amit M, Shariki C, Margulets V et al. Feeder layer- and serum-free culture of human embryonic stem cells. Biol Reprod 2004;70:837– 845. 11 Smith AG, Heath JK, Donaldson DD et al. Inhibition of pluripotential embryonic stem cell differentiation by purified polypeptides. Nature 1988;336:688 – 690. 12 Hovatta O, Mikkola M, Gertow K et al. A culture system using human foreskin fibroblasts as feeder cells allows production of human embryonic stem cells. Hum Reprod 2003;18:1404 –1409. 13 Koivisto H, Hyva¨rinen M, Stro¨mberg A-M et al. Cultures of human embryonic stem cells: serum replacement medium or serum-containing media and the effect of basic fibrolast growth factor. Reprod Biomed Online 2004;9:330 –337. 14 Inzunza J, Sahlen S, Holmberg K et al. Comparative genomic hybridization and karyotyping of human embryonic stem cells reveals the occurrence of an isodicentric X chromosome after long-term cultivation. Mol Hum Reprod 2004;10:461– 466. 15 Hamalainen H, Meissner S, Lahesmaa R. Signaling lymphocytic activation molecule (SLAM) is differentially expressed in human Th1 and Th2 cells. J Immunol Methods 2000;242:9 –19.

Molecular Signature of hESCs in Serum-Free Medium Microarray Center at Turku Centre for Biotechnology for providing facilities for the microarray analysis. We thank the personnel of the IVF Unit of Karolinska University Hospital, Huddinge, for their support in the stem cell research. R.L. and H.S. were supported by the Academy of Finland/JDRF, H.S. by Finnish Cultural foundations and NorFA, and O.H. by the Swedish Research Council/JDRF.

DISCLOSURES The authors indicate no potential conflicts of interest.

16 Bhattacharya B, Miura T, Brandenberger R et al. Gene expression in human embryonic stem cell lines: unique molecular signature. Blood 2004;103:2956 –2964. 17 Sato N, Meijer L, Skaltsounis L et al. Maintenance of pluripotency in human and mouse embryonic stem cells through activation of Wnt signaling by a pharmacological GSK-3-specific inhibitor. Nat Med 2004; 10:55– 63. 18 Sato N, Sanjuan IM, Heke M et al. Molecular signature of human embryonic stem cells and its comparison with the mouse. Dev Biol 2003;260:404 – 413. 19 Sperger JM, Chen X, Draper JS et al. Gene expression patterns in human embryonic stem cells and human pluripotent germ cell tumors. Proc Natl Acad Sci U S A 2003;100:13350 –13355. 20 Niwa H. Molecular mechanism to maintain stem cell renewal of ES cells. Cell Struct Funct 2001;26:137–148. 21 Nishimoto M, Fukushima A, Okuda A et al. The gene for the embryonic stem cell coactivator UTF1 carries a regulatory element which selectively interacts with a complex composed of Oct-3/4 and Sox-2. Mol Cell Biol 1999;19:5453–5465. 22 Duval D, Reinhardt B, Kedinger C et al. Role of suppressors of cytokine signaling (Socs) in leukemia inhibitory factor (LIF)-dependent embryonic stem cell survival. FASEB J 2000;14:1577–1584. 23 Schuringa JJ, van der Schaaf S, Vellenga E et al. LIF-induced STAT3 signaling in murine versus human embryonal carcinoma (EC) cells. Exp Cell Res 2002;274:119 –129. 24 Nakashima K, Yanagisawa M, Arakawa H et al. Synergistic signaling in fetal brain by STAT3-Smad1 complex bridged by p300. Science 1999; 284:479 – 482. 25 Morrisey EE, Tang Z, Sigrist K et al. GATA6 regulates HNF4 and is required for differentiation of visceral endoderm in the mouse embryo. Genes Dev 1998;12:3579 –3590. 26 Hay DC, Sutherland L, Clark J et al. Oct-4 knockdown induces similar patterns of endoderm and trophoblast differentiation markers in human and mouse embryonic stem cells. STEM CELLS 2004;22:225–235. 27 Chen RH, Ebner R, Derynck R. Inactivation of the type II receptor reveals two receptor pathways for the diverse TGF-beta activities. Science 1993;260:1335–1338. 28 Markowitz S, Wang J, Myeroff L et al. Inactivation of the type II TGF-beta receptor in colon cancer cells with microsatellite instability. Science 1995;268:1336 –1338. 29 Lai J, Chien J, Staub J et al. Loss of HSulf-1 up-regulates heparinbinding growth factor signaling in cancer. J Biol Chem 2003;278:23107– 23117.

Skottman, Stromberg, Matilainen et al. 30 Yamamoto M, Saijoh Y, Perea-Gomez A et al. Nodal antagonists regulate formation of the anteroposterior axis of the mouse embryo. Nature 2004;428:387–392. 31 Salvador JM, Hollander MC, Nguyen AT et al. Mice lacking the p53effector gene Gadd45a develop a lupus-like syndrome. Immunity 2002; 16:499 –508.

167 40 Adamson ED, Minchiotti G, Salomon DS. Cripto: a tumor growth factor and more. J Cell Physiol 2002;190:267–278. 41 Cai J, Cheng A, Luo Y et al. Membrane properties of rat embryonic multipotent neural stem cells. J Neurochem 2004;88:212–226.

32 Edelman GM, Jones FS. Outside and downstream of the homeobox. J Biol Chem 1993;268:20683–20686.

42 Huntriss J, Hinkins M, Oliver B et al. Expression of mRNAs for DNA methyltransferases and methyl-CpG-binding proteins in the human female germ line, preimplantation embryos, and embryonic stem cells. Mol Reprod Dev 2004;67:323–336.

33 Cortes F, Deschaseaux F, Uchida N et al. HCA, an immunoglobulin-like adhesion molecule present on the earliest human hematopoietic precursor cells, is also expressed by stromal cells in blood-forming tissues. Blood 1999;93:826 – 837.

43 Kershaw DB, Beck SG, Wharram BL et al. Molecular cloning and characterization of human podocalyxin-like protein. Orthologous relationship to rabbit PCLP1 and rat podocalyxin. J Biol Chem 1997;272: 15708 –15714.

34 Fujiwara H, Tatsumi K, Kosaka K et al. Human blastocysts and endometrial epithelial cells express activated leukocyte cell adhesion molecule (ALCAM/CD166). J Clin Endocrinol Metab 2003;88:3437–3443.

44 Ben-Shushan E, Thompson JR, Gudas LJ et al. Rex-1, a gene encoding a transcription factor expressed in the early embryo, is regulated via Oct-3/4 and Oct-6 binding to an octamer site and a novel protein, Rox-1, binding to an adjacent site. Mol Cell Biol 1998;18:1866 –1878.

35 Niwa H, Miyazaki J, Smith AG. Quantitative expression of Oct-3/4 defines differentiation, dedifferentiation or self-renewal of ES cells. Nat Genet 2000;24:372–376. 36 Carter MG, Hamatani T, Sharov AA et al. In situ-synthesized novel microarray optimized for mouse stem cell and early developmental expression profiling. Genome Res 2003;13:1011–1021. 37 Chambers I, Colby D, Robertson M et al. Functional expression cloning of Nanog, a pluripotency sustaining factor in embryonic stem cells. Cell 2003;113:643– 655.

45 Caricasole AA, van Schaik RH, Zeinstra LM et al. Human growthdifferentiation factor 3 (hGDF3): developmental regulation in human teratocarcinoma cell lines and expression in primary testicular germ cell tumours. Oncogene 1998;16:95–103. 46 Clark AT, Rodriguez RT, Bodnar MS et al. Human STELLAR, NANOG, and GDF3 genes are expressed in pluripotent cells and map to chromosome 12p13, a hotspot for teratocarcinoma. STEM CELLS 2004; 22:169 –179.

38 Mitsui K, Tokuzawa Y, Itoh H et al. The homeoprotein Nanog is required for maintenance of pluripotency in mouse epiblast and ES cells. Cell 2003;113:631– 642.

47 Spassov DS, Jurecic R. Mouse Pum1 and Pum2 genes, members of the Pumilio family of RNA-binding proteins, show differential expression in fetal and adult hematopoietic stem cells and progenitors. Blood Cells Mol Dis 2003;30:55– 69.

39 Pesce M, Anastassiadis K, Scholer HR. Oct-4: lessons of totipotency from embryonic stem cells. Cells Tissues Organs 1999;165:144 –152.

48 Tsuda M, Sasaoka Y, Kiso M et al. Conserved role of nanos proteins in germ cell development. Science 2003;301:1239 –1241.

See www.StemCells.com for supplemental material available online.