SEARCH

SEARCH BY CITATION

Keywords:

  • Embryonic stem cell;
  • ERas;
  • Pseudogene;
  • Polyadenylation;
  • Evolution

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The ERas gene is expressed in mouse embryonic stem (ES) cells and promotes their in vitro proliferation and tumorigenicity. We analyzed the expression of the human ERas gene in human ES cells by reverse transcription–polymerase chain reaction (RT-PCR) and serial analysis of gene expression but could not detect a full-length coding transcript. Sequence analysis predicted a premature polyadenylation signal for the human ERas transcript, which we confirmed by 3′ RACE analysis. By RT-PCR, we identified a truncated noncoding transcript in human ES cells that is downregulated during differentiation, suggesting conserved tissue specificity of the promoter region. Previous reports and expressed sequence tag databases indicate that orthologues of this gene are expressed in other mammals, including the mouse, dog, and cow, which suggests that it became a silenced pseudogene relatively recently in mammalian evolution. In addition to the premature polyadenylation site, both the human and chimpanzee ERas genes include typical Alu-S retrotransposon insertions that could also influence expression at this locus. The lack of ERas expression in human ES cells suggests that they could have significantly different tumorigenic properties than mouse ES cells.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Embryonic stem (ES) cells are pluripotent cells that maintain the ability to differentiate to derivatives of all three embryonic germ layers, even after prolonged culture in the undifferentiated state [13]. Most gene products known to be involved in maintaining the ES cell state, such as Oct4 [4, 5], fibroblast growth factor 4 [6, 7], FoxD3 [8, 9], Sox2 [10], and Nanog [11, 12], have been functionally characterized only in mouse ES cells. However, there are significant differences between mouse and human ES cells in cell cycle time, growth factor requirements, and cell-surface marker expression that presumably reflect species-specific differences in early development [13, 14]. In most cases it is not yet clear what molecular mechanisms are responsible for these differences.

The mouse ERas gene is a recently identified gene that supports tumorigenic growth of mouse ES cells by producing a constitutively active Ras protein [15]. Disruption of the ERas gene in mouse ES cells by homologous recombination results in a significantly reduced proliferation rate and a reduced tumorigenic potential without loss of pluripotency [15]. The human ERas gene was initially reported as a processed pseudogene (HRasp, Ha-Ras2) [16, 17], with several base substitutions in coding regions, but was more recently described as potentially encoding a functional human ERas protein [15]. However, the expression of the ERas gene remained unclear.

Recent transcriptome analysis has failed to detect ERas gene expression in human ES cells [18, 19]. Here we examined the expression of human ERas gene in human ES cells in detail and identified a truncated ERas gene product resulting from a premature polyadenylation signal upstream of its coding sequence. These differences suggest that the tumorigenic potential of human ES cells could differ fundamentally from mouse ES cells and that predicting the behavior of transplanted human ES cells from the behavior of transplanted mouse ES cells is probably inappropriate.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Cell Culture

Human ES cells (H1 and H9) were cultured as previously described either on mouse embryonic fibroblasts (MEFs) or in medium conditioned by MEFs [3, 20]. HEK 293 cells were cultured in Dulbecco's modified Eagle's medium with 10% fetal calf serum. To induce human ES cell differentiation, cells in feeder-free conditions were treated for 10 days with 10 μM retinoic acid in unconditioned media or with 0.1 μg/ml BMP4 (Upstate, Waltham, MA, http://www.upstate.com) in conditioned media [21].

Reverse Transcription–Polymerase Chain Reaction

mRNA was prepared with Trizol (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) and a Dyna-bead kit (Dynal Bio-tech, Oslo, Norway, http://www.dynalbiotech.com) according to the manufacturers' protocols. Reverse transcription–polymerase chain reaction (RT-PCR) was performed with Titan one-tube RT-PCR kit (Roche, Indianapolis, http://www.diagnostics.roche.com) with or without templates (poly A RNA from human ES cells or human genomic DNA). For one RT-PCR reaction, each poly A RNA corresponding to approximately 0.1–1 μg of total RNA or 1 μg of genomic DNA was used as the template.

The primers used for β-actin were CCAAGGCCAACCGCGAGAAGATGAC and AGGGTACATGGTGGTGCCGCCAGAC, which give rise to an amplified fragment of 587 bp. The primers used for oct-4 were AGAAGAGGATCACCCTGGGAT and AGAACCACACTCGGACCACAT, which give rise to an amplified fragment of 330 bp. The primers used for ERas coding sequence were ATGGAGCTGCCAACAAAGCCTGGCA and TTCAGGCCACAGAGCAGCCACAGT, which give rise to an amplified fragment of 702 bp. The primers used for ERas upstream region were TAACTAACACTAATTGACCAC and TTTATTGGGAACCTACTGTGT, which give rise to an amplified fragment of 703 bp. The RT-PCR reaction conditions were as follows: 1 cycle of 30 minutes of RT at 55°C and 3 minutes of denaturation at 95°C, 21 to 38 cycles of 1 minute of denaturation at 95°C, 1 minute of annealing at 55°C (for oct-4 and ERas upstream) or 72°C (for β-actin and ERas), and 2 minutes of extension at 72°C. In some cases, the initial RT was omitted to check the DNA contamination. The PCR products were subjected to 1.0% or 1.8% agarose gel electrophoresis and visualized by staining with ethidium bromide. The identification of PCR products was confirmed by DNA sequencing.

Sequence and Library Analysis

We used the SHES2 library (http://www.transcriptomes.org/) for the serial analysis of gene expression (SAGE) [22] to analyze the gene expression profile of human ES cells. Genomic sequences and expressed sequence tag (EST) clone information were obtained from the GenBank database (http://www.ncbi.nlm.nih.gov/) in NCBI. The chimpanzee genomic sequence was obtained from the Genome Sequencing Center of Washington University in St. Louis (http://www.genome.wustl.edu/). Sequence analysis was performed with the Blast search tool in NCBI and Genetyx-Mac software (Genetyx Corporation, Tokyo). Interspersed repeat elements were analyzed using the RepeatMasker software (Institute for Systems Biology, Seattle, http://www.repeatmasker.org/).

Polyadenylation Signal Analysis by 3′ RACE

Polyadenylation signal candidates (A-1, A-2, and A-3) in the ERas upstream region were synthesized to test for activity. The nucleotides to produce A-1 were GGAAGATTAAATTATAATTTAAAAGCCTGGCATGAACTGAGCACCTTCTGCATGCTTATCACCCGCTACTGACATATAGTTCACTTTTTTTTTTTTTTTTTTTandGCAAAAAAAAAAAAAAAAAAAAGTGAACTATATGTCAGTAGCGGGTGATAAGCATGCAGAAGGTGCTCAGTTCATGCCAGGCTTTTAAATTATAATTTAATCT. The nucleotides to produce A-2 were GGAACAGTAGGTTCCCAATAAATGTATGTTGAATAATTAATAAAGACTACATTCTAATGATAATGGCTAACAGACCACGTGATTTACATATTTATTTGGTTTATTATATTAT and GCAATAATATAATAAACCAAATAAATATGTAAATCACGTGGTCTGTTAGCCATTATCATTAGAATGTAGTCTTTATTAATTATTCAACATACATTTATTGGGAACCTACTGT. Th e nucleotides to produce A-3 were GGACAATAAATGTTGAAATAATGACACCCCACTGTCTCCTTGCCCTCAAATGGTCTCCCCTAACGTATCCCCTGTTGTCTTGCTTCTTCTCTTC and GCAGAAGAGAAGAAGCAAGACAACAGGGGATACGTTAGGGGAGACCATTTGAGGGCAAGGAGACAGTGGGGTGTCATTATTTCAACATTTATTG. The annealed nucleotides were subcloned into between two Sap I sites of pNeoEGFP plasmid (Clontech, Palo Alto, CA, http://www.clontech.com) to produce pNeoA-1EGFP, pNeoA-2EGFP, and pNeoA-3EGFP, respectively. The plasmids were transfected into HEK 293 cells using the FuGENE 6 reagent (Roche), and the mRNAs were recovered 2 days later. 3′ RACE analysis was performed with RACE Core Set, 3′-Full kit (TaKaRa Mirus Bio, Madison, WI, http://www.takaramirusbio.com) with a forward primer of TTAATACGACTCACTATAGG. The sequences of 3′ RACE products, around the polyadenylation site, were analyzed using a primer of GATGGAAGCCGGTCTTGTCGA.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Full-Length ERas mRNA Expression Is Not Detectable in Human ES Cells

We examined ERas expression by RT-PCR by using primers for the coding region, using poly A RNA from human ES cells as the template. Although the expression of β-actin or oct-4 gene was easily detectable by RT-PCR, we never detected the expression of ERas gene in several RT-PCR conditions (Fig. 1). However, we easily amplified the ERas coding sequence by the RT-PCR reaction, lacking an intron, when we used genomic DNA as the template (Fig. 1). This result is consistent with the previous reports that the ERas gene exists in the human genome [1517] but is not expressed in human ES cells [18, 19], suggesting that the expression of this gene is different between human and mouse.

thumbnail image

Figure Figure 1.. Analysis of the ERas gene expression in human embryonic stem (ES) cells. The expressions of β-actin and oct-4 genes are detectable by reverse transcription–polymerase chain reaction using mRNA from H1 and H9 human ES cells. ERas gene expression is not detected, whereas the target sequence is easily amplified when the genomic DNA (H1g and H9g) is used as the reaction template. The negative control (−) indicates no template in the reaction.

Download figure to PowerPoint

Analysis of the Transcriptional Activity on Human ERas Locus by SAGE and EST Clone Screening

To examine the transcriptional activity of ERas locus in human ES cells, we also checked a deep SAGE library of human ES cells (line H9) for 22 potential sense tags around the ERas locus (Fig. 2; Table 1). SAGE tags for ES cell markers oct-4 and nanog and SAGE tags for X-linked genes hprt1 (hypoxanthine phosphoribosyltransferase 1) and pgk1 (phosphoglycerate kinase 1) were present, but no full-length (17-base) SAGE tags representing the ERas gene were present in the library. When we reduced the tag length to 13 bases, to follow the incompletely cloned minor tags, we identified a tag (Fig. 2, tag h) located approximately 0.7 kbp upstream of the ERas coding sequence. We detected no tags in the antisense orientation around the ERas locus. These data suggest that this locus is expressed weakly, if at all, in human ES cells.

thumbnail image

Figure Figure 2.. Genomic structures of the ERas locuses in the X chromosome of human, mouse, and dog. The ERas genes are located between the HDAC 6 (histone deacetylase 6) and PCSK1N (proprotein convertase subtilisin/kexin type 1 inhibitor) genes in these species. The structure of the chimpanzee and rat ERas locus is highly conserved to the human and mouse locus, respectively. The locations of potential serial analysis of gene expression (SAGE) tags are indicated in the box, used for the human ERas analysis. The corresponding sequence to the BX100754 expressed sequence tag clone is indicated by an arrow. A1, A2, and A3 indicate the locations of the predicted polyadenylation signals. The locations of reverse transcription–polymerase chain reaction primers are indicated by the small arrows, used to check the expressions of the coding and the upstream region of the ERas gene. The inserted Alu elements in human ERas locus and the exons (E1 and E2) of mouse ERas gene are also indicated.

Download figure to PowerPoint

Table Table 1.. Serial analysis of gene expression (SAGE) tag hit in the human embryonic stem cell library
SAGE tags ofERaslocusFull (17)Short (13)
a: AGAAAATGGGTGCTTGC00
b: GTAGCTTCTTGGCAGAA00
c: GCCACCTTATGATCCCA00
d: TGGTTGCCATAGCTGCA00
e: TCCACTCCCCCAACCGC00
f: CTTATCACCCGCTACTG00
g: AGGGCTACAGTTTGTTG00
h: TCCTTGCCACACAGTAG01
i: GGGAGGAACTGCTAATG00
j: GGGATTCTTTTCTGGGT00
k: CAAATTGGAAGGTGATC00
l: GAGCTGCCAACAAAGCC00
m: GAGCCCTTCCTTCCAGG00
n: CCGCTGCTGCAGCCCTC00
o: AGATCCAGAGGGTCCAG00
p: GCAAGGTCCTGTAGGGA00
q: AGACCCCTGAGGCAACT00
r: GGAGTGCCTAATGTTGC00
s: CCGAGTGTTGTGTGGGT00
t: GCCACAAATCAGCTGCC00
u: TGTGCCTGCGACATCCC00
v: TGGATGTCTTCCCCTCC00
Oct4: TATCACTTTTTTCTTAA364 
Nanog: AGTACTACTTTAGTTGG29 
HPRT1: TATCTTCTAAGAATTTT14 
PGK1: GAGGAAGGCTCTGTTCC15 
Library: SHES-2(Nla III) 
Total tag: 468040  

Identification of a Premature Polyadenylation Signal in the Human ERas Locus

In examining EST databases for evidence of human ERas expression, we identified a clone (GenBank accession no. BX100754) of a polyadenylated, 280-bp fragment of the noncoding human ERas upstream region obtained from a human colon cancer sample (Fig. 2). Interestingly, the incompletely cloned SAGE tag (Fig. 2, tag h) of the ERas locus also corresponds to this region (Fig. 2; Table 1). These data suggested the existence of a premature upstream polyadenylation signal in the human ERas locus.

We searched for possible polyadenylation signals in the ERas upstream region and found three candidates (A-1, A-2, and A-3) having typical AATAAA or ATTAAA motifs with a following GU/U-rich element (Figs. 2, 3A). The A-1 element contains the SAGE tag f. The A-2 element includes the polyadenylation site of the BX100754 EST clone and the SAGE tag h. The A-3 element maps between the SAGE tags k and l. Each element was cloned into a pNeoEGFP plasmid between the strong cytomegalovirus promoter and the typical SV40 polyadenylation signal (Fig. 3B) and transfected into HEK293 cells. The resulting transcripts were analyzed by 3′ RACE. As shown in Figure 3C, the A-2 element functioned effectively as a polyadenylation signal. However, the A-1 element demonstrated no activity, and the A-3 element demonstrated only weak activity for inducing premature polyadenylation. By sequencing, we checked the polyadenylation site associated with the A-2 element and found that it corresponded to the polyadenylation site of the BX100754 EST clone (Fig. 3A).

thumbnail image

Figure Figure 3.. Analysis of the premature polyadenylation signal on the upstream of human ERas.(A): The sequences of the polyadenylation signal candidates. The polyadenylation signal (poly A signal) and the following GU/U-rich element are indicated by the solid or broken lines, respectively. The predicted serial analysis of gene expression (SAGE) tags (tags f and h) are also indicated by double lines. The polyadenylated sites in the A-2 and A-3 elements are boxed. The black arrowheads indicate the major polyadenylation sites in each element, and the white arrowhead indicates the polyadenylated site of the BX100754 expressed sequence tag clone. (B): Diagram of the constructs used for the analysis of the polyadenylation inducing ability of A-1, A-2, and A-3 elements. These elements were inserted into the pNeoEGFP plasmid between the cytomegalovirus (CMV) promoter and the SV40 polyadenylation signal (SV40 poly A) to produce the pNeoA-1EGFP, pNeoA-2EGFP, and pNeoA-3EGFP. The IVS represents an artificial intron. The location of the forward primer for the 3′ RACE analysis is indicated by a small arrow. (C): The result of the 3′ RACE analysis on the HEK 293 cells. The cells were transfected with pUC119 (control), pNeoEGFP, pNeoA-1EGFP, pNeoA-2EGFP, and pNeoA-3EGFP, respectively, and the expressed mRNA was analyzed 2 days after the transfection. The result without template mRNA, lane (−), is also displayed. The upper bands (∼1.6 kbp) and the lower bands (∼0.6 kbp) correspond to the fully transcribed and the premature polyadenylated products, respectively.

Download figure to PowerPoint

The Human ERas Promoter Is Weakly Active in Human ES Cells and Is Downregulated During Differentiation

The premature polyadenylation signal upstream of the human ERas coding region suggested to us that the ERas promoter could still be active in human ES cells but that a truncated transcript would be produced that would have been missed in our previous analysis. We thus used RT-PCR analysis to examine the region upstream of the premature polyadenylation signal (Fig. 2) and indeed detected expression in human ES cells (Fig. 4A). After the induction of differentiation by retinoic acid or BMP4 treatment, the ERas upstream region expression was downregulated (Fig. 4B), suggesting that this promoter still may maintain some tissue specificity comparable with the mouse ERas gene [15].

thumbnail image

Figure Figure 4.. (A): Reverse transcription(RT)– polymerase chain reaction analysis on the premature product of human ERas gene. The poly A RNA from the H1 and H9 human embryonic stem cells was analyzed with (RT +) or without (RT −) the RT procedure. (B): The effect of cell differentiation on the expression of the premature mRNA of human ERas gene. To induce differentiation, the H9 cells were treated with 10 μM of retinoic acid (RA) or 0.1 μg/ml of BMP-4 for 10 days (lane +). The control cells (lane −) were cultured under normal conditions. Oct-4 and β-actin expression was also analyzed.

Download figure to PowerPoint

Comparative Analysis on ERas Gene

Our results indicate that the full-length ERas gene is not transcribed in human ES cells, but previous results have shown that it is both transcribed and physiologically significant in mouse ES cells [15]. We were therefore interested in whether this gene is transcribed in other mammalian species. Using the Blast homology search program, we searched the EST clones for ERas orthologues in other species and found clones from Canis familiaris (dog) from the testis (GenBank: CO609560, CO598019, and CO600209) and Bos taurus (cattle) from the pooled library of multiple tissues, including liver, lung, hypothalamus, pituitary, and placenta (GenBank: CB454861 and CB456439) and from the adult brain (GenBank: CO891744) that correspond to the previously described v-Ha-ras 3 gene [23]. We also found the genomic sequences of each ERas locus of human (GenBank: AF196971), mouse (GenBank: NT_039700), rat (GenBank: NW_048035), dog (GenBank: AAEX01047891), and chimpanzee (chimpanzee genomic project Contig files of 1804.5 and 1804.6). Each of the ERas genes is located on the X chromosome. The neighboring 5′ and 3′ genes, histone deacetylase 6 (HDAC 6) and proprotein convertase subtilisin/kexin type 1 inhibitor (PCSK1N), respectively, are also conserved among these species (Fig. 2). The ERas amino acid sequences of each mammal, predicted from the genomic and cDNA sequences, are well conserved among these species (the arrangement is indicated in Fig. 5).

thumbnail image

Figure Figure 5.. The alignment on the predicted ERas amino acid sequences of five mammals (mus, mouse; rat, rat; hom, human; can, dog; bos, cattle). The human and chimpanzee ERas sequences are identical. The cattle sequence is incomplete at the N-terminal end (indicated by X). The amino acids conserved among the species are marked by red characters.

Download figure to PowerPoint

In both the human and chimpanzee ERas genes, there are two typical Alu-S retrotransposon insertions in the upstream region, near the possible promoter region and the conserved premature polyadenylation signal of A-2 (Fig. 2). The Alu-S subfamily is thought to have been amplified in ancestral primate genomes from approximately 48–30 million years ago [2426]. These elements can affect adjacent gene expressions [24, 27], and their presence could diminish ERas promoter activity. The Alu-S insertions, combined with the acquired premature polyadenylation before the divergence from chimpanzee, would ensure the complete silencing on the human ERas gene.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In this report, we confirmed the lack of full-length ERas gene expression in human ES cells but identified a truncated transcript of the 5′ noncoding region that is terminated by a premature polyadenylation signal. The expression of the ERas gene in species as divergent as mice, dogs, and cows suggests that its inactivation in higher primates is a relatively recent event in mammalian evolution. Disruption of the mouse ERas gene by homologous recombination resulted in impaired growth and reduced tumorigenicity of mouse ES cells but did not impair normal development [15]. It is tempting to speculate that some of the differences in proliferation rate or other properties of mouse and human ES cells are related to the differences of expression of this constitutively active ERas gene, although mouse and human ES cells differ significantly in other pathways also [19]. However, transient transfection of human ES cells with an expression vector placing ERas under the control of the human elongation factor 1 promoter induced significant cell death, and to date we have been unable to establish stable ERas expressing human ES cells. An antioncogenic role of activated Ras by the induction of apoptosis has been reported [2830], so this is not a surprising result, but it does imply broader differences between mouse and human ES cells in cell-cycle or apoptotic control. Thus, although the physiological significance of the species difference in ERas expression is not yet clear, it does suggest that the behavior of transplanted mouse ES cells will not accurately predict the behavior of transplanted human ES cells and that the use of primate models will be more appropriate to demonstrate the safety of human ES cell–based therapies.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank Dr. J. Yu, Dr. Y. Kim, and Ms. D. Faupel in our laboratory for technical assistance and manuscript preparation. We appreciate and thank Dr. M. Marra and Dr. C. Eaves in the Genomic Science Center, British Columbia Cancer Agency for the analysis on the human ES cell SAGE database SHES2. This research was supported by funds from the University of Wisconsin Foundation.

Disclosures

J.A.T. owns stock in and within the past 2 years served as an officer or member of the Board of Cellular Dynamics International.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References