TRANSLATIONAL AND CLINICAL RESEARCH

Transcription

1 TRANSLATIONAL AND CLINICAL RESEARCH Mesenchymal Stem Cells Shape Microglia Effector Functions Through the Release of CX3CL1 DEBORA GIUNTI, a BENEDETTA PARODI, a CESARE USAI, b LAURA VERGANI, c SIMONA CASAZZA, a SANTINA BRUZZONE, d,e GIANLUIGI MANCARDI, a,e ANTONIO UCCELLI a,e,f a Department of Neurosciences, Ophthalmology and Genetics; b Institute of Biophysics, National Research Council, Genoa, Italy; c DipTERIS, University of Genoa, Genoa, Italy; d Department of Experimental Medicine, Section of Biochemistry; e Center of Excellence for Biomedical Research, University of Genoa, Genoa, Italy; f Advanced Biotechnology Center (ABC), Genoa, Italy Key Words. Mesenchymal stem cells Microglia Chemokines Neuroprotection Immunomodulation Disclosure of potential conflicts of interest is found at the end of this article. ABSTRACT Mesenchymal stem cells (MSC) display a remarkable ability to modulate the immune response and protect the central nervous system mainly through the release of soluble factors in a paracrine fashion, affecting the functional behavior of cells in the tissues. Here we investigated the effect of the interaction between MSC and microglia in vitro, and we dissected the molecular and cellular mechanisms of this crosstalk. We demonstrated that MSC impair microglia activation by inflammatory cues through the inhibition of the expression and release of inflammatory molecules and stress-associated proteins. We showed that MSC significantly increase microglial expression and release of molecules associated with a neuroprotective phenotype such as CX3CR1, nuclear receptor 4 family, CD200 receptor, and insulin growth factor 1. Interestingly, MSC can enhance functional changes on microglia as depicted by the increase of intracellular calcium concentration and phagocytic activity. This last event is associated with an increased expression of triggering receptor expressed on myeloid cells-2, an innate immune receptor involved in phagocytosis in the absence of inflammation. The observed effects on CX3CR1-expressing microglia are due to the release of CX3CL1 by MSC, driven by inflammatory signals, as demonstrated by the reversal of the observed results when CX3CL1 expression was silenced in MSC or its release was blocked. Finally, we showed that exogenous CX3CL1 induce phenotypic and functional changes of microglia similar to those induced by MSC. These findings demonstrate that MSC instruct, through the release of CX3CL1, microglia responsiveness to proinflammatory signals by modulating constitutive calming receptors, typically expressed by steady-state microglia thus switching microglia from a detrimental phenotype to a neuroprotective one. STEM CELLS 2012;30: INTRODUCTION Mesenchymal stem cells (MSC) are stromal progenitors of mesodermal cells also referred as multipotent mesenchymal stromal cells, since their true stemness, that is, on division one daughter cell remains stem and the other is able to replenish a whole tissue compartment, has yet to be demonstrated [1]. In any case, MSC demonstrate classic adult stem cell multipotency in that they are capable of differentiating in vitro and in vivo to all mesenchymal lineages, including adipose, bone, cartilage, muscle, and myelosupportive stroma [2]. MSC initially attracted interest for their ability to differentiate into multiple cellular phenotypes both in culture and in vivo. However, recent observations indicate that, following i.v. administration, most cells are rapidly entrapped in the lungs and only a few engraft into injured tissues where they display negligible transdifferentiation capacity [3 5]. It has been demonstrated that MSC could provide an ideal cell source for repair of injured central nervous system (CNS) due to their striking therapeutic plasticity based on the capacity of modulating the immune response and on a wide range of bystander effects in the target tissues [6]. Indeed MSC regulate effector functions of cells of the adaptive and innate immunity through the release of soluble factors such as prostaglandin E2 (PGE2), nitric oxide (NO), indoleamine 2,3- dioxygenase (IDO), and others [7] leading to the induction of peripheral tolerance [8]. Moreover, MSC can protect axons and improve neuronal survival [4, 9, 10] releasing antiapoptotic [11] or antioxidant molecules [12]. It has been also reported that MSC may foster endogenous neurogenesis [13] and oligodendrogenesis [14 16] not only by direct release of trophic factors but also by stimulating glial cells to secrete Author contributions: D.G. and A.U.: designed research; D.G., B.P., C.U., L.V., S.C., and S.B.: performed research; D.G., B.P., G.M., and A.U.: analyzed data; D.G. and A.U.: wrote the paper; A.U.: provided financial support to the study. Correspondence: Antonio Uccelli, M.D., Department of Neurosciences Ophthalmology and Genetics, University of Genoa, Via De Toni 5, Genoa, Italy. Telephone: þ ; Fax: ; auccelli@neurologia.unige.it Received January 13, 2012; Revised June 6, 2012; accepted for publication June 19, 2012; first published online in STEM CELLS EXPRESS July 20, VC AlphaMed Press /2012/$30.00/0 doi: /stem.1174 STEM CELLS 2012;30:

2 Giunti, Parodi, Usai et al molecules leading to neuron survival and proliferation of the endogenous neural precursor cells [17]. These preclinical studies together indicate that MSC are bestowed with several features leading to amelioration of experimental autoimmune encephalomyelitis (EAE), a model of human multiple sclerosis, through mechanisms leading to inhibition of the pathogenic immune response and support of neural repair chiefly mediated by the release of soluble molecules in a bystander fashion [7]. Several studies showed that microglia play a critical role as resident immunocompetent cells within the CNS. They are highly active in their resting state, continually surveying the CNS microenvironment with extremely motile processes and protrusions. Furthermore, blood brain barrier disruption, brain injuries, systemic infections, and chronic diseases can induce the immediate and focal activation of microglia, switching its behavior from patrolling to shielding of the injured site [18, 19]. Upon activation, microglia produce a variety of effector molecules that have been closely associated with the pathogenesis of multiple sclerosis and neurodegenerative diseases [20 22]. Microglia reacts to infections and other insults activating Toll like receptors (TLRs), such as TLR4, by releasing proinflammatory molecules in the attempt to clear pathogens. Microglia can also be involved in the maintenance of CNS homeostasis phagocytizing apoptotic bodies and cellular debris and concur in the tissue repair through the release of neuroprotective molecules [22]. An attractive approach to treat neurological diseases where inflammation associates with neurodegeneration lays in the possibility of modifying the behavior of microglia switching their functional phenotype from a detrimental to a protective one [22, 23]. Microglia with neuroprotective features have been associated with an increased surface expression of the fractalkine receptor CX3CR1 [24] and may act through the nuclear receptor (NR) 4 family (NURR1) pathway to protect neurons by suppressing inflammatory gene expression [25]. In this study, we addressed the in vitro effect of MSC on microglia, and we dissected the molecular and cellular mechanisms of these interactions demonstrating that MSC can switch microglia from a detrimental behavior dominated by the release of proinflammatory molecules to a neuroprotective phenotype associated with the production of anti-inflammatory and trophic factors. Moreover, we showed that MSC induce functional changes on microglia as depicted by modifications in intracellular calcium concentration and phagocytic activity. Finally, we provided compelling evidence that CX3CL1 released by MSC plays a major role in inducing these beneficial effects on microglia. METHODS MSC In Vitro Culture and Expansion Bone marrow-derived MSC were isolated from 6-8-week-old C57BL/6J mice (Harlan Laboratories, San Pietro al Natisone (UD), Italy, as described elsewhere [8]. In brief, marrow cells, flushed out from tibias and femurs, were plated in 75 cm 2 tissue culture flasks at the concentration of cells per centimetre square using Murine Mesencult as medium (Stem Cell Technologies, Vancouver, BC, Canada, Cells were cultured in plastic plates as adherent cells and kept in a humidified 5% CO 2 incubator at 37 C, refreshing medium every 3 days for approximately 6 weeks when cells reached 80% confluence. Following treatment with 0.05% trypsin solution containing 0.02% EDTA (Euroclone, Milan, Italy, marrow cells were plated in 75 cm 2 flask at the density of cells. Mature MSC, obtained after four to five passages in culture, were defined by the expression of CD9, Sca-1, CD73, and CD44 and the lack of the hematopoietic markers CD45, CD34, and CD11b on their surface. The phenotype was analyzed using the following monoclonal antibodies directed against mouse surface markers: phycoerythrin cyanine 5 (PE Cy5) conjugated rat anti-mouse CD45, PE-conjugated rat anti-mouse Sca-1, purified rat anti-mouse CD9, PE-conjugated rat anti-mouse CD73 (all purchased from BD Pharmingen, San Diego, CA, Fluoroscein isothiocyanate anti-rat IgG1/2a (BD PharMingen) was used as secondary reagent for an indirect staining of CD9 positive cells. Analysis was performed using a FACSCalibur flow cytometer, and data were analyzed with Cell Quest software (Becton Dickinson Immunocytometry Systems, Franklin Lakes, NJ, Microglia Cultures The murine microglial cell line N9, originally developed by Prof. P. Ricciardi-Castagnoli and kindly provided by Prof. E. Zocchi, Genoa, Italy, was obtained by immortalization of E13 mouse embryonic cultures with the 3RV retrovirus carrying an activated v- myc oncogene that is similar to primary microglia in producing substantial amounts of NO and various cytokines after stimulation [26]. Cells were grown in Iscove s Modified Dulbecco s Medium (IMDM) containing 25 mm HEPES and L-glutamine, supplemented with 5% fetal calf serum (FCS), 100 IU/ml penicillin, 100 lg/ml streptomycin, and 50 nm b mercaptoethanol and were cultured in a humidified 5% CO 2 atmosphere at 37 C. Primary microglial cells were prepared from cerebral cortex of newborn C57BL/6J mice. Following removal of meninges, brain tissue was trypsinized for 15 minutes and, after mechanical dissociation, cell suspension was washed and plated on poly-l-lysine coated (10 lg/ml, Sigma Aldrich, St. Louis, MO, flasks (75 cm 2 ). Mixed glial cells were cultured in Basal Medium Eagle (BME) (Invitrogen, Carlsbad, CA, supplemented with 10% FCS (Invitrogen), 2 mm glutamine (Sigma Aldrich), and 100 lm gentamicin sulfate (Sigma Aldrich). After days, microglial cells were harvested from mixed primary glial cultures by mild shaking and collected by centrifugation. Purity of microglia was analyzed by flow cytometry using a CD11b antibody (> 98% surface expression) (R&D Systems, Minneapolis, MN, com). For microglia activation, 1 lg/ml lipopolysaccharide (LPS, Sigma Aldrich) or 10 ng/ml interferon c (IFNc) (R&D Systems) or 2.5 lg/ml non methylated CP oligonucleotides 2006 (CpG 2006) (5 0 -TCGTCGTTTTGTCGTTTTGTCGTT-3 0 ), as described by [27] (Tib Molbiol, Genoa, Italy, was added to the culture medium for 24 hours. N9 and N9-activated cell lines were cocultured for 24 hours in the presence or absence of MSC at 1:1 ratio ( cells) in a six-well plate in a transwell system (BD FALCON Cell Culture Inserts, BD Biosciences, San Diego, CA, Cells Proliferation Cell proliferation was measured by the flow-cytometric detection of the intracellular incorporation of 5,6 carboxyfluorescein diacetate succinimidyl ester (CFSE, Molecular Probes, Invitrogen) LPS-activated N9 cells were incubated for 4 days with 1 lm CFSE in the presence of absence of MSC at the same concentration. After this period, cells were harvested and cell proliferation was assessed by flow cytometry displaying CFSE dilution. RNA Isolation and Real Time Quantification Total RNA was isolated from N9 cells and primary microglia using TRIzol reagent (Invitrogen) according to the manufacturers instructions. First strand cdna was synthesized from 1 lg of total RNA using 60 lm random examer-primer, 20 U Transcriptor Reverse Transcriptase, 20 U Protector RNase Inhibitor, and 1 mm dntps in a final volume of 20 ll (Transcriptor First Strand cdna Synthesis Kit, Roche, Basel, Switzerland, Real-time polymerase chain reaction (PCR) reactions were performed in LightCycler 480 (Roche) in duplicate in a final volume of 20 ll containing 50 ng

3 2046 MSC Lead Microglia to a Neuroprotective Phenotype cdna, 1 ll of each primer pair 20 lm (TIB Mol Biol), 10 ll of LightCycler 480 SYBR Green I Master Mix (Roche). The hypoxanthine-guanine phosphoribosyltransferase (HPRT) was used as housekeeping gene to normalize the expression data. In RT-PCR experiments, we utilized primer pairs for the following genes: inducible Nitric Oxide Synthase (inos), heme oxygenase 1 (HOX1), metallothionein-1 and 2 (MT1 and MT2), poly(adpribose) polymerase-1 (PARP-1), tumor necrosis factor (TNF), interleukin1b (IL1b), insulin growth factor 1 (IGF1), fractalkine receptor (CX3CR1), CD200 receptor (CD200R), nuclear receptor (NR) 4 family (NURR1), prostaglandin E 2 receptor (EP2), prostaglandin E 4 receptor (EP4), triggering receptor expressed on myeloid cells-2 (TREM2), and fractalkine (CX3CL1) (Supporting Information Table S1). The relative gene expression of target genes in comparison with the HPRT reference gene was conducted following the comparative CT threshold method. The normalized expression was then expressed as relative quantity of mrna (fold induction) with respect to the control sample. Cytokines, NO Metabolites, and CX3CL1 Quantification Quantitative analysis of the levels of IGF1, IL1b, and TNF was performed by enzyme-linked immunosorbent assay (ELISA) using commercial available kits (R&D Systems) on supernatants derived from 24 hours transwell cultures. The concentration of nitrate and nitrite (NO metabolites) was determined by using a colorimetric reaction with the Griess reagent (total nitric oxide and nitrite/nitrate parameter assay kit, R&D Systems). Quantitative analysis of CX3CL1 was performed by ELISA utilizing the Quantikine Mouse CX3CL1/fractalkine Immunoassay (R&D Systems) on supernatants derived from MSC cultured for 24 hours in the presence of either 5 ng/ml IFNc or 1 lg/ml LPS and in the presence of both stimuli according to the manufacturer s instructions. In parallel, the quantitative expression of CX3CL1, in MSC activated as previously described, was analyzed by real time PCR. Determination of the Intracellular Cyclic ADP Ribose Levels Adherent N9 cells ( ) were incubated for 24 hours in IMDM with or without (control) LPS in the absence or presence of MSC. Cells were recovered by trypsin treatment, washed once with Krebs Ringers buffer, and cell pellets were lysed in 300 ll of deionized water and neutralized with perchloric acid (0.6 M final concentration). After centrifugation (1,500g for 10 minutes), the cyclic ADP Ribose (cadpr) content was measured on the neutralized cell extracts by means of a highly sensitive enzymatic cycling assay [28] and expressed as pmol/mg of cell protein. Fluorimetric Determination of the Intracellular Calcium Concentration (Ca 21 i) Adherent N9 cells ( ) were cultured on 20-mm diameter coverslips with or without LPS in the presence or absence of the same concentration of MSC for 24 hours. At the end of LPS treatment, cells were incubated with Fura-2-acetoxymethyl ester (Fura 2 AM) for 45 minutes at 37 C, and the Ca 2þ i was measured on an inverted microscope (Zeiss IM35, Zeiss S.p.A., Arese (MI), Italy, as described elsewhere [29]. Phagocytosis Assay N9 cells, at a concentration of ,wereculturedon35-mm glass bottom cell culture dishes with or without LPS in the presence or absence of the same concentration of MSC for 24 hours. At the end of LPS treatment, cells were incubated with 10 ll of red dyed fluoresbrite microspheres (Polysciences, Inc., Eppelheim, Germany, for 30 minutes at 37 C. The phagocytic activity was then stopped by adding 2 ml of icecold phosphate buffered saline (PBS), cells were washed twice with ice-cold PBS, and 2 ml of medium was added. Microglia was analyzed by confocal microscopy (TCS SP5, LEICA Microsystems GmbH, Milan, Italy, and the area of the maximal confocal section of the cells was measured. Phagocytosis was quantified as percent of this area occupied by the microspheres as described elsewhere [30]. Blocking of CX3CL1 Produced by MSC MSC were stably transfected with small interfering RNA (sirna) for CX3CL1 using Lipofectamine 2000 transfection reagent (Invitrogen). The most efficient target sequence for RNA interference was selected among four sequences provided by Qiagen (Hilden, Germany, TTGAAGTTGTT- GATCCTTTA-3 0 ) (Supporting Information Fig. S1). For the Ca 2þ i and phagocytosis experiments requiring a longer blocking of fractalkine-produced by MSC, cells were treated with a CX3CL1 monoclonal neutralizing antibody for 24 hours (antimouse CX3CL1 antibody, R&D Systems). Exogenous Administration of CX3CL1 LPS-activated N9 cells were cultured in the presence of 5 ng/ml recombinant mouse CX3CL1 (R&D Systems) for 24 hours, and the expression of TNF, IL1b, CX3CR1, NURR1, EP2, and TREM2 genes was measured by real time PCR. The Ca 2þ i and phagocytic activity of LPS-activated microglia in the presence of CX3CL1 was measured as described before. Statistical Analysis Statistical analysis was performed on independent samples, namely N9 with or without LPS cultured in the presence of MSC versus cells cultured alone (controls), using unpaired two-tailed Student s t test, which takes into account mean values and standard deviation. Analysis for repeated measures was performed by ANOVA. RESULTS MSC Affect the Expression and Release of Effector Molecules by Microglia We first investigated the effect of MSC on the expression and production of different inflammatory factors produced by microglia when activated by LPS. To this end, we utilized the immortalized murine microglial cell line N9, which has been widely utilized in the literature as cell line capable of producing large amount of cytokines following activation [26]. We observed a significant decrease of inos and TNF mrna expression when LPS-stimulated N9 cells were cultured in the presence of MSC (p <.05 and p <.01, respectively) (Fig. 1). Next, we investigated the influence of MSC on the production of NO metabolites (nitrites and nitrates) and the secretion of TNF by activated microglia. Resting microglia produced relatively small amount of nitrite and nitrate after 24 hours in culture (4 lm and 3 lm, respectively) compared to the amount produced following LPS stimulation (9 12 lm) (Supporting Information Fig. S2). When microglia was activated by LPS in the presence of MSC, production of NO metabolites significantly decreased, similar to the levels observed in resting microglia (5 6 lm, p <.05) (Supporting Information Fig. S2). Similarly, MSC significantly reduced TNF production by activated microglia after 24 hours of coculture (p <.01) (Supporting Information Fig. S2). Surprisingly, the expression and production of the proinflammatory cytokine IL1b by microglia activated in the presence of MSC was significantly upregulated compared to control cells (N9þLPS without MSC) (p <.001 and p <.05, respectively) (Fig. 1 for PCR experiments and Supporting Information Fig. S2 for ELISA). Next, we asked whether MSC could regulate the expression of molecules involved in the cellular stress response such as MTs, HOX1, and PARP, which are upregulated following LPS activation. As shown in Figure 1, the expression of the

4 Giunti, Parodi, Usai et al Figure 1. MSC inhibit the expression of effector molecules by microglia. MSC were cultured with LPS-activated microglia (gray bars) at 1:1 ratio ( cells), and the expression of inos, TNF, IL1b, and markers of stress-related response molecules were analyzed by real time polymerase chain reaction as described in Materials and Methods. Black bars represent LPS-activated N9, and white bars represent microglia in resting condition. Results are shown as mean 6 SD of at least three independent experiments. *, p <.05; **, p <.01 by t test. Abbreviations: HOX1, heme oxygenase 1; IL1b, interleukin1b; inos, inducible nitric oxide synthase; LPS, lipopolysaccharide; MSC, mesenchymal stem cell; MT1 and MT2, metallothionein isoform 1 and isoform 2; PARP, poly(adp-ribose) polymerase-1; TNF, tumor necrosis factor. two non-neuronal MT isoforms, MT1 and MT2, were strikingly downregulated by MSC compared to the levels observed in LPS-activated microglia (p <.001). Similarly, the expression of HOX1 and PARP, a typical marker of cellular damage and apoptosis was strikingly reduced compared to LPS-stimulated control cells (p <.001). In order to assess whether results obtained utilizing the cell line N9 are representative of what could be observed using microglia from primary cultures, we stimulated primary microglia with LPS as described above and measured gene expression and protein release of selected target molecules upon MSC coculture. Similarly to what observed using N9 cells, MSC significantly inhibited the upregulation and release of TNF and enhanced the expression and release of IL1b (Supporting Information Fig. S3). MSC Increase Expression and Release of Neuroprotective Molecules by Microglia We addressed the possible effects of MSC on three receptors associated with microglia neuroprotective phenotype such as CX3CR1, NURR1, and CD200R. As shown in Figure 2, the mrna expression of CX3CR1, NURR1, and CD200R was significantly downregulated in LPS-activated microglia (p <.05), while the presence of MSC significantly upregulated their expression (p <.05). Similar trend was observed for the trophic factor IGF1 that was downregulated in LPS-activated cells and upregulated by coculture with MSC (p <.05). Moreover, this effect of MSC was confirmed by the increased release of IGF1 induced by the presence of MSC (p <.05) as shown in Supporting Information Figure S2. Next, we analyzed the effect of MSC on activated N9 microglia expression of the PGE 2 receptors EP2 and EP4, whose activation have been suggested to promote cell survival in several models of tissue damage [31, 32]. We observed that the expression of EP2 was significantly increased when LPS-activated N9 cells were cultured in the presence of MSC (p <.05, Fig. 2); MSC were also able to upregulate the expression of EP4 but only restoring its expression value to that observed in N9 in resting condition (p <.05, Fig. 2). Similarly to what observed when MSC were cocultured with LPS-activated N9 cells, MSC were also capable of inducing upregulation of the neuroprotective genes CX3CR1, NURR1, and IGF1 in LPS-activated primary microglia (Supporting Information Fig. S3). The similarity of results obtained with primary microglia and the N9 cell line suggests that the latter is a robust and reproducible tool to evaluate the effect of MSC on microglia. Therefore, from now on, experiments were carried out exclusively on N9 cells. MSC Modulate the Profile of Genes Expression of Microglia when Activated by CpG and IFNc Next, we assessed whether MSC can have an impact also on gene expressed by microglia when activated by other proinflammatory stimuli such as IFNc, a cytokine known to enhance microglial phagocytosis [33] and CpG, a TLR9 agonist, which regulates the microglia immune response [34]. Overall, MSC displayed a similar ability to downregulate TNF, inos, and HOX1 expressed by microglia also when activated by IFNc and CpG. Similarly to what observed when microglia was activated by LPS, MSC were able to significantly upregulate the expression of IGF1, CX3CR1, and NURR1 (Supporting Information Fig. S4). These results demonstrate that MSC can act on functions of microglia regardless the way it is activated. MSC Induce Functional Changes on Microglia Without Affecting Its Proliferation Capacity In order to assess whether MSC could affect the activation state of microglia, we investigated the proliferative activity, intracellular calcium concentration, and phagocytic capacity of N9 cells stimulated with LPS in the absence or in the presence of MSC. MSC were not able to interfere with proliferation of LPS-activated N9 cells as demonstrated by the fact

5 2048 MSC Lead Microglia to a Neuroprotective Phenotype Figure 2. MSC increase microglia expression of neuroprotective molecules. MSC were cultured with LPS-activated microglia (gray bars) at 1:1 ratio ( cells), and the expression of CX3CR1, NURR1, CD200R, IGF1, EP2, and EP4 was quantified by real time polymerase chain reaction. MSC (gray bars) induced a significant upregulation of neuroprotective molecules on activated microglia compared to control (N9þLPS, black bars). MSC were also able to upregulate the expression of EP4 but only restoring its expression value to that observed in N9 in resting condition (white bars). Results are shown as mean 6 SD of at least three independent experiments. *, p <.05 by t test. Abbreviations: CD200R, CD200 receptor; CX3CR1, fractalkine receptor; EP2, prostaglandin E 2 receptor; EP4, prostaglandin E 4 receptor; IGF1, insulin growth factor 1; LPS, lipopolysaccharide; MSC, mesenchymal stem cell; NURR1, nuclear receptor 4 family. Figure 3. MSC induce functional changes in activated microglia. MSC significantly increased the intracellular Ca 2þ concentration in LPS-activated microglia (gray bar) compared to the levels measured in resting (white bar) and LPS-activated (black bar) microglia as determined by fluorimetric quantification (left upper panel). Coculture with MSC also enhanced phagocytic activity of LPS-activated microglia (gray bar) compared to activated microglia only (black bar) and resting microglia (white upper bar) as measured at confocal microscopy by incorporation of red dyed fluoresbrite microspheres (middle panel and lower panels). The expression of TREM2, quantified by real time PCR, was increased by the coculture of activated microglia with MSC (gray bars) compared to control activated N9 cells (black bars) and resting microglia (white bar) (right upper panel). Results are shown as mean 6 SD of at least three independent experiments. The images were representative of results obtained in three independent experiments. *, p <.01; **, p <.05 by t test. Abbreviations: LPS, lipopolysaccharide; MSC, mesenchymal stem cell; TREM2, triggering receptor expressed on myeloid cells-2. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] that microglia proliferated at 96 hours at the same rate observed in the presence or absence of MSC (Supporting Information Fig. S5). As indicated in Figure 3, MSC significantly increased the basal Ca 2þ concentration of LPS-activated microglia and were able to enhance their phagocytic activity (p <.001). cadpr is a potent and universal Ca 2þ mobilizer, which was demonstrated to mediate the LPS-induced Ca 2þ increase in N9 cells [35]. In line with the finding that MSC increased the Ca 2þ i of LPS-activated microglia, we observed a significant increase of the intracellular concentration of cadpr when LPS-activated N9 were cultured in the presence of MSC compared to control (0.12 pmol/mg vs pmol/mg, p <.05). Then, we addressed whether MSC can affect the expression on activated microglia of TREM2, which was recently demonstrated to play a role in the maintenance of CNS homeostasis without inflammation [30]. As shown in Figure 3, we detected a

6 Giunti, Parodi, Usai et al Figure 4. Silencing of CX3CL1 in MSC modulates the expression of effector genes in microglia upon activation. MSC were transfected with small interfering RNA for CX3CL1 for 24 hours and utilized in the culture with N9 in the presence of LPS as described in Materials and Methods. The gene expression of TNF, CX3CR1, NURR1, IL1b, EP2, and IGF1 was measured by real time polymerase chain reaction on LPS-activated N9 in the presence of MSC (light gray bars) or in the presence of MSC silenced for CX3CL1 (dark gray bars). White bars represent microglia in resting condition and black bars microglia activated with LPS. Results are shown as mean 6 SD of three independent experiments. *, p <.05; **, p <.01 by t test. Abbreviations: CX3CR1, fractalkine receptor; EP2, prostaglandin E 2 receptor; IL1b, interleukin1b; IGF1, insulin growth factor 1; LPS, lipopolysaccharide; MSC, mesenchymal stem cell; NURR1, nuclear receptor 4 family; TNF, tumor necrosis factor. significant upregulation of TREM2 expression by activated microglia cultured in the presence of MSC (p <.05). All together these findings suggest that MSC exposure impacts on microglia activation state affecting intracellular Ca 2þ signaling and phagocytic capacity. MSC Exert Beneficial Effect on Microglia Through the Release of CX3CL1 We next investigated the role of the chemokine fractalkine/ CX3CL1 and its specific receptor CX3CR1 that are constitutively expressed in several region of the CNS and are reported to mediate neuron microglial interaction, synaptic transmission, and neuronal protection from toxic insults [36]. Recent data showed that augmenting CX3CR1 signaling protect against microglial neurotoxicity [24]. As we observed that MSC can enhance CX3CR1 expression in LPS-stimulated microglia (Fig. 2), we hypothesized that MSC could exert their effect on microglia through the release of CX3CL1. We demonstrated that MSC express and release only minimal levels of CX3CL1 in the steady state condition but expression and release were significantly augmented following stimulation with two proinflammatory molecules such as IFNc and LPS (Supporting Information Fig. S5). To address the possibility that MSC-produced CX3CL1 has a major impact on microglia functions, we examined the effect of CX3CL1 silencing in MSC using the sirna technique. Indeed, when we utilized MSC transfected with sirna for CX3CL1 in the coculture with activated microglia, we detected a partial restoration of TNF expression compared to LPS-stimulated N9 in the presence of MSC (p <.01) (Fig. 4). More importantly, silencing of CX3CL1 resulted in a striking inhibition of the upregulation observed for those molecules such as IL1b, CX3CR1, NURR1, and EP2 whose expression was increased

7 2050 MSC Lead Microglia to a Neuroprotective Phenotype Figure 5. CX3CL1 secreted by MSC can induce functional changes on activated microglia. MSC were cultured with the same concentration of N9 cells ( cells) in the presence of 1 lg/ml of LPS and with (dark gray bars) or without (gray bars) 10 lg/ml of a blocking monoclonal anti-cx3cl1 antibody for 24 hours. Ca 2þ i, phagocytic activity and TREM2 mrna expression were measured as described in Materials and Methods. Blocking of release of CX3CL1 by MSC (dark gray bar) significantly inhibited the increase of intracellular Ca 2þ concentration observed when microglia was activated with LPS in the presence of MSC (light gray bar) as determined by fluorimetric quantification (left upper panel). Similarly, CX3CL1 blockade on MSC in coculture with N9 (dark gray bar) reverted the enhanced phagocytic activity (middle panel and lower panels) and the upregulation in the expression of TREM2, quantified by real time polymerase chain reaction (Right upper panel), of LPSactivated microglia observed in the presence of MSC (light gray bar). Control LPS-activated N9 cells are depicted as black bars while resting microglia is shown as white bar. Results are shown as mean 6 SD of at least three independent experiments. *, p <.01; **, p <.05 by t test. The images were representative of results obtained in three independent experiments. Abbreviations: acx3cl1, anti-cx3cl1 antibody; LPS, lipopolysaccharide; MSC, mesenchymal stem cell; TREM2, triggering receptor expressed on myeloid cells-2. in activated microglia following coculture with MSC (Fig. 4). In contrast, IGF1 was the only gene for which silencing of CX3CL1 in MSC did not vanished the effect of wild-type MSC on LPS-activated microglia (p <.01, Fig. 4), possibly suggesting that, in microglia, the IGF1 pathway is CX3CL1 independent. To further validate these results, we assessed whether blocking of CX3CL1 could functionally affect microglia activation utilizing a monoclonal anti-mouse CX3CL1 antibody (R&D Systems), as described in Materials and Methods. We demonstrated that blockade of CX3CL1 on MSC treated with the anti-cx3cl1 antibody for 24 hours resulted in the downregulation of the basal Ca 2þ concentration (p <.001), phagocytic activity (p <.001), and TREM2 RNA expression (p <.05) of stimulated microglia in the presence of treated MSC (Fig. 5). Exogenous CX3CL1 Mimics the Effects on N9 Cells Observed Following Microglia Exposure to MSC In order to confirm the key role of CX3CL1 in modulating function and activity of the LPS-activated microglia, we repeated some of the above-reported experiments by adding the recombinant CX3CL1 protein (R&D Systems) directly to microglia cultures during LPS treatment. The results reported in Figure 6 demonstrated that, in the presence of a concentration of CX3CL1 similar to that produced by IFNc and LPSactivated MSC (5 ng/ml, Supporting Information Fig. S5), LPS-stimulated microglia expressed lower levels of TNF compared to control cells (N9þLPS) (p <.05). Conversely, when we measured the mrna expression of both IL1b and of the neuroprotective genes under analysis such as CX3CR1, NURR1, and EP2, we demonstrated that CX3CL1 led to the upregulation of all genes as we observed when LPS-activated microglia was cultured in the presence of MSC (p <.05). Finally, we showed that CX3CL1 can affect the functional state of microglia, as shown by the significant increase of the Ca 2þ i (p <.001), phagocytic activity (p <.001), and TREM2 RNA expression on LPS-activated microglia (p <.05) (Fig. 6). All together these findings demonstrate that exogenous CX3CL1 can reproduce most of the MSC-mediated effects observed on activated microglia. DISCUSSION In this article, we demonstrate that MSC, a subset of progenitor cells displaying a remarkable ability to modulate effector functions of cells of the innate and adaptive immunity [7], can affect several features of microglia following activation with proinflammatory stimuli. Microglia, the resident macrophages of the CNS, are exquisitely sensitive to brain injury and disease, altering their morphology and phenotype in response to brain injuries and contributing to the maintenance of brain homeostasis [21] [22]. Upon activation, microglia have the capacity to release a large number of substances that can act detrimental or beneficial for the surrounding cells [37]. Various cytokines, inflammatory and also neuronal signaling molecules have been reported to actively control microglia function, thus instructing microglia toward a beneficial or detrimental phenotype. First, we observed that MSC reduce the increase of the expression of TNF, inos and oxidative stress-associated proteins, induced by LPS and other proinflammatory molecules, by microglia thus supporting their ability to inhibit the release of proinflammatory molecules typically occurring following classic activation as response to pathogens or during chronic brain inflammation. Second, we demonstrated that MSC can induce a significant upregulation of surface molecules playing a calming influence on microglia

8 Giunti, Parodi, Usai et al Figure 6. Exogenous CX3CL1 mimics the effect of mesenchymal stem cell on activated microglia. 5 ng/ml of recombinant CX3CL1 was added to LPS-activated N9 cells (gray bars), and the expression of TNF, IL1b, CX3CR1, NURR1, EP2, and TREM2 in microglia was analyzed by real time polymerase chain reaction while Ca 2þ i and phagocytic activity were measured as described in Materials and Methods and compared to LPS-activated N9 (black bars) and resting microglia (white bars). Results are shown as mean 6 SD of at least three independent experiments. *, p <.05; **, p <.05 by t test. Abbreviations: CX3CR1, fractalkine receptor; EP2, prostaglandin E 2 receptor; IL1b, interleukin1b; LPS, lipopolysaccharide; NURR1, nuclear receptor 4 family; TNF, tumor necrosis factor; TREM2, triggering receptor expressed on myeloid cells-2. and associated with a neuroprotective phenotype. Among these, CD200R expressed on microglia negatively regulates effector functions on myeloid cells [38], while NURR1 has been reported to suppress inflammatory genes expression in microglia [25]. Cardona et al. [24] reported that augmenting CX3CR1 signaling, through CX3CL1 binding, decreases microglia neurotoxicity. Recent data suggested also that activation of the PGE2 receptor EP2 promote cell survival in several models of tissue damage [31]. Interestingly, we detected a significant upregulation of IL-1b, a cytokine not only often

9 2052 MSC Lead Microglia to a Neuroprotective Phenotype involved in the inflammatory response but also considered to play a key role in promoting repair of the CNS, possibly inducing the synthesis and release of IGF by glial cells [39, 40]. The reported upregulation of the above-mentioned molecules suggests that MSC can instruct microglia responsiveness to proinflammatory signals such as LPS, IFNc, and CpG, by modulating constitutive calming receptors, typically expressed by steady-state microglia [22]. Of relevance, we demonstrated that MSC also induce functional modifications of microglia as depicted by the detection of enhanced intracellular calcium concentration and phagocytic activity without affecting the proliferation capacity of microglia. These results suggest that MSC appear to act on the ability of microglia to activate following LPS triggering and subsequently enter their executive phase by dissociating their capacity of releasing proinflammatory molecules from their phagocytic activity. This last event is associated with the acquisition of a neuroprotective phenotype by microglia as further confirmed by the upregulated expression of TREM2, an innate immune receptor expressed by microglia involved in phagocytosis in the absence of inflammation [30]. More importantly, we demonstrated that, following stimulation with LPS and IFNc, two molecules involved in CNS inflammatory events, MSC modulate microglia effector functions through the release of CX3CL1. Several soluble immunosuppressive factors have been reported to be involved in MSC-mediated immunoregulation, either produced constitutively by MSC or released following crosstalk with target cells. Examples of molecules belonging to the latter group are IDO and PGE2, which are released by MSC when licensed by IFNc produced by target cells [41, 42]. The ability of CX3CL1 released by MSC to promote the switch of microglia from a detrimental, neurotoxic phenotype dominated by the release of proinflammatory molecules to a beneficial, neuroprotective phenotype associated with the production of anti-inflammatory and trophic factors, enhanced expression of calming receptors, increased intracellular calcium concentrations, and phagocytosis was confirmed by experiments where CX3CL1 was added exogenously to microglia activated with LPS in the absence of MSC. These results are in line with previous studies reporting on the ability of CX3CR1 expressing microglia to exert striking neuroprotective functions through the binding to CX3CL1 released by neurons [24], which leads to enhancing intracellular calcium concentration [36] and increased phagocytic activity [43]. These data demonstrate that MSC, through the release of CX3CL1, may enable microglia to dissociate phagocytic activity from the release of proinflammatory molecules [44] thus enhancing its homeostatic role in the maintenance of a healthy CNS, possibly through an efficient removal of apoptotic cells and phagocytosis of cellular debris and fostering a proregenerative environment within the CNS [30, 43]. These results are also in line with the current evidence that MSC act mainly through the release of soluble factors in a paracrine fashion, possibly affecting the functional behavior of tissue-resident cells. Interestingly, MSC can induce dendritic cells to acquire a tolerogenic phenotype in vitro and in vivo [45] and educate alternatively activated macrophages [46] which have been demonstrated to be essential players in CNS healing [47]. Recently, Zhou et al. [48] demonstrated that MSC are able to inhibit LPS-stimulated microglia activation and the production of inflammatory factors through diffusible molecules and, within an inflammatory environment, can significantly increase the production of neurotrophic factors, possibly involved in tissue repair. In addition, MSC exert neuroprotective effects on dopaminergic neurons via an antiinflammatory mechanism mediated by the modulation of microglia activation [49]. More importantly, a recent study demonstrated that MSC alternatively activate microglia, thus enhancing its migratory features and ability of eliminating amyloid b[50]. These findings demonstrate that MSC are capable of functionally affecting myeloid cells behavior including microglia which is induced toward a protective phenotype, possibly accounting for some of the beneficial effects that have been associated with the therapeutic plasticity of MSC in EAE and other experimental models of neurological diseases [51]. Overall these results further confirm the incumbent translation of MSC cell therapy from experimental models to human neurological diseases where microglia may play a major pathogenic role [52]. ACKNOWLEDGMENTS This study was supported by grants to A.U. from the Italian Foundation for Multiple Sclerosis (FISM), the Italian Ministry of Health (Ricerca Finalizzata), the Regione Liguria, the Italian Ministry of University and Scientific Research (MIUR), the Progetto Limonte, and the Fondazione CARIGE. We thank Dr. E. Polazzi for technical assistance on Primary microglial cells isolation. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST The authors indicate no potential conflict of interests. REFERENCES 1 Javazon EH, Beggs KJ, Flake AW. Mesenchymal stem cells: Paradoxes of passaging. Exp Hematol 2004;32: Pittenger MF, Mackay AM, Beck SC et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284: Lee RH, Pulin AA, Seo MJ et al. 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 2009;5: Gerdoni E, Gallo B, Casazza S et al. Mesenchymal stem cells effectively modulate pathogenic immune response in experimental autoimmune encephalomyelitis. Ann Neurol 2007;61: Phinney DG, Prockop DJ. Concise review: Mesenchymal stem/multipotent stromal cells: The state of transdifferentiation and modes of tissue repair Current views. Stem Cells 2007;25: Uccelli A, Pistoia V, Moretta L. Mesenchymal stem cells: A new strategy for immunosuppression? Trends Immunol 2007;28: Uccelli A, Moretta L, Pistoia V. Mesenchymal stem cells in health and disease. Nat Rev Immunol 2008;8: Zappia E, Casazza S, Pedemonte E et al. Mesenchymal stem cells ameliorate experimental autoimmune encephalomyelitis inducing T cell anergy. Blood 2005;106: Zhang J, Li Y, Lu M et al. Bone marrow stromal cells reduce axonal loss in experimental autoimmune encephalomyelitis mice. J Neurosci Res 2006;84: Kassis I, Grigoriadis N, Gowda-Kurkalli B et al. Neuroprotection and immunomodulation with mesenchymal stem cells in chronic experimental autoimmune encephalomyelitis. Arch Neurol 2008;65:

10 Giunti, Parodi, Usai et al Crigler L, Robey RC, Asawachaicharn A et al. Human mesenchymal stem cell subpopulations express a variety of neuro-regulatory molecules and promote neuronal cell survival and neuritogenesis. Exp Neurol 2006;198: Lanza C, Morando S, Voci A et al. Neuroprotective mesenchymal stem cells are endowed with a potent antioxidant effect in vivo. J Neurochem 2009;110: Munoz JR, Stoutenger BR, Robinson AP et al. Human stem/progenitor cells from bone marrow promote neurogenesis of endogenous neural stem cells in the hippocampus of mice. Proc Natl Acad Sci USA 2005;102: Akiyama Y, Radtke C, Kocsis J. Remyelination of the rat spinal cord by transplantation of identified bone marrow stromal cells. J Neurosci 2002;22: 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: Bai L, Lennon DP, Eaton V et al. Human bone marrow-derived mesenchymal stem cells induce Th2-polarized immune response and promote endogenous repair in animal models of multiple sclerosis. Glia 2009;57: Li Y, Chen J, Chen XG et al. Human marrow stromal cell therapy for stroke in rat: Neurotrophins and functional recovery. Neurology 2002; 59: Nimmerjahn A, Kirchhoff F, Helmchen F. Resting microglial cells are highly dynamic surveillants of brain parenchyma in vivo. Science 2005;308: Rivest S. Regulation of innate immune responses in the brain. Nat Rev Immunol 009;9: Jack C, Ruffini F, Bar-Or A, Antel JP. Microglia and multiple sclerosis. J Neurosci Res 2005;81: Ransohoff RM, Perry VH. Microglial physiology: Unique stimuli, specialized responses. Annu Rev Immunol 2009;27: Saijo K, Glass CK. Microglial cell origin and phenotypes in health and disease. Nat Rev Immunol 2011;11: Schwartz M, Butovsky O, Bruck W et al. Microglial phenotype: Is the commitment reversible? Trends Neurosci 2006;29: Cardona AE, Pioro EP, Sasse ME et al. Control of microglial neurotoxicity by the fractalkine receptor. Nat Neurosci 2006;9: Saijo K, Winner B, Carson CT et al. A Nurr1/CoREST pathway in microglia and astrocytes protects dopaminergic neurons from inflammation-induced death. Cell 2009;137: Righi M, Mori L, De Libero G et al. Monokine production by microglial cell clones. Eur J Immunol 1989;19: Jarrossay D, Napolitani G, Colonna M et al. Specialization and complementarity in microbial molecule recognition by human myeloid and plasmacytoid dendritic cells. Eur J Immunol 2001;31: Graeff R, Lee HC. A novel cycling assay for cellular cadp-ribose with nanomolar sensitivity. Biochem J 2002;361(pt 2): Zocchi E, Daga A, Usai C et al. Expression of CD38 increases intracellular calcium concentration and reduces doubling time in HeLa and 3T3 cells. J Biol Chem 1998;273: Takahashi K, Rochford CD, Neumann H. Clearance of apoptotic neurons without inflammation by microglial triggering receptor expressed on myeloid cells-2. J Exp Med 2005;201: Jiang J, Ganesh T, Du Y et al. Neuroprotection by selective allosteric potentiators of the EP2 prostaglandin receptor. Proc Natl Acad Sci USA 2010;107: Liu D, Wu L, Breyer R et al. Neuroprotection by the PGE2 EP2 receptor in permanent focal cerebral ischemia. Ann Neurol 2005;57: Chan A, Magnus T, Gold R. Phagocytosis of apoptotic inflammatory cells by microglia and modulation by different cytokines: Mechanism for removal of apoptotic cells in the inflamed nervous system. Glia 2001;33: Butchi NB, Du M, Peterson KE. Interactions between TLR7 and TLR9 agonists and receptors regulate innate immune responses by astrocytes and microglia. Glia 2010;58: Franco L, Bodrato N, Moreschi I et al. Cyclic ADP-ribose is a second messenger in the lipopolysaccharide-stimulated activation of murine N9 microglial cell line. J Neurochem 2006;99: Harrison JK, Jiang Y, Chen S et al. Role for neuronally derived fractalkine in mediating interactions between neurons and CX3CR1- expressing microglia. Proc Natl Acad Sci USA 1998;95: Kettenmann H, Hanisch UK, Noda M, Verkhratsky A. Physiology of microglia. Physiol Rev 2011;91: Jenmalm MC, Cherwinski H, Bowman EP et al. Regulation of myeloid cell function through the CD200 receptor. J Immunol 2006;176: Fan J, Wojnar MM, Theodorakis M et al. Regulation of insulin-like growth factor (IGF)-I mrna and peptide and IGF-binding proteins by interleukin-1. Am J Physiol 1996;270(3 pt 2):R Mason JL, Suzuki K, Chaplin DD et al. Interleukin-1beta promotes repair of the CNS. J Neurosci 2001;21: Krampera M, Cosmi L, Angeli R et al. Role for interferon-gamma in the immunomodulatory activity of human bone marrow mesenchymal stem cells. Stem Cells 2006;24: Aggarwal S, Pittenger MF. Human mesenchymal stem cells modulate allogeneic immune cell responses. Blood 2005;105: Noda M, Doi Y, Liang J et al. Fractalkine attenuates excito-neurotoxicity via microglial clearance of damaged neurons and antioxidant enzyme heme oxygenase-1 expression. J Biol Chem 2011;286: Hughes MM, Field RH, Perry VH et al. Microglia in the degenerating brain are capable of phagocytosis of beads and of apoptotic cells, but do not efficiently remove PrPSc, even upon LPS stimulation. Glia 2010;58: Chiesa S, Morbelli S, Morando S et al. Mesenchymal stem cells impair in vivo T-cell priming by dendritic cells. Proc Natl Acad Sci USA 2011;108: Kim J, Hematti P. Mesenchymal stem cell-educated macrophages: A novel type of alternatively activated macrophages. Exp Hematol 2009; 37: Shechter R, London A, Varol C et al. Infiltrating blood-derived macrophages are vital cells playing an anti-inflammatory role in recovery from spinal cord injury in mice. PLoS Med 2009;6:e Zhou C, Zhang C, Chi S et al. Effects of human marrow stromal cells on activation of microglial cells and production of inflammatory factors induced by lipopolysaccharide. Brain Res 2009;1269: Kim YJ, Park HJ, Lee G et al. Neuroprotective effects of human mesenchymal stem cells on dopaminergic neurons through anti-inflammatory action. Glia 2009;57: Lee JK, Schuchman EH, Jin HK et al. Soluble CCL5 derived from bone marrow-derived mesenchymal stem cells and activated by amyloid-beta ameliorates Alzheimer s disease in mice by recruiting bone marrow-induced microglia immune responses. Stem Cells 2012;30: Uccelli A, Mancardi G. Stem cell transplantation in multiple sclerosis. Curr Opin Neurol 2010;23: Uccelli A, Laroni A, Freedman MS. Mesenchymal stem cells for the treatment of multiple sclerosis and other neurological diseases. Lancet Neurol 2011;10: See for supporting information available online.