- Email: [email protected]
Galectin-3 secreted by human umbilical cord blood-derived mesenchymal stem cells reduces amyloid-β42 neurotoxicity in vitro
FEBS Letters 584 (2010) 3601–3608
journal homepage: www.FEBSLetters.org
Galectin-3 secreted by human umbilical cord blood-derived mesenchymal stem cells reduces amyloid-b42 neurotoxicity in vitro Ju-Yeon Kim a,b, Dong Hyun Kim a, Dal-Soo Kim a, Ji Hyun Kim a, Sang Young Jeong a, Hong Bae Jeon a, Eun Hui Lee b, Yoon Sun Yang a, Wonil Oh a, Jong Wook Chang a,* a b
Biomedical Research Institute, MEDIPOST Co., Ltd., Seoul 137-874, Republic of Korea Department of Physiology, College of Medicine, The Catholic University of Korea, Republic of Korea
a r t i c l e
i n f o
Article history: Received 20 April 2010 Revised 13 July 2010 Accepted 16 July 2010 Available online 24 July 2010 Edited by Jesus Avila Keywords: hUCB-MSC Amyloid-b Galectin-3 Neuroprotection
a b s t r a c t In this study, we found that expression and secretion of galectin-3 (GAL-3) were upregulated by amyloid-b42 (Ab42) exposure in human umbilical cord blood-derived mesenchymal stem cell (hUCB-MSC) without cell death. Ab42-exposed rat primary cortical neuronal cells co-treated with recombinant GAL-3 were protected from neuronal death in a dose-dependent manner. hUCB-MSCs were cocultured with Ab42-exposed rat primary neuronal cells or the neuroblastoma cell line, SHSY5Y in a Transwell chamber. Coculture of hUCB-MSCs reduced cell death of Ab42-exposed neurons and SH-SY5Y cells. This neuroprotective effect of hUCB-MSCs was reduced significantly by GAL-3 siRNA. These data suggested that hUCB-MSC-derived GAL-3 is a survival factor against Ab42 neurotoxicity. Ó 2010 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
1. Introduction Since human umbilical cord blood-derived mesenchymal stem cells (hUCB-MSC) is capable of differentiating into mesodermal origin, such as bone, cartilage, and adipose, our institute has conducting a phase III clinical trial for cartilage regeneration (ClinicalTrials.gov Identifier: NCT01041001) under approval of the Korea Food and Drug Administration (KFDA). In their allogeneic reactions, hUCB-MSC is known to be stable immunologically, due to the absence of expression of major histocompatibility complex (MHC) class II and T-cell co-stimulatory factors [1]. In addition, hUCB-MSC suppresses allogeneic responses of lymphocytes under mixed lymphocyte reaction conditions [2]. Therefore, hUCB-MSC transplantation should be considered safe for use in stem cell therapy. Recently, therapeutic effects of hUCB-MSC on brain disease have been reported. For example, glioma-tropism of hUCB-MSC was proposed as a vehicle carrying suicide gene for glioma therapy [3,4]. In addition, hUCB-MSC transplantation improved a Niemann–Pick type C (NP-C) disease mouse model via modulation of neuroinflammation [5]. Collectively, these data suggested that hUCB-MSC should be regarded as a fascinating source for cell therapy for neurological disorders.
* Corresponding author. Fax: +82 2 475 1991. E-mail addresses: [email protected], [email protected] (J.W. Chang).
Alzheimer’s disease (AD) is an incurable neurodegenerative disease that is characterized by loss of neurons by amyloid-b (Ab) neurotoxicity [6]. AD progresses by gliosis and neuronal and synaptic loss, causing progressive loss of memory and cognitive function [7]. Recently, transplantation of rat or mouse neural stem cells (NSC) [8,9] and mouse bone marrow MSC resulted in improved cognition in transgenic AD mice [10]. In this study, we found for the first time that human UCB-MSC has a neuroprotective effect in vitro against Ab42 toxicity via secretion of galectin-3. Galectin-3 (GAL-3), known as a neuroprotective factor, was actively secreted by Ab42 exposure in hUCB-MSC. Our data demonstrated potential therapeutic effects of hUCB-MSC via secretion of beneficial factors, including GAL-3. 2. Materials and methods 2.1. Preparation of hUCB-MSC This study was approved by the Institutional Review Board of MEDIPOST Co., Ltd. Umbilical cord blood was collected from umbilical veins after neonatal delivery with informed consent of the pregnant mothers. MSCs were isolated by separation of mononuclear cells (MNC) using a Ficoll-Hypaque solution (d = 1.077 g/ cm3; Sigma). Following transfer to a-minimum essential medium (a-MEM; Gibco) supplemented with fetal bovine serum (FBS; Gibco), MNCs were seeded in culture flasks at 5 105 cells/cm2. Cells
0014-5793/$36.00 Ó 2010 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2010.07.028
3602
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
maintained in humidified 5% CO2 at 37 °C formed colonies of spindle-shaped cells. At 50% confluence, cells were harvested after treatment with 0.25% (w/v) trypsin–EDTA (Gibco) and were reseeded for expansion [11].
by cell count using i-Solution software (IMT i-Solution INC). Percentage of MAP2-positive cell was calculated by the ratio of number of MAP2-positive cells to DAPI-positive total cell number. Cell viability (%) was analyzed by a ratio of number of green colored cells to total number of cell in a LIVE/DEADÒ staining.
2.2. Cell culture and treatment 2.5. Western blot analysis All procedures for rat primary cortical and hippocampal neuron culture were conducted in accordance with institutional guidelines under the approved protocols. Pregnant Sprague–Dawley rats were purchased from Orientbio (Korea). Brain tissues were dissected from embryonic day 14 rat cortex and hippocampus tissues, and cells were mechanically dissociated in Ca2+/Mg2+-free Hank’s balanced salt solution. Initial cell seeding density was 3 103/cm2. Cells were allowed to proliferate in the presence of 20 ng/ml of basic fibroblast growth factor (bFGF) in serum-free Neurobasal media (Invitrogen). Cell proliferation was maintained in Neurobasal media containing bFGF for 4 days in order to reach 80% confluence prior to induction of differentiation by withdrawal of bFGF. The differentiation was induced in Neurobasal media without B27 supplement for 10 days. B27 supplement was not used in all culture procedures. During differentiation period, Neurobasal media was changed in every 2 days. After 10 days differentiation, cells were stained by anti-MAP2, GFAP, CD11b and nestin antibodies to evaluate percentage of neuron, astrocyte, microglia and neural progenitor cell. Percentage of each cell type was shown in Supplementary Fig. 1. Ab42 treated neuronal cells were cocultured with hUCB-MSC or treated with GAL-3 protein (R&D Biosystems) for 24 h. Rat primary neuronal cells or SH-SY5Y cells were treated with 10 lM of Ab42 (Sigma), and human recombinant GAL-3 for 24 h. Ab42 was solubilized in phosphate-buffered saline (PBS) and the stock was stored at 4 °C for 3 days and divided to aliquots at 80 °C. GAL-3 was dissolved in PBS containing 0.1% bovine serum albumin (BSA) and aliquoted at 80 °C. As a control, MRC5 human fibroblast cells (ATCC) were cultured under culture conditions that were identical to those of hUCB-MSC. hUCB-MSC (8 104 cells/cm2) were cocultured in an upper chamber (pore size: 1 lm) with Ab42-exposed rat primary neuronal cells or SH-SY5Y cells in a Transwell chamber (Falcon). 2.3. Real-time PCR For total RNA isolation, cells were lysed using TRIzol Reagent (Invitrogen). mRNA was reversely transcribed to cDNA. Amplification was performed with the Lightcycler real-time PCR (Roche Diagnostics) system using Lightcycler TaqMan Master mix (Roche), sequence-specific probes (Roche), and the following specific primers: human Gal-3 forward 50 -gagcctaccctgccactg-30 , reverse 50 -aggcaaaggcaggttataagg-30 ; RNA polymerase II forward 50 -cctccaggtggttcagagg-30 , reverse 50 -agttccaacaatggctaccg-30 was used for the control gene. Data analysis of relative gene expression was performed with the DDCt technique using the reference gene as a calibrator. 2.4. Immunostaining and LIVE/DEAD staining for quantitative analysis Cultured cells were fixed in 4% paraformaldehyde/0.15% picric acid in phosphate-buffered saline and were incubated with each anti-antibody. The following primary antibodies were used: tubulin b III (Sigma), microtubule-associated protein 2 (MAP2) (Millipore), glial fibrillary acidic protein (GFAP) (Millipore), nestin (Chemicon) and cleaved caspase-3 (Cell signaling). The image was analyzed using a fluorescent microscope (Nikon) and a confocal microscope (ZEISS, LSM 510 META). For objective assessment of cell death, a LIVE/DEADÒ Viability/Cytotoxicity kit (Invitrogen) was utilized according to instructions. All quantitation was performed
Cell extracts were prepared by ultrasonication (Branson) in buffer containing 9.8 M urea, 4% CHAPS, 130 mM DTT, 40 mM Tris–Cl, and 0.1% SDS. Secreted proteins of conditioned media were precipitated in ice-cold acetone at 20 °C for overnight. After centrifugation, precipitated proteins were solubilized in the same buffer for cell extracts. Protein amount was measured by the Bradford assay (Bio-RAD). The protein extract (20 lg) and precipitated protein from media (15 lg) were loaded on SDS–PAGE for separation of proteins. Proteins were transferred to a nitrocellulose membrane (NC) and then incubated with anti-GAL-3 (R&D Biosystem) or cleaved caspase-3 or caspase-3 (Cell signaling) and b-tubulin (Sigma) antibody. Protein was visualized by Enhanced Chemiluminescence (ECL) (GE Healthcare). The analysis of caspase-3 activation in Western blot was analyzed by QuantityOne software (Bio-RAD). 2.6. siRNA treatment For knock down of GAL-3 expression, control siRNA or siRNA for GAL-3 (Bioneer, Korea) was mixed with Lipofectamine™2000 (Invitrogen). hUCB-MSC were treated with this mixture for 6 h in antibiotic-free media containing 2% FBS. After 6 h, siRNA treated cells were kept for 48 h in serum-free media for evaluation of reduced secretion level of GAL-3. In coculture with Ab42-exposed SH-SY5Y cells, hUCB-MSC treated with each siRNA for 24 h were trypsinized, and detached 8 104 cells/cm2 cells were added to the upper chamber of the Transwell for additional 24 h. 2.7. ELISA For measurement of GAL-3 in conditioned media, media were filtered by syringe filter (0.2 lm, Millipore). Secreted GAL-3 was measured using the DuoSet ELISA Development System (R&D Systems) following a recommended manual. Results were analyzed at 450 nm using a microplate reader, VERSAmax (Molecular Devices).
3. Results 3.1. Expression and secretion of GAL-3 were upregulated by Ab42 in hUCB-MSC Since RT-PCR analysis revealed that the level of GAL-3 mRNA was increased by Ab42 exposure in hUCB-MSC (data not shown), we performed real-time PCR for further analysis. Compared to control cells, the transcriptional level of GAL-3 was upregulated by an average of 6.2-fold in the Ab42-exposed hUCB-MSC group (Fig. 1A). For analysis of the protein level of GAL-3 by Ab42 exposure, cell lysates and conditioned media of hUCB-MSC were collected in a time-dependent manner and analyzed by Western blot analysis. Although GAL-3 does not contain a signal peptide for secretion, secretion of GAL-3 by non-classical pathways has been reported [12]. Intracellular and extracellular GAL-3 was upregulated by Ab42 treatment in a time-dependent manner (Fig. 1B). b-Tubulin of Western blot and stained membrane by Ponceau S were loading control for lysate and media in Western blot analysis, respectively. In order to measure concentration of secreted GAL-3 by Ab42 treatment, the same sample of Fig. 1B was analyzed by ELISA. Concentration of secreted GAL-3 by
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
3603
Fig. 1. Ab42 induced upregulation of GAL-3 expression and secretion in hUCB-MSC. (A) hUCB-MSC was treated with Ab42 for 24 h in serum-free media. Total RNA was isolated from each cell and GAL-3 expression was analyzed by real-time RT-PCR. (B) Cell lysates and concentrated conditioned media were separated by SDS–PAGE and then immunoblotted with anti-GAL-3 antibody. Tubulin was analyzed for evaluation of loading amount in cell lysate. The Ponceau S staining of NC membrane was used for loading control of media. (C) Concentrated each medium was analyzed by ELISA according to a recommended instruction. (D) For analysis of the effect of Ab42 on hUCB-MSC and rat primary neuronal cells, cells exposed to Ab42 for 24 h underwent LIVE/DEAD staining. Red and green colors indicate nuclei of dead cells and live cells, respectively. (E) Cell viability (%) was analyzed by the ratio of the number of green colored cells to the total number of cells. Value represents mean ± SE of three independent experiments. *P < 0.05 vs MSC without Ab42. **P < 0.05 vs neuron without Ab42.
Ab42 was increased to average 0.49 ng/ml at 24 h (Fig. 1C). Although Ab42 induced cell death in rat primary cortical neuronal cell, it had no effect on viability of hUCB-MSC in LIVE/DEAD staining (Fig. 1D and E). A green/red color indicated live cell/nu-
clei of dead cells (Fig. 1D). After LIVE/DEAD staining, percentage of cell viability was quantitated by cell counting (Fig. 1E). These data suggested that Ab42 induced upregulation and secretion of GAL-3 in hUCB-MSC without cell death.
3604
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
3.2. Treatment of GAL-3 protected Ab42-mediated neurotoxicity in rat primary neuronal cells Since GAL-3 has been proposed as a neuronal survival factor [13], we tested whether recombinant GAL-3 treatment rescues Ab42 neurotoxicity or not. Ab42-exposed rat primary cortical neu-
ronal cells underwent simultaneous dose-dependent treatment with recombinant GAL-3. For analysis of surviving neurons under Ab42 neurotoxicity, neurons were selected by immunostaining with anti-MAP2 antibody (Fig. 2A). MAP2-positive neurons were counted for survival as rated by treatment with GAL-3 protein (Fig. 2B). Recombinant GAL-3 significantly protected cell death in
Fig. 2. Treatment of recombinant GAL-3 protected Ab42 neurotoxicity of rat primary neurons. Differentiated neuronal cells were obtained by the procedure described in Materials and Methods. Primary cortical neurons exposed to Ab42 were co-treated with recombinant GAL-3 in a dose-dependent manner. (A) After 24 h, immunostaining with anti-MAP2 antibody was performed for calculation of the percentage of surviving neurons. The image was captured by confocal microscope. Percentage of MAP2positive cells was shown in (B). Value represents mean ± SE of three independent experiments. *P < 0.05 vs Ab42-exposed cortical neurons. (C) Cell extracts were collected from duplicate cells of (A) and analyzed by Western blot analysis with anti-caspase-3. (D) The result of Western blot was quantitated by densitometric analysis. The graph indicated the ratio of cleaved form to full-length capspase-3. Value represents mean ± SE of three-independent experiments. (E) Sucrose (100 mM) and lactose (100 mM) were preincubated with GAL-3 recombinant protein (500 pg/ml) for 20 min. This mixture was treated to SH-SY-5Y cells with Ab42 for 24 h. each cell was stained by LIVE/ DEAD staining and the percentage of dead cell was analyzed by cell counting. **P < 0.05 vs Ab42-exposed cortical neuron. ***P < 0.05 vs Ab42-exposed SH-SY5Y with GAL-3 protein. Ct indicates cortical neurons.
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
3605
Fig. 3. Coculture of hUCB-MSC protected Ab42-exposed rat primary neurons and SH-5Y5Y cells. For evaluation of the neuroprotective effect of hUCB-MSC, hUCB-MSC was cocultured simultaneously with Ab42-exposed rat primary neuronal cells (A–C) or SH-5Y5Y cells, human neuroblastoma cells (D–F) for 24 h, respectively. As a control, nonrelated human fibroblasts were also cocultured. (A) After coculture, rat primary hippocampal neuronal cells were stained by anti-tubulin bIII antibody. (B) Percentage of tubulin bIII positive neuron was calculated. (C) Protection of cleavage of caspase-3 in rat primary cortical neurons by hUCB-MSC was analyzed by immunostaining. In the merged image, the red color indicates cleaved caspase-3 and the blue color shows DAPI staining. (D) In coculture, Ab42-exposed SH-SY5Y cells were cocultured with hUCB-MSC for 24 h and analyzed by LIVE/DEAD staining. (E) Percentage of dead cells was analyzed under each condition. (F) Cell lysates from (D) were analyzed by Western blot analysis with anticaspase-3 antibody. (G) The result of Western blot was quantitated by densitometric analysis. The graph indicated that a ratio of cleaved form to full-length capspase-3. Value represents mean ± SE of three-independent experiments. ***P < 0.05 vs Ab42-exposed SH-SY5Y cells. Hippo and Ct indicated hippocampal and cortical neurons, respectively. Value of (B) and (E) represent mean ± SE of each three-independent experiments. *P < 0.05 vs Ab42-exposed hippocampal neurons. **P < 0.05 vs Ab42-exposed SH-SY5Y.
3606
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
rat primary neurons by Ab42 exposure in a dose-dependent manner. In this experiment, the counting of DAPI-positive cells showed that GAL-3 treatment did not affect cell proliferation (Supplementary Fig. 2). In addition, cleavage of caspase-3 was monitored by
Western blot analysis and a ratio of cleaved form to full-length caspase-3 was analyzed by densitometric analysis. Treatment with GAL-3 reduced activation of caspase-3 in a dose-dependent manner by Ab42 exposure (Fig. 2C and D). In order to know a role of su-
Fig. 4. Knock down of GAL-3 by siRNA reduced neuroprotection effect of hUCB-MSC. (A) To set up conditions for knock down of GAL-3, hUCB-MSC was treated with each siRNA for 48 h in the presence of Ab42, and secreted GAL-3 was measured by ELISA. (B) As a control, hUCB-MSC treated with CON-siRNA or GAL-3 siRNA for 24 h were cocultured simultaneously with Ab42-exposed SH-SY5Y cells for another 24 h in a Transwell chamber. After LIVE/DEAD staining, percentage of surviving cells was calculated. (C) Duplicated cell lysates were analyzed by anti-caspase-3 antibody. The precursor caspase-3 (32 kDa) and the cleaved form (17 kDa) were detected by Western blot analysis. (D) The graph indicated a ratio of cleaved form to full-length capspase-3. Value represent mean ± SE of three-independent experiments. ***P < 0.05 vs Ab42-exposed CON siRNA SH-SY5Y. (E) Duplicated cells were immunostained by anti-cleaved caspase-3 antibody. Red (caspase-3) and blue (DAPI) colors were merged. Value of (A) and (B) represent mean ± SE of each three-independent experiments *P < 0.05 vs CON siRNA and **P < 0.05 vs Ab42-exposed SH-SY5Y.
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
gar-binding domain on GAL-3 for neuroprotection, recombinant GAL-3 (500 pg/ml) was preincubated with sucrose and lactose (each 100 mM). This mixture was treated to Ab42-exposed neuroblastoma cell line, SH-SY5Y cells. If carbohydrate-recognition domain is involved in neuroprotection, lactose and sucrose could be competitive [13]. To analyze percentage of dead cells, cells were stained by LIVE/DEAD staining. In percentage of dead cell, neuroprotection of GAL-3 increased significantly by sugars treatment (Fig. 2E). These data suggested that carbohydrate-recognition domain may participate in negative regulation for neuroprotection on GAL-3. Collectively, these data suggested that GAL-3 could protect Ab42-mediated neurotoxicity. 3.3. Coculture of hUCB-MSC protected Ab42 induced cell death of SHSY5Y and rat primary neurons Since we observed a survival effect of GAL-3 against Ab42 neurotoxicity and GAL-3 was actively secreted from hUCB-MSC by Ab42 exposure, hUCB-MSCs, or, as a control, non-related human lung fibroblast cells, MRC5 were cocultured simultaneously with Ab42-exposed rat primary hippocampal neuronal cells for 24 h cells in a Transwell chamber. Since pore size of the upper chamber is 1 lm, there is no cell contact. However, secreted proteins can pass through pores between hUCB-MSC and primary neuronal cells. Following Ab42 exposure, we examined cell death rate by immunostaining with anti-tubulin bIII (Fig. 3A). Tubulin bIII positive cells were counted for analysis of survival neurons (Fig. 3B). Ab42-exposed rat primary neurons showed typical cell death patterns, including alteration of neurite; however, these cells sustained viability dramatically by coculture of hUCB-MSC. However, MRC5 were not protective against neuronal cell death in a Transwell chamber. In addition, this event was also confirmed with rat primary cortical neuronal cells by immunostaining of anti-caspase-3 (cleaved form). As expected, cleaved caspase-3 (red color) by Ab42 neurotoxicity was reduced significantly by coculturing with hUCB-MSC, but not with MRC5 (Fig. 3C). For further study involving knock down of GAL-3 by siRNA, we conducted the same experiment with SH-SY5Y human neuroblastoma cells because rat primary neuronal cells were mixed neuronal and glial cells. We performed coculture of hUCB-MSC with Ab42-exposed SH-SY5Y cells (Fig. 3D). From LIVE/DEAD staining, we observed similar data on rat primary neuronal cells (Fig. 3E). Cleaved caspase-3 was reduced dramatically by coculturing of hUCB-MSC in Western blot analysis (Fig. 3F and G). These data suggested that hUCB-MSC rescued Ab42 toxicity in cortical and hippocampal neurons. 3.4. Knock down of GAL-3 abolished the neuroprotective effect of hUCB-MSC Since we observed a neuroprotective effect against Ab42-exposed SH-SY5Y cells by hUCB-MSC coculture (Fig. 3) and Ab42 induced GAL-3 expression in hUCB-MSC (Fig. 1), we attempted to inhibit expression of GAL-3 in hUCB-MSC. We tested inhibition of GAL-3 by three different siRNAs, and reduced secretion of GAL-3 was confirmed by ELISA. Three siRNAs demonstrated significant and efficient inhibition of GAL-3 secretion in the presence of Ab42 (Fig. 4A). hUCB-MSC were treated with siRNA1 and the three mixed siRNAs for 24 h, and were then cocultured simultaneously for an additional 24 h with Ab42-exposed SH-SY5Y cells, respectively, for determination of the neuroprotective role of GAL-3 in hUCB-MSC. In LIVE/DEAD staining, control siRNA treated hUCBMSC protected Ab42 neurotoxicity; however, this effect was reduced significantly by GAL-3 siRNA(s) treatment (Fig. 4B). Each duplicated cell was analyzed by immunostaining and Western blot analysis with anti-capspase-3 antibody (Fig. 4C and E). Similar results of LIVE/DEAD staining were confirmed by analysis of caspase-
3607
3. A ratio of cleaved form to full-length capspase-3 was analyzed by densitometric analysis according to Figs. 2 and 3 (Fig. 4D). Above data were confirmed again by the staining of cleaved caspase-3 (Fig. 4E). These data suggested that the neuroprotective effect of hUCB-MSC was mediated via GAL-3.
4. Discussion In the early stage of stem cell therapy, regeneration medicine was the primary interest. In the point of view of regeneration medicine for AD therapy using stem cells, overcoming AD seems impossible due to widespread loss of neurons in the brain. Recently, Block et al. reported on a paracrine effect of stem cells; BM-MSC secreted STC-1 for anti-apoptosis in UV-irradiated skin fibroblasts [14]. In addition, hBM-MSC improved cardiac function after myocardial infarction via secreted TSG-6 [15]. These data emphasized that stem cells are robust sources, not only for regeneration, but also for secretion of survival factors for damaged cells. In our data, we proved that hUCB-MSC secrete beneficial factors, including GAL-3, under pathological conditions such as Ab42 exposure. Although Ab42 induced potential neuronal cell death, Ab42 stimulated secretion of GAL-3 as a neuroprotective factor in hUCB-MSC without cell death (Fig. 1). GAL-3 is a multifunctional protein involved in adhesion, growth, and intracellular and nuclear transcription [16,17]. Although GAL-3 dose not contain signal peptides for secretion, it can be detected outside the cell [12,18]. In particular, functions of the extracellular form of GAL-3 have been reported in cell activation, TCR signaling, cell adhesion, migration, and IL-5 secretion, depending on cell type [19,20]. Secretion of GAL-3 seems to be enhanced by certain stimuli, because a low level of GAL-3 was detected by Western blot analysis in conditioned media of naïve hUCB-MSC (Fig. 1B). However, after Ab42 exposure, GAL-3 was strongly detected in conditioned media of hUCB-MSC. Similar data, such as calcium ionophore induced secretion of GAL-3, has been reported [21]. According to findings from a previous report, when GAL-3 was immobilized as a substratum, GAL-3 promoted notable neural cell adhesion and neurite outgrowth [13]. Furthermore, these activities were inhibited by polyclonal antibody and additive sugar including sucrose and lactose for GAL-3. This paper suggested that GAL-3 is a survival factor for neurons via cell adhesion and neurite outgrowth. In our study, recombinant GAL-3 treatment protected Ab42 neurotoxicity in a dose-dependent manner (Fig. 2). Since GAL-3 contains domain for carbohydrate-binding, we tested the involvement of this domain (Fig. 2E). The rate of cell death by Ab42 was more declined by preincubation of GAL-3 with sugar than that of GAL-3 protein alone. These data indicated that neuroprotection of GAL-3 could be regulated by the binding of intracellular sugar containing protein. Therefore, we should further study using truncated GAL-3 to characterize a critical domain for neuroprotection on GAL-3 and try to find intracellular GAL-3 binding partner containing carbohydrate for negative regulation. Since hUCB-MSC secreted GAL-3 under Ab42, we expected that coculture of hUCB-MSC may rescue Ab42 neurotoxicity. Our expectation was confirmed in rat primary hippocampal and cortical neurons using a coculture system (Fig. 3). Moreover, inhibition of GAL-3 secretion by siRNA reduced the neuroprotective effect of hUCB-MSC with neuronal cells (Fig. 4). In previous report [13], GAL-3 served as a substratum for cell adhesion and neurite outgrowth. Since adhesion cell requires attachment to survive because interaction between receptor and substratum activates intracellular signaling [22]. Under pathological conditions of Ab, secreted GAL-3 from hUCB-MSC should stimulate neuron survival through its counter receptors and support for neural cell adhesion and outgrowth for cell maintenance against Ab42 exposure. Stem cell transplantation has been attempted in
3608
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
AD-transgenic mice for evaluation of its possibility for use in stem cell therapy. Lee et al. reported that mouse BM-MSC transplantation reduced Ab deposit and improved memory deficit in an AD mice model [10]. Although several new drugs targeting reduction of Ab level have failed in clinical trials for improvement of declined cognition [23], stem cell transplantation showed a therapeutic effect on memory deficit via Ab removal. This result may be due to simultaneous action of diverse secretory factors from stem cells, including GAL-3, under pathological conditions. When stem cells were transplanted in AD mice, stem cell-derived factors participated not only in Ab removal but also in anti-apoptosis, inflammation, and neurogenesis. Therefore, we can observe combined therapeutic effects in a stem cell transplanted animal study. Currently, we are investigating functions of 14 secreted proteins from hUCB-MSC under Alzheimer’s pathological conditions in order to understand the molecular mechanism of the therapeutic effect by hUCB-MSC transplantation (unpublished data). Among 14 proteins, GAL-3 is one of the secreted proteins with active involvement in anti-apoptosis against Ab42 neurotoxicity. Although neuroprotective effect of GAL-3 alone was almost same with that of hUCBMSC coculture (Fig. 2), cell therapy with hUCB-MSC has several advantages. The first, GAL-3 can protect Ab42 neurotoxicity however it cannot recover neuron loss in AD. Transplanted hUCBMSC differentiated into neuron and then improved neurological deficit in disease model [24–26]. These data suggested that transplanted hUCB-MSC could participate in functional regeneration of damaged area. The second, hUCB-MSC has ability to modulate inflammation. Inflammation is associated with many neurodegenerative diseases including Alzheimer’s disease [27]. Many results suggested that inhibition of inflammation will be able to reverse or slow progression of AD. Human MSC including BM-MSC and UCB-MSC can perform anti-inflammatory action on inflammatory diseases [28–30]. Collectively, we found that hUCB-MSC protected Ab42 neurotoxicity via secretion of GAL-3. Based on these data, we suggest the potential of hUCB-MSC as a new source of stem cell therapy for treatment of Alzheimer’s disease. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.febslet.2010.07.028. References [1] Wang, M., Yang, Y., Yang, D., Luo, F., Liang, W., Guo, S. and Xu, J. (2009) The immunomodulatory activity of human umbilical cord blood-derived mesenchymal stem cells in vitro. Immunology 126, 220–232. [2] Oh, W., Kim, D.S., Yang, Y.S. and Lee, J.K. (2008) Immunological properties of umbilical cord blood-derived mesenchymal stromal cells. Cell Immunol. 251, 116–123. [3] Kim, D.S., Kim, J.H., Lee, J.K., Choi, S.J., Kim, J.S., Jeun, S.S., Oh, W., Yang, Y.S. and Chang, J.W. (2009) Overexpression of CXC chemokine receptors is required for the superior glioma-tracking property of umbilical cord blood-derived mesenchymal stem cells. Stem Cells Dev. 18, 511–519. [4] Kim, S.M., Lim, J.Y., Park, S.I., Jeong, C.H., Oh, J.H., Jeong, M., Oh, W., Park, S.H., Sung, Y.C. and Jeun, S.S. (2008) Gene therapy using TRAIL-secreting human umbilical cord blood-derived mesenchymal stem cells against intracranial glioma. Cancer Res. 68, 9614–9623. [5] Lee, H., Jin, H.K. and Bae, J.S. (2010) Human umbilical cord blood-derived mesenchymal stem cells improve neurological abnormalities of Niemann–Pick type C mouse by modulation of neuroinflammatory condition. J. Vet. Med. Sci. [Epub ahead of print].
[6] Mattson, M.P. (2004) Pathways towards and away from Alzheimer’s disease. Nature 430, 631–639. [7] Smith, M.A. (1998) Alzheimer disease. Int. Rev. Neurobiol. 42, 1–54. [8] Yamasaki, T.R., Blurton-Jones, M., Morrissette, D.A., Kitazawa, M., Oddo, S. and LaFerla, F.M. (2007) Neural stem cells improve memory in an inducible mouse model of neuronal loss. J. Neurosci. 27, 11925–11933. [9] Blurton-Jones, M., Kitazawa, M., Martinez-Coria, H., Castello, N.A., Muller, F.J., Loring, J.F., Yamasaki, T.R., Poon, W.W., Green, K.N. and LaFerla, F.M. (2009) Neural stem cells improve cognition via BDNF in a transgenic model of Alzheimer disease. Proc. Natl. Acad. Sci. USA 106, 13594–13599. [10] Lee, J.K., Jin, H.K., Endo, S., Schuchman, E.H., Carter, J.E. and Bae, J.S. (2010) Intracerebral transplantation of bone marrow-derived mesenchymal stem cells reduces amyloid-beta deposition and rescues memory deficits in Alzheimer’s disease mice by modulation of immune responses. Stem Cell 28, 329–343. [11] Jin, H.J., Nam, H.Y., Bae, Y.K., Kim, S.Y., Im, I.R., Oh, W., Yang, Y.S., Choi, S.J. and Kim, S.W. (2010) GD2 expression is closely associated with neuronal differentiation of human umbilical cord blood-derived mesenchymal stem cells. Cell. Mol. Life Sci. [Epub ahead of print]. [12] Nickel, W. (2003) The mystery of nonclassical protein secretion. A current view on cargo proteins and potential export routes. Eur. J. Biochem. 270, 2109–2119. [13] Pesheva, P., Kuklinski, S., Schmitz, B. and Probstmeier, R. (1998) Galectin-3 promotes neural cell adhesion and neurite growth. J. Neurosci. Res. 54, 639– 654. [14] Block, G.J., Ohkouchi, S., Fung, F., Frenkel, J., Gregory, C., Pochampally, R., DiMattia, G., Sullivan, D.E. and Prockop, D.J. (2009) Multipotent stromal cells are activated to reduce apoptosis in part by upregulation and secretion of stanniocalcin-1. Stem Cells 27, 670–681. [15] Lee, R.H., Pulin, A.A., Seo, M.J., Kota, D.J., Ylostalo, J., Larson, B.L., SemprunPrieto, L., Delafontaine, P. and Prockop, D.J. (2009) Intravenous hMSCs improve myocardial infarction in mice because cells embolized in lung are activated to secrete the anti-inflammatory protein TSG-6. Cell Stem Cell 5, 54–63. [16] Hsu, D.K., Chen, H.Y. and Liu, F.T. (2009) Galectin-3 regulates T-cell functions. Immunol. Rev. 230, 114–127. [17] Yang, R.Y., Hsu, D.K. and Liu, F.T. (1996) Expression of galectin-3 modulates Tcell growth and apoptosis. Proc. Natl. Acad. Sci. USA 93, 6737–6742. [18] Hughes, R.C. (1999) Secretion of the galectin family of mammalian carbohydrate-binding proteins. Biochim. Biophys. Acta 1473, 172–185. [19] Stowell, S.R., Qian, Y., Karmakar, S., Koyama, N.S., Dias-Baruffi, M., Leffler, H., McEver, R.P. and Cummings, R.D. (2008) Differential roles of galectin-1 and galectin-3 in regulating leukocyte viability and cytokine secretion. J. Immunol. 180, 3091–3102. [20] Cortegano, I., del Pozo, V., Cardaba, B., de Andres, B., Gallardo, S., del Amo, A., Arrieta, I., Jurado, A., Palomino, P., Liu, F.T. and Lahoz, C. (1998) Galectin-3 down-regulates IL-5 gene expression on different cell types. J. Immunol. 161, 385–389. [21] Sato, S. and Hughes, R.C. (1994) Control of Mac-2 surface expression on murine macrophage cell lines. Eur. J. Immunol. 24, 216–221. [22] Grossmann, J. (2002) Molecular mechanisms of ‘‘detachment-induced apoptosis–Anoikis”. Apoptosis 7, 247–260. [23] Holmes, C., Boche, D., Wilkinson, D., Yadegarfar, G., Hopkins, V., Bayer, A., Jones, R.W., Bullock, R., Love, S., Neal, J.W., Zotova, E. and Nicoll, J.A. (2008) Long-term effects of Ab42 immunisation in Alzheimer’s disease: follow-up of a randomised, placebo-controlled phase I trial. Lancet 372, 216–223. [24] Lim, J.Y., Park, S.I., Oh, J.H., Kim, S.M., Jeong, C.H., Jun, J.A., Lee, K.S., Oh, W., Lee, J.K. and Jeun, S.S. (2008) Brain-derived neurotrophic factor stimulates the neural differentiation of human umbilical cord blood-derived mesenchymal stem cells and survival of differentiated cells through MAPK/ERK and PI3K/ Akt-dependent signaling pathways. J. Neurosci. Res. 86, 2168–2178. [25] Chung, D.J., Choi, C.B., Lee, S.H., Kang, E.H., Lee, J.H., Hwang, S.H., Han, H., Lee, J.H., Choe, B.Y., Lee, S.Y. and Kim, H.Y. (2009) Intraarterially delivered human umbilical cord blood-derived mesenchymal stem cells in canine cerebral ischemia. J. Neurosci. Res. 87, 3554–3567. [26] Tio, M., Tan, K.H., Lee, W., Wang, T.T. and Udolph, G. (2010) Roles of db-cAMP, IBMX and RA in aspects of neural differentiation of cord blood derived mesenchymal-like stem cells. PLoS One 5, e9398. [27] Glass, C.K., Saijo, K., Winner, B., Marchetto, M.C. and Gage, F.H. (2010) Mechanisms underlying inflammation in neurodegeneration. Cell 14, 918–934. [28] Aggarwal, S. and Pittenger, M.F. (2005) Human mesenchymal stem cells modulate allogeneic immune cell responses. Blood 105, 1815–1822. [29] Oh, J.Y., Kim, M.K., Shin, M.S., Lee, H.J., Ko, J.H., Wee, W.R. and Lee, J.H. (2008) The anti-inflammatory and anti-angiogenic role of mesenchymal stem cells in corneal wound healing following chemical injury. Stem cells 26, 1047–1055. [30] Kim, J.Y., Jeon, H.B., Yang, Y.S., Oh, W. and Chang, J.W. (2010) Application of human umbilical cord blood-derived mesenchymal stem cells in disease models. World J. Stem Cells 2, 34–38.
journal homepage: www.FEBSLetters.org
Galectin-3 secreted by human umbilical cord blood-derived mesenchymal stem cells reduces amyloid-b42 neurotoxicity in vitro Ju-Yeon Kim a,b, Dong Hyun Kim a, Dal-Soo Kim a, Ji Hyun Kim a, Sang Young Jeong a, Hong Bae Jeon a, Eun Hui Lee b, Yoon Sun Yang a, Wonil Oh a, Jong Wook Chang a,* a b
Biomedical Research Institute, MEDIPOST Co., Ltd., Seoul 137-874, Republic of Korea Department of Physiology, College of Medicine, The Catholic University of Korea, Republic of Korea
a r t i c l e
i n f o
Article history: Received 20 April 2010 Revised 13 July 2010 Accepted 16 July 2010 Available online 24 July 2010 Edited by Jesus Avila Keywords: hUCB-MSC Amyloid-b Galectin-3 Neuroprotection
a b s t r a c t In this study, we found that expression and secretion of galectin-3 (GAL-3) were upregulated by amyloid-b42 (Ab42) exposure in human umbilical cord blood-derived mesenchymal stem cell (hUCB-MSC) without cell death. Ab42-exposed rat primary cortical neuronal cells co-treated with recombinant GAL-3 were protected from neuronal death in a dose-dependent manner. hUCB-MSCs were cocultured with Ab42-exposed rat primary neuronal cells or the neuroblastoma cell line, SHSY5Y in a Transwell chamber. Coculture of hUCB-MSCs reduced cell death of Ab42-exposed neurons and SH-SY5Y cells. This neuroprotective effect of hUCB-MSCs was reduced significantly by GAL-3 siRNA. These data suggested that hUCB-MSC-derived GAL-3 is a survival factor against Ab42 neurotoxicity. Ó 2010 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
1. Introduction Since human umbilical cord blood-derived mesenchymal stem cells (hUCB-MSC) is capable of differentiating into mesodermal origin, such as bone, cartilage, and adipose, our institute has conducting a phase III clinical trial for cartilage regeneration (ClinicalTrials.gov Identifier: NCT01041001) under approval of the Korea Food and Drug Administration (KFDA). In their allogeneic reactions, hUCB-MSC is known to be stable immunologically, due to the absence of expression of major histocompatibility complex (MHC) class II and T-cell co-stimulatory factors [1]. In addition, hUCB-MSC suppresses allogeneic responses of lymphocytes under mixed lymphocyte reaction conditions [2]. Therefore, hUCB-MSC transplantation should be considered safe for use in stem cell therapy. Recently, therapeutic effects of hUCB-MSC on brain disease have been reported. For example, glioma-tropism of hUCB-MSC was proposed as a vehicle carrying suicide gene for glioma therapy [3,4]. In addition, hUCB-MSC transplantation improved a Niemann–Pick type C (NP-C) disease mouse model via modulation of neuroinflammation [5]. Collectively, these data suggested that hUCB-MSC should be regarded as a fascinating source for cell therapy for neurological disorders.
* Corresponding author. Fax: +82 2 475 1991. E-mail addresses: [email protected], [email protected] (J.W. Chang).
Alzheimer’s disease (AD) is an incurable neurodegenerative disease that is characterized by loss of neurons by amyloid-b (Ab) neurotoxicity [6]. AD progresses by gliosis and neuronal and synaptic loss, causing progressive loss of memory and cognitive function [7]. Recently, transplantation of rat or mouse neural stem cells (NSC) [8,9] and mouse bone marrow MSC resulted in improved cognition in transgenic AD mice [10]. In this study, we found for the first time that human UCB-MSC has a neuroprotective effect in vitro against Ab42 toxicity via secretion of galectin-3. Galectin-3 (GAL-3), known as a neuroprotective factor, was actively secreted by Ab42 exposure in hUCB-MSC. Our data demonstrated potential therapeutic effects of hUCB-MSC via secretion of beneficial factors, including GAL-3. 2. Materials and methods 2.1. Preparation of hUCB-MSC This study was approved by the Institutional Review Board of MEDIPOST Co., Ltd. Umbilical cord blood was collected from umbilical veins after neonatal delivery with informed consent of the pregnant mothers. MSCs were isolated by separation of mononuclear cells (MNC) using a Ficoll-Hypaque solution (d = 1.077 g/ cm3; Sigma). Following transfer to a-minimum essential medium (a-MEM; Gibco) supplemented with fetal bovine serum (FBS; Gibco), MNCs were seeded in culture flasks at 5 105 cells/cm2. Cells
0014-5793/$36.00 Ó 2010 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2010.07.028
3602
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
maintained in humidified 5% CO2 at 37 °C formed colonies of spindle-shaped cells. At 50% confluence, cells were harvested after treatment with 0.25% (w/v) trypsin–EDTA (Gibco) and were reseeded for expansion [11].
by cell count using i-Solution software (IMT i-Solution INC). Percentage of MAP2-positive cell was calculated by the ratio of number of MAP2-positive cells to DAPI-positive total cell number. Cell viability (%) was analyzed by a ratio of number of green colored cells to total number of cell in a LIVE/DEADÒ staining.
2.2. Cell culture and treatment 2.5. Western blot analysis All procedures for rat primary cortical and hippocampal neuron culture were conducted in accordance with institutional guidelines under the approved protocols. Pregnant Sprague–Dawley rats were purchased from Orientbio (Korea). Brain tissues were dissected from embryonic day 14 rat cortex and hippocampus tissues, and cells were mechanically dissociated in Ca2+/Mg2+-free Hank’s balanced salt solution. Initial cell seeding density was 3 103/cm2. Cells were allowed to proliferate in the presence of 20 ng/ml of basic fibroblast growth factor (bFGF) in serum-free Neurobasal media (Invitrogen). Cell proliferation was maintained in Neurobasal media containing bFGF for 4 days in order to reach 80% confluence prior to induction of differentiation by withdrawal of bFGF. The differentiation was induced in Neurobasal media without B27 supplement for 10 days. B27 supplement was not used in all culture procedures. During differentiation period, Neurobasal media was changed in every 2 days. After 10 days differentiation, cells were stained by anti-MAP2, GFAP, CD11b and nestin antibodies to evaluate percentage of neuron, astrocyte, microglia and neural progenitor cell. Percentage of each cell type was shown in Supplementary Fig. 1. Ab42 treated neuronal cells were cocultured with hUCB-MSC or treated with GAL-3 protein (R&D Biosystems) for 24 h. Rat primary neuronal cells or SH-SY5Y cells were treated with 10 lM of Ab42 (Sigma), and human recombinant GAL-3 for 24 h. Ab42 was solubilized in phosphate-buffered saline (PBS) and the stock was stored at 4 °C for 3 days and divided to aliquots at 80 °C. GAL-3 was dissolved in PBS containing 0.1% bovine serum albumin (BSA) and aliquoted at 80 °C. As a control, MRC5 human fibroblast cells (ATCC) were cultured under culture conditions that were identical to those of hUCB-MSC. hUCB-MSC (8 104 cells/cm2) were cocultured in an upper chamber (pore size: 1 lm) with Ab42-exposed rat primary neuronal cells or SH-SY5Y cells in a Transwell chamber (Falcon). 2.3. Real-time PCR For total RNA isolation, cells were lysed using TRIzol Reagent (Invitrogen). mRNA was reversely transcribed to cDNA. Amplification was performed with the Lightcycler real-time PCR (Roche Diagnostics) system using Lightcycler TaqMan Master mix (Roche), sequence-specific probes (Roche), and the following specific primers: human Gal-3 forward 50 -gagcctaccctgccactg-30 , reverse 50 -aggcaaaggcaggttataagg-30 ; RNA polymerase II forward 50 -cctccaggtggttcagagg-30 , reverse 50 -agttccaacaatggctaccg-30 was used for the control gene. Data analysis of relative gene expression was performed with the DDCt technique using the reference gene as a calibrator. 2.4. Immunostaining and LIVE/DEAD staining for quantitative analysis Cultured cells were fixed in 4% paraformaldehyde/0.15% picric acid in phosphate-buffered saline and were incubated with each anti-antibody. The following primary antibodies were used: tubulin b III (Sigma), microtubule-associated protein 2 (MAP2) (Millipore), glial fibrillary acidic protein (GFAP) (Millipore), nestin (Chemicon) and cleaved caspase-3 (Cell signaling). The image was analyzed using a fluorescent microscope (Nikon) and a confocal microscope (ZEISS, LSM 510 META). For objective assessment of cell death, a LIVE/DEADÒ Viability/Cytotoxicity kit (Invitrogen) was utilized according to instructions. All quantitation was performed
Cell extracts were prepared by ultrasonication (Branson) in buffer containing 9.8 M urea, 4% CHAPS, 130 mM DTT, 40 mM Tris–Cl, and 0.1% SDS. Secreted proteins of conditioned media were precipitated in ice-cold acetone at 20 °C for overnight. After centrifugation, precipitated proteins were solubilized in the same buffer for cell extracts. Protein amount was measured by the Bradford assay (Bio-RAD). The protein extract (20 lg) and precipitated protein from media (15 lg) were loaded on SDS–PAGE for separation of proteins. Proteins were transferred to a nitrocellulose membrane (NC) and then incubated with anti-GAL-3 (R&D Biosystem) or cleaved caspase-3 or caspase-3 (Cell signaling) and b-tubulin (Sigma) antibody. Protein was visualized by Enhanced Chemiluminescence (ECL) (GE Healthcare). The analysis of caspase-3 activation in Western blot was analyzed by QuantityOne software (Bio-RAD). 2.6. siRNA treatment For knock down of GAL-3 expression, control siRNA or siRNA for GAL-3 (Bioneer, Korea) was mixed with Lipofectamine™2000 (Invitrogen). hUCB-MSC were treated with this mixture for 6 h in antibiotic-free media containing 2% FBS. After 6 h, siRNA treated cells were kept for 48 h in serum-free media for evaluation of reduced secretion level of GAL-3. In coculture with Ab42-exposed SH-SY5Y cells, hUCB-MSC treated with each siRNA for 24 h were trypsinized, and detached 8 104 cells/cm2 cells were added to the upper chamber of the Transwell for additional 24 h. 2.7. ELISA For measurement of GAL-3 in conditioned media, media were filtered by syringe filter (0.2 lm, Millipore). Secreted GAL-3 was measured using the DuoSet ELISA Development System (R&D Systems) following a recommended manual. Results were analyzed at 450 nm using a microplate reader, VERSAmax (Molecular Devices).
3. Results 3.1. Expression and secretion of GAL-3 were upregulated by Ab42 in hUCB-MSC Since RT-PCR analysis revealed that the level of GAL-3 mRNA was increased by Ab42 exposure in hUCB-MSC (data not shown), we performed real-time PCR for further analysis. Compared to control cells, the transcriptional level of GAL-3 was upregulated by an average of 6.2-fold in the Ab42-exposed hUCB-MSC group (Fig. 1A). For analysis of the protein level of GAL-3 by Ab42 exposure, cell lysates and conditioned media of hUCB-MSC were collected in a time-dependent manner and analyzed by Western blot analysis. Although GAL-3 does not contain a signal peptide for secretion, secretion of GAL-3 by non-classical pathways has been reported [12]. Intracellular and extracellular GAL-3 was upregulated by Ab42 treatment in a time-dependent manner (Fig. 1B). b-Tubulin of Western blot and stained membrane by Ponceau S were loading control for lysate and media in Western blot analysis, respectively. In order to measure concentration of secreted GAL-3 by Ab42 treatment, the same sample of Fig. 1B was analyzed by ELISA. Concentration of secreted GAL-3 by
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
3603
Fig. 1. Ab42 induced upregulation of GAL-3 expression and secretion in hUCB-MSC. (A) hUCB-MSC was treated with Ab42 for 24 h in serum-free media. Total RNA was isolated from each cell and GAL-3 expression was analyzed by real-time RT-PCR. (B) Cell lysates and concentrated conditioned media were separated by SDS–PAGE and then immunoblotted with anti-GAL-3 antibody. Tubulin was analyzed for evaluation of loading amount in cell lysate. The Ponceau S staining of NC membrane was used for loading control of media. (C) Concentrated each medium was analyzed by ELISA according to a recommended instruction. (D) For analysis of the effect of Ab42 on hUCB-MSC and rat primary neuronal cells, cells exposed to Ab42 for 24 h underwent LIVE/DEAD staining. Red and green colors indicate nuclei of dead cells and live cells, respectively. (E) Cell viability (%) was analyzed by the ratio of the number of green colored cells to the total number of cells. Value represents mean ± SE of three independent experiments. *P < 0.05 vs MSC without Ab42. **P < 0.05 vs neuron without Ab42.
Ab42 was increased to average 0.49 ng/ml at 24 h (Fig. 1C). Although Ab42 induced cell death in rat primary cortical neuronal cell, it had no effect on viability of hUCB-MSC in LIVE/DEAD staining (Fig. 1D and E). A green/red color indicated live cell/nu-
clei of dead cells (Fig. 1D). After LIVE/DEAD staining, percentage of cell viability was quantitated by cell counting (Fig. 1E). These data suggested that Ab42 induced upregulation and secretion of GAL-3 in hUCB-MSC without cell death.
3604
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
3.2. Treatment of GAL-3 protected Ab42-mediated neurotoxicity in rat primary neuronal cells Since GAL-3 has been proposed as a neuronal survival factor [13], we tested whether recombinant GAL-3 treatment rescues Ab42 neurotoxicity or not. Ab42-exposed rat primary cortical neu-
ronal cells underwent simultaneous dose-dependent treatment with recombinant GAL-3. For analysis of surviving neurons under Ab42 neurotoxicity, neurons were selected by immunostaining with anti-MAP2 antibody (Fig. 2A). MAP2-positive neurons were counted for survival as rated by treatment with GAL-3 protein (Fig. 2B). Recombinant GAL-3 significantly protected cell death in
Fig. 2. Treatment of recombinant GAL-3 protected Ab42 neurotoxicity of rat primary neurons. Differentiated neuronal cells were obtained by the procedure described in Materials and Methods. Primary cortical neurons exposed to Ab42 were co-treated with recombinant GAL-3 in a dose-dependent manner. (A) After 24 h, immunostaining with anti-MAP2 antibody was performed for calculation of the percentage of surviving neurons. The image was captured by confocal microscope. Percentage of MAP2positive cells was shown in (B). Value represents mean ± SE of three independent experiments. *P < 0.05 vs Ab42-exposed cortical neurons. (C) Cell extracts were collected from duplicate cells of (A) and analyzed by Western blot analysis with anti-caspase-3. (D) The result of Western blot was quantitated by densitometric analysis. The graph indicated the ratio of cleaved form to full-length capspase-3. Value represents mean ± SE of three-independent experiments. (E) Sucrose (100 mM) and lactose (100 mM) were preincubated with GAL-3 recombinant protein (500 pg/ml) for 20 min. This mixture was treated to SH-SY-5Y cells with Ab42 for 24 h. each cell was stained by LIVE/ DEAD staining and the percentage of dead cell was analyzed by cell counting. **P < 0.05 vs Ab42-exposed cortical neuron. ***P < 0.05 vs Ab42-exposed SH-SY5Y with GAL-3 protein. Ct indicates cortical neurons.
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
3605
Fig. 3. Coculture of hUCB-MSC protected Ab42-exposed rat primary neurons and SH-5Y5Y cells. For evaluation of the neuroprotective effect of hUCB-MSC, hUCB-MSC was cocultured simultaneously with Ab42-exposed rat primary neuronal cells (A–C) or SH-5Y5Y cells, human neuroblastoma cells (D–F) for 24 h, respectively. As a control, nonrelated human fibroblasts were also cocultured. (A) After coculture, rat primary hippocampal neuronal cells were stained by anti-tubulin bIII antibody. (B) Percentage of tubulin bIII positive neuron was calculated. (C) Protection of cleavage of caspase-3 in rat primary cortical neurons by hUCB-MSC was analyzed by immunostaining. In the merged image, the red color indicates cleaved caspase-3 and the blue color shows DAPI staining. (D) In coculture, Ab42-exposed SH-SY5Y cells were cocultured with hUCB-MSC for 24 h and analyzed by LIVE/DEAD staining. (E) Percentage of dead cells was analyzed under each condition. (F) Cell lysates from (D) were analyzed by Western blot analysis with anticaspase-3 antibody. (G) The result of Western blot was quantitated by densitometric analysis. The graph indicated that a ratio of cleaved form to full-length capspase-3. Value represents mean ± SE of three-independent experiments. ***P < 0.05 vs Ab42-exposed SH-SY5Y cells. Hippo and Ct indicated hippocampal and cortical neurons, respectively. Value of (B) and (E) represent mean ± SE of each three-independent experiments. *P < 0.05 vs Ab42-exposed hippocampal neurons. **P < 0.05 vs Ab42-exposed SH-SY5Y.
3606
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
rat primary neurons by Ab42 exposure in a dose-dependent manner. In this experiment, the counting of DAPI-positive cells showed that GAL-3 treatment did not affect cell proliferation (Supplementary Fig. 2). In addition, cleavage of caspase-3 was monitored by
Western blot analysis and a ratio of cleaved form to full-length caspase-3 was analyzed by densitometric analysis. Treatment with GAL-3 reduced activation of caspase-3 in a dose-dependent manner by Ab42 exposure (Fig. 2C and D). In order to know a role of su-
Fig. 4. Knock down of GAL-3 by siRNA reduced neuroprotection effect of hUCB-MSC. (A) To set up conditions for knock down of GAL-3, hUCB-MSC was treated with each siRNA for 48 h in the presence of Ab42, and secreted GAL-3 was measured by ELISA. (B) As a control, hUCB-MSC treated with CON-siRNA or GAL-3 siRNA for 24 h were cocultured simultaneously with Ab42-exposed SH-SY5Y cells for another 24 h in a Transwell chamber. After LIVE/DEAD staining, percentage of surviving cells was calculated. (C) Duplicated cell lysates were analyzed by anti-caspase-3 antibody. The precursor caspase-3 (32 kDa) and the cleaved form (17 kDa) were detected by Western blot analysis. (D) The graph indicated a ratio of cleaved form to full-length capspase-3. Value represent mean ± SE of three-independent experiments. ***P < 0.05 vs Ab42-exposed CON siRNA SH-SY5Y. (E) Duplicated cells were immunostained by anti-cleaved caspase-3 antibody. Red (caspase-3) and blue (DAPI) colors were merged. Value of (A) and (B) represent mean ± SE of each three-independent experiments *P < 0.05 vs CON siRNA and **P < 0.05 vs Ab42-exposed SH-SY5Y.
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
gar-binding domain on GAL-3 for neuroprotection, recombinant GAL-3 (500 pg/ml) was preincubated with sucrose and lactose (each 100 mM). This mixture was treated to Ab42-exposed neuroblastoma cell line, SH-SY5Y cells. If carbohydrate-recognition domain is involved in neuroprotection, lactose and sucrose could be competitive [13]. To analyze percentage of dead cells, cells were stained by LIVE/DEAD staining. In percentage of dead cell, neuroprotection of GAL-3 increased significantly by sugars treatment (Fig. 2E). These data suggested that carbohydrate-recognition domain may participate in negative regulation for neuroprotection on GAL-3. Collectively, these data suggested that GAL-3 could protect Ab42-mediated neurotoxicity. 3.3. Coculture of hUCB-MSC protected Ab42 induced cell death of SHSY5Y and rat primary neurons Since we observed a survival effect of GAL-3 against Ab42 neurotoxicity and GAL-3 was actively secreted from hUCB-MSC by Ab42 exposure, hUCB-MSCs, or, as a control, non-related human lung fibroblast cells, MRC5 were cocultured simultaneously with Ab42-exposed rat primary hippocampal neuronal cells for 24 h cells in a Transwell chamber. Since pore size of the upper chamber is 1 lm, there is no cell contact. However, secreted proteins can pass through pores between hUCB-MSC and primary neuronal cells. Following Ab42 exposure, we examined cell death rate by immunostaining with anti-tubulin bIII (Fig. 3A). Tubulin bIII positive cells were counted for analysis of survival neurons (Fig. 3B). Ab42-exposed rat primary neurons showed typical cell death patterns, including alteration of neurite; however, these cells sustained viability dramatically by coculture of hUCB-MSC. However, MRC5 were not protective against neuronal cell death in a Transwell chamber. In addition, this event was also confirmed with rat primary cortical neuronal cells by immunostaining of anti-caspase-3 (cleaved form). As expected, cleaved caspase-3 (red color) by Ab42 neurotoxicity was reduced significantly by coculturing with hUCB-MSC, but not with MRC5 (Fig. 3C). For further study involving knock down of GAL-3 by siRNA, we conducted the same experiment with SH-SY5Y human neuroblastoma cells because rat primary neuronal cells were mixed neuronal and glial cells. We performed coculture of hUCB-MSC with Ab42-exposed SH-SY5Y cells (Fig. 3D). From LIVE/DEAD staining, we observed similar data on rat primary neuronal cells (Fig. 3E). Cleaved caspase-3 was reduced dramatically by coculturing of hUCB-MSC in Western blot analysis (Fig. 3F and G). These data suggested that hUCB-MSC rescued Ab42 toxicity in cortical and hippocampal neurons. 3.4. Knock down of GAL-3 abolished the neuroprotective effect of hUCB-MSC Since we observed a neuroprotective effect against Ab42-exposed SH-SY5Y cells by hUCB-MSC coculture (Fig. 3) and Ab42 induced GAL-3 expression in hUCB-MSC (Fig. 1), we attempted to inhibit expression of GAL-3 in hUCB-MSC. We tested inhibition of GAL-3 by three different siRNAs, and reduced secretion of GAL-3 was confirmed by ELISA. Three siRNAs demonstrated significant and efficient inhibition of GAL-3 secretion in the presence of Ab42 (Fig. 4A). hUCB-MSC were treated with siRNA1 and the three mixed siRNAs for 24 h, and were then cocultured simultaneously for an additional 24 h with Ab42-exposed SH-SY5Y cells, respectively, for determination of the neuroprotective role of GAL-3 in hUCB-MSC. In LIVE/DEAD staining, control siRNA treated hUCBMSC protected Ab42 neurotoxicity; however, this effect was reduced significantly by GAL-3 siRNA(s) treatment (Fig. 4B). Each duplicated cell was analyzed by immunostaining and Western blot analysis with anti-capspase-3 antibody (Fig. 4C and E). Similar results of LIVE/DEAD staining were confirmed by analysis of caspase-
3607
3. A ratio of cleaved form to full-length capspase-3 was analyzed by densitometric analysis according to Figs. 2 and 3 (Fig. 4D). Above data were confirmed again by the staining of cleaved caspase-3 (Fig. 4E). These data suggested that the neuroprotective effect of hUCB-MSC was mediated via GAL-3.
4. Discussion In the early stage of stem cell therapy, regeneration medicine was the primary interest. In the point of view of regeneration medicine for AD therapy using stem cells, overcoming AD seems impossible due to widespread loss of neurons in the brain. Recently, Block et al. reported on a paracrine effect of stem cells; BM-MSC secreted STC-1 for anti-apoptosis in UV-irradiated skin fibroblasts [14]. In addition, hBM-MSC improved cardiac function after myocardial infarction via secreted TSG-6 [15]. These data emphasized that stem cells are robust sources, not only for regeneration, but also for secretion of survival factors for damaged cells. In our data, we proved that hUCB-MSC secrete beneficial factors, including GAL-3, under pathological conditions such as Ab42 exposure. Although Ab42 induced potential neuronal cell death, Ab42 stimulated secretion of GAL-3 as a neuroprotective factor in hUCB-MSC without cell death (Fig. 1). GAL-3 is a multifunctional protein involved in adhesion, growth, and intracellular and nuclear transcription [16,17]. Although GAL-3 dose not contain signal peptides for secretion, it can be detected outside the cell [12,18]. In particular, functions of the extracellular form of GAL-3 have been reported in cell activation, TCR signaling, cell adhesion, migration, and IL-5 secretion, depending on cell type [19,20]. Secretion of GAL-3 seems to be enhanced by certain stimuli, because a low level of GAL-3 was detected by Western blot analysis in conditioned media of naïve hUCB-MSC (Fig. 1B). However, after Ab42 exposure, GAL-3 was strongly detected in conditioned media of hUCB-MSC. Similar data, such as calcium ionophore induced secretion of GAL-3, has been reported [21]. According to findings from a previous report, when GAL-3 was immobilized as a substratum, GAL-3 promoted notable neural cell adhesion and neurite outgrowth [13]. Furthermore, these activities were inhibited by polyclonal antibody and additive sugar including sucrose and lactose for GAL-3. This paper suggested that GAL-3 is a survival factor for neurons via cell adhesion and neurite outgrowth. In our study, recombinant GAL-3 treatment protected Ab42 neurotoxicity in a dose-dependent manner (Fig. 2). Since GAL-3 contains domain for carbohydrate-binding, we tested the involvement of this domain (Fig. 2E). The rate of cell death by Ab42 was more declined by preincubation of GAL-3 with sugar than that of GAL-3 protein alone. These data indicated that neuroprotection of GAL-3 could be regulated by the binding of intracellular sugar containing protein. Therefore, we should further study using truncated GAL-3 to characterize a critical domain for neuroprotection on GAL-3 and try to find intracellular GAL-3 binding partner containing carbohydrate for negative regulation. Since hUCB-MSC secreted GAL-3 under Ab42, we expected that coculture of hUCB-MSC may rescue Ab42 neurotoxicity. Our expectation was confirmed in rat primary hippocampal and cortical neurons using a coculture system (Fig. 3). Moreover, inhibition of GAL-3 secretion by siRNA reduced the neuroprotective effect of hUCB-MSC with neuronal cells (Fig. 4). In previous report [13], GAL-3 served as a substratum for cell adhesion and neurite outgrowth. Since adhesion cell requires attachment to survive because interaction between receptor and substratum activates intracellular signaling [22]. Under pathological conditions of Ab, secreted GAL-3 from hUCB-MSC should stimulate neuron survival through its counter receptors and support for neural cell adhesion and outgrowth for cell maintenance against Ab42 exposure. Stem cell transplantation has been attempted in
3608
J.-Y. Kim et al. / FEBS Letters 584 (2010) 3601–3608
AD-transgenic mice for evaluation of its possibility for use in stem cell therapy. Lee et al. reported that mouse BM-MSC transplantation reduced Ab deposit and improved memory deficit in an AD mice model [10]. Although several new drugs targeting reduction of Ab level have failed in clinical trials for improvement of declined cognition [23], stem cell transplantation showed a therapeutic effect on memory deficit via Ab removal. This result may be due to simultaneous action of diverse secretory factors from stem cells, including GAL-3, under pathological conditions. When stem cells were transplanted in AD mice, stem cell-derived factors participated not only in Ab removal but also in anti-apoptosis, inflammation, and neurogenesis. Therefore, we can observe combined therapeutic effects in a stem cell transplanted animal study. Currently, we are investigating functions of 14 secreted proteins from hUCB-MSC under Alzheimer’s pathological conditions in order to understand the molecular mechanism of the therapeutic effect by hUCB-MSC transplantation (unpublished data). Among 14 proteins, GAL-3 is one of the secreted proteins with active involvement in anti-apoptosis against Ab42 neurotoxicity. Although neuroprotective effect of GAL-3 alone was almost same with that of hUCBMSC coculture (Fig. 2), cell therapy with hUCB-MSC has several advantages. The first, GAL-3 can protect Ab42 neurotoxicity however it cannot recover neuron loss in AD. Transplanted hUCBMSC differentiated into neuron and then improved neurological deficit in disease model [24–26]. These data suggested that transplanted hUCB-MSC could participate in functional regeneration of damaged area. The second, hUCB-MSC has ability to modulate inflammation. Inflammation is associated with many neurodegenerative diseases including Alzheimer’s disease [27]. Many results suggested that inhibition of inflammation will be able to reverse or slow progression of AD. Human MSC including BM-MSC and UCB-MSC can perform anti-inflammatory action on inflammatory diseases [28–30]. Collectively, we found that hUCB-MSC protected Ab42 neurotoxicity via secretion of GAL-3. Based on these data, we suggest the potential of hUCB-MSC as a new source of stem cell therapy for treatment of Alzheimer’s disease. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.febslet.2010.07.028. References [1] Wang, M., Yang, Y., Yang, D., Luo, F., Liang, W., Guo, S. and Xu, J. (2009) The immunomodulatory activity of human umbilical cord blood-derived mesenchymal stem cells in vitro. Immunology 126, 220–232. [2] Oh, W., Kim, D.S., Yang, Y.S. and Lee, J.K. (2008) Immunological properties of umbilical cord blood-derived mesenchymal stromal cells. Cell Immunol. 251, 116–123. [3] Kim, D.S., Kim, J.H., Lee, J.K., Choi, S.J., Kim, J.S., Jeun, S.S., Oh, W., Yang, Y.S. and Chang, J.W. (2009) Overexpression of CXC chemokine receptors is required for the superior glioma-tracking property of umbilical cord blood-derived mesenchymal stem cells. Stem Cells Dev. 18, 511–519. [4] Kim, S.M., Lim, J.Y., Park, S.I., Jeong, C.H., Oh, J.H., Jeong, M., Oh, W., Park, S.H., Sung, Y.C. and Jeun, S.S. (2008) Gene therapy using TRAIL-secreting human umbilical cord blood-derived mesenchymal stem cells against intracranial glioma. Cancer Res. 68, 9614–9623. [5] Lee, H., Jin, H.K. and Bae, J.S. (2010) Human umbilical cord blood-derived mesenchymal stem cells improve neurological abnormalities of Niemann–Pick type C mouse by modulation of neuroinflammatory condition. J. Vet. Med. Sci. [Epub ahead of print].
[6] Mattson, M.P. (2004) Pathways towards and away from Alzheimer’s disease. Nature 430, 631–639. [7] Smith, M.A. (1998) Alzheimer disease. Int. Rev. Neurobiol. 42, 1–54. [8] Yamasaki, T.R., Blurton-Jones, M., Morrissette, D.A., Kitazawa, M., Oddo, S. and LaFerla, F.M. (2007) Neural stem cells improve memory in an inducible mouse model of neuronal loss. J. Neurosci. 27, 11925–11933. [9] Blurton-Jones, M., Kitazawa, M., Martinez-Coria, H., Castello, N.A., Muller, F.J., Loring, J.F., Yamasaki, T.R., Poon, W.W., Green, K.N. and LaFerla, F.M. (2009) Neural stem cells improve cognition via BDNF in a transgenic model of Alzheimer disease. Proc. Natl. Acad. Sci. USA 106, 13594–13599. [10] Lee, J.K., Jin, H.K., Endo, S., Schuchman, E.H., Carter, J.E. and Bae, J.S. (2010) Intracerebral transplantation of bone marrow-derived mesenchymal stem cells reduces amyloid-beta deposition and rescues memory deficits in Alzheimer’s disease mice by modulation of immune responses. Stem Cell 28, 329–343. [11] Jin, H.J., Nam, H.Y., Bae, Y.K., Kim, S.Y., Im, I.R., Oh, W., Yang, Y.S., Choi, S.J. and Kim, S.W. (2010) GD2 expression is closely associated with neuronal differentiation of human umbilical cord blood-derived mesenchymal stem cells. Cell. Mol. Life Sci. [Epub ahead of print]. [12] Nickel, W. (2003) The mystery of nonclassical protein secretion. A current view on cargo proteins and potential export routes. Eur. J. Biochem. 270, 2109–2119. [13] Pesheva, P., Kuklinski, S., Schmitz, B. and Probstmeier, R. (1998) Galectin-3 promotes neural cell adhesion and neurite growth. J. Neurosci. Res. 54, 639– 654. [14] Block, G.J., Ohkouchi, S., Fung, F., Frenkel, J., Gregory, C., Pochampally, R., DiMattia, G., Sullivan, D.E. and Prockop, D.J. (2009) Multipotent stromal cells are activated to reduce apoptosis in part by upregulation and secretion of stanniocalcin-1. Stem Cells 27, 670–681. [15] Lee, R.H., Pulin, A.A., Seo, M.J., Kota, D.J., Ylostalo, J., Larson, B.L., SemprunPrieto, L., Delafontaine, P. and Prockop, D.J. (2009) Intravenous hMSCs improve myocardial infarction in mice because cells embolized in lung are activated to secrete the anti-inflammatory protein TSG-6. Cell Stem Cell 5, 54–63. [16] Hsu, D.K., Chen, H.Y. and Liu, F.T. (2009) Galectin-3 regulates T-cell functions. Immunol. Rev. 230, 114–127. [17] Yang, R.Y., Hsu, D.K. and Liu, F.T. (1996) Expression of galectin-3 modulates Tcell growth and apoptosis. Proc. Natl. Acad. Sci. USA 93, 6737–6742. [18] Hughes, R.C. (1999) Secretion of the galectin family of mammalian carbohydrate-binding proteins. Biochim. Biophys. Acta 1473, 172–185. [19] Stowell, S.R., Qian, Y., Karmakar, S., Koyama, N.S., Dias-Baruffi, M., Leffler, H., McEver, R.P. and Cummings, R.D. (2008) Differential roles of galectin-1 and galectin-3 in regulating leukocyte viability and cytokine secretion. J. Immunol. 180, 3091–3102. [20] Cortegano, I., del Pozo, V., Cardaba, B., de Andres, B., Gallardo, S., del Amo, A., Arrieta, I., Jurado, A., Palomino, P., Liu, F.T. and Lahoz, C. (1998) Galectin-3 down-regulates IL-5 gene expression on different cell types. J. Immunol. 161, 385–389. [21] Sato, S. and Hughes, R.C. (1994) Control of Mac-2 surface expression on murine macrophage cell lines. Eur. J. Immunol. 24, 216–221. [22] Grossmann, J. (2002) Molecular mechanisms of ‘‘detachment-induced apoptosis–Anoikis”. Apoptosis 7, 247–260. [23] Holmes, C., Boche, D., Wilkinson, D., Yadegarfar, G., Hopkins, V., Bayer, A., Jones, R.W., Bullock, R., Love, S., Neal, J.W., Zotova, E. and Nicoll, J.A. (2008) Long-term effects of Ab42 immunisation in Alzheimer’s disease: follow-up of a randomised, placebo-controlled phase I trial. Lancet 372, 216–223. [24] Lim, J.Y., Park, S.I., Oh, J.H., Kim, S.M., Jeong, C.H., Jun, J.A., Lee, K.S., Oh, W., Lee, J.K. and Jeun, S.S. (2008) Brain-derived neurotrophic factor stimulates the neural differentiation of human umbilical cord blood-derived mesenchymal stem cells and survival of differentiated cells through MAPK/ERK and PI3K/ Akt-dependent signaling pathways. J. Neurosci. Res. 86, 2168–2178. [25] Chung, D.J., Choi, C.B., Lee, S.H., Kang, E.H., Lee, J.H., Hwang, S.H., Han, H., Lee, J.H., Choe, B.Y., Lee, S.Y. and Kim, H.Y. (2009) Intraarterially delivered human umbilical cord blood-derived mesenchymal stem cells in canine cerebral ischemia. J. Neurosci. Res. 87, 3554–3567. [26] Tio, M., Tan, K.H., Lee, W., Wang, T.T. and Udolph, G. (2010) Roles of db-cAMP, IBMX and RA in aspects of neural differentiation of cord blood derived mesenchymal-like stem cells. PLoS One 5, e9398. [27] Glass, C.K., Saijo, K., Winner, B., Marchetto, M.C. and Gage, F.H. (2010) Mechanisms underlying inflammation in neurodegeneration. Cell 14, 918–934. [28] Aggarwal, S. and Pittenger, M.F. (2005) Human mesenchymal stem cells modulate allogeneic immune cell responses. Blood 105, 1815–1822. [29] Oh, J.Y., Kim, M.K., Shin, M.S., Lee, H.J., Ko, J.H., Wee, W.R. and Lee, J.H. (2008) The anti-inflammatory and anti-angiogenic role of mesenchymal stem cells in corneal wound healing following chemical injury. Stem cells 26, 1047–1055. [30] Kim, J.Y., Jeon, H.B., Yang, Y.S., Oh, W. and Chang, J.W. (2010) Application of human umbilical cord blood-derived mesenchymal stem cells in disease models. World J. Stem Cells 2, 34–38.