TISSUE-SPECIFIC STEM CELLS A Nuclear Magnetic Resonance Biomarker for Neural Progenitor Cells: Is It All Neurogenesis? ¨ ,a FRANCISCO J. RIVERA,a PAUL RAMM,a,b SEBASTIEN COUILLARD-DESPRES,a SONJA PLOTZ a a MONIKA KRAMPERT, BERNADETTE LEHNER, WERNER KREMER,b ULRICH BOGDAHN,a HANS R. KALBITZER,b LUDWIG AIGNERa,c a
Department of Neurology and bInstitute for Biophysics and Physical Biochemistry, University of Regensburg, Regensburg, Germany; cInstitute of Molecular Regenerative Medicine, Paracelsus Private Medical University Salzburg, Salzburg, Austria Key Words. Imaging • Neural stem cells • Mesenchymal stem cells • Lipid droplets
ABSTRACT In vivo visualization of endogenous neural progenitor cells (NPCs) is crucial to advance stem cell research and will be essential to ensure the safety and efficacy of neurogenesisbased therapies. Magnetic resonance spectroscopic imaging (i.e., spatially resolved spectroscopy in vivo) is a highly promising technique by which to investigate endogenous neurogenesis noninvasively. A distinct feature in nuclear magnetic resonance spectra (i.e., a lipid signal at 1.28 ppm) was recently attributed specifically to NPCs in vitro and to neurogenic regions in vivo. Here, we demonstrate that although this 1.28ppm biomarker is present in NPC cultures, it is not specific for the latter. The 1.28-ppm marker was also evident in mesenchy-
mal stem cells and in non-stem cell lines. Moreover, it was absent in freshly isolated NPCs but appeared under conditions favoring growth arrest or apoptosis; it is initiated by induction of apoptosis and correlates with the appearance of mobile lipid droplets. Thus, although the 1.28-ppm signal cannot be considered as a specific biomarker for NPCs, it might still serve as a sensor for processes that are tightly associated with neurogenesis and NPCs in vivo, such as apoptosis or stem cell quiescence. However, this requires further experimental evidence. The present work clearly urges the identification of additional biomarkers for NPCs and for neurogenesis. STEM CELLS 2009; 27:420 – 423
Disclosure of potential conflicts of interest is found at the end of this article.
INTRODUCTION The existence of neural progenitor cells (NPCs) and neurogenesis in the adult human central nervous system [1–3] has generated enormous interest in the field of regenerative medicine. Neurogenesis can be modulated, and thus, it may provide a basis for cellular and functional brain repair [4]. Methods and devices for in vivo imaging of neurogenesis are urgently needed, since they will facilitate the development of neurogenesis-based therapies. In animals, optical bioluminescence imaging of neurogenesis has recently been achieved in transgenic reporter mice [5]. The use of transgenic approaches, however, is restricted to animals and is not appropriate for humans. Here, nuclear magnetic resonance (NMR) spectroscopy offers the opportunity to investigate molecular compositions of cells and tissues in physiological environments noninvasively and thus constitutes a promising approach to visualizing endogenous neurogenesis in humans. Manganas et al. [6] recently described a specific lipid signal (1.28 ppm) in NMR spectra of cultured NPCs in vitro. According to this study, this peak was absent or significantly lower in mature neurons and astrocytes, as well as in non-neural stem/progenitor cells, but present in NPCs [6]. Moreover, this
signal was detected by magnetic resonance spectroscopic imaging (MRSI) in vivo in the hippocampus of rodents and humans and decreased with aging [6]. Here, we revisit the NPC specificity of the 1.28-ppm biomarker by NMR spectroscopy using different cell types, including NPCs, and different culture conditions. Moreover, we associate the 1.28-ppm signal with mobile lipid droplets that appear during cell death.
MATERIALS
AND
METHODS
Cell Cultures NPCs from mouse fetal forebrains (embryonic day 13–15.5) were isolated and grown as neurospheres as described [7, 8]. Cos7 cells and mesenchymal stem cells (MSCs) were cultivated as described [9, 10]. For Cos7 and MSC experiments, controls were seeded at a density of 6 ⫻ 103 cells per cm2; treated cells lacked regular medium changes and were grown densely until de visu postconfluence. Immunocytochemical analyses were performed as described [7]. Apoptosis was induced with 50 –100 M Menadion (SigmaAldrich Chemie GmbH, Mu¨nchen, Germany, http://www. sigmaaldrich.com) for 5 hours at culture day 5 for NPCs and with 25 M Menadion for 18 hours at control conditions for MSCs and
Author contributions: P.R.: collection and/or assembly of data, data analysis and interpretation, manuscript writing; S.C.-D. and W.K.: data analysis and interpretation; S.P.: collection and/or assembly of data; F.J.R., M.K., and B.L.: collection and/or assembly of data, data analysis and interpretation; U.B. and H.R.K.: conception and design; L.A.: conception and design, data analysis and interpretation, manuscript writing. Correspondence: Ludwig Aigner, Ph.D., Institute of Molecular Regenerative Medicine, Paracelsus Medical University, Strubergasse 21, A-5020 Salzburg, Austria. Telephone: 43-662-442002-1280; Fax: 43-662-442002-1209; e-mail:
[email protected] Received August 19, 2008; accepted for publication October 26, 2008; first published online in STEM CELLS EXPRESS November 6, 2008. ©AlphaMed Press 1066-5099/2009/$30.00/0 doi: 10.1634/stemcells.2008-0816
STEM CELLS 2009;27:420 – 423 www.StemCells.com
Ramm, Couillard-Despres, Plo¨tz et al.
421
Figure 1. The 1.28-ppm nuclear magnetic resonance (NMR) biomarker is not specific for NPCs. (A): Phase-contrast microscopic images of NPCs at different days in culture. Single cells and small spheres gradually merge into big neurospheres within a period of 2 weeks. Scale bar ⫽ 200 m. (B): Nestin (green) staining of adherent NPCs. 4,6-Diamidino-2-phenylindole (DAPI) (blue) was used to visualize cell nuclei. Scale bar ⫽ 50 m. (C): Quantification of the Nestin staining. The percentage of Nestin-positive cells is depicted for NPC cultures grown for 5, 8, 11, and 14 days. Note that the Nestin-positive percentage did not increase during the measurement series; n ⫽ 3. (D): Densitometric analysis of Western blot for Nestin, normalized against actin. Again, note that the relative content of Nestin protein did not increase over the 14 days in culture. (E): NMR spectra of NPCs at 5, 8, 11, and 14 days in culture. The 1.28-ppm peak (dotted line) was clearly absent at day 5. From day 8 to day 14 the peak emerged and increased. Assigned signals were as follows: agarose (m, 3.8 –3.3 ppm), choline containing compounds (s, 3.2 ppm), acetone (s, 2.2 ppm), acetate (s, 1.9 ppm), alanine (d, 1.47 ppm), lactate (d, 1.31 ppm), fatty acid methylene chain (s, 1.28 ppm), ethanol (t, 1.17 ppm), isopropanol (d, 1.16 ppm), amino acid methyls (1.05– 0.9 ppm), and macromolecular methyls (1.05– 0.8 ppm). (F): Quantification of the 1.28-ppm peak. The deconvoluted peak ((CH2)lip) was integrated and normalized to the (CH3)m.m. as described in Materials and Methods; n ⱖ 4. (G): Quantitative analysis of the 1.28-ppm intensity in different cell types and culture conditions. Shown are NPCs (days 5 and 14; n ⱖ 4), forebrain prep. (n ⫽ 3), and Cos7 cells (n ⫽ 6) and MSCs (n ⫽ 3) each at de visu log-phase (control) and postconfluence. (H): Nile Red (green) staining in 5- and 14-day-old NPC cultures. DAPI (blue) was used to stain all nuclei present in the cultures. Scale bar ⫽ 15 m. Statistical analysis was as described; ⴱ, p ⬍ .05; ⴱⴱ, p ⬍ .01; ⴱⴱⴱ, p ⬍ .001. Abbreviations: a.u., arbitrary units; (CH2)lip, fatty acid methylene signal; (CH3)m.m., macromolecular methyl signals; MSC, mesenchymal stem cell; NPC, neural progenitor cell; Forebrain prep., freshly prepared fetal (embryonic day 13.5) mouse forebrains.
Cos7 cells. Lactate dehydrogenase (LDH) assays were carried out following the operating instructions (Promega, Madison, WI, http:// www.promega.com). Immunostainings of dissociated and seeded cells, Western blot of neurospheres, and fluorescence-activated cell sorting analyses of cells were performed as described [7, 11, 12]. Nile Red staining was done as described [13].
NMR Spectroscopy For NMR spectroscopy 1–10 million cells per sample were washed twice in phosphate-buffered saline (PBS) and embedded in ultralow gelling point agarose (Sigma-Aldrich; 1% agarose in PBS solution containing 10% heavy water and 40 M dimethyl-silapentanesulfonate [DSS]) to avoid inhomogeneous distributions and sedimentations inside the 5-mm NMR tubes (Norell Inc., Landisville, NJ, http://www.nmrtubes.com). Samples were cooled to 5°C, and NMR measurement started within 15 minutes thereafter. During measurement, the temperature was kept at 5°C. Measurements were performed at high resolution with 1H-NMR Bruker Avance 600 and 800 MHz spectrometers using a gradient-based water suppression pulse sequence [14]. Sixty-four scans with 65,536 data points and a 4.7-second repetition time were accumulated, followed by an exponential line broadening of 0.3 Hz. After Fourier transformation the spectra were phase- and baseline-corrected manually. DSS was used
www.StemCells.com
as an internal reference standard (0 ppm). The quantification was obtained by deconvolution of the spectral region of interest between 1.50 and 0.82 ppm. Briefly, singlets, douplets, and triplets at the chemical shifts of lactate, alcohols, amino acids, fatty acid methylene, macromolecular methylene, and methyl were fitted using a standard Matlab (The MathWorks, Natick, MA, http://www. mathworks.com) fitting routine. The resulting fatty acid methylene peak around 1.28 ppm and the macromolecular methyl signals between 1.05 and 0.80 ppm, which were used for normalization to cell density, were thus fully separated from overlapping multiplets of small molecules. This evaluation procedure rather underestimates the 1.28-ppm peak intensity at high values because of the nonnegligible fatty acid methyl contribution to the macromolecular methyl intensity used for normalization and is thus accurate for verification of significant fatty acid signal increases. Significance was tested as usual with two-tailed heteroscedastic Student’s t test (ⴱ, p ⬍ .05; ⴱⴱ, p ⬍ .01; ⴱⴱⴱ, p ⬍ .001). Values are mean ⫾ SD.
RESULTS
AND
DISCUSSION
First, we analyzed NPCs prepared from developing mouse forebrain (embryonic day [E] 13–E15.5), which were grown as
NMR of Stem Cells
422
Figure 2. The 1.28-ppm peak correlates with growth arrest and apoptosis. (A): Cell cycle investigation of NPCs using fluorescence-activated cell sorting. The cell distribution shifted to G0/G1 phase during the culture period. n ⫽ 3. (B): LDH assay outcome of the conditioned media compared with the 1.28-ppm peak intensity of NPCs cultured for 5, 8, 11, and 14 days. Both the LDH value and the 1.28-ppm peak intensity increased upon extended culturing. (C): The correlation between the 1.28-ppm peak intensity and the LDH value. r2 ⫽ 0.5416; n ⫽ 4. (D): Apoptosis induction experiment using Menadion. For NPCs (n ⱖ 4), MSCs (n ⱖ 2), and Cos7 cells (n ⫽ 6) the 1.28-ppm peak intensities at de visu log-phase (control) and after addition of Menadion (50 – 100 M for NPCs and 25 M for Cos7 and MSCs) and incubation for a further 5 hours (NPCs) and 18 hours (Cos7 and MSCs) are shown. ⴱ, p ⬍ .05; ⴱⴱ, p ⬍ .01; ⴱⴱⴱ, p ⬍ .001. Abbreviations: a.u., arbitrary units; (CH2)lip, fatty acid methylene signal; (CH3)m.m., macromolecular methyl signals; LDH, lactate dehydrogenase; MSC, mesenchymal stem cell; NMR, nuclear magnetic resonance; NPC, neural progenitor cell.
neurospheres [7, 8], for the presence of the 1.28-ppm signal in NMR spectra. These cells proliferated rapidly and readily generated growing spheres within the 2 weeks of analysis (Fig. 1A). NPC cultures contained a high content of neural stem cells indicated by the expression of Nestin (Fig. 1B). The percentage of Nestin-expressing cells and the overall levels of Nestin did not increase during the culture period (Fig. 1C, 1D). Surprisingly, the 1.28-ppm biomarker was absent in freshly dissociated fetal mouse brain, which is known to contain a high percentage of neural stem cells [7]. It was absent in cultures at day 5, emerged at day 8, and increased until day 14 (Fig. 1E). Normalized quantifications of the 1.28-ppm signal intensity demonstrate a significant increase along with the duration of NPCs in culture (Fig. 1F). Taken together, this suggests that the 1.28-ppm signal does not correlate with the abundance of neural stem cells in the NPC culture but rather with the progressive appearance of another cell type or of a certain cell-physiological status. To address the specificity of the 1.28-ppm marker with respect to stemness and/or neural lineage, the presence and intensity of this signal was investigated in MSCs as an example of a non-neural stem cell and in Cos7 cells as an example of a non-neural and non-stem cell type of cell. Although the 1.28ppm peak was virtually absent in low-density proliferating logphase MSC and Cos7 cultures, it became prominent in postconfluent cultures (Fig. 1G). This suggests that the 1.28-ppm peak is not specific for neural stem cell identity but might be associated with a physiological condition that occurs with postconfluence culture conditions. Manganas et al. associated the 1.28-ppm peak with the lipid fraction of cells [6]. Indeed, numerous reports suggest that the 1.28-ppm peak correlates with a group of cellular lipid molecules called NMR-visible mobile lipids (reviewed in [15]). The accumulation of mobile lipids leading to a 1.28-ppm NMR signal has already been described in numerous cell types, but not in stem cells, and associated with the induction of apoptosis and/or growth arrest [16 –24]. The mobile lipids are identified microscopically through Nile Red staining [13]. In our experiments, Nile Red labeling revealed the absence of such lipids in NPCs grown for 5 days. However, Nile Red-positive cells increased in number and in staining intensity over time in culture (Fig. 1H).
Typically, growth arrest and/or cell death increases with high cell density, confluence, or medium starvation. Thus, we analyzed the distribution of the cells within the cycle phases and monitored the amount of cell death in NPCs of different time points in culture. Flow cytometry demonstrated a shift in the cell populations from G2/M to G0/G1 phase upon longer culture period, indicating a partial growth arrest of the cultures (Fig. 2A). Thus, the 1.28-ppm signal might correlate with quiescence of cells. Alternatively, and as recently hypothesized [25], the 1.28-ppm peak might correlate with cell death. LDH assays, used as a correlate for cell death, indicated increasing levels of cell death in NPC cultures over time (Fig. 2B). This increase strongly correlated with the 1.28 peak intensity (r2 ⫽ 0.5416) (Fig. 2B, 2C). To further strengthen this observation, different cell types were treated with Menadion, a substance commonly used to induce cell death [26, 27]. Treatment of NPCs, Cos7 cells, and MSCs with Menadion significantly increased the 1.28-ppm peak (Fig. 2D), suggesting that the appearance of the 1.28-ppm peak also correlates with physiological events that take place during cell death. In vivo, high levels of apoptosis are specifically found in neurogenic brain regions, such as the hippocampus [28]. The regional high level of apoptosis might be at the origin of the 1.28-ppm peak detected by MRSI in the human hippocampus but not in the cortex [6]. However, considering the relatively low number of apoptotic NPCs in the entire hippocampus and the voxel size that was sampled in the study by Manganas et al. [6], the possibility of detecting the 1.28-ppm signal seems to be extremely low. With the aim of modeling such a scenario, we constructed an additive spectrum approximating the apoptotic/healthy cell ratio and added nine spectra of day 5 NPCs to one spectrum of day 14 NPCs. As a result, no 1.28-ppm signal was detected, indicating that the signal-to-noise ratio was too high to allow the detection of the 1.28-ppm signal.
CONCLUSION We have found a strong correlation of the 1.28-ppm peak with culture conditions favoring growth arrest and apoptosis. Our results demonstrate that the 1.28-ppm signal is not specific for neural stem or progenitor cells. Nevertheless, since cell death and quiescence are typically associated with neuro-
Ramm, Couillard-Despres, Plo¨tz et al. genic regions, the 1.28-ppm MRSI might still be a useful approach to visualizing such areas. The validity of the 1.28ppm signal as a measure for neurogenesis, however, undoubtedly requires additional in vitro and in vivo experiments. The present work clearly prompts the identification of neurogenesis-specific NMR peaks or patterns to offer the opportunity for future in vivo imaging of neurogenesis in the healthy and diseased human brain.
ACKNOWLEDGMENTS This work was supported by the Bavarian State Ministry of Sciences, Research and the Arts (ForNeuroCell grant), by the Ger-
REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Eriksson PS, Perfilieva E, Bjork-Eriksson T et al. Neurogenesis in the adult human hippocampus. Nat Med 1998;4:1313–1317. Be´dard A, Parent A. Evidence of newly generated neurons in the human olfactory bulb. Brain Res Dev Brain Res 2004;151:159 –168. Curtis MA, Kam M, Nannmark U et al. Human neuroblasts migrate to the olfactory bulb via a lateral ventricular extension. Science 2007;315: 1243–1249. Ming GL, Song H. Adult neurogenesis in the mammalian central nervous system. Annu Rev Neurosci 2005;28:223–250. Couillard-Despres S, Finkl R, Winner B et al. In vivo optical imaging of neurogenesis: Watching new neurons in the intact brain. Mol Imaging 2008;7:28 –34. Manganas LN, Zhang X, Li Y et al. Magnetic resonance spectroscopy identifies neural progenitor cells in the live human brain. Science 2007; 318:980 –985. Karl C, Couillard-Despres S, Prang P et al. Neuronal presursor specific activity of a human doublecortin regulatory sequence. J Neurochem 2005;92:264 –282. Wachs FP, Couillard-Despres S, Engelhardt M et al. High efficacy of clonal growth and expansion of adult neural stem cells. Lab Invest 2003;83:949 –962. Couillard-Despres S, Uyanik G, Ploetz S et al. Mitotic impairment by doublecortin is diminished by doublecortin mutations found in patients. Neurogenetics 2004;5:83–93. Rivera FJ, Couillard-Despres S, Pedre X et al. Mesenchymal stem cells instruct oligodendrogenic fate decision on adult neural stem cells. STEM CELLS 2006;24:2209 –2219. Couillard-Despres S, Quehl E, Altendorfer K et al. Human in vitro reporter model of neuronal development and early differentiation processes. BMC Neurosci 2008;9:31. Wachs FP, Winner B, Couillard-Despres S et al. Transforming growth factor-beta1 is a negative modulator of adult neurogenesis. J Neuropathol Exp Neurol 2006;65:358 –370. Greenspan P, Mayer EP, Fowler SD. Nile red: A selective fluorescent stain for intracellular lipid droplets. J Cell Biol 1985;100:965–973. Araníbar N, Ott KH, Roongta V et al. Metabolomic analysis using optimized NMR and statistical methods. Anal Biochem 2006;355:62–70.
www.StemCells.com
423
many Federal Ministry of Education and Research (Bundesministerium fu¨r Bildung und Forschung Grants 0312134 and 01GG0706), and by the European Union FP6 (Project Diagnostic Molecular Imaging LSHB-CT-2005-512146). F.J.R. is a fellow of the Alexander von Humboldt Foundation, Georg Forster Program.
DISCLOSURE
POTENTIAL CONFLICTS OF INTEREST
OF
The authors indicate no potential conflicts of interest.
15 Hakuma¨ki JM, Kauppinen RA. 1H NMR visible lipids in the life and death of cells. Trends Biochem Sci 2000;25:357–362. 16 Blankenberg FG, Storrs RW, Naumovski L et al. Detection of apoptotic cell death by proton nuclear magnetic resonance spectroscopy. Blood 1996;87:1951–1956. 17 Di Vito M, Lenti L, Knijn A et al. 1H NMR-visible mobile lipid domains correlate with cytoplasmic lipid bodies in apoptotic T-lymphoblastoid cells. Biochim Biophys Acta 2001;1530:47– 66. 18 Al-Saffar NM, Titley JC, Robertson D et al. Apoptosis is associated with triacylglycerol accumulation in Jurkat T-cells. Br J Cancer 2002;86:963–970. 19 Brindle KM. Detection of apoptosis in tumors using magnetic resonance imaging and spectroscopy. Adv Enzyme Regul 2002;42:101–112. 20 Iorio E, Di Vito M, Spadaro F et al. Triacsin C inhibits the formation of 1H NMR-visible mobile lipids and lipid bodies in HuT 78 apoptotic cells. Biochim Biophys Acta 2003;1634:1–14. 21 Quintero M, Cabanas ME, Arus C. A possible cellular explanation for the NMR-visible mobile lipid (ML) changes in cultured C6 glioma cells with growth. Biochim Biophys Acta 2007;1771:31– 44. 22 Delikatny EJ, Lander CM, Jeitner TM et al. Modulation of MR-visible mobile lipid levels by cell culture conditions and correlations with chemotactic response. Int J Cancer 1996;65:238 –245. 23 Barba I, Mann P, Cabanas ME et al. Mobile lipid production after confluence and pH stress in perfused C6 cells. NMR Biomed 2001;14: 33– 40. 24 Rosi A, Grande S, Luciani AM et al. (1H) MRS studies of signals from mobile lipids and from lipid metabolites: Comparison of the behavior in cultured tumor cells and in spheroids. NMR Biomed 2004;17:76 –91. 25 Friedman SD. Comment on “Magnetic resonance spectroscopy identifies neural progenitor cells in the live human brain”. Science 2008;321:640, author reply 640. 26 Gerasimenko JV, Gerasimenko OV, Palejwala A et al. Menadioneinduced apoptosis: Roles of cytosolic Ca(2⫹) elevations and the mitochondrial permeability transition pore. J Cell Sci 2002;115:485– 497. 27 Braun S, Krampert M, Bodo E et al. Keratinocyte growth factor protects epidermis and hair follicles from cell death induced by UV irradiation, chemotherapeutic or cytotoxic agents. J Cell Sci 2006;119:4841– 4849. 28 Biebl M, Cooper CM, Winkler J et al. Analysis of neurogenesis and programmed cell death reveals a self- renewing capacity in the adult rat brain. Neurosci Lett 2000;291:17–20.
