Selenium Supplementation Restores the Antioxidative Capacity and Prevents Cell Damage in Bone Marrow Stromal Cells In Vitro

Abstract

Bone marrow stromal cells (BMSCs) and other cell populations derived from mesenchymal precursors are developed for cell-based therapeutic strategies and undergo cellular stress during ex vivo procedures. Reactive oxygen species (ROS) of cellular and environmental origin are involved in redox signaling, cumulative cell damage, senescence, and tumor development. Selenium-dependent (glutathione peroxidases [GPxs] and thioredoxin reductases [TrxRs]) and selenium-independent (superoxide dismutases [SODs] and catalase [CAT]) enzyme systems regulate cellular ROS steady state levels. SODs process superoxide anion to hydrogen peroxide, which is subsequently neutralized by GPx and CAT; TrxR neutralizes other ROS, such as peroxinitrite. Primary BMSCs and telomerase-immortalized human mesenchymal stem cells (hMSC-TERT) express GPx1–3, TrxR1, TrxR2, SOD1, SOD2, and CAT. We show here that in standard cell cultures (5%–10% fetal calf serum, 5–10 nM selenite), the activity of antioxidative selenoenzymes is impaired in hMSC-TERT and BMSCs. Under these conditions, the superoxide anion processing enzyme SOD1 is not sufficiently stimulated by an ROS load. Resulting oxidative stress favors generation of micronuclei in BMSCs. Supplementation of selenite (100 nM) restores basal GPx and TrxR activity, rescues basal and ROS-stimulated SOD1 mRNA expression and activity, and reduces ROS accumulation in hMSC-TERT and micronuclei generation in BMSCs. In conclusion, BMSCs in routine cell culture have low antioxi-dative capacity and are subjected to oxidative stress, as indicated by the generation of micronuclei. Selenite supplementation of BMSC cultures appears to be an important countermeasure to restore their antioxidative capacity and to reduce cell damage in the context of tissue engineering and transplantation procedures.

Introduction

Bone marrow stromal cells (BMSCs) can give rise to differentiation of terminally differentiated mesenchymal cells, such as osteoblasts, chondrocytes, myotubes, and adipocytes. Mesenchymal stem cells (MSCs) can give rise to a transient amplifying cell pool, which in the context of many reviews and experiments is also termed MSCs, but probably should rather be called mesenchymal precursors or a similar term. They can be readily isolated from various sources and are developed for cell-based regenerative therapies, such as tissue engineering of bone and cartilage [1–5]. Precursor cells of any kind are as well used as target cells for gene therapy [6–8]. In vivo and especially ex vivo procedures required for both tissue engineering and gene therapy expose cells to considerable oxidative stress. Oxidative stress causes damage of the genome and proteome and promotes senescence, aging, and tumorigenesis [9–23]. The biological relevance of oxidative stress for the survival and self-renewal capacity of in this case hematopoetic stem cells has recently been stressed by the fact that in stem cells defective in the cell cycle checkpoint activator ataxia telangiectasia mutated [24], the self renewal capacity can be enhanced by antioxidative substances such as N-acetyl-cysteine treatment but not by telomerase overproduction.

Reactive oxygen species are produced within a cell by several enzyme systems involved in respiratory burst reactions (e.g., NADPH oxidases), by NO synthesis, and by several signaling cascades. The products generated are involved in procedures of hormone synthesis (e.g., thyroid hormones and steroid hormones [9]), vessel relaxation, redox regulation of multiple promoters and—if not sufficiently compartmentalized or in case of spillover—in damage of the genome and proteome of cells [10–11]. SOD synthesizes hydrogen peroxide from superoxide radicals. Isoenzymes contain copper, zinc (SOD1 and the extracellular isoenzyme SOD3), iron, manganese (SOD2), and nickel (bacteria) in their active centers [25]. Deficiency, as well as overproduction, of SOD1 is associated with diseases such as anemia, cystic fibrosis, and amyotrophic lateral sclerosis [26]; SOD2 knockout in Drosophila is lethal early after hatching or causes early adult onset mortality due to enhanced sensitivity against various stress mediators, oxygen, and toxins [27–32]. SOD1 and SOD2 cannot substitute for each other [33]. SOD1 mRNA is constitutively expressed and regulated by hydrogen peroxide and other substances. Identified promoter response elements comprise sites for Sp-1, C/EBP, Egr-1, WT-1, AP2, a xenobiotic response element, and an antioxidant response element [25, 34–40].

Hydrogen peroxide produced by SOD enzymes from superoxide anion and water is neutralized to H₂O by glutathione peroxidases (GPxs), thioredoxin reductases (TrxRs), and catalase. Four GPx enzymes and three TrxR enzymes are selenium-dependent, carrying selenocysteine in their active centers. Selenocysteine, the 21st amino acid, is encoded by the opal stop codon UGA, which in the presence of a 3′ hairpin loop structure is translated as selenocysteine and incorporated into proteins [9]. Selenium deficiency leads to translation of truncated proteins. There is a well-characterized protein- and organ-related hierarchy in selenium incorporation and retention; for example, TrxR ranks high compared with cellular GPx (cGPx; GPx1) and plasma GPx (pGPx; GPx3) at the protein level (due to the affinity of selenocysteine insertion sequence [SECIS] binding proteins to the individual SECIS element of a selenoprotein), and brain and testis rank high compared with other organs [9]. The expression of various selenoproteins is modulated by selenite both at the transcription and the translation level [41–44]. Recently, the human selenoprotein genome has been described to comprise 25 human selenoproteins as indicated by a whole genome screen searching for SECIS elements within the 3′-untranslated regions of open reading frames [42].

Formation of micronuclei is a biomarker of DNA damage. The micronucleus assay has emerged as one of the preferred methods to assess chromosome damage, and the method is now widely applied for population monitoring of genetic damage (e.g., for studying nutrigenomics and chromosomal instability, to assess the individual oxidative burden in kidney insufficiency and long-term hemodialysis, for screening of chemicals for genotoxic potential, and for the prediction of interindividual variations in radiosensitivity [12–16]). The connection between oxidative load and the formation of micronulei is being fostered by a recent report in which overexpression of the antioxidant thioredoxin (the main substrate of TrxR) in fibroblasts of patients suffering from Fanconi anemia (a disorder in which high concentrations of superoxides are responsible for DNA damage) protects cells from mitomycin C-induced micronuclei formation [17, 18]. Cumulative damage contributes to aging processes of the organism and to cellular fail-safe programs such as cellular senescence, as has been indirectly shown in transgenic animals, in which overexpression of single and combined components of antioxidative systems influenced longevity [19–23].

We have previously reported on the expression of various selenoproteins in bone cells and cells of the myelomonocytic pathway of differentiation [45–47], but little is known about the expression and role of antioxidant systems in stem cells or in cells of the transient proliferating pool derived from them, or of their effectivity to scavenge the genome and proteome of these cells during ex vivo procedures. We now used BMSC cultures and the recently established telomerase-immortalized human mesenchymal stem cell line (hMSC-TERT) [48] to demonstrate that their antioxidative capacity under standard (selenium-deficient) cell culture conditions is impaired and to show biological consequences in terms of intracellular production of reactive oxygen species and formation of micronuclei. Selenite supplementation of culture media was capable of restoring the anti-oxidative capacity of BMSCs and of reducing intracellular ROS production and stress-related generation of micronuclei.

Materials and Methods

Cell Culture

Media for cell culture and fetal calf serum (FCS) were obtained from PAA Laboratories (Linz, Austria, http://www.paa.at). hMSC-TERT cells were produced and cultured as described [48]. Cells were cultured in Earle's minimum essential medium containing 10% FCS. BMSCs were isolated from bone marrow from different donors and cultivated up to seven passages by a standardized protocol [5, 49]. Bone marrows were recovered in informed consent from the explanted femoral heads of patients undergoing elective hip arthroplasty. The procedure was approved by the local Ethics Committee of the University of Würzburg. Briefly, bone marrow preparations were washed with medium (Dulbecco's modified Eagle's medium [DMEM]-F12 supplemented with 10% FCS, penicillin/streptomycin, and 50 μg/mlascorbate[Sigma-Aldrich,St.Louis,http://www.sigmaaldrich.com]), and centrifuged at 100 rpm for 5 minutes. The pellet was reconstituted in medium and washed four times, and the supernatants of the washing steps containing the released cells were collected. Cells were centrifuged and cultivated at a density of 3 × 10⁸ cells per 150-cm² culture flask at 37°C in 5% CO₂. Adherent cells were washed after 2 days and cultivated until confluency with or without 100 nM sodium selenite supplementation (Sigma-Aldrich). Upon establishment of the protocol in our laboratory, we again characterized the BMSC population obtained by fluorescence-activated cell sorting (FACS) analyses and were able to exclude expression of hematopoetic markers (CD14 and CD45; data not shown). The chondrogenic and osteogenic differentiation potential of these BMSCs was repeatedly characterized [50]. Starting with passage 1 after confluence of the first 150-cm² flask cell numbers were determined, an equal number of selenium-deficient and selenium-supplemented cells were seeded into new 150-cm² flasks (1.8 × 10⁵ to 1.8 × 10⁶), and cell doubling rates were calculated. THP1 cells (human monocytic leukemia cells) were grown in RPMI 1640 medium containing 10% FCS, and human fetal osteoblasts (hFOBs; obtained from T. Spelsberg [51]) and HepG2 cells (human hepatocarcinoma cells) were cultivated in DMEM-F12 containing 10% FCS. All cells were grown at 37°C, (except hFOB cells, which were grown at 34°C), in a humidified atmosphere consisting of 5% CO₂ and 95% air.

GPx Assay

Cells were sonicated in buffer containing 250 mM sucrose, 20 mM HEPES, and 1 mM EDTA, pH 7.4. GPx activity was assayed by the method of Dreher using tertiary butyl hydroperoxide as the substrate, as described previously [52]. Cytosols of the cells were added to the reaction mixture in a final volume of 1 ml (0.1 M Tris, 0.5 mM EDTA, pH 8, 200 mM NADPH [Roche Diagnostics, Mannheim, Germany, http://www.roche-applied-science.com], 2 mM glutathione, 1 U/ml glutathione reductase). The reaction was started by the addition of 7 μM tertiary butyl hydroperoxide. After an initial incubation period of 1–1.5 minutes, the oxidation of NADPH was registered at 340 nm, within the linear range of the reaction for 3 minutes. The activity of GPx was expressed as nanomoles of NADPH oxidized per minutes and mg of protein. Unspecific NADPH oxidation was measured by the complete inhibition of GPx by the addition of 100 mM of the GPx inhibitor mercaptosuccinate to the incubation mixture before starting the reaction. The background values were subtracted from the results obtained. Each measurement was obtained in duplicate. Protein content of cell extracts was determined by using Rotiquant Protein Assay (Carl Roth GmbH, Karlsruhe, Germany, http://www.carl-roth.de). All chemicals were obtained from Sigma-Aldrich.

TrxR Assay

TrxR activity was measured as described previously [53]. Cells were sonicated in 1 mM EDTA, 50 mM Tris, pH 7.5. TrxR activity was determined in a volume of 0.5 ml containing 100 mM potassium phosphate, pH 7.0, 10 mM EDTA, 0.2 mM NADPH, and 2.5 mM 5,5′-dithio-bis nitrobenzoic acid. Reactions were started by adding 100 μl of cell extract. The change in absorption was monitored at 412 nm for 2 minutes. All chemicals were obtained from Sigma-Aldrich. Protein content of cell extracts was determined using Rotiquant Protein Assay (Carl Roth GmbH). The activity of TrxR was expressed as milliunits per milligram of protein.

SOD Assay

For the determination of SOD activity, hMSC-TERT cells were cultivated with or without selenite supplementation. Cells were harvested in 50 mM potassium phosphate buffer (pH 7.8) and sonicated, and cytosols were used for the SOD assay. We used the bathocuproine sulfonate-nitroblue tetrazolium (NBT) test described by Beauchamp et al. [54, 55] and Spitz and Oberley [56]. The main principle of the test is the reduction of NBT to blue formazan by superoxide anions, which can be measured at 560 nm at room temperature. The underlying reaction is a competition between SOD and NBT for O₂⁻ radicals, which are generated by a xanthine-xanthine-oxidase system. The rate of NBT reduction in the absence of SOD is taken as the reference value. In a final volume of 1 ml, 50 mM potassium phosphate buffer, pH 7.8, 1 mM diethylenetriaminepentaacetic acid, 1 U of catalase, 56 nM NBT, 0.1 mM xanthine, and 50 nM bathocuproine disulfonate disodium salt were mixed, and reaction was started by adding 20 μl of cell extract. To differentiate between SOD1 and SOD2 activity, samples were incubated in an additional reaction with 5 mM NaCN for 45 minutes to block SOD1 activities. All chemicals were obtained from Sigma-Aldrich. The background values were subtracted from the results obtained. Each measurement was obtained in duplicate. Protein content of cell extracts was determined by using the Rotiquant Protein Assay (Carl Roth GmbH).

Reverse Transcription-Polymerase Chain Reaction

Total RNA was isolated from hMSC-TERT cells, primary BM-SCs obtained from different donors, hFOB, HepG2, and THP1 cells as controls using the NucleoSpin RNA II kit (Macherey-Nagel, Düren, Germany, http://www.macherey-nagel.de) according to the manufacturer's instructions. Two micrograms of total RNA was reverse-transcribed with Moloney murine leukemia virus reverse transcriptase (Promega, Mannheim, Germany, http://www.promega.com) in a volume of 20 μl. For polymerase chain reaction (PCR), 1 μl of cDNA was used as a template in a volume of 50 μl. Taq DNA polymerase was obtained from Qiagen (Hilden, Germany, http://www1.qiagen.com). DNA fragments were amplified by using a standard protocol. PCR conditions were as follows: 30 seconds at 94°C, 30 seconds at annealing temperature, 60 seconds at 72°C; 40 cycles. The sequences of the primers used, annealing temperatures, MgCl₂ concentrations, and the sizes of the PCR products are listed in Table 1.

Real-Time PCR

For monitoring SOD1 mRNA expression in hMSC-TERT cells and primary BMSCs and to confirm its regulation by selenite supplementation and H₂O₂, a real-time PCR protocol was established. Cells were cultivated with or without 100 nM selenite for at least 1 week and stimulated with or without 50 μM H₂O₂ for up to 8 h. One microliter of cDNA was used for SOD1 and EF1α amplification as a housekeeping gene. PCR conditions were as follows: 30 seconds at 94°C, 30 seconds at annealing temperature, 60 seconds at 72°C (see Table 1 for PCR conditions). Real-time PCR was performed with the DNA Engine Opticon system (MJ Research, Waltham, MA, http://www.mjr.com) using SYBR Green (Biozym Scientific GmbH, Hessisch Oldendorf, Germany, http://www.biozym.com) as fluorescent dye. For quantification and statistical analyses, SOD1 mRNA expression was normalized to the expression levels of the housekeeping gene EF1α using the relative expression software tool (REST) [57]. Specificity of SOD1 amplicons were confirmed by melting curve analyses.

Western Blot

HMSC-TERT cells were cultivated with or without 100 nM selenite and stimulated with or without 50 μM H₂O₂ as described above. Cells were harvested in homogenization buffer (50 mM Tris, 1 mM EDTA, 1 mM Pefabloc, 1 mM dithiothreitol). The protein content was determined using the Rotiquant (Carl Roth GmbH) protein assay. Twenty micrograms of protein were boiled for 5 minutes in SDS-polyacrylamide gel electrophoresis buffer (100 mM Tris, pH 6.8, 7.5% glycerol, 1% SDS, 0.025% bromphenol blue) and separated by SDS gel electrophoresis. Proteins were transferred to Optitran BA S 85 membranes (Schleicher and Schuell, Dassel, Germany, http://www.whatman.com). The membranes were treated with a buffer containing 0.1% Tween 20, 2% horse serum, 2.5% bovine serum albumin (BSA), 2.5% milk powder in phosphate-buffered saline (PBS) for 2 hours to inhibit nonspecific binding. Then, the membranes were incubated in 0.1% Tween 20, 1% horse serum, and 1% milk powder in PBS with a monoclonal antibody against SOD1 (Santa Cruz Biotechnology Inc., Heidelberg, Germany, http://www.scbt.com). Membranes were washed with 10 mM Tris, pH 7.5, 140 mM NaCl, 2 mM EDTA, 0.1% Triton X-100, 1% horse serum, 1% BSA, and 1% milk powder, followed by incubation with horseradish peroxidase-conjugated goat anti-mouse IgG antibody (Sigma-Aldrich) using a solution containing 0.1% Tween 20, 1% horse serum, 1% BSA, and 1% milk powder in PBS. The expression of SOD1 was detected by using the enhanced chemiluminescence system (Amersham Biosciences, GE Healthcare Life Sciences, Freiburg, Germany, http://www.amersham.com).

Determination of Reactive Oxygen Species

In hMSC-TERT cells cultivated with or without 100 nM selenite, the accumulation of reactive oxygen species (ROS) was determined using 2′7′-dichlorodihydrofluorescin diacetate (H₂DCF-DA) (Invitrogen, Karlsruhe, Germany, http://www.invitrogen.com), a nonfluorescent lipid permeable compound that is oxidized by intracellular ROS to form the impermeable fluorescent compound DCF [58–60]. Cells were treated with 30μM dichlorofluorescein diacetat (DCFH-DA) for 30 minutes, trypsinized, and washed twice with PBS, 1% BSA. DCF fluorescence was determined by FACS analyses using a FACS cytometer (BD LSR I; Becton, Dickinson and Company, Heidelberg, Germany, http://www.bd.com) at an excitation wavelength of 488 nm and a 530 nm emission filter bandpassed for DCF. For analyzing the flow cytometry histograms, the free software WinMDI version 2.8 (Scripps Research Institute Cytometry Software, http://facs.scripps.edu/software.html) was used.

Micronucleus Assay

Primary BMSCs were cultivated with or without 100 nM selenite on glass slides for up to four passages until confluence. Cells were fixed with methanol at −20°C for 4 hours. For staining of the nuclei and micronuclei, cells were incubated with staining solution (0.006% acridine orange [Sigma-Aldrich] in Sørensen buffer [15 mM Na₂HPO₄, 15 mM KH₂PO₄]) for 3 minutes, washed twice with Sørensen buffer for 5 minutes each, and mounted for microscopy [12–14]. Cells were analyzed with a Zeiss Axioskop 2 fluorescence microscope (Carl Zeiss, Göttingen, Germany, http://www.zeiss.com), and nuclei and micronuclei were counted in a blinded fashion. Each MSC population was analyzed by evaluating three sets of 1,000 nuclei. Photographs were taken with an AxioCam MRc camera and the AxioVision 4.0 software (Carl Zeiss).

Results

Expression Profile of Antioxidative Enzyme Systems in Primary BMSCs and hMSC-TERT Cells

Both primary BMSCs obtained from different subjects and hMSC-TERT cell cultures expressed a broad variety of antioxidative enzyme systems and selenoproteins, as determined by reverse transcription-PCR analyses (Fig. 1, lanes 2–6). As a control, the genes of interest were also amplified from THP1 (monocytic), hFOB, and HepG2 (hepatocarcinoma) cells (lanes 7–9). The selenoproteins TrxR1 and TrxR2, cGPx, phospholipid-hydroperoxide glutathione peroxidase, and the nonselenoproteins SOD1, SOD2, and catalase were expressed in equal amounts in all samples analyzed. The pGPx could be detected in different preparations of primary BMSCs (lanes 3–6) and in the control sample HepG2. In the more osteogenic differentiated hFOB cell line, the amplified signal was weaker, and no amplicons, or a very faint band, appeared when using cDNA from hMSC-TERT cells (lanes 2 and 7). BMSCs express selenoprotein P (SeP) in a high amount (lane 3–6), comparable to the expression in liver cells (lane 8), whereas hMSC-TERT and hFOB cells show much weaker signals (lanes 2 and 7). The gastrointestinal glutathione peroxidase was only expressed in HepG2 cells (lane 8) but was undetectable in BMSCs and hMSC-TERT cells (lane 2–6). The selenoproteins thioredoxin glutathione reductase (TGR, TrxR3) and glutathione peroxidase 6 (GPx6) could not be detected in all samples (data not shown). As a control, the housekeeping gene EF1α was amplified.

GPx and TrxR Activity in hMSC-TERT Cells and BMSCs

When cultured in the presence of 100 nM sodium selenite, both the activities of GPx and TrxR were considerably enhanced in BMSCs and hMSC-TERT cells. BMSCs show basal GPx activities ranging between 50 and 80 mU/mg protein, varying between different subjects (Fig. 2A). The activity could be enhanced 1.8-fold by cultivating cells in the presence of 100 nM selenite. The stimulation varied between different subjects. The basal GPx activity of hMSC-TERT cells is much lower and ranges between 6 and 7 mU/mg protein. Activity could be stimulated twofold in the presence of selenite.

BMSCs showed basal TrxR activity between 9 and 100 mU/mg protein, varying between different subjects (Fig. 2B). Activity could be enhanced 1.5-fold by cultivating the cells in the presence of 100 nM selenite, with a variation between different subjects. HMSC-TERT cells show a basal TrxR activity of 22 mU/mg protein. A 1.4-fold stimulation could be also determined after selenite supplementation.

ROS Accumulation in hMSC-TERT Cells

To obtain reproducible results, avoid individual biases, and get sufficient amounts of cells, we used hMSC-TERT cells to measure the accumulation of ROS in selenium adequate (100 nM selenite) versus selenium-deficient cultures. The determination of ROS accumulation in hMSC-TERT cells showed a higher ROS load in selenite-deficient cells (Fig. 2C, left panel, gray line) compared with selenite-supplemented cells (black line). The selenium-deficient hMSC-TERT cells showed a DCFH-DA fluorescence intensity median of 14.75 (Fig. 2C, right panel, gray bar), which was significantly reduced to 6.03 in selenium-supplemented cells (black bar).

SOD1 Regulation by Selenite and H₂O₂ in Primary BMSCs and hMSC-TERT Cells

By performing a dot blot screening procedure on the mRNA expression of antioxidative enzymes in selenium-deficient and selenium-adequate primary BMSCs (BMSC73) and hMSC-TERT cultures, we realized a marked stimulation of SOD1 mRNA species in both cell types by selenite supplementation (data not shown). As a consequence, we measured SOD1 mRNA in primary BMSCs and hMSC-TERT, and protein and activity of SOD1 in hMSC-TERT, during H₂O₂ stimulation in the presence and absence of 100 nM selenite. mRNA levels were determined by quantitative PCR, protein expression was determined by Western blotting, and semiquantitative densitometry and SOD1 activity was measured by using an enzymatic assay. Under selenium deficiency, SOD1 mRNA expression was modulated slightly, by 50 μ M H₂O₂ after 0.5–8 hours of treatment in hMSC-TERT cells (Fig. 3A, gray bars). SOD1 mRNA expression was significantly higher in cells cultured under adequate selenite supplementation (Fig. 3A, black bars). Under basal conditions, without H₂O₂ treatment, SOD1 mRNA content of selenite-supplemented cells was 1.75-fold higher than in untreated cells. After 1 hour of stimulation of cells with H₂O₂, SOD1 mRNA was markedly enhanced under adequate selenite supplementation compared with selenium-deficient cells and compared with untreated cells.

In selenite-deficient primary BMSC271 and BMSC280, SOD1 mRNA expression was not modulated by 50 μM H₂O₂ treatment after 0.5–7 hours and 0.5–1 hours, respectively (Fig. 3B, gray bars). When BMSC271 and BMSC280 were supplemented with 100 nM selenite and stimulated with 50 μM H₂O₂, SOD1 mRNA expression was significantly enhanced time-dependently and at any time point, with a maximal twofold stimulation after 0.5 and 4 hours in BMSC271 and a maximal 1.5-fold stimulation after 1 hour in BMSC280 (Fig. 3B, black bars). Values were normalized to the expression of the housekeeping gene EF1α.

SOD1 protein expression as detected by Western blotting was enhanced in hMSC-TERT cells cultivated under selenite supplementation and stimulated with 50 μM H₂O₂ for 8 hours (Fig. 3C, black bar). In cells grown under selenium-deficient conditions, SOD1 expression was independent of H₂O₂ treatment.

SOD1 activity was measured in hMSC-TERT cells cultivated with or without selenite supplementation and stimulation with 10, 25, and 50 μM H₂O₂ for 8 hours (Fig. 3D). In selenite-deficient cells, SOD1 activity was only slightly modulated by H₂O₂ treatment and ranged between 70 and 100 U/mg protein (gray bars). Under adequate selenite supplementation, hMSC-TERT cells showed a twofold enhancement of SOD1 activity after 8 hours of treatment with 10 μM H₂O₂ (black bars). Higher doses of H₂O₂ could not further stimulate SOD1 activity.

Micronucleus Assay in Primary BMSCs

The occurrence of micronuclei (MN) as a biomarker for DNA damage was analyzed in primary BMSCs cultivated with or without 100 nM selenite. BMSCs of three donors (BMSC274, BMSC276, and BMSC278) in different passages were investigated by determining the number of MN per 1,000 cells. Of each sample, three sets of 1,000 cells were analyzed in a blinded fashion. The results are summarized in Table 2. In cells cultivated under adequate selenite concentrations, the number of MN was reduced by a mean of 40% (28.8%–58.3%) compared with cells cultivated under selenite deficiency. MN frequency varied considerably between different donors. Figure 4 shows micronuclei formation in BMSC276 cultivated under selenium-deficient conditions.

Cell Proliferation and Population Doubling

We did not find a significant difference in population doubling time between selenium-supplemented and control cultures (data not shown). As is known from the literature, primary BMSC culture could be maintained for three to seven passages independent of donor age but with a great deal of variation among individual donors. The cell doubling rates starting with passage 1 varied between 50 and 280 hours.

Discussion

Adult stem cells are rare cells with low proliferation capacity, which according to present hypotheses reside in stem cell niches and give rise to a transient amplifying cell pool that differentiates, thereby regenerating the respective tissues [61]. Mesenchymal stem cells can be isolated from bone marrow stroma and other sources, and the transient amplifying pool derived from it can be differentiated towards various mesenchymal pathways, for example, osteogenesis, chondrogenesis, adipogenesis and others. We used the immortalized cell line hMSC-TERT and primary cultures of BMSCs to demonstrate that cells cultured in vitro under conditions of low selenium supply show symptoms of oxidative stress and that supplementation of selenium represents an adequate countermeasure. The immortalized cell system was used to get sufficient, reproducible, and unbiased cell material, propagated under long-term low selenium culture conditions to yield significant results. The translation of the results obtained with this system into primary cultures is of course hampered by donor differences in expression levels of antioxi-dative systems and in nutritional selenium uptake. This is probably why differences in cellular TrxR activity in individual donor cell populations were not always significant if basal activity was individually high (Fig. 2B), whereas GPx activity was always significantly enhanced in all samples after selenium supplementation, probably due to its lower rank in protein hierarchy for selenium incorporation [9]. This indicates that—albeit with individual variability—cellular antioxidative capacity rapidly decreases ex vivo and/or is primarily low in donors at risk.

There is ample evidence and consensus that reactive oxygen species are indispensable for redox signaling, hormone synthesis, and intracellular killing of bacteria [9, 43, 46], but if not appropriately controlled by neutralization and compartmentalization, they may damage DNA, proteins, and lipids. Prevention and repair of damage by ROS is mediated in a concerted action of antioxidative systems and DNA repair mechanisms. Anything beyond their capacities may cause irreversible damage followed by senescence or tumor promotion [19, 62–66].

As we could show both in immortalized and in primary cells, basal and ROS-stimulated expression of SOD1 is diminished under conditions of selenium deficiency in standard cell culture systems. Thus, these cells produce less hydrogen peroxide at the expense of superoxide anion accumulation; for example, they harbor an increased risk of cell damage [67]. The SOD1 promoter hydrogen peroxide response element is located between bases −533 and −520 [68] and may be blocked under these conditions for unknown reasons, but it is effective after selenite supplementation. However, this mechanism avoids additional hydrogen peroxide production—albeit at the risk of superoxide anion accumulation—in a situation in which the neutralization capacity for hydrogen peroxide is impaired due to low GPx activity.

Having clearly demonstrated these biochemical phenomena, the question of biological consequences was evaluated applying an ROS production assay in hMSC-TERT cells. Again, in this case, the immortalized cells were used to get reproducible results and sufficient amounts of cells, and the difference was highly significant in selenium-supplemented versus control cells. The telomerase transfection overcomes fail-safe programs, and as we have shown previously, these cells in fact harbor mutations in culture if maintained under high proliferation pressure (Ink4a/ARF, KRAS), which can lead to tumor formation in nude mice [69]. Using the micronuclei formation assay in primary cells, we could also demonstrate that in fact significant amounts of micronuclei are formed in culture and that selenium supplementation reduced the number of micronuclei up to 58%. The amount of 30–50 micronuclei per 1,000 cells found in some cultures is comparable to reports about micronuclei formation in lymphocytes in patients on chronic hemodialysis [12].

Stem cells are widely used in research to elucidate molecular mechanisms of differentiation and to establish cell-based strategies of tissue repair, tissue engineering, and transplantation. Thus it appears to be very important to care for the integrity of the genome and proteome of these cells for various reasons, for example, quality and survival of stem cell preparations. As we can show here, micronuclei formation as readout of damage of BMSCs ex vivo can be effectively reduced and the antioxidative capacity enhanced by the simple means of adding adequate amounts of selenium to the culture medium. Selenium deficiency has been described to cause cell death in vitro in Jurkat cells [70], and several specific stem cell media are supplemented with selenite for empirical reasons [71]. Our data deliver a molecular explanation.

In the setting of primary BMSC cultures, it is obvious that we can control for the oxidative load and that endogenous fail-safe programs do work since the population doubling ceases by time and cells stop proliferating in vitro in higher passages. Further translational research is warranted to ever optimize quality and security issues in cell-based therapeutic strategies.

Summary

We show here that under standard cell culture conditions, primary BMSCs have low antioxidative capacity and show symptoms of cellular stress, such as formation of micronuclei. The simple means of selenium supplementation in cultures enhances mechanisms of selenium-dependent and selenium-independent ROS scavenging, restores their antioxidative capacity, and effectively reduces micronuclei formation. We therefore conclude that selenite supplementation should be part of good medical practice in cell cultures used for tissue engineering, genetic engineering, and transplantation to optimize scavenging mechanisms of ex vivo-manipulated stem cells.

Table Table 1.. Primers used for reverse transcription polymerase chain reaction

Table Table 2.. MN frequency per 1,000 cells in primary bone marrow stromal cells cultivated with or without 100 nM selenite

Figure Figure 1.. Reverse transcription-polymerase chain reaction expression profile of antioxidative enzyme systems in hMSC-TERT (lane 2) and primary BMSCs (lanes 3–6). BMSCs were obtained from four different subjects (BMSC53, BMSC73, BMSC113, and BMSC158). cDNAs from THP1 (human monocytic leukemia), hFOB, and HepG2 (hepato-carcinoma) cells were used for comparison and positive controls (lanes 7–9). Experiments were done in triplicates, and a representative result is shown. Abbreviations: BMSC, bone marrow stromal cell; CAT, catalase; cGPx, cellular glutathione peroxidase; GI-GPx, gastrointestinal GPx; hFOB, human fetal osteoblast; hMSC, human mesenchymal stem cell; pGPx, plasma GPx; PH-GPx, phospholipid-hydroperoxide GPx; SeP, selenoprotein P; SOD, superoxide dismutase; TrxR, thioredoxin reductase.

Figure Figure 2.. Selenoenzyme activity in hMSC-TERT and primary BMSC and ROS accumulation in hMSC-TERT. GPx (A) and TrxR activity (B) of hMSC-TERT and primary BMSCs isolated from different subjects. Cells were cultivated with (black bars) or without (gray bars) 100 nM selenite supplementation. Activity assays were done in triplicates, and the results are expressed as mean ± SEM. Statistical analyses were performed by using analysis of variance (*, p.< 05;**, p < .005). (C): Reactive oxygen species (ROS) accumulation in hMSC-TERT cells. Left: Cells were cultivated in the presence (black line) and absence (gray line) of 100 nM selenite. DCFH-DA fluorescence of cells was determined by fluorescence-activated cell sorting cytometry as an indicator of ROS accumulation within the cells. A representative result is shown. Right: Median of fluorescence intensity of selenium-deficient (gray bar) and selenium-supplemented (black bar) hMSC-TERT cells. The experiments were repeated three times and done in triplicates. The results are shown as mean ± SEM. Statistical analyses were performed by using analysis of variance (**, p = .005). Abbreviations: BMSC, bone marrow stromal cell; DCFH-DA, dichloro-fluorescein diacetat; GPx, glutathione peroxidase; hMSC-TERT, telomerase-immortalized human mesenchymal stem cell; n.s., not significant; TrxR, thioredoxin reductase.

Figure Figure 3.. SOD1 expression in hMSC-TERT and primary BMSCs and activity in hMSC-TERT cells cultivated with or without 100 nM selenite and stimulated with H₂O₂. (A): Real-time PCR amplification of SOD1 in hMSC-TERT cultivated with (black bars) or without (gray bars) 100 nM selenite supplementation and H₂O₂ stimulation for 0.5–8 h. SOD1 expression is normalized to EF1α as a housekeeping gene. Experiments were done in triplicates; the results are expressed as mean ± SEM. Statistical analyses were performed by using the relative expression software tool (REST) [57] (*, p < .05). (B): Real-time PCR amplification of SOD1 in primary BMSC271 and BMSC280 with (black bars) or without (gray bars) 100 nM selenite supplementation and H₂O₂ stimulation for 0.5–7 h (BMSC271) and 0.5 and 1 h (BMSC280), respectively. SOD1 expression is normalized to EF1α as a housekeeping gene. Statistical analyses were performed by using the REST [57] (α, p < .05). (C): Western blot detection of SOD1 in hMSC-TERT cells cultivated with or without 100 nM selenite supplementation and stimulation with or without 50 μM H₂O₂ for 8 h. Signals were quantified by optical densitometry. Experiments were done three times, and a representative experiment is shown. White bar: No selenite, no H₂O₂; light gray bar: no selenite, 50 μM H₂O₂; dark gray bar: 100 nM selenite, no H₂O₂; black bar: 100 nM selenite, 50 μM H₂O₂. (D): SOD1 activity in hMSC-TERT cells cultivated with (black bars) or without (gray bars) 100 nM selenite supplementation and stimulation with H₂O₂ with 10, 25, and 50 μM for 8 h. Experiments were done four times, and the results are expressed as mean ± SEM. Statistical analyses were performed by using analysis of variance (*, p < .05). Abbreviations: BMSC, bone marrow stromal cell; hMSC-TERT, telomerase-immortalized human mesenchymal stem cell; PCR, polymerase chain reaction; SOD, superoxide dismutase.

Figure Figure 4.. Micronuclei formation in BMSC276 cultivated under selenium-deficient conditions. Arrows mark micronuclei in four different cells (magnification ×400).

Disclosures

The authors indicate no potential conflicts of interest.

Acknowledgements

We thank Prof. Dr. Katja Becker-Brandenburg, Giessen, Germany, for helpful discussions; Dr. Norbert Schütze, Würzburg, Germany, for providing the hFOB cell line; Dr. Ulrich Nöth and Martina Regensburger, Würzburg, Germany, for the preparation of primary MSC cultures; and Kristin Kobras, Würzburg, Germany, for help with the micronucleus assay. This work was supported by German Research Society Priority Program 1087 and Research Training Group 639.