Stem Cells
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

First published online December 22, 2005
Stem Cells Vol. 24 No. 3 March 2006, pp. 662 -670
doi:10.1634/stemcells.2005-0552; www.StemCells.com
© 2006 AlphaMed Press

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Table
Right arrow All Versions of this Article:
2005-0552v1
24/3/662    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bowman, T. V.
Right arrow Articles by Goodell, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bowman, T. V.
Right arrow Articles by Goodell, M. A.
STEM CELL GENETICS AND GENOMICS

Differential mRNA Processing in Hematopoietic Stem Cells

Teresa V. Bowmana,b, Andrew J. McCooeya, Akil A. Merchanta,c, Carlos A. Ramosa,c, Patricia Fonsecad, Alan Poindextere, Steven B. Bradfutea,f, Daniela M. Oliveiraa, Rahshaana Greena, Yayun Zhenga, Kathyjo A. Jacksona,g, Stuart M. Chambersa,b, Shannon L. McKinney-Freemana,f, Kevin G. Norwooda, Gretchen Darlingtond,h, Preethi H. Gunaratned,i, David Steffend,e,i, Margaret A. Goodella,b,d,f,g

a Cell and Gene Therapy Center,
b Cell and Molecular Biology Program,
c Departments of Medicine and
d Molecular and Human Genetics,
e Bioinformatics Research Center,
f Departments of Immunology,
g Pediatrics, and
h Molecular and
i Cellular Biology, Baylor Human Genome Sequencing Center, Baylor College of Medicine, Houston, Texas, USA

Key Words. RNA processing • Hematopoietic stem cells • cDNA library • Stem cell activation

Correspondence: Margaret A. Goodell, Ph.D., Center for Cell and Gene Therapy, Baylor College of Medicine, N1030, One Baylor Plaza, Houston, Texas 77030, USA. Telephone: 713-798-1265; Fax: 713-798-1230; e-mail: goodell{at}bcm.tmc.edu

Received November 8, 2005; accepted for publication December 14, 2005.

   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Hematopoietic stem cells (HSCs) maintain tissue homeostasis by rapidly responding to environmental changes. Although this function is well understood, the molecular mechanisms governing this characteristic are largely unknown. We used a sequenced-based strategy to explore the role of both transcriptional and post-transcriptional regulation in HSC biology. We characterized the gene expression differences between HSCs, both quiescent and proliferating, and their differentiated progeny. This analysis revealed a large fraction of sequence tags aligned to intronic sequences, which we showed were derived from unspliced transcripts. A comparison of the biological properties of the observed spliced versus unspliced transcripts in HSCs showed that the unspliced transcripts were enriched in genes involved in DNA binding and RNA processing. In addition, levels of unspliced message decreased in a transcript-specific fashion after HSC activation in vivo. This change in unspliced transcript level coordinated with increases in gene expression of splicing machinery components. Combined, these results suggest that post-transcriptional regulation is important in HSC activation in vivo.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Regenerative organs like the hematopoietic system must maintain equilibrium in cell death, proliferation and differentiation. Hematopoietic stem cells (HSCs) are at the root of the maintenance of homeostasis in the blood system. Hallmarks of HSC behavior that allow for tight regulation of equilibrium are their abilities to undergo self-renewal division as well as produce cells of all differentiated blood lineages in a controlled and rapidly inducible fashion.

To understand the molecular mechanisms controlling stem cell behavior, several groups have taken large-scale genomics approaches to characterize their gene expression profiles [1–9]. Some of these studies have noted promiscuous gene expression in HSCs [1, 10, 11]. This promiscuity could give HSCs access to transcriptional programs of differentiated blood cells, thus allowing for rapid response after an insult.

Although these observations have been made with regard to transcription in HSCs, we know virtually nothing of the global post-transcriptional regulation that must occur. Splicing kinetics, mRNA export, and mRNA stability are dynamic processes regulated by many factors, including cell cycle status. Over the past several years, links have been drawn between splicing factors and cell cycle regulators in eukaryotes [12, 13]. Specifically, in studies examining the role of post-transcriptional mRNA regulation in various cell lines, it has been noted that certain transcripts exist as pre-mRNA during resting stages of cell cycle and undergo maturation as the cell transits the cell cycle [14–16]. Although these studies highlighted the importance of mRNA regulation in gene expression, they were performed on a limited number of transcripts and only examined the effects in vitro. In our current study, we examined the global effects of post-transcriptional regulation on highly purified murine HSCs.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Cell Isolation
For suppressive subtraction hybridization (SSH)-concatenation cDNA sequencing (CCS) libraries, bone marrow (BM) was collected from femurs and tibias of 8-week-old C57Bl/6 mice. For analysis of gene expression over an HSC activation time course, BM was collected from mice 0, 3, or 6 days after 5-fluorouracil (5FU) treatment. 5FU was administered intravenously at a single dose of 150 mg/kg total body weight (Sigma, St. Louis, http://www.sigmaaldrich.com). BM was stained with Hoechst 33342 to identify HSCs according to the side population (SP) [17]. Cells were stained with anti-Sca1-phycoerythrin (PE) (E13–161.7) and anti-Gr1-fluorescein isothiocyanate (RB6–8C5) (BD Pharmingen, San Diego, http://www.bdbiosciences.com/pharmingen). BM was sorted for the SP profile, Sca1 positivity, and Gr1 negativity on a MoFlo (Cytomation, Fort Collins, CO, http://www.dako.dk). To eliminate any changes that could be introduced by cell sorting, both BM and HSC were sorted from the same cell BM preparation (Fig. 1AGo). One million non-SP cells (all BM cells excluding SP) were sorted, and approximately 70,000 SP cells were isolated with a purity of 97%. In addition, one out of 10 Sca1+ SP cells shows engraftment potential in a transplantation setting, indicating high functional purity [18, 19].


Figure 1
View larger version (24K):
[in this window]
[in a new window]
  		Figure 1. Schematic of experimental design. (A): Cells were sorted from total BM isolated from the tibias and femurs of adult mice. HSCs were sorted based on Hoechst efflux and expression of the cell surface protein Sca1. Sorted BM included all marrow cell types excluding SP. (B): RNA was then isolated from sorted cell population. To enrich for rare transcripts unique to each cell type, SSH was performed. Four comparisons were made, yielding four libraries: 1) CST-HSC: HSC vs. BM; 2) CST-BM: BM vs. HSC; 3) CST-QuHSC: HSC vs. proliferating HSC; and 4) CST-ProHSC: HSC vs. quiescent HSC. (C): After SSH, enriched cDNAs were subjected to CCS. First, the cDNA are cut into smaller fragments with four common-cutting restriction enzymes (cut sites are noted by red arrows). Second, fragments are concatemerized. Third, concatemers are shotgun-sequenced (red half arrows mark sequencing start sites). (D): After sequence reads are acquired, they are mapped onto the mouse genome using the BLAT algorithm. Blue and green lines represent sequence tags mapped onto genes. The blue tag represents a spliced tag, whereas the green tag represents an unspliced tag that spans an exon-intron boundary. Once mapped, the tags are annotated. Part 2 displays a prototypic UCSC genome browser view of a mapped sequence tag to the gene CTLA2-{alpha}. (E): Based on the annotation information, tags were binned into spliced and unspliced transcript groups. These groups were then analyzed for any potential differences in biology and behavior after HSC activation. Abbreviations: BM, bone marrow; CCS, concatenation cDNA sequencing; GO, Gene Ontology; HSC, hematopoietic stem cell; RT-PCR, reverse transcription-polymerase chain reaction; SSH, suppressive subtraction hybridization; UCSC, University of California Santa Cruz; qHSC, quiescent HSC; pHSC, proliferating HSC; CST, CCS tag.

 
SSH and CCS
Total RNA was isolated from sorted HSCs and BM using TRIreagent (Sigma) according to manufacturer’s instructions. To remove any contaminating genomic DNA, RNA samples were treated with DNaseI (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) according to the manufacturer’s instructions. To detect any contaminating genomic DNA, PCR was performed on each RNA sample in the absence of reverse transcription (no-RT). No PCR product for the housekeeping gene ß-actin was detected after 45 cycles of PCR. SMART cDNA synthesis kit (BD Clontech, Palo Alto, CA, http://www.clontech.com) was used to make full-length cDNA. After first-strand synthesis, cDNA was amplified with long distance PCR for 25 cycles to obtain enough material for SSH. SSH was performed as previously described using the PCR-Select kit (BD Clontech) [20]. Briefly, cDNA was digested with RsaI to produce more uniformly sized fragments for hybridization. Two rounds of subtraction were performed, yielding differentially expressed cDNA with different 5'- and 3'-tags, which were later used for suppression PCR. After two rounds of PCR selection, cDNA products were digested with EagI and NotI to remove the adaptors and NlaIII and MboI to generate smaller tags. The resulting tags were concatenated using T4 DNA ligase to create large concatemers, which were sheared into smaller fragments through nebulization. The resulting inserts were end-repaired and shotgun-sequenced in the Human Genome Sequencing Center (HGSC) as described [21].

Sequencing Analysis
The sequence reads from above were processed through several steps to convert them into gene hits. The reads (fasta and quality files) were trimmed by the standard BCM-HGSC pipeline [22]. The concatenated tags in these reads were recovered by "electronic digestion," splitting them at EagI, NotI, NlaIII, and MboI restriction sites. Repeat sequences were identified with Repeat-Masker [23, 24] using the -mus and -xsmall options to mask rodent repeat sequences. Sequence tags were aligned to the Mm3 build of the mouse genome [25] using the BLAST-like alignment tool (BLAT) program [26] with the options -qMask = lower, -out = pslx -noHead, -q = rna, -t = dna, -minIdentity = 95, -minMatch = 4. Next, Transcriptomatic, a software tool designed to resolve CCS tags (CST), was used to identify "chimeric" tags and to assign a probability-like score to identify the best match in cases in which a tag aligns to multiple sites in the genome. Details of this process will be published elsewhere. These deconvoluted tags have been submitted to GenBank and can be downloaded from http://www.liver-HSC.scgap.org/txom.

Genome hits were then mapped to genes, gene predictions, and transcripts by matching the genomic coordinates to downloaded copies of the University of California Santa Cruz (UCSC) genome browser annotation tables [27]. Transcriptomatic was used to assign a single best hit for each tag. For this purpose, any match to an exon sequence is considered more interesting than any match to an intron sequence, where any match that contains more than five bases of intron sequence is considered an intron match. Within the set of exon or intron matches, the relative interest of the different gene sets is (knownGene > mgcGenes > refFlat > all_mrna > ensGene > twinscan > sgpGene > softberryGene > geneid > genscan). Simple ad hoc scripts were used to count the total number of intron and exon bases present in the genes matched and in the tags. All scripts used in this analysis are available upon request from David Steffen (steffen{at}bcm.tmc.edu).

Real-Time RT-PCR
For validation of differential expression, total RNA was isolated from HSCs and BM using RNAqueous (Ambion, Austin, TX, http://www.ambion.com) and digested with DNaseI (Invitrogen) according to manufacturers’ instructions. Approximately 4–5 pg of total RNA was isolated per HSC. First-strand synthesis was performed with random primers and Superscript II reverse transcriptase (Invitrogen). Each real-time PCR experiment was performed twice from separate cell isolations. Within each experiment, PCR reactions were done in duplicate. For each RNA sample, genomic DNA contamination was determined by PCR on a no-RT control for the housekeeping gene ß-actin. After reverse transcription (RT), multiplexed real-time PCR reactions were performed using gene-specific TaqMan probe sets (supplemental online Table 1). As an endogenous control, 18S rRNA probes were used in concert with gene expression assays specific for our gene-of-interest (GOI). Standard cycling conditions were used for a total of 60 cycles. Fold changes were calculated using the {Delta}{Delta}Ct method (Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com).

For assessment of spliced and unspliced mRNA levels, real-time RT-PCR was also used. Primers were designed using Primer Express (Applied Biosystems) (supplemental online Table 1). Each primer pair was compared with the mouse genome to ensure specificity for the GOI. PCR products were detected by SYBR green fluorescence. ß-Actin primers were used as positive controls for PCR, and the resulting PCR product was used to normalize GOI-specific products for cross-isolation comparison. The average ß-actin threshold cycle (Ct) values for each time point were 27.9 ± 1.4 for day 0, 25.4 ± 2.6 for day 3, and 25.2 ± 0.77 for day 6. Cycling parameters used were standard for SYBR green analysis. Dissociation curve analysis was used to determine PCR product specificity.

Gene Ontology Analysis
Gene Ontology (GO) analysis was performed using EASE version 2.0 (http://david.niaid.nih.gov/david/ease.html), a stand-alone software package that allows annotation and statistical analysis of gene lists [28]. Using EASE version 2.0, we compared the GO content of each of the spliced/unspliced gene lists (CST-HSC unspliced, CST-HSC spliced, CST-BM unspliced, and CST-BM spliced) to the GO content of All Genes list, which is comprised of all genes in CST-HSC and CST-BM represented by greater than or equal to five hits. A total of 108 genes were analyzed. To determine whether a spliced/unspliced gene list was enriched over the All Genes list in a particular GO category, a probability score was assessed using Fischer’s exact test and a Bonferroni multiple testing correction was applied. All GO categories were analyzed. On average, 15 GO groups were shown to be significantly enriched (p < .1) per comparison.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
Generation and Sequencing of an HSC-Specific cDNA Library
To examine HSC gene expression in an unbiased fashion, we chose a sequencing strategy as opposed to a microarray platform. Because we aimed to identify transcripts unique to HSCs, we employed SSH, which enriches for rare transcripts (Fig. 1BGo) [20]. Two subtraction strategies were employed, HSC versus BM and quiescent HSC versus proliferating HSC. The HSC versus BM subtraction was designed to separate transcripts expressed in both mature blood cells and HSCs from transcripts expressed solely in HSCs, resulting in an HSC-enriched library (CST-HSC) and a BM-enriched library (CST-BM). The quiescent HSC-versus-proliferating HSC subtraction was designed to identify transcripts differentially expressed between quiescent (CST-QuHSC) and activated HSCs (CST-ProHSC), thus identifying genes important for HSC self-renewal. The RNA for the HSC libraries was isolated from HSCs purified from adult BM according to Hoechst 33342 dye efflux, SP, and Sca1+ characteristics [17, 29]. Activated HSCs were isolated from mice 6 days after an injection with the pyrimidine analog 5FU, which kills cycling hematopoietic cells. This in turn stimulates quiescent HSCs to proliferate, thus inducing replenishment of the depleted blood system. CST-BM was built from RNA isolated from adult BM, which contains cells from all blood lineages.

The libraries were constructed using SSH [20] followed by CCS (Fig. 1CGo) [30]. CCS is a sequencing procedure in which cDNA tags are concatenated, resulting in a chimeric concatemer representing many transcripts, followed by shotgun sequencing. In this method, differentially expressed cDNA from two populations are compared through cDNA sequence tags [31]. CCS allows for sequencing of multiple sequence tags from a single sequence read, up to 10 tags per read, permitting more efficient and cost-effective sequencing [32]. Sequence tags generated from SSH-CCS represent a sampling of the major portion of a transcript, and therefore this technique has the potential to obtain information on alternative splicing and intron inclusions. This procedure has the advantage over serial analysis of gene expression (SAGE), which inspects only 3'-ends of transcripts. Furthermore, the increased size of the cDNA tags (~150–250 base pairs [bp]) over SAGE tags (~10–20 bp) increases the specificity, and therefore the accuracy, of mapping these tags reliably on the genome. However, quantitative information with regard to relative gene expression level is lost as a result of SSH, requiring additional techniques to examine transcript abundance.

After CCS, concatenated sequences from single chimeric reads were resolved into individual tags by aligning sequence reads with the mouse genome (Build Mm3, February 2003) using the BLAT algorithm (Fig. 1DGo) [26]. A software tool termed Transcriptomatic (http://www.liver-HSC.scgap.org/txom) was built to automate the various steps needed to map sequence tags onto the genome and present the results in a useful viewing format. Using Transcriptomatic, tags were converted into gene hits by comparing them with the UCSC genome annotation tables (http://genome.ucsc.edu/goldenPath/gbdDescriptions). The identity of a sequence tag was determined using multiple criteria (see Materials and Methods). CST-HSC was composed of 178 clones containing 532 tags representing 394 nonredundant genes, whereas CST-BM consisted of 1361 clones containing 4748 tags representing 1423 nonredundant genes. CST-QuHSC was composed of 570 clones containing 1145 tags representing 568 nonredundant genes, whereas CST-ProHSC consisted of 570 clones containing 802 tags representing 548 nonredundant genes. All of these data can be found at http://www.liver-HSC.scgap.org/txom and have been deposited in GenBank.

Because the SSH-CCS procedure does not reveal information regarding relative expression, alternative techniques were employed. Semi-quantitative real-time RT-PCR was performed to validate the enriched expression of CST-HSC transcripts in HSCs as well as to quantitate their differential expression between HSCs and BM (Fig. 1EGo). Relative expression differences were determined using semi-quantitative RT-PCR using Taqman Expression Assays (Applied Biosystems), which are primer/probe sets that amplify products across exon-exon junctions. Genes selected for validation were represented in CST-HSC multiple times and were either not present in CST-BM or present at a lower level. All genes tested were verified to be expressed in HSCs at a higher level than in BM (Table 1Go).


View this table:
[in this window]
[in a new window]
  		Table 1. Real-time RT-PCR validation of CST-HSC transcripts

 
Expression of Unspliced Pre-mRNA Observed in HSCs
When we examined the distribution of sequence tags within the architecture of genes, we found a significant proportion of tags mapping within introns. This phenomenon was observed in all four libraries but was greatest in CST-HSC, suggesting that it may result from an HSC-specific post-transcriptional regulation (52% in CST-HSC vs. 8%–28% in other three libraries). We ruled out contaminating genomic DNA as the major source of intronic sequences due to the low percentage of tags within intergenic regions (3.5%), despite the fact that approximately 75% of the mouse genome is intergenic [25]. In addition, we tested the RNA used to construct the libraries for genomic DNA contamination. We performed PCR using primers to detect the housekeeping gene ß-actin to assess the presence of genomic DNA. In the absence of reverse transcription (no-RT), the only possible template for PCR in an RNA sample is derived from genomic DNA. No detectable signal was found in the no-RT RNA after 45 cycles of PCR, indicating no significant amount of genomic DNA contamination (data not shown). These observations suggested that another source of DNA or RNA was responsible for the large percentage of intronic sequences.

Another source of intron-containing DNA or RNA present in a cell is unspliced pre-mRNA. Therefore, we wanted to determine what fractions of each gene in CST-HSC and CST-BM were represented by spliced and unspliced mRNA. To ascertain these fractions, we examined the distribution of intronic and exonic tags within each gene observed in CST-HSC and CST-BM. For this analysis, only genes represented by greater than or equal to five tags were examined. Genes represented by only exonic tags were deemed 100% spliced; genes represented by only intronic tags were deemed 100% unspliced. We observed that most genes identified in our libraries (~85%) appeared to be completely spliced or completely unspliced. This result indicates that we rarely observed transcripts in their intermediate state. We thus hypothesized that the spliced and unspliced transcripts might represent biologically different gene groups.

Next we used the GO annotations (http://www.geneontology.org) to analyze the content of the list of spliced and unspliced genes to determine any functional difference between the spliced and unspliced gene lists (Fig. 1EGo). The GO is a semi-hierarchical, controlled vocabulary, which describes gene products in terms of their associated biological processes, cellular components, and molecular functions organized from more general to more specific [33]. In CST-HSC, we found the unspliced gene list to be enriched in the GO categories "DNA binding" (twofold) and "ribonucleoprotein complex" (fivefold) compared with the All Genes list (Fig. 2Go). Some of the genes enriched in these categories include ecotropic viral integration site 1 (Evi1), hepatic leukemia factor (HLF), GATA zinc finger domain containing 2B (GATAd2b), translin, and splicing factor arg/ser-rich 12 (SFRS12). "Catabolism" (3.4-fold) and "carbohydrate metabolism" (3.6-fold) were enriched in the spliced gene list compared with the All Genes list (Fig. 2Go). Some of the genes enriched in these categories include lysozyme, phosphoglycerate mutase I (PGAM1), and matrix metallopeptidase-8 (MMP8). These enrichments were not observed when comparing spliced and unspliced gene lists from CST-BM, suggesting that this is an HSC-specific trend. These observations suggest a biological relevance for the unspliced messages observed in HSCs for transcriptional and post-transcriptional regulation.


Figure 2
View larger version (33K):
[in this window]
[in a new window]
  		Figure 2. GO analysis revealed biological differences between spliced and unspliced transcripts in HSCs. Bar graph depicting the percentage of genes in each gene list within the indicated GO group. All Genes denotes all genes in CST-HSC and CST-BM represented by greater than or equal to five tags. A total of 108 genes fit these criteria. *, no genes in the CST-HSC unspliced gene list found in the GO category "Catabolism." #, enriched in CST-HSC unspliced gene list compared with "All Genes" (p < .02), +, enriched in CST-HSC spliced gene list compared with All Genes (p < .09). All other GO category comparisons on this graph showed an insignificant enrichment (p > .27). Abbreviations: BM, bone marrow; GO, Gene Ontology; HSC, hematopoietic stem cell; CST, CCS tag.

 
Unspliced Transcripts Are Enriched in Quiescent HSCs
In other studies analyzing gene expression in various mammalian cell lines, it was observed that the levels of several transcripts increased when cells transit from to G0 G1/S upon stimulation [14–16, 34]. The mechanisms producing the cell cycle-dependent changes in transcript levels were found not to be at the level of transcription but at the level of mRNA regulation. Because normal HSCs are largely in G0 [35], this may account for our observation that a large portion of mRNA in HSCs is unspliced. Thus, we followed pre-mRNA levels in HSCs over a 5FU-induced activation time course, during which HSCs are induced to proliferate and then return to quiescence in vivo (Fig. 1EGo). Induction of HSC proliferation occurs in a time-dependent manner with proliferation peaking at 6 days after 5FU treatment. HSCs isolated 0, 3, or 6 days after activation were used to determine variations in unspliced and spliced message levels.

In our analysis, we detected spliced mRNAs using exon-exon primers that spanned the intervening intron; unspliced mRNAs were detected by exon-intron primers (Fig. 3AGo) [36]. For each gene tested, three or more primer pairs were used that assayed two distinct exon-exon pairs and their connecting introns. The selected genes were chosen based on one of two criteria. They were previously shown to be controlled by post-transcriptional regulation in response to an activating signal in a pituitary cell line (Pitx2 and Cyclin D2) [37] or shown through microarray analysis to exhibit different expression behaviors after HSC activation (CTLA2-{alpha}, c-Myc, and Mcm3) [6].


Figure 3
View larger version (31K):
[in this window]
[in a new window]
  		Figure 3. Direct examination of spliced and unspliced transcripts reveals changes in pre-mRNA levels after HSC activation. (A): Diagram of strategy to detect and distinguish spliced versus unspliced mRNA. (B): Graphs showing the real-time polymerase chain reaction (PCR) results of spliced versus unspliced mRNA expression after hematopoietic stem cell (HSC) activation. Plotted values are normalized to ß-actin. *, Pitx2 not detected in HSCs at day 0. For each time point, the experiment was performed twice on separate cell isolations. For each cell isolation, real-time reverse transcription (RT)-PCR was performed in duplicate. Abbreviations: 5FU, 5-fluorouracil; Ct, threshold cycle; HSC, hematopoietic stem cell.

 
Our PCR results for spliced and unspliced products indicated that differences between pre-mRNA and mature mRNA do occur in HSCs. Levels of Pitx2 and CTLA2-{alpha} spliced and unspliced products changed coordinately when HSCs went into cycle (Fig. 3BGo; Table 2Go). Levels of both spliced and unspliced fractions of the c-Myc transcript showed some downregulation when HSCs were activated (Fig. 3BGo). However, the negative fold-change in message levels from day 0 to day 3 after activation was greater in the unspliced message than in the spliced message (-6 vs. -1.22), suggesting differential regulation on the c-Myc pre-mRNA and mature mRNA (Table 2Go). Because the level of pre-mRNA decreased to a greater extent than the mature transcript, mRNA stability, as well as splicing, may be a form of post-transcriptional regulation used. Cyclin D2, a gene important to the G1-to-S transition, showed differences between spliced and unspliced messages after activation. The cyclin D2 spliced message did not change dramatically over the time course, whereas the unspliced message displayed a continual decrease in levels (Fig. 3BGo). Consistent with an increase in splicing with activation, Mcm3, a gene important to DNA replication, showed vastly different patterns of spliced and unspliced messages after HSC activation. Mcm3 spliced message exhibited an increase, whereas a decrease in unspliced Mcm3 message in cycling HSCs was observed. These data suggest a non-transcriptional role for the cell cycle–dependent change in Mcm3 message, potentially at the level of pre-mRNA splicing. This form of post-transcriptional regulation could serve a biological role in HSC activation.


View this table:
[in this window]
[in a new window]
  		Table 2. Fold changes of transcripts from days 0–3 after hematopoietic stem cell activation

 
Expression of mRNA Processing Genes in Quiescent Versus Proliferating HSCs
Because our results suggested a role for mRNA processing in HSC activation, we asked whether gene expression changes in mRNA splicing machinery also occurred after HSC activation. In the literature, there are reports of select transcripts existing as high-abundance unspliced message during G0 of cell cycle and then rapidly undergoing splicing when the cell enters the cell cycle [14, 15]. Associated with these observations was the recent finding that major components of pre-mRNA splicing and export machinery show an increase in gene expression concomitant with entry into the cell cycle [34]. We hypothesized that the levels of mRNA processing factors would be low in resting HSCs but should increase in response to activation. To examine this possibility, we surveyed the expression of these genes over an HSC activation time course using microarray data from Venezia et al. [6]. The time course studied is similar to the one described in the above section. Genes that showed highest gene expression from days 2–6 were considered to increase expression with proliferation state (Fig. 4Go). Time-dependent expression patterns for each gene were determined by fitting smooth curves to the expression profiles using regression analysis followed by analysis of variance to identify genes with significant time patterns [6]. We analyzed more than 150 probe sets that were in a GO category involved in splicing or RNA processing (1.2% of total genes on Affymetrix MGU74Av2 array; Santa Clara, CA, http://www.affymetrix.com). Only 33 genes exhibited a significant time-dependent gene expression pattern (p < .05). Of these, 67% (22/33 genes) exhibited an increase in gene expression affiliated with increases in HSC proliferation (3% of the total genes showing an increased gene expression pattern) (Fig. 4Go). Although this enrichment is not large, it is nonrandom according to a {chi}2 test (p < .01). These data indicate a link between high levels of unspliced message observed and the low levels of mRNA processing observed in resting HSCs.


Figure 4
View larger version (35K):
[in this window]
[in a new window]
  		Figure 4. Splicing factors show an activation-dependent expression increase in hematopoietic stem cell (HSCs). Plots displaying the time-dependent gene expression profile of the splicing factor genes over a time course of HSC activation and recovery. All genes shown have an r2 > 0.7 and p < .05 [6]. For each time point, microarrays were performed in duplicate. The three graphs display the pattern of gene expression and are separated based on the overall intensity of the gene expression value. Abbreviation: 5FU, 5-fluorouracil.

 

   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
Improvements in sequencing and microarray technologies have propelled gene expression profiling into the forefront of genomics research. Microarrays permit one to simultaneously and reproducibly assay the expression levels of tens of thousands of genes, but preclude novel gene discovery. In contrast, sequencing studies assess gene expression of a particular cell type in an unbiased fashion that has the potential to uncover new transcripts with each discrete experiment. In addition, when using a tag-based sequencing strategy, such as CCS, which samples through the length of the transcript rather than at one end, we can obtain additional information about differential regulation and alternative splicing.

Moreover, the completed genomes allow investigators to analyze genomic data at a deeper level. Through comparison with the mouse genome, we found pre-mRNA (as indicated by a high prevalence of intronic material in our cDNA) to be abundant in HSCs. Pre-mRNA was also detected in several other HSC datasets but was not recognized as such due to the lack of the mouse genome [3, 7]. Other studies analyzing human cDNA libraries did recognize pre-mRNA as a major component but dismissed the presence of this population as uninformative [38, 39].

Regulation of mRNA processing and stability is a key post-transcriptional mechanism to control mature transcript levels and may be crucial for transcriptome homeostasis. The data from our study suggest that regulation at the level of pre-mRNA might be a form of control used in HSCs. Analysis of changes in pre-mRNA levels in HSCs over an activation time course revealed rapid changes in unspliced mRNA levels that coordinate with changes in proliferation state. In addition, we observed increases in expression of splicing factors after activation. These findings suggest that changes in mRNA processing and stability may be key factors in HSC activation control. Our data also suggest that after a stress signal (5FU), selective processing of transcripts (i.e., c-Myc, Cyclin D2, and Mcm3) required during cell division can occur. Our work here has shown a correlation between expression of splicing factors, splicing, and cell proliferation in HSCs.

A previous study, which examined the changes in SerpinF1 transcript levels in quiescent and proliferating fibroblast cell lines, showed that control of the abundance of this transcript was completely regulated post-transcriptionally, underscoring the potential importance of mRNA processing and stability in regulating transcript levels [16]. In addition, this regulation was found to be dependent on the proliferative status of the cell line. Studies on other transcripts revealed a similar correlation between transcript stability, processing, and proliferative state [14, 15, 34]. Activation of specific pro-proliferative signaling pathways, such as Wnt/ß-catenin (a known mediator of HSC proliferation), has also been linked to post-transcriptional mRNA control [37]. These previous studies highlighted a role for post-transcriptional regulation in a transcript-specific manner.

Some potential mechanisms for how post-transcriptional regulation might modify a cellular response in a transcription-independent manner have been proposed. One commonality to post-transcriptional regulation is the modification of the levels of proteins involved with a particular cellular response. Some examples of post-transcriptional regulatory mechanisms include nuclear retention [40] and intron retention [41, 42]. The general mechanism of nuclear retention is the sequestration of select noncoding regulatory poly(A) RNAs in distinct subnuclear regions termed paraspeckles. The nuclear retained RNAs can be alternative splice forms of other protein coding transcripts. The noncoding RNA regulates the translation of the corresponding protein coding transcript. Intron retention also titrates the total amount of protein produced. By retaining an intron, the transcript is not fully processed and ready for translation. Thus, the amount of fully processed transcript, not total transcript, is the important factor for determining protein production.

The relevance of nuclear retention on stress response was shown in a study characterizing the paraspeckle localized poly(A) RNAs [40]. Here, they showed the expression of an alternative spliceform from the mCAT2 gene that produced a regulatory nuclear-retained poly(A)RNA. The mCAT2 protein is critical for the rapid response needed after a stress signal to modulate L-arginine uptake for the NO pathway. This reported mCAT2 regulatory RNA, CTN-RNA, could regulate the translation of mCAT2 in a rapid stress-inducible fashion. CTN-RNA was sequestered in the nucleus until a stress response signal was detected. The stress signal induced translocation of CTN-RNA into the cytoplasm and led to increased mCAT2 protein production.

Intron retention is another potential mechanism of stress responsive post-transcriptional regulation. One example is the UPR-induced splicing of the stress-responsive transcription factor HAC1 in yeast [43]. Prior to a stress response, HAC1 RNA is present in an unspliced form. Upon activation of the unfolded protein response, HAC1 RNA is then spliced by the endoribonuclease IRE1, translated into Hac1 protein, and then in turn activates the transcription of stress-responsive genes. On a more global level, intron retention was found to be a form of alternative splicing [42]. In Arabidopsis, greater than 17% of alternatively spliced transcripts contained a retained intron. In addition, the majority of these intron-retaining transcripts were found to have a role in biological processes involved in responding to stress and external stimuli. Recently, alternative splicing was shown to be a component in HSC gene expression [44]. The results from these works implicate intron retention as an important component of post-transcriptional stress response. Our data suggest a high preponderance of intron-retaining transcripts in HSCs that could be important in cellular responses to stress such as treatment with the chemotherapeutic drug 5FU. The effects of the observed transcript-specific intron retention on protein levels and stress responsiveness in HSCs need to be explored further to determine the exact mechanism of action.

The high levels of unspliced transcripts found in HSCs might also serve as a source of intron-derived noncoding regulatory RNAs. The emerging evidence for functional roles of noncoding RNAs derived from introns of known genes stresses the importance of identifying and characterizing such transcripts. The importance of noncoding RNAs has recently become more apparent with the discovery of microRNAs, which are believed to be natural mediators of RNA interference. MicroRNAs can mediate cellular responses by altering the stability of a transcript [45, 46]. Recently, it has been shown that the majority of human microRNAs are derived from the introns of protein-coding transcripts [47]. Because our HSC libraries are enriched for intronic sequences of protein-coding transcripts, these data offer a searchable set of potential HSC-enriched microRNA precursors. Moreover, regulation of mRNA stability and microRNA production may be related processes. Continued exploration of the role both processes have in HSC function will be important to our understanding of transcriptional and post-transcriptional regulation and their involvement in HSC fate decisions.


   DISCLOSURES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The authors indicate no potential conflicts of interest.


   ACKNOWLEDGMENTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
This work was funded by grants from the NIH (CA78792 and DK63588). M.A.G. is a scholar of the Leukemia and Lymphoma Society. A.A.M. is a Molecular Medicine Scholar. We thank Richard Gibbs for critical reading of the manuscript. We thank Mike Cubbage and Chris Threeton for excellent flow cytometry assistance, Lisa White at the Baylor Core Microarray Facility, and the staff at the Baylor Human Genome Sequencing Center.


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Akashi K, He X, Chen J et al. Transcriptional accessibility for genes of multiple tissues and hematopoietic lineages is hierarchically controlled during early hematopoiesis. Blood 2003;101:383–389.[Abstract/Free Full Text]

  2. Ivanova NB, Dimos JT, Schaniel C et al. A stem cell molecular signature. Science 2002;298:601–604.[Abstract/Free Full Text]

  3. Phillips RL, Ernst RE, Brunk B et al. The genetic program of hematopoietic stem cells. Science 2000;288:1635–1640.[Abstract/Free Full Text]

  4. Ramalho-Santos M, Yoon S, Matsuzaki Y et al. "Stemness": Transcriptional profiling of embryonic and adult stem cells. Science 2002;298:597– 600.[Abstract/Free Full Text]

  5. Terskikh AV, Easterday MC, Li L et al. From hematopoiesis to neuropoiesis: Evidence of overlapping genetic programs. Proc Natl Acad Sci U S A 2001;98:7934–7939.[Abstract/Free Full Text]

  6. Venezia TA, Merchant AA, Ramos CA et al. Molecular signatures of proliferation and quiescence in hematopoietic stem cells. PLoS Biol 2004;2:e301.[CrossRef][Medline]

  7. Sharov AA, Piao Y, Matoba R et al. Transcriptome analysis of mouse stem cells and early embryos. PLoS Biol 2003;1:E74.[Medline]

  8. Kiel MJ, Yilmaz OH, Iwashita T et al. SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell 2005;121:1109–1121.[CrossRef][Medline]

  9. Forsberg EC, Prohaska SS, Katzman S et al. Differential expression of novel potential regulators in hematopoietic stem cells. PLoS Genet 2005;1:e28.

  10. Miyamoto T, Iwasaki H, Reizis B et al. Myeloid or lymphoid promiscuity as a critical step in hematopoietic lineage commitment. Dev Cell 2002;3:137–147.[CrossRef][Medline]

  11. Hu M, Krause D, Greaves M et al. Multilineage gene expression precedes commitment in the hemopoietic system. Genes Dev 1997; 11:774–785.[Abstract/Free Full Text]

  12. Ben-Yehuda S, Dix I, Russell CS et al. Genetic and physical interactions between factors involved in both cell cycle progression and pre-mRNA splicing in Saccharomyces cerevisiae. Genetics 2000;156:1503–1517.[Abstract/Free Full Text]

  13. Lundgren K, Allan S, Urushiyama S et al. A connection between pre-mRNA splicing and the cell cycle in fission yeast: cdc28+ is allelic with prp8+ and encodes an RNA-dependent ATPase/helicase. Mol Biol Cell 1996;7:1083–1094.[Abstract]

  14. Ash J, Ke Y, Korb M et al. Introns are essential for growth-regulated expression of the mouse thymidylate synthase gene. Mol Cell Biol 1993;13:1565–1571.[Abstract/Free Full Text]

  15. Darville MI, Rousseau GG. E2F-dependent mitogenic stimulation of the splicing of transcripts from an S phase-regulated gene. Nucleic Acids Res 1997;25:2759–2765.[Abstract/Free Full Text]

  16. Coljee VW, Rotenberg MO, Tresini M et al. Regulation of EPC-1/PEDF in normal human fibroblasts is posttranscriptional. J Cell Biochem 2000; 79:442–452.[CrossRef][Medline]

  17. Goodell MA, Brose K, Paradis G et al. Isolation and functional properties of murine hematopoietic stem cells that are replicating in vivo. J Exp Med 1996;183:1797–1806.[Abstract/Free Full Text]

  18. Osawa M, Hanada K, Hamada H et al. Long-term lymphohematopoietic reconstitution by a single CD34-low/negative hematopoietic stem cell. Science 1996;273:242–245.[Abstract]

  19. Camargo FD, Chambers SM, Drew E et al. Hematopoietic stem cells do not engraft with absolute efficiencies. Blood 2006;107:501–507.[Abstract/Free Full Text]

  20. Diatchenko L, Lau YF, Campbell AP et al. Suppression subtractive hybridization: A method for generating differentially regulated or tissue-specific cDNA probes and libraries. Proc Natl Acad Sci U S A 1996; 93:6025–6030.[Abstract/Free Full Text]

  21. Richards S, Liu Y, Bettencourt BR et al. Comparative genome sequencing of Drosophila pseudoobscura: Chromosomal, gene, and cis-element evolution. Genome Res 2005;15:1–18.[Abstract/Free Full Text]

  22. Havlak P, Chen R, Durbin KJ et al. The Atlas genome assembly system. Genome Res 2004;14:721–732.[Abstract/Free Full Text]

  23. Jurka J. Repbase update: A database and an electronic journal of repetitive elements. Trends Genet 2000;16:418–420.[CrossRef][Medline]

  24. Smit A, Hubley R, Green P. RepeatMasker Open-3.0. 1996–2004. Available at http://repeatmasker.org. Accessed July 14, 2005.

  25. Waterston RH, Lindblad-Toh K, Birney E et al. Initial sequencing and comparative analysis of the mouse genome. Nature 2002;420:520–562.[CrossRef][Medline]

  26. Kent WJ. BLAT–the BLAST-like alignment tool. Genome Res 2002;12: 656–664.[Abstract/Free Full Text]

  27. Karolchik D, Baertsch R, Diekhans M et al. The UCSC Genome Browser Database. Nucleic Acids Res 2003;31:51–54.[Abstract/Free Full Text]

  28. Hosack DA, Dennis G Jr, Sherman BT et al. Identifying biological themes within lists of genes with EASE. Genome Biol 2003;4:R70.[CrossRef][Medline]

  29. Spangrude GJ, Heimfeld S, Weissman IL. Purification and characterization of mouse hematopoietic stem cells. Science 1988;241:58–62.[Abstract/Free Full Text]

  30. Gunaratne PH, Wu JQ, Garcia AM et al. Concatenation cDNA sequencing for transcriptome analysis. C R Biol 2003;326:971–977.[Medline]

  31. Qiu J, Gunaratne P, Peterson LE et al. Novel potential ALL low-risk markers revealed by gene expression profiling with new high-throughput SSH-CCS-PCR. Leukemia 2003;17:1891–1900.[CrossRef][Medline]

  32. Gunaratne PH. Advances in genome biology and technology. Expert Rev Mol Diagn 2004;4:757–760.[CrossRef][Medline]

  33. Ashburner M, Ball CA, Blake JA et al. Gene ontology: Tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 2000;25:25–29.[CrossRef][Medline]

  34. Pryor A, Tung L, Yang Z et al. Growth-regulated expression and G0-specific turnover of the mRNA that encodes URH49, a mammalian DExH/D box protein that is highly related to the mRNA export protein UAP56. Nucleic Acids Res 2004;32:1857–1865.[Abstract/Free Full Text]

  35. Arai F, Hirao A, Ohmura M et al. Tie2/angiopoietin-1 signaling regulates hematopoietic stem cell quiescence in the bone marrow niche. Cell 2004;118:149–161.[CrossRef][Medline]

  36. Clark TA, Sugnet CW, Ares M Jr. Genomewide analysis of mRNA processing in yeast using splicing-specific microarrays. Science 2002; 296:907–910.[Abstract/Free Full Text]

  37. Briata P, Ilengo C, Corte G et al. The Wnt/beta-catenin->Pitx2 pathway controls the turnover of Pitx2 and other unstable mRNAs. Mol Cell 2003;12:1201–1211.[CrossRef][Medline]

  38. Sorek R, Safer HM. A novel algorithm for computational identification of contaminated EST libraries. Nucleic Acids Res 2003;31:1067–1074.[Abstract/Free Full Text]

  39. Kampa D, Cheng J, Kapranov P et al. Novel RNAs identified from an in-depth analysis of the transcriptome of human chromosomes 21 and 22. Genome Res 2004;14:331–342.[Abstract/Free Full Text]

  40. Prasanth KV, Prasanth SG, Xuan Z et al. Regulating gene expression through RNA nuclear retention. Cell 2005;123:249–263.[CrossRef][Medline]

  41. Mansilla A, Lopez-Sanchez C, de la Rosa EJ et al. Developmental regulation of a proinsulin messenger RNA generated by intron retention. EMBO Rep 2005.

  42. Ner-Gaon H, Halachmi R, Savaldi-Goldstein S et al. Intron retention is a major phenomenon in alternative splicing in Arabidopsis. Plant J 2004;39:877–885.[CrossRef][Medline]

  43. Sidrauski C, Walter P. The transmembrane kinase Ire1p is a site-specific endonuclease that initiates mRNA splicing in the unfolded protein response. Cell 1997;90:1031–1039.[CrossRef][Medline]

  44. Pritsker M, Doniger TT, Kramer LC et al. Diversification of stem cell molecular repertoire by alternative splicing. Proc Natl Acad Sci U S A 2005;102:14290–14295.[Abstract/Free Full Text]

  45. Gazzani S, Lawrenson T, Woodward C et al. A link between mRNA turnover and RNA interference in Arabidopsis. Science 2004;306: 1046–1048.[Abstract/Free Full Text]

  46. Lim LP, Lau NC, Garrett-Engele P et al. Microarray analysis shows that some microRNAs downregulate large numbers of target mRNAs. Nature 2005;433:769–773.[CrossRef][Medline]

  47. Rodriguez A, Griffiths-Jones S, Ashurst JL et al. Identification of mammalian microRNA host genes and transcription units. Genome Res 2004;14:1902–1910.[Abstract/Free Full Text]





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Table
Right arrow All Versions of this Article:
2005-0552v1
24/3/662    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bowman, T. V.
Right arrow Articles by Goodell, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bowman, T. V.
Right arrow Articles by Goodell, M. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
STEM CELLS THE ONCOLOGIST CME ALPHAMED PRESS JOURNALS
http://www.stemcellsportal.com/