Hypoxia Causes Downregulation of Mismatch ... - Wiley Online Library

14 downloads 59 Views 2MB Size Report
Mismatch repair • Hypoxia-inducible factor • Microsatellite instability • Hypoxia • Cancer stem cells ... genomic instability in the murine and human stem cells.
THE STEM CELL NICHE Hypoxia Causes Downregulation of Mismatch Repair System and Genomic Instability in Stem Cells FRANCISCO JAVIER RODR I´GUEZ-JIM E´ NEZ, VICTORIA MORENO-MANZANO, RUT LUCAS-DOMINGUEZ, ´ JOS E´ -MAR I´A SANCHEZ -PUELLES Farmacologı´a Molecular, Centro de Investigacio´n Prı´ncipe Felipe, Valencia, Spain Key Words. Mismatch repair • Hypoxia-inducible factor • Microsatellite instability • Hypoxia • Cancer stem cells

ABSTRACT The DNA mismatch repair (MMR) system maintains genomic integrity by correcting replication errors: its malfunction causes genomic instability in several tumor types. Hypoxia-inducible factor-1␣ (HIF1␣), the major regulator of the processes that occur in hypoxia and certain epigenetic events downregulate the expression of MMR genes in cancer cells. However, there is a lack of information regarding MMR regulation and the genetic stability of stem cells under hypoxic conditions. The expression of the MMR system is downregulated in murine and human stem cells cultured in hypoxia, which correlates with lower DNA repair activity in neural stem cells. We observed, through the use of short hairpin loop RNAi expression constructs, that HIF1␣ positively regulated MLH1 and MSH6 when the C17.2 neural stem cells were exposed to short-term hypoxia. However, in prolonged

exposure to oxygen depletion, the reduced transcriptional activation of MMR genes was directed by specific epigenetic events. Chromatin immunoprecipitation experiments showed a hypoacetylated/hypermethylated histone H3 and lower SP1 binding within MLH1 and MSH6 adjacent promoter regions. Treatment with the histone deacetylase inhibitor trichostatin A increased histone H3 acetylation and SP1 occupancy and enhanced MMR expression. Sequencing of microsatellite markers revealed genomic instability in the murine and human stem cells grown under hypoxia. Thus, the present article reports, for the first time in the stem cell field, experimental data that indicate that hypoxic niches are an environment in which stem cells might undergo genomic instability, which could lie at the origin of subpopulations with cancer stem cell properties. STEM CELLS 2008;26:2052–2062

Disclosure of potential conflicts of interest is found at the end of this article.

INTRODUCTION An increasing body of research has shown that certain cancers contain multipotent neural stem cell-like cancer cells with tumor-initiating ability, which may originate this malignancy [1–3]. Stem cells may undergo oncogenic transformation due to mutations, probably influenced by their environment, which leads them to become cancer stem cells, with uncontrolled and unlimited proliferation [4]. Neural stem cells exist within a physiological hypoxia (1%–5% O2) in both embryonic and adult brains [5]. Hypoxia-inducible factor-1 (HIF1) is a crucial regulator of tumor cell adaptation to hypoxic stress and is the key element responsible for the regulation, under low oxygen tension, of many genes involved in important biological processes, such as glycolysis, proliferation, and angiogenesis [6]. HIF1 is a heterodimer composed of an oxygen-regulated ␣ subunit and a constitutively expressed ␤ subunit. Overexpression of HIF1␣ because of intratumoral hypoxia determines major physiological pathways in tumor growth and neovascularization [7]. High

HIF1␣ levels contribute to tumor progression and are a marker of aggressive disease in several tumor types [8, 9]. Moreover, hypoxia is responsible for diminished DNA repair and, therefore, high mutagenesis [10, 11], as well as for enrichment of mismatch repair (MMR)-deficient cells with augmented microsatellite instability (MSI) [12]. MutS (MSH2, MSH3, MSH6) and two MutL homologues (MLH1, PMS2) make up the MMR family responsible for the DNA eukaryote repair process. The MutS␣ complex (MSH2–MSH6) is responsible for the repair of base-base mispairs, whereas MutS␤ (MSH2–MSH3) repairs larger insertions/deletions. Both complexes recruit the MutL heterodimer to initiate downstream repair events [13]. Mutation and loss of the MMR machinery is associated with increased genomic instability, shown in increased mutation rates, especially at microsatellite loci [14]. MSI is considered a hallmark of nonpolyposis colorectal cancer [15] and is also present in a significant proportion of other cancer types [16, 17]. However, the role of MMR deficiency in the pathogenesis of brain tumors remains controversial [18]. During the carcinogenic process, epigenetic alterations involving altered methylation and chro-

Author contributions: F.J.R.-J.: conception and design, data analysis and interpretation, manuscript writing, laboratory managing; V.M.-M.: conception and design, collection and assembly of data, data analysis; R.L.-D.: provision of study material, collection and assembly of data; J.-M.S.-P.: administrative support, financial support, conception and design, interpretation of data, manuscript writing, laboratory managing, final approval of manuscript. Correspondence: Jose´-Marı´a Sa´nchez-Puelles, Ph.D., Department Farmacologı´a Molecular, Fundacio´n Valenciana Centro de Investigacio´n Prı´ncipe Felipe, 46012 Valencia, Spain. Telephone: 34963289680; Fax: 34963289701; e-mail: [email protected]; or Francisco Javier Rodrı´guezJime´nez, Farmacologı´a Molecular, Centro de Investigacio´n Prı´ncipe Felipe, Valencia, Spain. Telephone: ⫹34963289680; Fax: 34963289701; e-mail: [email protected] Received December 10, 2007; accepted for publication May 18, 2008; first published online in STEM CELLS EXPRESS May 29, 2008. ©AlphaMed Press 1066-5099/2008/$30.00/0 doi: 10.1634/stemcells.2007-1016

STEM CELLS 2008;26:2052–2062 www.StemCells.com

Rodrı´guez-Jime´nez, Moreno-Manzano, Lucas-Dominguez et al. matin remodeling by histone modification lead to the functional loss of critical genes, such as tumor suppressor or DNA repair genes. Therefore, MLH1, but not MSH2, MSH3, or MSH6, can be silenced by promoter hypermethylation, a mechanism underlying the presence of the MSI in gastric [19] and endometrial carcinomas [20]. Aberrant CpG island methylation in the promoter region is associated with transcriptionally repressive chromatin and seems to be linked to the deacetylation of histones. Deacetylation may be important in the initial silencing of transcription, usually accompanied by aberrant methylation [21]. Both epigenetic events play a crucial role in stem cell identity and tumorigenesis [22]. It has become increasingly clear that epigenetic mechanisms regulate the access of certain transcription factors to their binding sites. In fact, deacetylation and trimethylation of histone H3 on lysine 9 (H3K9Me3) impair SP1 occupancy [23]. In addition, CpG island methylation prevents SP1 binding [24] and vice versa [25, 26]. There is a wide body of research concerning the epigenetic modifications that occur in tumor cells. However, little is known about the effects that altered epigenetic events may produce on the expression of DNA repair genes and the genomic integrity of stem cells under hypoxic conditions. In the research reported here, we used C17.2 cells, which are considered as operationally defined neural stem cells (NSC) [27, 28] and a promising vector for the treatment of regenerative diseases [29]. C17.2 cells differentiate toward replacement of neurons [30], as a therapeutic tool in Parkinson’s disease [29, 31] and in spinal cord regeneration [32]. They migrate specifically toward an advancing neoplasm and trail islands of tumor cells migrating away from the tumor mass [33]. Similar to C17.2, a human neural stem cell line (IhNSC) immortalized by v-myc, endowed with the properties of human NSC, has recently been reported as a suitable cell line for developing assays that are essential for diagnoses and cell therapy studies [34]. Here we demonstrate, for the first time in the stem field, the repression of the MMR system in murine and human stem cells under hypoxia. We show that HIF1␣ is not directly involved in the repression of MMR in C17.2, as described for cancer cells. Our results show that MLH1 and MSH6 transcriptional downregulation in hypoxia are associated with hypoacetylated and hypermethylated histone H3 that impair SP1 binding in their promoter regions. The deregulated MMR system, caused by hypoxia, may contribute to generating genomic instability in stem cells, which leads to malignant transformation into cancer stem cells.

MATERIALS

AND

METHODS

Mouse Neural and Human Mesenchymal Stem Cell Culture Stem cells were cultured in an undifferentiated state. C17.2 were grown in high-glucose Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum, 5% horse serum, 1% glutamine (2 mM) and 1% penicillin/streptomycin/fungizone [27]. Primary neurosphere culture was obtained from olfactory bulb of embryonic CD1 mice (13.5–14.5 days post coitum) by mechanical dissociation, in line with the ethical and legal rules of the European Commission (no. L358, ISSN 0738-6978). Neurospheres were identified, isolated, and expanded using the neurosphere assay [35] and cultured as a suspension in DMEM/Ham’s F-12 medium, supplemented with 100 U/ml penicillin, 100 ␮g/ml streptomycin, 2 mM L-glutamine, 5 mM Hepes buffer, 0.125% NaHCO3, 0.6% glucose, 0.025 mg/ml insulin, 80 ␮g/ml apotransferrin, 16 nM progesterone, 60 ␮M putrescence, 24 nM sodium selenite, 4 ␮g/ml bovine serum albumin, 0.7 U/ml heparin, 20 ng/ml epidermal growth factor (Sigma-Aldrich, St. Louis, http://www. sigmaaldrich.com), and 20 ng/ml fibroblast growth factor 2 (Sigma-

www.StemCells.com

2053

Aldrich) for 7 days. Hypoxia was created by culturing cells in a chamber (In vivo2 400; Ruskinn Life Sciences, Pencoed, Bridgen, U.K., http://www.ruskinn.com) flushed with premixed gas consisting of 1% O2, 5% CO2, and 94% N2 at 37°C. Human bone marrow mesenchymal stem cells (BMMSC) derived from a single 19-yearold healthy donor were purchased from Inbiobank (San Sebastian, Spain, http://www.inbiobank.org). Dental pulp stem cells (DPSC) came from normal molars freshly extracted for orthodontic reasons from young donors. Both cell types were cultured in DMEM supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin. Murine neural stem cells were cultured for 3, 6, 9, 24, and 48 hours in hypoxic conditions, and human mesenchymal stem cells were cultured for 3 and 48 hours. Cells were harvested for analysis after treatment. The effects of hypoxia (created by using an In Vivo2 400 chamber) for cell viability were calculated by the propidium iodide staining method, and no significant death was observed (Molecular Probes, Eugene, OR, http://probes.invitrogen.com). Cell death produced during the above-mentioned time course was determined by the detection of propidium iodide emission with the FC 500 Series Flow Cytometry System (Beckman Coulter, Fullerton, CA, http://www.beckmancoulter.com).

RNA and DNA Isolation Total RNA or genomic DNA was extracted by using NucleoSpin RNA/protein or NucleoSpin Blood (Macherey Nagel, Du¨ren, Germany, http://www.mn-net.com), respectively, in line with the manufacturer’s instructions.

Real-Time Polymerase Chain Reaction (TaqMan) Total RNA obtained from murine C17.2 cells and neurospheres or human BMMSC and DPSC was independently reverse-transcribed using TaqMan reverse transcription reagents (Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com), following the manufacturer’s instructions. It was analyzed by the ABI Prism 5700 Sequence Detection System instrument and software (Applied Biosystems). Each reaction was carried out in duplicate in three independent experiments. As template, 40 ng of cDNA from target and housekeeping genes was prepared in separate tubes for each TaqMan reaction. MGB Assay-on-Demand TaqMan probes (Applied Biosystems) are shown in supplemental online Table 1. The comparative threshold cycle (CT) method was used to calculate the relative expression [36]. For quantification of gene expression, the target gene value normalized to the expression of an endogenous reference (GAPDH) was designated ⌬CT (⌬CT ⫽ CT [test gene] ⫺ CT [GAPDH]). ⌬CT for hypoxic samples was then subtracted from the ⌬CT for normoxic samples, to generate ⌬⌬CT (⌬⌬CT ⫽ ⌬CT [hypoxia] ⫺ ⌬CT [normoxia]). The mean of these ⌬⌬CT measurements was used to calculate the fold change in gene expression (2⫺⌬⌬CT). Representative results were presented as the mean ⫾ SE. For chromatin immunoprecipitation (ChIP) experiments, specific primers used in real-time polymerase chain reaction (PCR) are listed in supplemental online Table 1.

Western Blotting Analysis Cells were collected and proteins were extracted by using NucleoSpin RNA/protein (Macherey Nagel), according to the manufacturer’s instructions. Equal amounts of protein extracts (50 ␮g) were loaded onto a 10% SDS-polyacrylamide gel and resolved by standard SDS-polyacrylamide gel electrophoresis. Proteins were electrophoretically transferred onto polyvinylidene difluoride membranes. Membranes were blocked with 3% skim milk in Tris Buffered Saline Tween 20 for 90 minutes and probed overnight with specific antibodies against HIF1␣ at a 1:200 dilution (NeoMarkers; Lab Vision, Fremont, CA, http://www.labvision.com), EGLN3 at a 1:1,000 dilution (Novus Biologicals, Inc., Littleton, CO, http:// www.novusbio.com), MSH6 at a 1:500 dilution (BD Pharmingen, San Diego, http://www.bdbiosciences.com/index_us.shtml), MLH1 at a 1:500 dilution (BD Pharmingen), and MSH2 at a 1:500 dilution (Calbiochem, San Diego, http://www.emdbiosciences.com). ␤-Actin at a 1:5,000 dilution (Sigma-Aldrich) and glyceraldehyde-3phosphate dehydrogenase (GAPDH) at a 1:10,000 dilution (Trevigen, Gaithersburg, MD, http://www.trevigen.com) were used as

2054

loading controls. Subsequently, membranes were incubated with rabbit anti-mouse or donkey anti-rabbit horseradish peroxidaseconjugated secondary antibody (1:5,000) (Sigma-Aldrich). Blots were viewed by the ECL (Amersham Biosciences, Buckinghamshire, U.K., http://www.amersham.com) detection system. Relative protein levels were quantitated by densitometry with Quantity One software (Bio-Rad, Hercules, CA, http://www.bio-rad.com). Results were standardized, using ␤-actin as the reference.

Short Hairpin Loop RNAi Expression Constructs for HIF1␣ in C17.2 Cells Short hairpin loop RNA interference expression constructs (shRNAi) target sequences against mHIF1␣ (GenBank NM_010431) were designed by the Dharmacon siDesign Center (Dharmacon, Inc., Lafayette, CO, http://www.dharmacon.com). Expression cassettes for shRNAi target sequences were formed by the previous 17 nucleotides for directional cloning, 19 nucleotide sequences homologous to the target gene of interest (supplemental online Table 1, uppercase letters), a loop of 9 nucleotides (lowercase letters), 19 nucleotides for the complementary sequence of the antisense strand (uppercase letters), and the H1 promoter obtained from the pSUPER vector (OligoEngine, Seattle, WA, http://www.oligoengine.com). The target sequences were amplified by the primers shown in supplemental online Table 1. We used 50 ng of pSUPER vector as template and 2.5 U of Pfu DNA polymerase (Stratagene, La Jolla, CA, http://www.stratagene.com). The PCR cycling conditions were 95°C for 2 minutes; 30 cycles of 95°C for 30 seconds, 50°C for 30 seconds, and 72°C for 1 minute; and a final cycle of 72°C for 10 minutes. Amplified sequences were cloned in pENTRY vector (Invitrogen, Carlsbad, CA, http://www.invitrogen.com), and after purification of positive clones, target sequences were included in the pA70 destination vector, derived from pSUPER-Retro. Plasmid DNA transfections were performed with HEK293 cells at 60%– 70% confluence by addition of 5 ␮g of retroviral vector (pA70Scramble or pA70-shRNAi target) and 5 ␮g of ecotropic vector in 1.5 ml of Opti-MEM I Reduced Serum Medium (Invitrogen, Paisley, U.K., http://www.invitrogen.com). As reference, we used the pA70 retroviral vector containing a scramble sequence (supplemental online Table 1). Then 48 ␮l of Lipofectamine 2000 (Invitrogen) with 1.5 ml of Opti-MEM I Reduced Serum Medium and DNA were mixed and incubated for 20 minutes at room temperature before transfections. Supernatants from HEK293 cells containing reconstituted viruses were added to C17.2. Positive infected cells were selected with puromycin (2 ␮g/ml) for 3 days. Cells that survived selection were expanded and analyzed for target expression by TaqMan analysis or Western blotting.

Chromatin Immunoprecipitation Chromatin immunoprecipitation analysis used ChIP-It Express (Active Motif, Rixensart, Belgium, http://www.activemotif.com), in line with the manufacturer’s instructions. Cells were grown to 80% confluence, and proteins were cross-linked to DNA by addition of 1% formaldehyde to the cells for 10 minutes at room temperature. Cells were collected by centrifugation for 10 minutes at 2,500 rpm. Pellets were resuspended in ice-cold lysis buffer, incubated on ice for 30 minutes, and transferred to an ice-cold Dounce homogenizer. Nuclei were centrifuged at 5,000 rpm for 10 minutes at 4°C and incubated with 50 ␮l of working stock of enzymatic shearing cocktail (200 U/ml) for 10 minutes at 37°C. After centrifuging at 10,000 rpm at 4°C for 10 minutes, 10% of supernatant was used to quantify the amount of DNA present in the starting material (input). ChIP reactions were performed by adding 25 ␮l of protein G magnetic beads, 10 ␮l of ChIP buffer 1, 15 ␮g of sheared chromatin, 1 ␮l of protease inhibitor cocktail, and 10 ␮g of antibody. In each case, samples were incubated with 10 ␮g of anti-acetyl-histone H3 (Upstate, Charlottesville, VA, http://www. upstate.com), anti-H3K9Me3 (Abcam, Cambridge, U.K., http://www. abcam.com), or anti-SP1 (Upstate) antibodies. An isotype-matched antibody was used as control for nonspecific binding. Samples were incubated in a rolling shaker overnight at 4°C. Magnetic beads were washed and incubated with elution buffer. The protein DNA crosslinking was reversed by incubating tubes at 65°C for 2.5 hours and subsequently treated with proteinase K. Target DNA was quantified by real-time PCR (SYBR Green, Applied Biosystems) at least three times

Hypoxia Causes Genomic Instability in Stem Cells in two independent experiments. The fraction of immunoprecipitated DNA was obtained as percentage input ⫽ 2(Ctinput ⫺ CtChIP) ⫻ 100%. Relative occupancy was calculated as a ratio of specific signal over background. Results from hypoxia were referred to normoxia. Analysis in silico of mMLH1 and mMSH6 promoter regions used the Genomatix (Ann Arbor, MI, http://www.genomatix.de) bioinformatics software portal. We searched for putative binding sites for SP1 in the adjacent 5⬘-upstream regions and designed primers (supplemental online Table 1) that amplified the promoter regions shown in Figure 3.

Trichostatin A and 5ⴕ-AZA-2ⴕ-Deoxycytidine Treatment C17.2 cells were split 24 hours before different treatments with 300 nM trichostatin A (TSA) (Sigma-Aldrich), 5⬘-AZA-2⬘-deoxycytidine (5AZA-dC) (1, 5, 15 ␮M) (Sigma-Aldrich), a combination of 300 nM TSA and 15 ␮M 5AZA-dC, or phosphate-buffered saline as control. C17.2-treated cells were incubated for 24 hours under hypoxic conditions. Neurospheres, BMMSC, and DPSC were exposed to hypoxia and the corresponding treatment for 48 hours. The concentrations and timing of 5AZA-dC and TSA treatment were based on similar published studies [37, 38].

In Vitro DNA Repair Activity DNA repair activity was evaluated by the G/T mismatch binding protein DNA repair kit (Active Motif), according to the manufacturer’s instructions. C17.2 cells and neurospheres were grown in hypoxia for 24 and 48 hours, respectively. Briefly, cell extracts (2.5 ␮g) were diluted in lysis buffer and added to a coated plate with immobilized DNA molecule containing a G/T mismatch that is recognized by MSH6. The plate was incubated for 1 hour at room temperature with mild agitation. After washing steps, MSH6 antibody was added and incubated for 1 hour at room temperature. After incubation with horseradish peroxidase (HRP)-conjugated antibody for 1 hour at room temperature, the wells were washed three times, and the colorimetric reaction was obtained after addition of developing solution. Absorbance was measured on a spectrophotometer at 450 nm.

Host Cell Replication Error Assay The pZCA29 vector [39] was kindly provided by Dr. T.M. Ruenger (Department of Dermatology, Boston University School of Medicine). The plasmid carries the LacZ gene, interrupted by 29 CA repeats. These inserts inactivate LacZ by a ⫹1 frameshift. Mutations in the plasmid lead to a normal reading frame or to a ⫺1 frameshift that causes blue coloration. Four million C17.2 cells were transfected with 2 ␮g of pZCA29 by lipofection using the Lipofectamine 2000 reagent (Invitrogen). Starting 24 hours after transfection, the cells were incubated under normoxia or hypoxia for 2 days. Replicated DNA was recovered from the transfected cells on day 4. Unreplicated input plasmid DNA was removed by digestion with DpnI. Recovered pZCA29 plasmid was introduced into Escherichia coli DH10B (Invitrogen) by electrotransformation and then selected on Luria-Bertani agar plates containing 100 ␮g/ml ampicillin for selection, as well as isopropyl-␤-D-thiogalactopyranoside and 5-bromo-4-chloro-3-indolyl ␤-galactosidase. The total numbers of white and blue colonies were counted, and the mutation frequency was calculated as the mean of blue colonies divided by the mean of the total number of colonies.

Microsatellite Instability Analysis Genomic DNA was extracted from mouse C17.2 cells, neurospheres, and human mesenchymal cells (BMMSC and DPSC), after different incubation times under hypoxic conditions (1% O2). DNA was used as template in a multiplex PCR. The amplified targets from C17.2 and neurospheres are nucleotide repeat DNA markers that include the microsatellite loci mBAT26, mBAT37, mBAT59, mBAT67, and D15Mit93. For human samples, the amplified markers were hBAT25, hBAT26, hD5S346, hD17S250, hD2S123, hD11S904, hD9S171, hTBP, hRB, and htp53Alu. PCR products were amplified using conditions and specific fluorescently labeled primers described previously [40, 41]. Fluorescently labeled PCR products were analyzed by ABI 3730xl DNA Analyzer (Applied

Rodrı´guez-Jime´nez, Moreno-Manzano, Lucas-Dominguez et al.

2055

Figure 1. Expression profile of mismatch repair (MMR) in murine and human stem cells under hypoxic conditions. Graphics show fewer MMR transcripts in C17.2 cells (A), neurospheres (B), BMMSC (C), and DPSC (D) under Hx than under Nx (basal level), obtained by real-time quantitative polymerase chain reaction analysis. EGLN3 is included as an Hx-induced target-control gene. Results were standardized by the housekeeping gene GAPDH. mRNA levels were calculated by the 2⫺⌬⌬CT method. Results were obtained from three independent experiments. Panels illustrate protein expression levels of MMR (MLH1, MSH6, MSH2), HIF1␣, and its target EGLN3, monitored by Western blotting. ␤-Actin was included as loading control. (E): MSH6 expression in C17.2 cells was treated with different concentrations of the Hx-mimetic agents DFO and cobalt chloride (CoCl2) for 3 hours. Abbreviations: BMMSC, bone marrow mesenchymal stem cells; DFO, desferoxamine; DPSC, dental pulp stem cells; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; h, hours; Hx, hypoxia; Nx, normoxia.

Biosystems), in line with the manufacturer’s instructions. MSI was analyzed by comparison of hypoxia-treated samples with nontreated samples by GeneMapper 3.7 software (Applied Biosystems).

Statistical Analysis Statistical comparisons were assessed by Student’s t test. All p values were derived from a two-tailed statistical test, using the SPSS 11.5 software (SPSS, Chicago, http://www.spss.com). A p value ⬍.05 was considered statistically significant.

RESULTS Downregulation of the MMR System in Stem Cells Under Hypoxia We investigated the possibility that hypoxia could compromise cellular DNA repair processes by downregulation of MMR expression in a mouse multipotent neural precursor cell line (C17.2), in embryonic neurospheres, and in human mesenchymal stem cells (BMMSC and DPSC). To evaluate the influence of hypoxia on the MMR expression level, mouse and human stem cells were exposed to 1% O2 at different times under hypoxic conditions or incubated in 20% O2 under normoxia. Viability of cells cultured in hypoxia was examined during the time course, and no toxicity was observed (data not shown). We evaluated, by real-time PCR using TaqMan probes (supplemental online Table 1), the expression level of the HIF1␣-target gene EGLN3 as positive control of hypoxic conditions [42]. As expected, upreguwww.StemCells.com

lated expression of EGLN3 mRNA was detected after 3 hours of hypoxic treatment, whereas HIF1␣ transcripts were constitutive, as widely reported in the literature (data not shown). Results were standardized to the housekeeping gene GAPDH. For the evaluated stem cell models, transcriptional downregulation for the MMR system occurred at 3–48 hours of hypoxic exposure (Fig. 1A–1D). However, high MMR protein expression was still present at 3 hours of hypoxia, suggesting a delay in the effects of such transcriptional silencing. Examination of protein levels in C17.2 cells showed that EGLN3, MSH6, and MLH1 increased at 3 hours, coinciding with HIF1␣ maximum expression (Fig. 1A). Protein expression of MSH2, MSH6, and MLH1 decreased at 24 and 48 hours of hypoxia exposure, when reduction of HIF1␣ became clearer. MMR proteins were also studied in murine neurospheres and human stem cell models to further develop the results obtained in the C17.2 cell line. Regardless of the stem cell model, the higher protein expression level of MLH1 and MSH6 is consistent with HIF1␣ and decreased accordingly. Chemically induced hypoxia is widely used. C17.2 cells treated with different concentrations of the hypoxia-mimetic agents desferoxamine and cobalt chloride (CoCl2) for 3 hours increased the expression of HIF1␣ and MSH6 protein in a dose-dependent manner (Fig. 1E).

Involvement of HIF1␣ in MMR Regulation Under Hypoxia The results described above led us to investigate whether HIF1␣ was responsible for such an effect. We first knocked down

2056

Hypoxia Causes Genomic Instability in Stem Cells

Figure 2. Effect of shRNAi against mHIF1␣ on MMR expression in C17.2 cells at short Hx. C17.2 cells were transduced with Scr sequence-shRNAi (Scr-shRNAi) or constructs containing specific shRNAi sequences for mHIF1␣. (A): Knockdown expression of HIF1␣ (left) and EGLN3 (right) after shRNAi treatment was examined by TaqMan analysis after 3 and 24 h of Hx. (B): MLH1 and MSH6 mRNA expression in C17.2 cells was modified by Scr-shRNAi or HIF1␣-shRNAi and was grown for 3 or 24 h in Hx. (C): Protein levels of HIF1␣, EGLN3, MLH1, MSH6, and ␤-actin (as loading control) in C17.2-Scr-shRNAi and C17.2-HIF1␣-shRNAi cells were compared with basal conditions after 3 and 24 h of Hx (Nx Scr-shRNAi). Statistical comparisons were performed with Student’s t test; ⴱ, p ⬍ .05; ⴱⴱ, p ⬍ .01. Abbreviations: h, hours; Hx, hypoxia; MMR, mismatch repair; Nx, normoxia; Scr, Scramble; shRNAi, short hairpin loop RNA interference.

mHIF1␣, using shRNAi. The ablation in C17.2 cells stably transduced with retrovirally mediated shRNAi for mHIF1␣ was documented by TaqMan analysis and Western blotting. Efficiency of shRNAi-induced silencing was examined after 3 and 24 hours of hypoxia by quantification of mRNA expression of HIF1␣ and its target gene EGLN3 (Fig. 2A). After 3 hours of hypoxia exposure and HIF1␣-shRNAi modification, EGLN3, MLH1, and MSH6 mRNA expression was significantly lower than in C17.2-Scramble-shRNAi cells (EGLN3, p ⬍ .01, Fig. 2A; MLH1, p ⬍ .05 and MSH6, p ⬍ .01, Fig. 2B). Evaluation of their protein expression also showed lower levels in C17.2-HIF1␣-shRNAi cells after 3 hours of hypoxia (Fig. 2C). These results suggest that rapid induction of HIF1␣ under hypoxia conditions might positively regulate the expression of the MMR system in C17.2 neural stem cells. However, recovery of MMR expression was not observed in ablated C17.2-HIF1␣-shRNAi cells after prolonged exposures to hypoxia. Therefore, unlike the data published for cancer cells [43], the results obtained with immortalized C17.2 neural stem cells show that HIF1␣ may not be involved in MMR repression. This suggests that other mechanisms may repress its expression.

Specific Epigenetic Events Regulate MMR Expression in Neural Stem Cells During Hypoxia Hypermethylation of the CpG islands is a major cause of decreased expression of MLH1 in tumors [20], although gene silencing of MLH1 was previously demonstrated to be closely linked to reduced acetylation of histones in cancer cells [44]. Therefore, we investigated the occurrence of aberrant hypermethylation of MSH6

and MLH1 CpG islands in C17.2, grown under normoxia or hypoxia, by bisulfite sequencing PCR assay, as described in the supplemental online data. A set of primers, described in supplemental online Table 1, was designed to cover 547 (MSH6) or 586 (MLH1) base pairs of their CpG-rich region. At least six clones of each condition were sequenced, and we observed no significant differences of methylation in their CpG islands between normoxia and hypoxia (supplemental online Fig. 1). We examined by ChIP the acetylation status of histone (H3) in the promoter region of MLH1 and MSH6 in C17.2 cells grown under normoxia or hypoxia. Immunoprecipitated DNA was evaluated by SYBR Green using different primers (supplemental online Table 1). Values of the amplified DNA from immunoprecipitated chromatin were standardized to the corresponding total input DNA. A negative control antibody was included to detect nonspecific binding and amplification. Hypoxic changes were quantified and referred to normoxia (Fig. 3, baseline). We found hypoacetylated histone H3 after 24 hours of hypoxic exposure (Fig. 3A) and observed that histone H3 acetylation was enhanced by treatment with the deacetylase inhibitor TSA (300 nM). Moreover, levels of the repressive chromatin modification mark of H3K9Me3 increased at the mentioned loci (Fig. 3B).

Prevalence of Histone Deacetylase Inhibitor (TSA) over Methyltransferase Inhibitor (5AZA-dC) in the Regulation of MMR Genes Under Hypoxia in Primary Murine and Human Stem Cells To validate the results obtained in the C17.2 cell line, which indicate the role of epigenetic suppression of MMR genes in

Rodrı´guez-Jime´nez, Moreno-Manzano, Lucas-Dominguez et al.

2057

Figure 3. ChIP analysis of histone H3 modifications and SP1 binding for MLH1 and MSH6 promoter regions in hypoxic C17.2 cells. Specific relative immunoprecipitation changes were calculated by real-time quantitative polymerase chain reaction (SYBR Green) as a ratio of specific signal over background in Hx and compared with Nx. Representative results were obtained in three independent experiments. Primers used to examine MLH1 and MSH6 promoter regions are indicated in upper panels (named from A to M). (A): Changes of AcH3 were evaluated for each promoter in C17.2 cells after 24 hours of Hx or treatment with TSA (0.3 ␮M) immediately prior to hypoxic exposure. (B): Enrichment of K9H3Me at 24 hours of Hx. (C): SP1 binding sites close to transcription start points are depicted at the top. Diminished relative occupancy of SP1 in MLH1 and MSH6 promoter regions of C17.2 hypoxic cells and recovery by treatment with TSA (0.3 ␮M) are shown. ⴱ, p ⬍ .05; ⴱⴱ, p ⬍ .01; ⴱⴱⴱ, p ⬍ .001; determined by Student’s t test for each pair of primers. Abbreviations: AcH3, acetylated histone H3; ChIP, chromatin immunoprecipitation; Hx, hypoxia; K9H3Me, histone H3–K9 trimethylated; Nx, normoxia; TSA, trichostatin A.

neural stem cells, we examined this result in primary cultured murine and human stem cells. For this purpose, embryo neurospheres from mice and human mesenchymal stem cells—from bone marrow and dental pulp—were used. Neurospheres, BMMSC, and DPSC were treated with methylation inhibitor (5AZA-dC) or TSA and immediately exposed to hypoxia. For C17.2 cells, TSA significantly prevented downregulation of MutS (MSH2, MSH3, MSH6, p ⬍ .01) and MutL (PMS2, p ⬍ .01; MLH1, p ⬍ .05) genes under hypoxic conditions (Fig. 4A). Treatment with 5AZA-dC produced minimal or no effect on MMR mRNA levels. Similar results were obtained after protein level evaluation (Fig. 4A, lower panels). The analysis of CpG island methylation and H3 acetylation at the mentioned loci was consistent with 5AZA-dC and TSA treatment. Combination of 5AZA-dC and TSA did not show additional synergetic inhibition of the MMR system (data not shown). Examination of MLH1 protein expression after 5AZA-dC and TSA treatments showed comparable results in neurospheres, DPSC, and BMMSC (Fig. 4B, 4C, 4D, respectively), although a slight increase in MLH1 might be detected in DPSC after 5AZA-dC addition. www.StemCells.com

SP1 Binding Within MLH1 and MSH6 Promoters Depends on Histone H3 Acetylation/Methylation Status We analyzed the occupancy of SP1 at binding sites predicted in silico within the MSH6 and MLH1 promoters. The specific ChIP signals obtained after immunoprecipitation with anti-SP1 antibody showed reduced SP1 enrichment at 24 hours of hypoxia (Fig. 3C) in C17.2 cells, a result that correlated well with the lower histone H3 acetylation obtained. To clarify whether or not SP1 binding was affected by H3 acetylation status, we treated C17.2 cells with the histone deacetylase inhibitor TSA (300 nM) and detected significantly higher SP1 interactions (Fig. 3C) within MSH6 and MLH1 promoter regions in the hypoxic cells exposed to it.

DNA Repair Activity We evaluated whether MMR downregulation was associated with a deficiency of DNA repair activity by measuring MSH6 activation in C17.2 cells and neurospheres grown under normoxia and hypoxia conditions. Nuclear protein extracts from

2058

Hypoxia Causes Genomic Instability in Stem Cells

Figure 4. Effects of TSA or 5AZA-dC treatment on MMR expression in stem cells under Hx. (A): Upper panels show the results after treatment with TSA (300 nM) (left panels) or methyltransferase inhibitor 5AZA-dC (1, 5, and 15 ␮M) (right panels) on mRNA levels of MutS (MSH2, MSH3, and MSH6) and MutL (MLH1, PMS2) members. The results obtained by Western blotting after the mentioned treatments, on protein levels of MLH1 and MSH6, are also shown. ⴱ, p ⬍ .05; ⴱⴱ, p ⬍ .01; determined by Student’s t test. (B–D): MLH1 protein expression in neurospheres (B), in BMMSC (C), and in DPSC (D) after Hx for 48 hours, Hx ⫹ 5AZA-dC (15 ␮M) for 48 hours, or Hx ⫹ TSA (300 nM) for 48 hours. Protein quantification was done by densitometry analyses. Abbreviations: 5AZA-dC, 5⬘-AZA-2⬘-deoxycytidine; BMMSC, bone marrow mesenchymal stem cells; DPSC, dental pulp stem cells; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; h, hours; Hx, hypoxia; Nx, normoxia; TSA, trichostatin A.

normoxic and hypoxic samples were obtained, and 2.5 ␮g was added to a plate on which a linear oligonucleotide containing a G/T mismatch was immobilized. MSH6 present in nuclear extract binds specifically to this oligonucleotide. The primary antibody anti-MSH6 recognized an epitope on MSH6 protein that is accessible after DNA binding. Addition of a secondary HRP-conjugated antibody provided a colorimetric readout that was quantified at 450 nm by spectrophotometry. The results obtained showed good correlation between decreased MSH6 protein expression and lower repair activity in hypoxic samples of both the immortalized C17.2 cell line and murine primary neural stem cells (neurospheres) (Fig. 5A).

Analysis of Mutation Frequency Under Hypoxia DNA damage, measured as unrepaired mutation rate, was analyzed with plasmid shuttle vector pZCA29 containing out-offrame poly(CA) repeats interrupting a ␤-galactosidase reporter gene. Insertions or deletions in the repeat tract during replication of the vector in the recipient human cells could, therefore, result in correction of the reading frame of ␤-galactosidase. After 24 hours of transfection, C17.2 cells were grown for 2 days under hypoxia exposure. Plasmids were recovered at day 4 from the human cells and transduced into permissive bacteria. Figure 5B

shows the mutation frequency as an increase in ␤-galactosidasepositive clones in hypoxic samples.

Hypoxia Causes Genomic Instability in Stem Cells Since MMR inactivation [13, 15] or deregulation [43] induces genomic instability, we examined the status of microsatellite sequences of genomic DNA from murine and human stem cell models, grown under hypoxic versus normoxic conditions. The murine-specific markers [40] selected were mBAT26, mBAT37, mBAT59, mBAT67, and D15Mit93. The amplified markers for human samples were hBAT25, hBAT26, hD5S346, hD17S250, hD2S123, hD11S904, hD9S171, hTBP, hRB, and htp53Alu [41]. We detected additional fragments in C17.2 cells and in primary murine and human stem cells, which indicated unrepaired replication errors in stem cell samples exposed to hypoxia (Fig. 6). Thus, stem cells in a hypoxic environment suffer genomic instability, which is closely linked in the literature to the acquisition of a malignant cell phenotype and cancer progression [15, 45].

DISCUSSION Hypoxia causes a wide variety of physiological changes, most of which are governed by HIF1␣, a key regulator that

Rodrı´guez-Jime´nez, Moreno-Manzano, Lucas-Dominguez et al.

Figure 5. Evaluation of the DNA mutation frequency and DNA repair activity of neural stem cells under Hx. (A): Evaluation of DNA repair activity in C17.2 cells and neurospheres under Hx. MSH6 (GTBP) binds to an immobilized linear oligonucleotide containing a G/T mismatch. MSH6 was immunodetected and the binding was quantified by detection of a colorimetric reaction at 450 nm. DNA repair activity of MSH6 was monitored in duplicate for samples from Nx, Hx (24 hours for C17.2 or 48 hours for neurospheres), or Hx ⫹ TSA (300 nM). Units of OD are shown for each condition. The mean and SD for two independent experiments are given. (B): Host cell replication error assay was performed with the shuttle vector pZCA29. The mutation frequency in the plasmid pZCA29 after passage through the host cells is indicated as the frequency of a reversion to a LacZ gene with correct reading frame caused by unrepaired mutations. Plasmid DNA was isolated from the cells after 2 days of Hx and 3 days of transfection. The graph shows the frequency of blue bacterial colonies obtained after passage of pZCA29 in C17.2 cells and the subsequent bacterial transformation that shows the host cells’ microsatellite instability. The spontaneous mutation frequency of pZCA29 in E. coli DH10B is shown by the plotted line. The mean and the SD of three independent experiments are given. ⴱ, p ⬍ .05; ⴱⴱ, p ⬍ .01; ⴱⴱⴱ, p ⬍ .001; determined by Student’s t test. Abbreviations: Hx, hypoxia; Nx, normoxia; OD, optical density; TSA, trichostatin A.

triggers the transcription of genes involved in important processes, such as angiogenesis, cell survival, invasion, and glucose metabolism [6]. In cancer cells, HIF1␣ but not HIF2␣ is responsible for genomic instability, since it induces downregulation of MutS␣ (MSH2 and MSH6), but MLH1 remains unchanged [43]. Because of the presence of stem cells in hypoxic niches, the increasing interest in cancer stem cells as a possible cause of this malignancy and lack of knowledge about the MMR system in stem cells, we examined the expression profiles of MutS and MutL members along a representative time course of hypoxia exposure in different stem cell types. We extended this evaluation from murine to human stem cell models, and we confirmed general MMR transcriptional repression in hypoxia. However, www.StemCells.com

2059

HIF1␣, in parallel with MSH6 and MLH1 proteins, peaked after a short period of hypoxia (3 hours), indicating a possible delay in transcriptional decline. We investigated whether HIF1␣ was responsible for this upregulation, and our data showed a reliable decrease in the protein expression level of MLH1 and MSH6 in C17.2-HIF1␣-shRNAi-silenced cells after 3 hours of hypoxia. Therefore, HIF1␣ might be involved in the positive short-term regulation of hypoxia exposure. Nevertheless, HIF1␣ regulates not only MutS, as described previously [43], but also, as we show here for the first time, MutL members of MMR machinery. At 24 hours of hypoxia, C17.2-HIF1␣-shRNAi cells did not recover the expression of MLH1 and MSH6, indicating that HIF1␣ does not repress their expression. We also observed steady downregulation of the MMR system, which became clearer after 24 hours of hypoxia (Fig. 1), when less HIF1␣ protein was detected. All these findings suggest that other mechanisms apart from HIF1␣ might be involved in the transcriptional repression of MMR genes. Moreover, increased HIF1␣ does not always cause MMR deficiency [46]. Very recent publications demonstrate additional regulation mechanisms, such as transcriptional repression by DEC1 and DEC2 [47]. According to our results obtained with neural murine and human mesenchymal stem cells, MMR genes suffer downregulation by specific and dynamic epigenetic modifications, induced by oxygen depletion. Silencing of MMR genes is associated mainly with aberrant DNA methylation [20, 24, 48, 49]. Here we demonstrate that this aberrant promoter hypermethylation is not responsible for reduced expression of MLH1 and MSH6 in C17.2 neural stem cells, as does occur in cancer cells. It has been reported that genes that undergo an aberrant promoter hypermethylation of their CpG islands in cancer stem cells remain unmethylated in stem cells [50]. Some authors suggest a more dominant role for DNA methylation over histone deacetylase activity for maintenance of gene repression [37, 38, 51–53]. Nonetheless, MLH1 mRNA from fibroblast and cancer cells was downregulated under hypoxia, but TSA prevented its reduction by favoring histone acetylation [44, 54]. We confirmed a hypoacetylated/hypermethylated histone H3 and a prevalence of histone deacetylase inhibitor (TSA) over methyltransferase inhibitor (5AZA-dC) effects in the regulation of the MMR genes of C17.2 neural stem cells under hypoxia. TSA treatment also increased the expression level of MLH1 in primary cultured mouse neurospheres and human BMMSC and DPSC. Consistent with our findings, Cameron et al. indicate that histone deacetylation has a role when levels of DNA methylation are reduced [38]. In RASSF1A promoter, deacetylation and lysine 9 trimethylation of histone H3 impair binding of SP1 [22]. In other studies, HIF1␣ is recruited by SP1, which constitutively binds to the MSH2 and MSH6 gene promoters regardless of O2 concentration, but HIF1␣ interaction is the key element in their regulation [43]. In contrast, we noticed the diminished occupancy of SP1 at MSH6 and MLH1 promoter binding sites in C17.2 cells after prolonged hypoxia, when HIF1␣ was hardly detected. The increase in SP1 enrichment by TSA indicates that deacetylation of H3 impairs its binding. A hypoxic microenvironment creates a hypermutagenic state in cancer cells [55] and, in agreement with our findings, also leads to lower DNA repair activity in neural stem cells. We detected MSI, possibly due to MMR deficiency, in several hypoxic stem cells by inspection of various mouse and human markers. In vitro hypoxic preconditioning of embryonic stem cells has been put forward as a strategy for promoting cell survival and functional benefits after transplantation into the ischemic rat brain [56] or for increasing efficacy of human endothelial progenitor cells for therapeutic neovascularization

Hypoxia Causes Genomic Instability in Stem Cells

2060

Figure 6. Genomic instability in stem cells cultured in Hx. Genomic DNA was extracted from mouse C17.2 and neurospheres or from human BMMSC and DPSC at different incubation times under Nx and Hx conditions. Using specific fluorescently labeled primers, microsatellite instability (MSI) was analyzed for the nucleotide repeat DNA markers that include the mouse microsatellite loci mBAT26, mBAT37, mBAT59, mBAT67, and mD15Mit93. For human samples, the amplified markers were hBAT25, hBAT26, hD5S346, hD17S250, hD2S123, hD11S904, hD9S171, hTBP, hRB, and htp53Alu. Fluorescent polymerase chain reaction products were analyzed by an ABI 3730xl DNA automated sequencer (Applied Biosystems). Mononucleotide repeat markers amplified from various stem cells grown in Nx were compared with the corresponding matching hypoxic samples. Positive replication errors obtained by MSI analysis are shown in the graphs. Additional fragments (marked by continuous lines and arrows) or absent fragments (indicated by dotted lines and arrows) detected in hypoxic samples are distinguishable from the size of predominant alleles of control samples (Nx). Results of markers from mouse and human stem cells examined at different times are shown in the tables. A minus sign (⫺) denotes that the sample is MSI-negative; a plus sign (⫹), MSI-positive; a plus-or-minus sign (⫾), possibly MSI-positive. Abbreviations: BMMSC, bone marrow mesenchymal stem cells; d, days; DPSC, dental pulp stem cells; Hx, hypoxia; nd, not detected; Nx, normoxia.

[57]. Monitoring the status of DNA repair machinery in hypoxic conditions may contribute to better present and future use of cell therapies as promising nanomedicines.

CONCLUSION Hypoxia causes repression of the DNA repair system (MutS and MutL) in stem cells through epigenetic chromatin inactivation and reduced SP1 binding. Hypoxic niches are the physiological environment of many stem cell types and may be the origin of aberrant processes that cause genomic instability in incipient small subpopulations that conserve their self-renewal properties. Environmental conditions may also affect the vectors suggested in cell transplantation. The results given here alert researchers to the need to monitor vectors’ genetic integrity during the development and application of current regenerative therapies.

ACKNOWLEDGMENTS We are grateful to Miodrag Stojkovic for fruitful scientific discussions and critical reading of the manuscript. We gratefully acknowledge Augusto Silva for help with the culture of primary neural stem cells. We also thank Pablo Mateos for excellent technical support, Vanessa Bou for contribution to the ChIP experiments, and Marı´a Teresa Calvo for assistance with the primary stem cell culture. This research was funded by the Fondo de Investigaciones Sanitarias, Instituto de Salud Carlos III (Spain), projects PI051973 and CP05/00182.

DISCLOSURE

POTENTIAL CONFLICTS OF INTEREST

OF

The authors indicate no potential conflicts of interest.

Rodrı´guez-Jime´nez, Moreno-Manzano, Lucas-Dominguez et al.

2061

REFERENCES 30 1 2 3 4 5 6 7

8 9 10 11 12 13

14 15 16

17 18 19 20 21 22 23

24 25 26 27 28 29

Ignatova TN, Kukekov VG, Laywell ED et al. Human cortical glial tumors contain neural stem-like cells expressing astroglial and neuronal markers in vitro. Glia 2002;39:193–206. Singh SK, Hawkins C, Clarke ID et al. Identification of human brain tumour initiating cells. Nature 2004;432:396 – 401. Hemmati HD, Nakano I, Lazareff JA et al. Cancerous stem cells can arise from pediatric brain tumors. Proc Natl Acad Sci U S A 2003; 100:15178 –15183. Vescovi AL, Galli R, Reynolds BA. Brain tumour stem cells. Nat Rev Cancer 2006;6:425– 436. Zhu LL, Wu LY, Yew DT et al. Effects of hypoxia on the proliferation and differentiation of NSCs. Mol Neurobiol 2005;31:231–242. Semenza GL. Targeting HIF-1 for cancer therapy. Nat Rev Cancer 2003;3:721–732. Maxwell PH, Dachs GU, Gleadle JM et al. Hypoxia-inducible factor-1 modulates gene expression in solid tumors and influences both angiogenesis and tumor growth. Proc Natl Acad Sci U S A 1997;94: 8104 – 8109. Birner P, Schindl M, Obermair A et al. Expression of hypoxia-inducible factor 1alpha in epithelial ovarian tumors: Its impact on prognosis and on response to chemotherapy. Clin Cancer Res 2001;7:1661–1668. Birner P, Gatterbauer B, Oberhuber G et al. Expression of hypoxiainducible factor-1 alpha in oligodendrogliomas: Its impact on prognosis and on neoangiogenesis. Cancer 2001;92:165–171. Yuan J, Glazer PM. Mutagenesis induced by the tumor microenvironment. Mutat Res 1998;400:439 – 446. Yuan J, Narayanan L, Rockwell S et al. Diminished DNA repair and elevated mutagenesis in mammalian cells exposed to hypoxia and low pH. Cancer Res 2000;60:4372– 4376. Kondo A, Safaei R, Mishima M et al. Hypoxia-induced enrichment and mutagenesis of cells that have lost DNA mismatch repair. Cancer Res 2001;61:7603–7607. Scherer SJ, Avdievich E, Edelmann W. Functional consequences of DNA mismatch repair missense mutations in murine models and their impact on cancer predisposition. Biochem Soc Trans 2005;33: 689 – 693. Fishel R, Lescoe MK, Rao MR et al. The human mutator gene homolog MSH2 and its association with hereditary nonpolyposis colon cancer. Cell 1993;75:1027–1038. Aaltonen LA, Peltomaki P, Leach FS et al. Clues to the pathogenesis of familial colorectal cancer. Science 1993;260:812– 816. Chiaravalli AM, Furlan D, Facco C et al. Immunohistochemical pattern of hMSH2/hMLH1 in familial and sporadic colorectal, gastric, endometrial and ovarian carcinomas with instability in microsatellite sequences. Virchows Arch 2001;438:39 – 48. Shibata D, Mori Y, Cai K et al. RAB32 hypermethylation and microsatellite instability in gastric and endometrial adenocarcinomas. Int J Cancer 2006;119:801– 806. Eckert A, Kloor M, Giersch A et al. Microsatellite instability in pediatric and adult high-grade gliomas. Brain Pathol 2007;17:146 –150. Fleisher AS, Esteller M, Tamura G et al. Hypermethylation of the hMLH1 gene promoter is associated with microsatellite instability in early human gastric neoplasia. Oncogene 2001;20:329 –335. Esteller M, Levine R, Baylin SB et al. MLH1 promoter hypermethylation is associated with the microsatellite instability phenotype in sporadic endometrial carcinomas. Oncogene 1998;17:2413–2417. Baylin SB, Herman JG. DNA hypermethylation in tumorigenesis: Epigenetics joins genetics. Trends Genet 2000;16:168 –174. Shukla V, Vaissiere T, Herceg Z. Histone acetylation and chromatin signature in stem cell identity and cancer. Mutat Res 2008;637: 1–15. Strunnikova M, Schagdarsurengin U, Kehlen A et al. Chromatin inactivation precedes de novo DNA methylation during the progressive epigenetic silencing of the RASSF1A promoter. Mol Cell Biol 2005;25: 3923–3933. Clark SJ, Harrison J, Molloy PL. Sp1 binding is inhibited by (m)Cp(m) CpG methylation. Gene 1997;195:67–71. Macleod D, Charlton J, Mullins J et al. Sp1 sites in the mouse aprt gene promoter are required to prevent methylation of the CpG island. Genes Dev 1994;8:2282–2292. Brandeis M, Frank D, Keshet I et al. Sp1 elements protect a CpG island from de novo methylation. Nature 1994;371:435– 438. Snyder EY, Deitcher DL, Walsh C et al. Multipotent neural cell lines can engraft and participate in development of mouse cerebellum. Cell 1992; 68:33–51. Parker MA, Anderson JK, Corliss DA et al. Expression profile of an operationally-defined neural stem cell clone. Exp Neurol 2005;194: 320 –332. Liu WG, Lu GQ, Li B et al. Dopaminergic neuroprotection by neurturin-

www.StemCells.com

31

32

33

34

35

36

37 38 39 40 41 42

43 44 45 46 47

48 49 50 51 52

53

expressing c17.2 neural stem cells in a rat model of Parkinson’s disease. Parkinsonism Relat Disord 2007;13:77– 88. Snyder EY, Yoon C, Flax JD et al. Multipotent neural precursors can differentiate toward replacement of neurons undergoing targeted apoptotic degeneration in adult mouse neocortex. Proc Natl Acad Sci U S A 1997;94:11663–11668. Ourednik J, Ourednik V, Lynch WP et al. Neural stem cells display an inherent mechanism for rescuing dysfunctional neurons. Nat Biotechnol 2002;20:1103–1110. Teng YD, Lavik EB, Qu X et al. Functional recovery following traumatic spinal cord injury mediated by a unique polymer scaffold seeded with neural stem cells. Proc Natl Acad Sci U S A 2002;99: 3024 –3029. Aboody KS, Brown A, Rainov NG et al. Neural stem cells display extensive tropism for pathology in adult brain: Evidence from intracranial gliomas. Proc Natl Acad Sci U S A 2000;97:12846 –12851. De Filippis L, Lamorte G, Snyder EY et al. A novel, immortal, and multipotent human neural stem cell line generating functional neurons and oligodendrocytes. STEM CELLS 2007;25:2312–2321. Reynolds BA, Weiss S. Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system. Science 1992;255:1707–1710. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25:402– 408. Kondo Y, Shen L, Issa JP. Critical role of histone methylation in tumor suppressor gene silencing in colorectal cancer. Mol Cell Biol 2003;23: 206 –215. Cameron EE, Bachman KE, Myohanen S et al. Synergy of demethylation and histone deacetylase inhibition in the re-expression of genes silenced in cancer. Nat Genet 1999;21:103–107. Diem C, Runger TM. A novel plasmid shuttle vector for the detection and analysis of microsatellite instability in cell lines. Mutat Res 1998; 407:117–124. Bacher JW, Abdel Megid WM, Kent-First MG et al. Use of mononucleotide repeat markers for detection of microsatellite instability in mouse tumors. Mol Carcinog 2005;44:285–292. Dietmaier W, Wallinger S, Bocker T et al. Diagnostic microsatellite instability: Definition and correlation with mismatch repair protein expression. Cancer Res 1997;57:4749 – 4756. del Peso L, Castellanos MC, Temes E et al. The von Hippel Lindau/ hypoxia-inducible factor (HIF) pathway regulates the transcription of the HIF-proline hydroxylase genes in response to low oxygen. J Biol Chem 2003;278:48690 – 48695. Koshiji M, To KK, Hammer S et al. HIF-1alpha induces genetic instability by transcriptionally downregulating MutSalpha expression. Mol Cell 2005;17:793– 803. Mihaylova VT, Bindra RS, Yuan J et al. Decreased expression of the DNA mismatch repair gene Mlh1 under hypoxic stress in mammalian cells. Mol Cell Biol 2003;23:3265–3273. Kinzler KW, Vogelstein B. Lessons from hereditary colorectal cancer. Cell 1996;87:159 –170. Lehtonen HJ, Makinen MJ, Kiuru M et al. Increased HIF1 alpha in SDH and FH deficient tumors does not cause microsatellite instability. Int J Cancer 2007;121:1386 –1389. Nakamura H, Tanimoto K, Hiyama K et al. Human mismatch repair gene, MLH1, is transcriptionally repressed by the hypoxia-inducible transcription factors, DEC1 and DEC2. Oncogene 2008 [Epub ahead of print]. Herman JG, Umar A, Polyak K et al. Incidence and functional consequences of hMLH1 promoter hypermethylation in colorectal carcinoma. Proc Natl Acad Sci U S A 1998;95:6870 – 6875. Bearzatto A, Szadkowski M, Macpherson P et al. Epigenetic regulation of the MGMT and hMSH6 DNA repair genes in cells resistant to methylating agents. Cancer Res 2000;60:3262–3270. Ohm JE, McGarvey KM, Yu X et al. A stem cell-like chromatin pattern may predispose tumor suppressor genes to DNA hypermethylation and heritable silencing. Nat Genet 2007;39:237–242. Kawamoto K, Okino ST, Place RF et al. Epigenetic modifications of RASSF1A gene through chromatin remodeling in prostate cancer. Clin Cancer Res 2007;13:2541–2548. Francia G, Green SK, Bocci G et al. Down-regulation of DNA mismatch repair proteins in human and murine tumor spheroids: Implications for multicellular resistance to alkylating agents. Mol Cancer Ther 2005;4: 1484 –1494. Fahrner JA, Eguchi S, Herman JG et al. Dependence of histone modifications and gene expression on DNA hypermethylation in cancer. Cancer Res 2002;62:7213–7218.

Hypoxia Causes Genomic Instability in Stem Cells

2062

54 Kondo Y, Shen L, Yan PS et al. Chromatin immunoprecipitation microarrays for identification of genes silenced by histone H3 lysine 9 methylation. Proc Natl Acad Sci U S A 2004;101:7398 –7403. 55 Reynolds TY, Rockwell S, Glazer PM. Genetic instability induced by the tumor microenvironment. Cancer Res 1996;56:5754 –5757. 56 Theus MH, Wei L, Cui L et al. In vitro hypoxic preconditioning of

embryonic stem cells as a strategy of promoting cell survival and functional benefits after transplantation into the ischemic rat brain. Exp Neurol 2008;210:656 – 670. 57 Akita T, Murohara T, Ikeda H et al. Hypoxic preconditioning augments efficacy of human endothelial progenitor cells for therapeutic neovascularization. Lab Invest 2003;83:65–73.

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