Neuroprotective erythropoietin attenuates microglial activation, including morphological changes, phagocytosis, and cytokine production

Accepted Manuscript Research report Neuroprotective erythropoietin attenuates microglial activation, including morphological changes, phagocytosis, and cytokine production Tetsuya Tamura, Mineyoshi Aoyama, Seiko Ukai, Hiroki Kakita, Kazuya Sobue, Kiyofumi Asai PII: DOI: Reference:

S0006-8993(17)30093-8 http://dx.doi.org/10.1016/j.brainres.2017.02.023 BRES 45293

To appear in:

Brain Research

Received Date: Revised Date: Accepted Date:

4 November 2016 27 January 2017 21 February 2017

Please cite this article as: T. Tamura, M. Aoyama, S. Ukai, H. Kakita, K. Sobue, K. Asai, Neuroprotective erythropoietin attenuates microglial activation, including morphological changes, phagocytosis, and cytokine production, Brain Research (2017), doi: http://dx.doi.org/10.1016/j.brainres.2017.02.023

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Research Report Neuroprotective erythropoietin attenuates microglial activation, including morphological changes, phagocytosis, and cytokine production Tetsuya Tamuraa,b, Mineyoshi Aoyamaa,c,*, Seiko Ukaia, Hiroki Kakitaa,d, Kazuya Sobueb and Kiyofumi Asaia a

Department of Molecular Neurobiology, Nagoya City University Graduate School of

Medical Sciences, Nagoya, Japan b

Department of Anesthesiology and Intensive Care Medicine, Nagoya City University

Graduate School of Medical Sciences, Nagoya, Japan c

Department of Pathobiology, Nagoya City University Graduate School of

Pharmaceutical Sciences, Nagoya, Japan d

Department of Perinatal and Neonatal Medicine, Aichi Medical University, Nagakute,

Japan

*Corresponding author: Mineyoshi Aoyama M.D., Ph.D. Department of Pathobiology Nagoya City University Graduate School of Pharmaceutical Sciences 3-1, Tanabe-dori, Mizuho-ku, Nagoya 467-8603, Japan Phone: +81-52-836-3451

Fax: +81-52-836-3454

E-mail: [email protected]

1

E-mail addresses: [email protected] (T. Tamura), [email protected] (M. Aoyama), [email protected] (S. Ukai) [email protected] (H. Kakita) [email protected] (K. Sobue) [email protected] (K. Asai)

Abstract Erythropoietin (EPO), a hematopoietic hormonal cytokine induced in response to hypoxia, has neuroprotective effects. EPO receptor (EPOR) is expressed in microglia, resident immune cells in the brain. However, the effect of EPO on microglial activation is not clear. In the present study, we demonstrated that the EPOR is highly expressed in microglia, rather than in neurons or astrocytes, in in vitro experiments. Therefore, we investigated whether EPO could attenuate lipopolysaccharide (LPS)-mediated activation of microglia in vitro. The BV-2 microglial cell line was treated with LPS in the absence or presence of EPO. In the presence of EPO, microglial expression of LPS-induced inflammatory cytokine genes was significantly decreased. In addition, EPO suppressed the LPS-induced phagocytic activity of BV-2 cells towards fluorescent beads, as well as induction of inducible nitric oxide synthase. In in vivo experiments, EPO significantly 2

decreased the LPS-induced expression of inflammatory cytokine genes in mouse brains. Furthermore, morphological analysis of cortical microglia in the brains of mice stimulated with LPS revealed that combined treatment with EPO alleviated LPS-induced morphological changes in the microglia. These data indicate that EPO attenuates microglial activation, including morphological changes in vivo, phagocytosis in vitro, and the production of inflammatory cytokines in vivo and in vitro. Further investigation of EPO modulation of LPS-induced microglial activation may contribute to the development of novel neuroprotective therapies.

Highlights • Erythropoietin (EPO) has a neuroprotective effect. • EPO receptor (EPOR) expression is higher in microglia than in neurons. • EPO treatment attenuates microglial activation. • The neuroprotection of EPO may be mediated by regulation of microglial activation.

Keywords: erythropoietin (EPO); neuroprotective effect; microglial activation; phagocytosis; inflammatory cytokines; morphological changes.

1. Introduction Erythropoietin (EPO) is a 34 kDa glycoprotein hormone that is induced by hypoxia. The main site of EPO production is the fetal liver and the adult kidney. EPO supports the proliferation and differentiation of erythroid progenitor cells for survival 3

(Lin et al., 1996; Wu et al., 1995). EPO is also expressed in the brain. At the cellular level, intrinsic EPO expression is detected in neurons and astrocytes, and primarily in astrocytes (Masuda et al., 1994; Ruscher et al., 2002). The EPO receptor (EPOR) is expressed in the brain, and provides a signaling mechanism for the neuroprotective role of EPO in ischemic diseases and neurodegenerative diseases (Juul et al., 1998; Marti et al., 1996; Masuda et al., 1993; Morishita et al., 1997). At the cellular level, the EPOR is expressed in human and rodent neurons, astrocytes, and microglia (Chin et al., 2000; Nagai et al., 2001; Noguchi et al., 2007), and we reported previously that EPOR was expressed in mouse cultured oligodendrocyte precursor cells (OPC; Kato et al., 2011). EPO signaling from astrocytes to OPCs could prevent OPC damage under conditions of hypoxia/reoxygenation injury (Kato et al., 2011). Microglia are the resident macrophages in the brain and have innate immune functions in the central nervous system (CNS; Lawson et al., 1990). Microglial activation is associated with morphological changes, proliferation, motility, phagocytosis, and cytokine release (Boscia et al., 2009; Harrigan et al., 2008). Microglial activation states can be divided into classical and alternative activation states (Colton, 2009). Classically activated microglia are cytotoxic because they secrete proinflammatory cytokines and reactive oxygen species (ROS). In contrast, alternatively activated microglia induce anti-inflammatory cytokines and are involved in wound repair and debris clearance (Gordon, 2003). Phagocytosis of microbes is associated with inflammation, whereas the phagocytosis of apoptotic cells occurs in the absences of inflammation. Recognition of microbes induces a microglial phagocytic response, 4

which is associated with the release of pro-inflammatory cytokines (Neumann et al., 2009). Accordingly, phagocytosis with inflammation could be considered the result of classic microglial activation. Activated microglia are observed in nearly all kinds of neurological diseases, including neurodegenerative diseases, such as Alzheimer’s disease (AD; Prokop et al., 2013), Parkinson’s disease (PD; Long-Smith et al., 2009), and amyotrophic lateral sclerosis (ALS; Sargsyan et al., 2005), infectious and inflammatory diseases, such as multiple sclerosis (MS; Brown, 2001; Napoli and Neumann, 2010), stroke (Yenari et al., 2010), and traumatic (Loane and Byrnes, 2010) and radiation-induced (Chiang et al., 1993) brain injury. Because classical microglial activation is considered to be the major cause of these diseases, from a therapeutic viewpoint it is important to control classical microglial activation. In the present study, we chose to use a lipopolysaccharide (LPS)-stimulated model because LPS is known to induce classical microglial activation (Kobayashi et al., 2013). In the present study, we investigated whether the neuroprotective effects of EPO could attenuate microglial activation. To determine the effects of EPO on microglia, we examined the effects of EPO on: (i) the expression of inflammatory cytokines and the phagocytic activity of the BV-2 murine microglial cell line after LPS treatment in vitro, and (ii) in vivo cytokine production and morphological changes to microglia in the brains of mice after LPS treatment.

5

2. Results 2.1. Expression of EPOR in rat primary cultured cells The quantitative analyses of EPOR expression were performed on neurons, astrocytes, and microglia from primary cultures of rat cells in the present study. EPOR expression was found to be higher in microglia than in neurons or astrocytes (Fig. 1A; P < 0.01; n = 6). Immunocytochemical analysis revealed higher expression of EPOR on neurons and microglia (Fig. 1B).

2.2. Inflammatory cytokine expression in BV-2 cells after LPS and EPO treatment The effects of EPO on LPS-induced activation of microglia were assessed by measuring the expression of inflammatory cytokines in the murine microglial BV-2 cell line. We analyzed the time course of LPS-induced cytokine expression under control conditions and in the presence of EPO. Tumor necrosis factor-α (Tnfα) expression peaked 1 h after LPS treatment (Fig. 2A), whereas interleukin-1β (Il1b) and interleukin-6 (Il6) expression peaked 3 h (Fig. 2B) and 6 h (Fig. 2C) after LPS treatment, respectively. In addition, we examined the effect of EPO treatment on peak cytokine mRNA expression and found that EPO significantly reduced peak mRNA expression for all three inflammatory cytokines (Fig. 2A–C; P < 0.01; n = 6).

2.3. Effects of EPO on LPS-induced inducible nitric oxide synthase (iNOS) expression in BV-2 cells The effects of EPO on LPS-induced inducible nitric oxide synthase (iNOS) 6

mRNA expression in BV-2 cells were evaluated. In both untreated control and EPO-treated BV-2 cells, iNOS expression was very low or below the threshold of the assay detection. However, LPS induced iNOS gene expression in BV-2 cells, which peaked 6 h after LPS treatment, and this increase in iNOS expression was significantly reduced by EPO (Fig. 3A; P < 0.01; n = 6). Western blotting analysis revealed very faint iNOS protein expression in both the untreated control and EPO-treated groups. In contrast, significantly higher iNOS protein expression was seen in the LPS-treated group, which was reduced by EPO (Fig. 3B). Quantitative analysis revealed an approximate 3.3-fold increase in iNOS protein in LPS-treated BV-2 cells, which was significantly decreased by EPO (Fig. 3B; P < 0.01; n = 3). In addition, nitric oxide (NO) levels in the conditioned medium of untreated and treated BV-2 cells were evaluated as nitrite concentrations using the Griess assay. However, NO levels were very low and could not be reliably measured (data not shown).

2.4. Effects of LPS and EPO on phagocytotic activity of BV-2 cells Phagocytosis is an important physiological and pathological function of activated microglia (Kreutzberg, 1996). The phagocytosis of fluorescently labeled latex beads by BV-2 microglial cells was evaluated to determine whether EPO could suppress microglial phagocytotic activity induced by LPS. A small number of cells in the EPO-treated and untreated groups contained phagocytosed beads. The number of cells 7

containing phagocytosed beads increased after LPS stimulation, but this number was reduced by EPO (Fig. 4A). The phagocytotic activity of BV-2 microglial cells was quantified by counting the number of cells containing fluorescent beads. Phagocytosed beads were present in 17.3 ± 1.8% and 11.6 ± 1.0% of cells in the untreated control and EPO-treated groups, respectively, compared with 36.7 ± 0.6% of cells in the LPS-treated group (Fig. 4B; P < 0.01; n = 6). In contrast, phagocytosed beads were found in only 21.3 ± 1.9% of cells in the EPO+LPS-treated group (Fig. 4B; P < 0.01; n = 6).

2.5. Effects of LPS and EPO on expression of inflammatory cytokines in mouse brain The effects of EPO on inflammatory cytokine mRNA expressions were determined in the brains of adult male mice that had been administered a single i.p. injection of LPS. Peripheral LPS administration rapidly stimulates the production of circulating cytokines that can cross the blood-brain barrier (BBB). These cytokines can activate microglia in the brain, which then produce additional inflammatory cytokines (Qin et al., 2007). In the present study, LPS injection induced expression of Tnfα, Il1b, and Il6 in mouse brains. LPS-induced expression of Tnfα and Il1b mRNA was significantly reduced by prior i.p. injection of EPO (Fig. 5; P < 0.05; n = 6).

2.6. Effects of LPS and EPO on microglial morphology To determine EPO effects on microglia in vivo, microglial activation was analyzed in the parietal cortex of mouse brain in a control group, an EPO-treated group, 8

an LPS-treated group, and an EPO+LPS-treated group 6 h after LPS injection. Initially, the number of ionized calcium-binding adaptor protein 1 (Iba1) -positive microglia per area was counted in each group to assess microglial density. Mean cell density was 141.0 ± 9.7 and 135.5 ± 7.3 cells per 100000 µm2 in the untreated control and EPO-treated groups, respectively, compared with 105.1 ± 11.9 and 139.5 ± 3.4 cells per 100000 µm2 in the LPS-treated group and EPO+LPS-treated group, respectively. No significant differences in microglial density were found between groups. Next, microglial morphology was analyzed (Fig. 6). There were no significant differences in the number of mildly activated microglia (Stage 1) in the LPS-treated group compared with the control group. Interestingly, brains from LPS-treated mice had the largest number of microglia at higher stages of activation (Stages 2 and 3; Fig. 6A, B; P < 0.01 compared with the control group; n = 6). However, EPO treatment prior to LPS injection decreased the number of microglia at higher stages of activation (Stages 2 and 3; Fig. 6A, B; P < 0.01 compared with LPS treatment alone; n = 6). Finally, the sizes of the cell bodies of microglial cells were quantified by measuring the area of Iba1-positive cells using Developer Toolbox v1.9. Mean cell body size was 92.1 ± 4.1 and 96.5 ± 3.7 µm2 in the untreated control and EPO-treated groups, respectively, compared with 120.4 ± 4.3 µm2 in the LPS-treated group. However, EPO treatment of the mice prior to injection of LPS reduced mean cell body size to 93.6 ± 3.1 µm2 (Fig. 7; P < 0.01 compared with LPS alone; n = 304–591).

9

3. Discussion The possible role of EPO as a novel neuroprotective agent for the CNS has stimulated significant recent research interest (Chong et al., 2003). Unfortunately, the cellular mechanism that underlies the EPO/EPOR signaling pathway in the brain remains unclear. The present study has shown that EPO attenuates LPS-induced microglial activation. Our present study examined the quantitative comparison of cellular EPOR expression in neurons and glia using the primary culture cells in vitro, although the culture cells might be in artificial environments. Using quantitative RT-PCR, our study confirmed a higher expression of EPOR in microglia than in neurons in rat primary culture. Furthermore, we examined the expression of the intrinsic EPO gene in primary microglia. However, using quantitative RT-PCR, intrinsic EPO expression was not observed in primary microglia (data not shown). Subsequently, we examined the effect of EPO on microglial activation in vitro. BV-2 murine microglial cell line was used in the present study because BV-2 cells are the most frequently used substitute for primary microglia (Henn et al., 2009), and EPOR expression in BV-2 cells has been confirmed in previous reports (Wilms et al., 2009). The expression of the intrinsic EPO gene was also evaluated in BV-2 cells, but again quantitative RT-PCR failed to demonstrate intrinsic EPO expression in either naïve or LPS-treated BV-2 cells in the absence or presence of EPO (data not shown). There are many different ways to activate microglia, including hypoxia or cytokine treatment. In the present study we chose to use LPS (10 ng/mL) stimulation to activate microglia because most previous studies have used this 10

concentration of LPS. Using this concentration of LPS resulted in consistent activation of microglia. It remains contentious as to whether EPO can attenuate LPS-induced microglial activation (De et al., 2002; Mitkovski et al., 2015; Vairano et al., 2002; Wilms et al., 2009). In a previous study, EPO was not able to prevent the production of inflammatory mediators, including NO and inflammatory cytokines, in rat microglial cells, as well as in the BV-2 cell line (Wilms et al., 2009). However, in that study, microglia were activated by a higher dose of LPS (200 ng/mL) than that used in the present study (10 ng/mL). The higher dose of LPS may have overcome any effect of EPO on activated microglia. Conversely, other studies have focused on the phagocytic ability of reactive microglia. Apoptotic neurons have increased the externalization of phosphatidylserine (PS) on the outer leaflet of plasma membrane. The activated microglia have higher expressions of the PS receptor (PSR), and PS recognition is a key step in the process of phagocytosis of neurons by microglia (De et al., 2002). EPO treatment decreased microglial PSR expression in vitro (Chong et al., 2003), suggesting that EPO treatment suppressed the ability of microglial phagocytosis of dying neurons. In the present study, microglial phagocytosis was investigated by measuring the cellular phagocytosis of fluorescently labeled latex beads by BV-2 cells in vitro. EPO decreased the number of phagocytosing BV-2 microglia with and without LPS stimulation (Fig. 4). EPO decreased from 36.7 ± 0.6% to 21.3 ± 1.9% of cells with LPS stimulation, and from 17.3 ± 1.8% to 11.6 ± 1.0% of cells without LPS stimulation. EPO decreased a higher number of phagocytosing BV-2 microglia with LPS stimulation than microglia without LPS stimulation. We would have liked to clarify the mechanisms underlying the 11

anti-inflammatory effect of EPO in LPS-stimulated microglia, especially the signaling mechanisms involved, but the mechanisms were not clear. For example, there were no significant changes in phosphatidylinositol 3-kinase (PI3K) signaling as a downstream effector of EPO–EPOR signaling or in nuclear factor (NF)-κB signaling as a downstream effector of LPS/Toll-like receptor (TLR) signaling (data not shown). We also examined the possibility of the involvement of mitogen-activated protein kinase (MAPK) signaling, another downstream effector of the LPS/TLR signaling pathway, and observed partial suppression of Jun N-terminal kinase (JNK) signaling in LPS-stimulated microglia by EPO. However, these observations were not reproducible and could not be confirmed. EPO treatment of LPS-stimulated cells had no effect on p38 and ERK signaling. Finally, we investigated the effect of EPO on microglia activated by LPS in the brain in vivo. Brains from LPS-treated mice had the largest number of microglia at higher stages of activation. However, EPO treatment prior to LPS injection decreased the number of microglia at higher stages of activation (Figure 6). Furthermore, EPO treatment of mice prior to injection of LPS reduced mean cell body size (Figure 7). This value of the cell body seems higher than the values previously reported (Jinno et al., 2007; Kozlowski and Wimer, 2012; Torres-Platas et al., 2014). The values depend on the immunohistochemical staining methods and the analysis software settings including the cut-off intensity and the definition of the cell body. Attenuation of microglial activation was not observed when male mice were given a single i.p. injection of EPO simultaneously with LPS. However, when male mice were given a single i.p. injection 12

of EPO 24 h before LPS injection, there were more consistent morphological changes in the brain and suppression of inflammatory cytokine expression than those observed after simultaneous administration of LPS and EPO. The prior injection of EPO may allow sufficient time for EPO to cross the BBB and be distributed throughout the brain before the rapid distribution of LPS after its injection. Previously, we reported on the neuroprotective effect of intrinsic EPO under hypoxic conditions (Kato et al., 2011; Nagaya et al., 2014). However, this effect of intrinsic hypoxia-induced EPO was limited by treatment with the inflammatory cytokine tumor necrosis factor (TNF) α, which suppressed EPO expression in the brain and kidney (Nagaya et al., 2014). We considered whether intrinsic EPO could attenuate microglial activation in LPS-treated mice in the present study. However, the role of intrinsic EPO would not be significant, because the present study was not performed under hypoxic conditions and LPS administration would lead to the inflammatory reaction, including the release of TNFα, and consequently, intrinsic EPO expression would be suppressed in the brain and kidneys of LPS-treated mice. In conclusion, we have found that EPO treatment can attenuate LPS-induced microglial activity, including morphological changes in vivo, phagocytosis in vitro, and inflammatory cytokine production in vivo and in vitro. The neuroprotective effects of EPO in the brain may involve regulation of microglial activation. Investigating effective regulation of microglial activation by EPO may have important implications for the development of novel neuroprotective therapies for hypoxic–ischemic brain diseases, as well as neuroinflammatory diseases, including neurodegenerative diseases that are 13

associated with microglial activation. 4. Experimental Procedures 4.1. Animals and reagents All rats and mice were purchased from Japan SLC (Shizuoka, Japan). Animal experiment protocols were approved by the Animal Care and Use Committee, Nagoya City University Graduate School of Medical Sciences. LPS (O55:B5) was purchased from Sigma-Aldrich (St. Louis, MO, USA), and human recombinant EPO was purchased from KIRIN (Tokyo, Japan).

4.2. Primary glia and neuron cultures Primary microglia were prepared and maintained as previously described (Kakita et al., 2013). Briefly, the primary mixed glial cultures were prepared from postnatal day (PND1) 1 Wistar rats by dissociation. Dissociated cells were plated in 75-cm2 culture flasks (Corning Incorporated Life Sciences, Lowell, MA, USA). After 10 days, the cultures were shaken at 100 r.p.m. for 30 min to yield loosely attached cells. The detached cells were collected and re-plated on non-coated plates. Immunostaining with Iba1 antibodies (Wako, Osaka, Japan) revealed that microglia represented >95% of the cell population. Primary astrocytes were prepared and maintained as previously described (Nagaya et al., 2014). Briefly, the primary mixed glial cultures were prepared as described above for primary microglia cultures. After 10 days, the cultures were shaken overnight to remove cellular debris and loosely attached cells including microglia, 14

oligodendrocytes, and their early precursors. After shaking, the cultured components were subcultured on tissue culture dishes. The cell populations consisted of >95% astrocytes, which was determined by immunocytochemical staining with glial fibrillary acidic protein (GFAP) antibodies (DAKO, Glostrup, Denmark). Primary neuronal cultures were prepared and maintained as described previously (Nagaya et al., 2014). Briefly, the cerebral cortex was treated as described above for primary microglia and astrocyte cultures. Cells were dissociated on glass coverslips coated with BD Matrigel Basement Membrane Matrix Growth Factor Reduced (BD Bioscience, San Jose, CA, USA). Neurons were used for experiments after seven days. Immunostaining with microtubule-associated protein 2 (MAP2) antibodies (Millipore, Billerica, MA, USA) revealed that neurons represented >95% of the cell population.

4.3. Immunocytochemical staining of primary glia and neuron cultures Immunocytochemical staining was examined as previously described (Kakita et al., 2013). Cell cultures were washed with phosphate-buffered saline (PBS), followed by 3% paraformaldehyde in PBS for 30 min. After being washed with PBS, cells were incubated for 1 h with blocking solution. The primary antibodies used in this study were a rabbit polyclonal anti-EPOR antibody (1:100 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA) and each of the following cell marker antibodies: anti-ED1 monoclonal antibody (mAb; 1:100 dilution; Serotec, Oxford, UK), anti-GFAP mAb (1:100 dilution; Millipore), and anti-MAP2 mAb (1:100 dilution; Millipore). The 15

secondary antibodies used in this study were Alexa Flour 488 and Alexa Flour 594 (1:1000 dilution; Invitrogen) for 1 h at room temperature. The coverslips were mounted with ProLong Gold antifade reagent containing 4′,6′-diamidino-2-phenylindole (DAPI; Invitrogen), and visualized using an AX70 fluorescence microscope (Olympus, Tokyo, Japan).

4.4. BV-2 microglial cell line cultures The murine microglial cell line BV-2 was cultured and maintained at 37°C in a 5% CO2 humidified incubator at 21% O2. Cells were sub-cultured every third or fourth day. Cells were re-plated on 35 mm tissue culture dishes (3 × 10 5 cells/dish). After the culture medium had been replaced with fresh FBS-free medium, LPS (10 ng/mL; Sigma-Aldrich) was added to the culture medium. Recombinant EPO (10 mIU/mL; KIRIN) was added to the culture medium as required. Cells were maintained at 37°C in a 5% CO2 humidified incubator at 21% O2.

4.5. Quantitative reverse transcription–polymerase chain reaction Quantitative reverse transcription–polymerase chain reaction (RT-PCR) was performed as previously described (Nagaya et al., 2014). Briefly, total RNA was extracted with an RNeasy Mini Kit (Qiagen, Valencia, CA, USA). Reverse transcription was performed using random primers and Ready-To-Go You-Prime First Strand Beads (GE Healthcare Bio-Science, Piscataway, NJ, USA) or PrimeScript RT Master Mix (Takara Bio Inc., Otsu, Japan). The synthesized cDNA was subjected to PCR 16

amplification using SYBR Green Master Mix Reagents (Applied Biosystems) and the primer pairs given in Table 1. The amplification program was as follows: 95°C for 10 min for initial denaturation, and amplification for 40 cycles at 95°C for 15 s and 60°C for 1 min. The relative quantification of target genes was normalized against that of an endogenous control, namely β-actin (Actb) or glyceraldehyde-3-phosphate dehydrogenase (Gapdh), after confirming that the efficiency of amplification of cDNA of the target genes, Actb and Gapdh was the same.

4.6. Western blotting Western blotting analysis was performed as previously described (Hamano et al., 2015). Briefly, BV-2 cells were collected in a lysis buffer of 5 mM Tris-HCl, pH 7.4, containing 0.5% Nonidet P-40 (NP-40) and a protease inhibitor cocktail (Sigma). Protein content was estimated using a BCA protein assay reagent kit (Pierce Chemicals, Rockford, IL, USA). The equal quantities of protein samples were separated by 7.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, electrotransferred onto Immobilion-P membranes (Millipore), and probed with a polyclonal anti-iNOS antibody (1:1000 dilution; BD Bioscience). Blots were developed using appropriate horseradish peroxidase-conjugated secondary antibodies (1:5000 dilution; Amersham Biosciences, Piscataway, NJ, USA). Protein-antibody complexes were visualized using an enhanced chemiluminescence (ECL) method (Amersham Biosciences). Gel protein loading was normalized by stripping the blots and re-probing them with an anti-actin polyclonal antibody (1:1000 dilution; Sigma). 17

4.7. Phagocytosis assay The microglial phagocytosis was evaluated by imaging analysis with latex beads (1-µm diameter carboxylate-modified yellow-green fluorescent; Invitrogen), as previously described (Kakita et al., 2013). Briefly, BV-2 cells were re-plated on glass coverslips in 24-well tissue culture dishes (1 × 105 cells/dish). After treatment, latex beads that had been diluted in the same medium were added to each well to a final concentration of 5 × 10 7 beads per well and the samples were left for 15 min. Then, the BV-2 cells were thoroughly washed with PBS to remove any extracellular beads before being fixed in 3% paraformaldehyde. After washing again with PBS, the BV-2 cells were blocked for 1 h with blocking solution (3% BSA, 0.1% glycine in PBS). The cells were incubated with anti-Iba1 antibody (1:500; Wako) for 1 h. After the 1-h incubation, the cells were washed with PBS and incubated with fluorescent Alexa Flour 594-conjugated goat anti-rabbit IgG as the secondary antibody (1:1000 dilution; Invitrogen) for 1 h. Finally, the coverslips were mounted with ProLong Gold antifade reagent containing DAPI. BV-2 cells were examined under a NIKON A1R confocal microscope (Tokyo, Japan) to determine whether the fluorescent beads had been phagocytosed and were present inside the cells. The number of BV-2 cells (of 100 cells examined at a magnification of ×200) containing more than one fluorescence bead was counted and is expressed as a percentage. Cell counts were made by observers blinded to the treatment group, and counts from six fields were averaged. Observations were obtained from at least three independent experiments. 18

4.8. LPS and EPO treatment of mice Eight-week-old male ICR mice were given a single intraperitoneal (i.p.) injection of LPS (5 mg/kg) or an equivalent volume of PBS as a vehicle. LPS dose in the present study was determined based on the results of previous studies (Bossu et al., 2012; Qin et al., 2007). To assess the neuroprotective effects of EPO on the LPS-induced activation of microglia, mice were injected with a single dose of EPO (5000 IU/kg, i.p.) or an equivalent volume of vehicle 24 h before LPS injection to allow sufficient time for the EPO to penetrate the BBB. EPO dose in the present study was based on that used in a previous study (Hagemeyer et al., 2012). Brain tissues were collected from mice for mRNA and immunohistochemical analysis 6 h after injection of LPS.

4.9. Immunohistochemical staining of brain tissue Mice were deeply anesthetized with CO2 and transcardially perfused with normal saline to remove the blood followed by 4% paraformaldehyde. Brains were dissected out from all mice and the right hemisphere was fixed in 4% paraformaldehyde for two days before being cryoprotected with 30% sucrose overnight at 4°C. Tissue samples were embedded in Tissue-Tek OCT compound (Sakura Finetek, Tokyo, Japan) and frozen in liquid nitrogen. Brain tissues were cut into 10 µm sagittal sections and collected with a freezing stage microtome (Leica CM1800; Leica Instruments, Tokyo, Japan). Brain sections were incubated with anti-Iba1 antibody (Wako) overnight at 4°C. 19

After washing with PBS, the sections were incubated with an Alexa Flour 488-conjugated secondary antibody (1:1000 dilution; Invitrogen) for 1 h. The sections were mounted with ProLong Gold antifade reagent containing DAPI. Cells with the precise Iba1 staining of the cytoplasm and the DAPI staining of the nucleus were identified as microglia in the parietal cortex. Cell number was counted by observers blinded to the treatment groups. Cell density was measured as the number of cells per area (100000 µm2). Data were recorded from six fields and were used to determine microglial density. Activation of microglia in the parietal cortex was evaluated using the fractionation method of unbiased analysis, as previously described (Hutson et al., 2011). The morphological parameters as previously reported (Hutson et al., 2011) were modified to the stage score of microglial activation ranging from the resting stage (Stage 0) to the activated stages (Stages 1–3). The microglia morphologies are classified on the basis of complexity of processes, thickness, length, and cell body size (Hutson et al., 2011). Cells with the Iba1 staining of the entire body and the DAPI staining of the nucleus were identified as microglia. Maximal intensity projections were obtained using a NIKON A1R confocal microscope with a 40× objective, 512×512 pixels. Microglial activation was scored by observers blinded to the treatment groups. Observations were obtained from at least three independent experiments. Cell body sizes were measured using Developer Toolbox v1.9 (GE Healthcare Bio-Science), with “cell body” defined as the Iba1-stained area of cells with a DAPI-stained nucleus. Cell body area was measured as the Alexa Flour 488 (green)-stained area (µm2) of cells with a 20

DAPI-stained nucleus.

4.10. Statistical analysis Analysis of variance (ANOVA) was used to compare continuous data, followed by Bonferroni’s post hoc test, using SPSS for Windows version 13.0 (SPSS, Chicago, IL, USA). Data are expressed as the mean ± SEM. A value of P < 0.05 was considered statistically significant.

21

Author contributions TT: Conception and design, collection and assembly of data, data analysis and interpretation, manuscript writing, final approval of manuscript; MA: Conception and design, financial support, collection and assembly of data, data analysis and interpretation, manuscript writing, final approval of manuscript; SU: Collection and assembly of data, final approval of manuscript; HK: Collection and assembly of data, data analysis and interpretation, final approval of manuscript; KS: Conception and design, financial support, final approval of manuscript; KA: Conception and design, financial support, collection and assembly of data, data analysis and interpretation, manuscript writing, final approval of manuscript.

Acknowledgments The authors thank Mr. Hiroshi Takase for his technical assistance. This work was supported in part by Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science, KAKEN grant numbers 25461651 and 23591509.

References Boscia, F., Gala, R., Pannaccione, A., Secondo, A., Scorziello, A., Di Renzo, G., Annunziato, L., 2009. NCX1 expression and functional activity increase in

22

microglia invading the infarct core. Stroke 40, 3608–3617. Bossù, P., Cutuli, D., Palladino, I., Caporali, P., Angelucci, F., Laricchiuta, D., Gelfo, F., De Bartolo, P., Caltagirone, C., Petrosini, L., 2012. A single intraperitoneal injection of endotoxin in rats induces long-lasting modifications in behavior and brain protein levels of TNF-α and IL-18. J. Neuroinflammation 9, 101. Brown, D.R., 2001. Microglia and prion disease. Microsc. Res. Tech. 54, 71–80. Brown, G.C., Neher, J.J., 2010. Inflammatory neurodegeneration and mechanisms of microglial killing of neurons. Mol. Neurobiol. 41, 242–247. Chiang, C.S., McBride, W.H., Withers, H.R., 1993. Radiation-induced astrocytic and microglial responses in mouse brain. Radiother. Oncol. 29, 60–68. Chin, K., Yu, X., Beleslin-Cokic, B., Liu, C., Shen, K., Mohrenweiser, H.W., Noguchi, C.T., 2000. Production and processing of erythropoietin receptor transcripts in brain. Mol. Brain Res. 81, 29–42. Chong, Z.Z., Kang, J-Q., Maiese, K., 2003. Erythropoietin fosters both intrinsic and extrinsic neuronal protection through modulation of microglia, Akt1, Bad, and caspase-mediated pathways. Br. J. Pharmacol. 138, 1107–1118. Colton, C.A., 2009. Heterogeneity of microglial activation in the innate immune response in the brain. J. Neuroimmune. Pharmacol. 4, 399–418. De, S.R., Ajmone-Cat, M.A., Nicolini, A., Minghetti, L., 2002. Expression of phosphatidylserine receptor and down-regulation of pro-inflammatory molecule production by its natural ligand in rat microglial cultures. J. Neuropathol. Exp. Neurol. 61, 237–244. 23

Gordon, S., 2003. Alternative activation of macrophages. Nat. Rev. Immunol. 3, 23–35 Hagemeyer, N., Boretius, S., Ott, C., Von Streitberg, A., Welpinghus, H., Sperling, S., Frahm, J., Simons, M., Ghezzi, P., Ehrenreich, H., 2012. Erythropoietin attenuates neurological and histological consequences of toxic demyelination in mice. Mol. Med. 18, 628–635. Hamano, S., Mori, Y., Aoyama, M., Kataoka, H., Tanaka, M., Ebi, M., Kubota, E., Mizoshita, T., Tanida, S., Johnston, R.N., Asai, K., Joh, T., 2015. Oncolytic reovirus combined with trastuzumab enhances antitumor efficacy through TRAIL signaling in human HER2-positive gastric cancer cells. Cancer Lett. 356, 846-854. Harrigan, T.J., Abdullaev, I.F., Jourd'heuil, D., Mongin, A.A., 2008. Activation of microglia with zymosan promotes excitatory amino acid release via volume-regulated anion channels: the role of NADPH oxidases. J. Neurochem. 106, 2449–2462. Henn, A., Lund, S., Hedtjärn, M., Schrattenholz, A., Pörzgen, P., Leist, M., 2009. The suitability of BV2 cells as alternative model system for primary microglia cultures or for animal experiments examining brain inflammation. ALTEX 26, 83-94. Hutson, C.B., Lazo, C.R., Mortazavi, F., Giza, C.C., Hovda, D., Chesselet, M.F., 2011. Traumatic brain injury in adult rats causes progressive nigrostriatal dopaminergic cell loss and enhanced vulnerability to the pesticide paraquat. J. Neurotrauma 28, 1783–1801.

24

Jinno, S., Fleischer, F., Eckel, S., Schmidt, V., Kosaka, T., 2007. Spatial arrangement of microglia in the mouse hippocampus: a stereological study in comparison with astrocytes. Glia 55, 1334-1347. Juul, S.E., Anderson, D.K., Li, Y., Christensen, R.D., 1998. Erythropoietin and erythropoietin receptor in the developing human central nervous system. Pediatr. Res. 43, 40–49. Kakita, H., Aoyama, M., Nagaya, Y., Asai, H., Hussein, M.H., Suzuki, M., Kato, S., Saitoh, S., Asai, K., 2013. Diclofenac enhances proinflammatory cytokine-induced phagocytosis of cultured microglia via nitric oxide production. Toxicol. Appl. Pharm. 268, 99-105. Kato, S., Aoyama, M., Kakita, H., Hida, H., Kato, I., Ito, T., Goto, T., Hussein, M.H., Sawamoto, K., Togari, H., Asai, K., 2011. Endogenous erythropoietin from astrocyte protects the oligodendrocyte precursor cell against hypoxic and reoxygenation injury. J. Neurosci. Res. 10, 1566–1574. Kobayashi, K., Imagama, S., Ohgomori, T., Hirano, K., Uchimura, K., Sakamoto, K., Hirakawa, A., Takeuchi, H., Suzumura, A., Ishiguro, N., Kadomatsu, K., 2013. Minocycline selectively inhibits M1 polarization of microglia. Cell Death. Dis. 4, e525. Kozlowski, C., Weimer, R.M., 2012. An automated method to quantify microglia morphology and application to monitor activation state longitudinally in vivo. PLoS One 7, e31814. Kreutzberg, G.W., 1996. Microglia: a sensor for pathological events in the CNS. Trends 25

Neurosci. 19, 312–318. Lawson, L.J., Perry, V.H., Dri, P., Gordon, S., 1990. Heterogeneity in the distribution and morphology of microglia in the normal adult mouse brain. Neuroscience 39, 151–170. Lin, C.S., Lim, S.K., D’Agati, V., Costantini, F., 1996. Differential effects of an erythropoietin receptor gene disruption on primitive and definitive erythropoiesis. Genes Dev. 10, 154–164. Loane, D.J., Byrnes, K.R., 2010. Role of microglia in neurotrauma. Neurotherapeutics 7, 366–377. Long-Smith, C.M., Sullivan, A.M., Nolan, Y.M. 2009. The influence of microglia on the pathogenesis of Parkinson’s disease. Prog. Neurobiol. 89, 277–287. Marti, H.H., Wenger, R.H., Rivas, L.A., Straumann, U., Digicaylioglu, M., Henn, V., Yonekawa, Y., Bauer, C., Gassmann, M., 1996. Erythropoietin gene expression in human, monkey and murine brain. Eur. J. Neurosci. 8, 666–676. Masuda, S., Nagao, M., Takahata, K., Konishi, Y., Gallyas, F., Tabira, T., Sasaki, R., 1993. Functional erythropoietin receptor of the cells with neural characteristics. Comparison with receptor properties of erythroid cells. J. Biol. Chem. 268, 11208–11216. Masuda, S., Okano, M., Yamagishi, K., Nagao, M., Ueda, M. Sasaki, R., 1994. A novel site of erythropoietin production. Oxygen-dependent production in cultured rat astrocytes. J. Biol. Chem. 269, 19488–19493.

26

Mitkovski, M., Dahm, L., Heinrich, R., Monnheimer, M., Gerhart, S., Stegmüller, J., Hanisch, U.K., Nave, K.A., Ehrenreich, H., 2015. Erythropoietin dampens injury-induced microglial motility. J. Cereb. Blood Flow Metab. 35, 1233-1236. Miyachi, T., Asai, K., Tsuiki, H., Mizuno, H., Yamamoto, N., Yokoi, T., Aoyama, M., Togari, H., Wada, Y., Miura, Y., Kato, T., 2001. Interleukin-1beta induces the expression of lipocortin 1 mRNA in cultured rat cortical astrocytes. Neurosci. Res. 40, 53–60. Morishita, E., Masuda, S., Nagao, M., Yasuda, Y., Sasaki, R., 1997. Erythropoietin receptor is expressed in rat hippocampal and cerebral cortical neurons, and erythropoietin prevents in vitro glutamate-induced neuronal death. Neuroscience 76, 105–116. Nagai, A., Nakagawa, E., Choi, H.B., Hatori, K., Kobayashi, S., Kim, S.U., 2001. Erythropoietin and erythropoietin receptors in human CNS neurons, astrocytes, microglia, and oligodendrocytes grown in culture. J. Neuropathol. Exp. Neurol. 60, 386–392. Nagaya, Y., Aoyama, M., Tamura, T., Kakita, H., Kato, S., Hida, H., Saitoh, S., Asai, K., 2014. Inflammatory cytokine tumor necrosis factor-alpha suppresses neuroprotective endogenous erythropoietin from astrocytes mediated by hypoxia-inducible factor-2alpha. Eur. J. Neurosci. 40, 3620–3626. Napoli, I., Neumann, H., 2010. Protective effects of microglia in multiple sclerosis. Exp. Neurol. 225, 24–28. Neumann, H., Kotter, M.R., Franklin, R.J., 2009. Debris clearance by microglia: an 27

essential link between degeneration and regeneration. Brain 132, 288–295. Noguchi, C.T., Asavaritikrai, P., Teng, R., Jia, Y., 2007. Role of erythropoietin in the brain. Crit. Rev. Oncol. Hematol. 64, 159–171. Prokop, S., Miller, K.R., Heppner, F.L., 2013. Microglia actions in Alzheimer’s disease. Acta Neuropathol. 126, 461–477. Qin, L., Wu, X., Block, M.L., Liu, Y., Breese, G.R., Hong, J., Knapp, D.J., Crews, F.T., 2007. Systemic LPS causes chronic neuroinflammation and progressive neurodegeneration. Glia 55, 453–462. Ruscher, K., Freyer, D., Karsch, M., Isaev, N., Megow, D., Sawitzki, B., Priller, J., Dirnagl, U. Meisel, A., 2002. Erythropoietin is a paracrine mediator of ischemic tolerance in the brain: evidence from an in vitro model. J. Neurosci. 23, 10291–10301. Sargsyan, S.A., Monk, P.N., Shaw, P.J., 2005. Microglia as potential contributors to motor neuron injury in amyotrophic lateral sclerosis. Glia 51, 241–253. Torres-Platas, S.G., Comeau, S., Rachalski, A., Bo, G.D., Cruceanu, C., Turecki, G., Giros, B., Mechawar, N., 2014. Morphometric characterization of microglial phenotypes in human cerebral cortex. J. Neuroinflammation 11, 12. Vairano, M., Dello Russo, C., Pozzoli, G., Battaglia, A., Scambia, G., Tringali, G., Aloe-Spiriti, M.A., Preziosi, P., Navarra, P., 2002. Erythropoietin exerts anti-apoptotic effects on rat microglial cells in vitro. Eur. J. Neurosci. 16, 584-592. Walter, L., Neumann, H., 2009. Role of microglia in neuronal degeneration and regeneration. Semin. Immunopathol. 31, 513–525. 28

Wilms, H., Schwabedissen, B., Sievers, J., Lucius, R., 2009. Erythropoietin does not attenuate cytokine production and inflammation in microglia implications for the neuroprotective effect of erythropoietin in neurological diseases. J. Neuroimmunol. 212, 106-111. Wu, H., Liu, X., Jaenisch, R., Lodish, H.F., 1995. Generation of committed erythroid BFU-E and CFU-E progenitors does not require erythropoietin or the erythropoietin receptor. Cell 83, 59–67. Yenari, M.A., Kauppinen, T.M., Swanson, R.A. 2010. Microglial activation in stroke: therapeutic targets. Neurotherapeutics 7, 378–391.

29

Figure Legends Fig. 1. Expression of erythropoietin (EPO) receptor (EPOR) in primary cultures of rat neurons, astrocytes, and microglia. (A) EPOR mRNA levels were evaluated by quantitative RT-PCR and expression was normalized against that of the β-actin (Actb) gene. EPOR mRNA was highly expressed in microglia compared with neurons and astrocytes. Data are the mean ± SEM (n = 6 in each group). #P < 0.01 compared with neurons (one-way ANOVA with Bonferroni correction for all comparison pairs). (B) Neurons, astrocytes, and microglia were stained with antibodies against cell-specific markers (red) and EPOR (green). Neurons were stained with antibodies against the neuron cell marker β-tubulin (red), astrocytes were stained with antibodies against the astrocyte cell marker glial fibrillary acidic protein (GFAP; red), and microglia were stained with antibodies against the microglial cell marker ED1 (red). Nuclei were stained with 4′,6′-diamidino-2-phenylindole (DAPI; blue). Bar, 100 µm. Representative photomicrographs show that EPOR was highly expressed in microglia.

Fig. 2. Time course of changes in the expression of inflammatory cytokine genes. Expression of (A) tumor necrosis factor-α (Tnfα), (B) interleukin-1β (Il1b), and (C) interleukin-6 (Il6) was determined in BV-2 cells stimulated with lipopolysaccharide (LPS; open bars) or LPS plus EPO (filled bars). Data are the mean ± SEM (n = 6 in each group). #P < 0.01 compared with LPS alone at the same time point (two-way ANOVA with Bonferroni correction for all comparison pairs).

30

Fig. 3. Effects of EPO on inducible nitric oxide synthase (iNOS) expression in BV-2 cells. (A) Time course of changes in iNOS gene expression in BV-2 cells stimulated with LPS (open bars) or LPS plus EPO (filled bars). Data are the mean ± SEM (n = 6 in each group). (B) Western blot analysis of iNOS protein in BV-2 cells after 12 h treatment. Semiquantitative measurements of iNOS protein were made using ImageJ software (National Institutes of Health, Bethesda, MD, USA). Data are the mean ± SEM (n = 3). #P < 0.01 compared with the control group; **P < 0.01 compared with LPS alone (one-way ANOVA with Bonferroni correction for all comparison pairs).

Fig. 4. Effects of EPO on the phagocytic activity of BV-2 cells stimulated with LPS. (A) The phagocytic activity of BV-2 cells was measured 24 h after LPS treatment. Representative photomicrographs show BV-2 cells containing phagocytosed fluorescent beads (green) immunolabeled using an anti-ionized calcium-binding adaptor protein 1 (Iba1) antibody (red). Bar, 50 µm. (B) Number of BV-2 cells containing phagocytosed fluorescent beads 24 h after LPS treatment as a percentage of 100 cells evaluated. Nuclei were stained with DAPI (blue). Data are the mean ± SEM (n = 6). #P < 0.01 compared with the control group; **P < 0.01 compared with LPS alone (one-way ANOVA with Bonferroni correction for all comparison pairs).

Fig. 5. Effects of EPO on the expression of inflammatory cytokines and iNOS in the brains of mice stimulated with LPS. Tnfα, Il1b, Il6, and iNOS mRNA levels were evaluated by quantitative RT-PCR. Data are the mean ± SEM (n = 6 in each group). #P 31

< 0.01 compared with the control group; *P < 0.05 compared with LPS alone (one-way ANOVA with Bonferroni correction for all comparison pairs).

Fig. 6. Effects of EPO on microglial morphology in the parietal cortex of the brain of mice stimulated with LPS. (A) Representative photomicrographs showing microglial morphology after staining using antibodies against Iba1 (green). Nuclei were stained with DAPI (blue). Bar, 50 µm. Maximal intensity projections were obtained using a NIKON A1R confocal microscope with a 40× objective, 512×512 pixels. (B) Microglia were stained with antibodies against Iba1 (green) and categorized into four stages of activation, where Stage 0 is the resting stage and Stages 1–3 are activated stages. The number of microglial cells at each stage was counted in random samples. Data are the mean ± SEM (n = 6). #P < 0.01 compared with the control group; **P < 0.01 compared with LPS alone (one-way ANOVA with Bonferroni correction for all comparison pairs).

Fig. 7. Effects of EPO on microglial cell body size in the parietal cortex of the brain of mice stimulated with LPS. (A) Representative photomicrographs showing microglial cell body area (yellow) after staining using antibodies against Iba1 (green). Nuclei were enclosed in red after staining with DAPI (blue).

Bar, 10 µm. The line defines

microglial cell body area. Maximal intensity projections were obtained using a NIKON A1R confocal microscope with a 40× objective, 512×512 pixels. (B) The mean cell body area size in each treatment group. Data are the mean ± SEM (n = 304–591). #P < 0.01 compared with the control group; **P < 0.01 compared with LPS alone (one-way 32

ANOVA with Bonferroni correction for all comparison pairs).

33

34

35

36

37

38

39

Table 1. Primer pairs used for polymerase chain reaction amplification Gene

Primer

Rat Actb

Forward: 5′-TCATGAAGTGTGACGTTGACATCCGT-3′ Reverse: 5′-CCTAGAAGCATTTGCGGTGCAGGATG-3′

EPOR

Forward: 5′-TCTCACTGTTGCTGACTGTGCTG-3′ Reverse: 5′-AAGTTACCCTTGTGGGTGGTGAA-3′

Mouse Gapdh

Forward: 5′-TGTGTCCGTCGTGGATCTGA-3′ Reverse: 5′-TTGCTGTTGAAGTCGCAGGAG-3′

EPOR

Forward: 5′-GCTTCAGCGGATTCTGGAGTG-3′ Reverse: 5′-CGTCAGCAACAGCGAGATGAG-3′

Tnfα

Forward: 5′-TATGGCCCAGACCCTCACA-3′ Reverse: 5′-GGAGTAGACAAGGTACAACCCATC-3′

Il1b

Forward: 5′-TCCAGGATGAGGACATGAGCAC-3′ Reverse: 5′-GAACGTCACACACCAGCAGGTTA-3′

Il6

Forward: 5′-CAACGATGATGCACTTGCAGA-3′ Reverse: 5′-CTCCAGGTAGCTATGGTACTCCAGA-3′

iNOS

Forward: 5′-GCAGAGATTGGAGGCCTTGTG-3′ Reverse: 5′-GGGTTGTTGCTGAACTTCCAGTC-3′

Actb, β-actin; EPOR, erythropoietin receptor; Gapdh, glyceraldehyde-3-phosphate dehydrogenase; Tnfα, tumor necrosis factor-; Il1b, interleukin-1β; Il6, interleukin-6; iNOS,

40

inducible nitric oxide synthase.

41

Highlights • Erythropoietin (EPO) has a neuroprotective effect. • EPO receptor (EPOR) expression is higher in microglia than in neurons. • EPO treatment attenuates microglial activation. • The neuroprotection of EPO may be mediated by regulation of microglial activation.

42