Antagonizing bone morphogenetic protein 4 attenuates disease progression in a rat model of amyotrophic lateral sclerosis

Experimental Neurology 307 (2018) 164–179

Contents lists available at ScienceDirect

Experimental Neurology journal homepage: www.elsevier.com/locate/yexnr

Research Paper

Antagonizing bone morphogenetic protein 4 attenuates disease progression in a rat model of amyotrophic lateral sclerosis

T

Tomomi Shijoa, Hitoshi Waritaa, Naoki Suzukia, Kensuke Ikedaa, Shio Mitsuzawaa, Tetsuya Akiyamaa, Hiroya Onoa, Ayumi Nishiyamaa, Rumiko Izumia, Yasuo Kitajimab, ⁎ Masashi Aokia, a b

Department of Neurology, Tohoku University Graduate School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980-8574, Japan Department of Medicine and Science in Sports and Exercise, Tohoku University Graduate School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980-8574, Japan

A R T I C LE I N FO

A B S T R A C T

Keywords: ALS Amyotrophic lateral sclerosis Antisense oligonucleotide Astrocyte BMP Bone morphogenetic protein Motor neuron Neuroinflammation Non-cell autonomous SOD1

Amyotrophic lateral sclerosis (ALS) is an adult-onset, fatal neurodegenerative syndrome characterized by the systemic loss of motor neurons with prominent astrocytosis and microgliosis in the spinal cord and brain. Astrocytes play an essential role in maintaining extracellular microenvironments that surround motor neurons, and are activated by various insults. Growing evidence points to a non-cell autonomous neurotoxicity caused by chronic and sustained astrocytic activation in patients with neurodegenerative diseases, including ALS. However, the mechanisms that underlie the harmful effects of astrocytosis in patients with ALS remain unresolved. We focused on bone morphogenetic proteins as a major soluble factor that promotes astrocytogenesis and its activation in the adult spinal cord. In a transgenic rat model with ALS-linked mutant Cu/Zn superoxide dismutase gene, BMP4 was progressively up-regulated in reactive astrocytes of the spinal ventral horns, whereas the BMP-antagonist noggin was decreased in association with neuronal degeneration. Continuous intrathecal noggin supplementation after disease onset significantly ameliorated motor dysfunction symptoms, neurogenic muscle atrophy, and extended survival of symptomatic ALS model rats, despite lack of deterrence against neuronal death itself. The exogenous noggin inhibited astrocytic hypertrophy, astrocytogenesis, and neuroinflammation by inactivating both Smad1/5/8 and p38 mitogen-activated protein kinase pathways. Moreover, intrathecal infusion of a Bmp4-targeted antisense oligonucleotides and provided selective Bmp4 knockdown in vivo, which suppressed astrocyte and microglia activation, reproducing the aforementioned results by noggin treatment. Collectively, we clarified the involvement of BMP4 in the processes of excessive gliosis that exacerbate the disease progression of the ALS model rats. Our study demonstrated that BMP4, with its downstream signaling, might be a novel therapeutic target for disease-modifying therapies in ALS.

1. Introduction Amyotrophic lateral sclerosis (ALS) is an adult-onset, fatal neurodegenerative syndrome where patients develop progressive muscle atrophy and weakness, ultimately leading to respiratory failure, generally within a few years of onset (Tandan and Bradley, 1985). To date, there are no available curative therapies for improving the symptoms or halting the devastating effects of motor neuron disease. The pathological hallmark in ALS is prominent and systemic loss of motor neurons in the central nervous system (CNS). This process involves marked activation of astrocytes and microglia, similar to other neurodegenerative diseases such as Alzheimer's disease and Parkinson's disease (Maragakis

⁎

and Rothstein, 2006). Astrocytes are the major non-neuronal cells in the CNS and have a pivotal role in maintaining homeostasis of the extracellular microenvironment (Burda and Sofroniew, 2014). Under various pathological conditions, by both acute and chronic insults to the CNS, astrocytes develop enlarged cell bodies and processes, and strongly express intermediate filaments such as glial fibrillary acidic protein (GFAP), nestin, and vimentin (Ben Haim et al., 2015). When acute CNS injury occurs, perilesional astrocytes form neuroprotective glial scars, which prevent the injury from spreading to the intact regions. However, within the context of neurodegenerative conditions, chronically activated astrocytes have adverse neuronal effects. Growing evidence

Corresponding author at: Department of Neurology, Tohoku University Graduate School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980-8574, Japan. E-mail addresses: [email protected] (T. Shijo), [email protected] (H. Warita), [email protected] (N. Suzuki), [email protected] (K. Ikeda), [email protected] (S. Mitsuzawa), [email protected] (T. Akiyama), [email protected] (H. Ono), [email protected] (A. Nishiyama), [email protected] (R. Izumi), [email protected] (Y. Kitajima), [email protected] (M. Aoki). https://doi.org/10.1016/j.expneurol.2018.06.009 Received 17 March 2018; Received in revised form 25 May 2018; Accepted 15 June 2018 Available online 20 June 2018 0014-4886/ © 2018 Elsevier Inc. All rights reserved.

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

were also approved by the Ethics Committee of Tohoku University Graduate School of Medicine (No. 2017–1-652). The patients with ALS were diagnosed according to the revised El-Escorial criteria (Brooks et al., 2000). Participants' ages, sex, disease durations, CSF cell numbers and CSF protein levels are summarized in Supplementary Table S1. CSF was collected by lumbar puncture, and the specimens were immediately centrifuged at 1000 rpm for 15 min at 4 °C. The supernatant was then aliquoted in sterile 2-ml polypropylene tubes and stored at −80 °C.

suggests that, in patients with ALS and ALS animal models, the reactive astrocytes lose their homeostatic functions secondary to glutamate uptake deficiencies, while expressing high levels of pro-inflammatory chemokines and cytokines (Yamanaka and Komine, 2018). Collectively, regulating the pathological activation of astrocytes may suppress disease progression in patients with ALS. Resident CNS astrocytes have the ability to proliferate, in contrast to terminally differentiated neurons. In addition, throughout the mammalian spinal cord and brain, tissue-specific neural stem/progenitor cells, along with ventricular neuraxis and glial-restricted precursors dispersed throughout the parenchyma, are believed to exist, certainly in adulthood (Horner et al., 2000). Nearly any insult is likely to activate their proliferation and differentiation, producing astrocytes in vivo (Takahashi et al., 2003) (Lepore et al., 2008). Although mechanisms underlying insult-induced astrocytosis in the adult CNS are not fully elucidated, recent progress in developmental and stem cell biology has revealed a variety of extracellular molecules that promote astrocytogenesis and their subsequent maturation. Bone morphogenetic proteins (BMPs) are glycoproteins that belong to the transforming growth factor beta (TGFβ) superfamily. These proteins are indispensable for neural development (Katagiri and Watabe, 2016), as well as for bone and cartilage formation. Among the BMP families, BMP2 and 4 promote astrocytic differentiation from neural stem cells, while inhibiting both neuronal and oligodendrocytic differentiation. In the adult mammalian CNS, BMP2, 4, noggin, and their receptors are expressed (Miyagi et al., 2012) and are involved in pooling neural stem cells by modulating their proliferation and differentiation (Katagiri and Watabe, 2016). BMP2, 4, 6 and 7 are up-regulated under neuropathological conditions such as spinal cord injury (Setoguchi et al., 2004) (Cui et al., 2015) and Alzheimer's disease (Tang et al., 2009). These observations suggest that BMPs play unresolved, but essential, roles in adult astrocytosis. Considering the diverse cellular sources that facilitate astrocytic proliferation and differentiation, we focused on BMPs as potential endogenous regulators of astrocytosis. By investigating their temporal and spatial expressions in the adult rat spinal cord under an ALS-like neurodegenerative condition, we attempted to clarify possible relationships between BMP signaling and the activated astrocytes responsible for non-cell autonomous neurotoxicity to motor neurons.

2.2.2. Rat CSF collection The ALS Tg and non-Tg rats (n = 4 per group, 6 groups) were anesthetized with isoflurane and 1% halothane in a mixture of 30% oxygen and 70% nitrous oxide. After creating a midline incision of the skin over the third to fifth lumbar spinal processes, the fifth lumbar vertebra was laminectomized, and the dura mater was exposed. A pinhole was placed on the dura mater with a 24-gauge needle, and a polyethylene tube (PE10, Becton Dickinson, New Jersey, USA) was inserted into the subarachnoid space. Approximately 100 μl of the rat CSF was collected to the tube and was processed and stored in sterile 0.5-ml polypropylene tubes at −80 °C, along with the aforementioned human CSF. 2.3. Enzyme-linked immunosorbent assay (ELISA) To examine BMP4 and noggin in CSF, we used ELISA kits (Human BMP4 ELISA Kit [LS-F25876, lot 6133] and Mouse/Human/Rat NOG/ Noggin ELISA Kit [LS-F5920, lot 5844] for human CSF, Rat BMP4 ELISA Kit [LS-F26512, lot 6015] and LS-F5920 for rat CSF, Lifespan Biosciences, Washington, USA), according to the manufacturer's instructions. Human CSF samples were not diluted, and rat CSF samples were diluted at 1:10 with dilution buffer. Each sample was run in duplicate, together with freshly prepared standards. The chemiluminescent quantifications for the BMP4 ELISA kits and the absorbance measurements at 450 nm for the noggin ELISA kits were performed using a microplate reader (Varioscan Flash, Thermo Scientific, Massachusetts, USA). 2.4. Vivo-Morpholino® antisense oligonucleotide (ASO) The antisense sequences for Rattus norvegicus Bmp4 mRNA and negative control were designed as shown in Supplementary Table S2 (GeneTools, Oregon, USA). All morpholinos were obtained as prequantitated, sterile, salt-free, lyophilized solids in glass vials (GeneTools). Each ASO was reconstituted to a 0.5 mM solution with sterile water and stocked at room temperature (RT) before administration.

2. Materials and methods 2.1. Experimental animals Transgenic (Tg) female rats overexpressing an ALS-linked mutant human Cu/Zn superoxide dismutase gene (SOD1) were used (Nagai et al., 2001). The Tg rats ubiquitously express His46Arg mutant human SOD1 in the heterozygous state. The ALS Tg rats were divided into three groups: presymptomatic (approximately 20–21 weeks of age, no symptoms, n = 16), early symptomatic (approximately 24–25 weeks of age, mild unilateral hindlimb paresis, n = 16), and late symptomatic stage (approximately 28–29 weeks of age, complete paraplegia, n = 16). Age-matched non-transgenic wild-type littermates (non-Tg, n = 16 per group) served as controls. All experimental procedures were approved by the Animal Committee of the Tohoku University Graduate School of Medicine (No. 2014MdA-202) and were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 8023, revised 1978).

2.5. Continuous intrathecal administration We carefully observed the body weights (BW) and motor functions of the ALS Tg rats every other day, as reported previously (Shijo et al., 2018). We defined the onset of the motor neuron disease as the day we first observed unilateral hind limb paresis. The day after disease onset, recombinant human noggin protein (6057-NG, lot TNT181510A, TNT201601A, TNT1815091, and TNT2317013, R&D Systems, Minnesota, USA) or ASO was administered intrathecally using osmotic pumps (Durect Corporation, California, USA). The preparation of osmotic pumps and their installation via transplant surgery were described previously (Ishigaki et al., 2007). In brief, osmotic pumps (Model number 2002 for noggin and 1003D for ASO, Durect Corporation, California, USA) were filled to capacity with each solution (17.8 μg/kg/ day over 14 days for noggin and 0.40 μg/μl, 1.39 μl/h for 3 days for ASO) or vehicle (phosphate buffered saline [PBS] for noggin and physiological saline for ASO) using a filling needle. The quantity of noggin was determined with reference to a previous report (Matsuura et al., 2008). The quantity and rate of ASO administrations were determined

2.2. Cerebrospinal fluid (CSF) sample collection 2.2.1. Human CSF collection Six sporadic ALS (sALS) patients and six disease controls (patients hospitalized at Tohoku University Hospital from 2012 to 2013) were included. Informed written consent was obtained from all human participants for experimental use of the CSF. All procedures in this study 165

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

2.10. Immunofluorescence

by a pilot study using wild-type rats, which showed mild bilateral hindlimb paralysis and microglial activation in the spinal cord with 40 μg/day for one day of ASO (data not shown). For the infusion, we connected polyethylene tubes (PE60 and PE10, Becton Dickinson, New Jersey, USA). To attain a constant flow rate before use, the pumps were incubated overnight in sterile saline at 37 °C. For transplant surgery, the rats (n = 50 for noggin and n = 22 for ASO administration) were anesthetized using isoflurane, followed by 1% halothane in a mixture of 30% oxygen and 70% nitrous oxide. A skin incision was made and the fifth lumbar vertebra was laminectomized to identify the dura mater. Through a pinhole placed on the dura mater, the polyethylene catheter was inserted into the subarachnoid space approximately 3 cm rostrally. The pumps were removed at day 14 (Model 2002) or day 3 (Model 1003D). We confirmed that all indwelling pumps with catheters were placed properly by the time of their removal.

2.10.1. Spinal cord We dissected the spinal cord specimens and prepared the tissues for immunofluorescence analyses as described previously (Shijo et al., 2018). The deeply anesthetized rats (n = 43) were transcardially perfused with ice-cold heparinized saline followed by perfusion with 4% paraformaldehyde (PFA) in phosphate buffer (PB). The entire spinal cord was quickly removed, immersed in the same fixative for 4 h, cryoprotected using sucrose at 4 °C, frozen in O.C.T. compound (Sakura Finetek, Tokyo, Japan), and stored at −80 °C. Transverse 12-μm-thick sections were prepared using a cryostat (CM1950; Leica Instruments, Nussloch, Germany), collected on MAS-coated glass slides (Superfrost; Matsunami Glass, Osaka, Japan), air-dried at RT, and stored at −80 °C for further study. For immunofluorescence, nonspecific binding was blocked with 5% normal goat serum (Vector Laboratories, Burlingame, CA) and 0.1% triton X-100 (Wako Pure Chemical, Osaka, Japan) in PBS for 15 min after washing thoroughly in PBS. The slides were incubated overnight with primary antibodies (Supplementary Table 3) at 4 °C. After sufficient washing in PBS, the slides were incubated with appropriate secondary antibodies for 120 min at RT. After washing thoroughly in PBS, the sections were embedded in PermaFluor (Dako Agilent Technologies, Glostrup, Denmark). The slides were stored in the dark at 4 °C until further analyses.

2.6. 5-Bromo-2′-deoxyuridine (BrdU) administration To label newly generated cells in vivo, the noggin or vehicle-treated ALS Tg rats (n = 10) were subcutaneously administered BrdU (150 mg/ kg/day), a thymidine analog incorporated into the nuclei of S phase cells, for 7 days using osmotic pumps (2ML1, Durect). The pump was filled to capacity with BrdU solution (dissolved in 50:50 v/v mixture of dimethyl sulfoxide and propylene glycol), capped, and then incubated overnight in sterile saline at 37 °C. Subcutaneous implantation of the pump 2ML1 was performed at the time of intrathecal catheterization, and removed on day 7.

2.10.2. Skeletal muscle The right tibial anterior (TA) muscle was dissected to evaluate muscle fiber diameter. The rats (n = 10) were deeply anesthetized and transcardially perfused with ice-cold heparinized saline. The right TA muscle was dissected out and cleaned of excess fat, connective tissue, and tendons. The muscle was rapidly frozen in an O.C.T. compound using 2-methylbutane cooled with liquid nitrogen, and stored at −80 °C. The horizontal 10-μm-thick sections were cut on a cryostat, collected on MAS-coated glass slides, air-dried, and stored at −80 °C for further study. For immunofluorescence, the tissues were fixed in 4% PFA in PB for 10 min at RT, then in methanol for 10 min at −20 °C, washed in PBS, and blocked with 5% normal goat serum in PBS for 15 min. The slides were incubated overnight at 4 °C with primary antibodies for laminin (Supplementary Table 3) overnight 4 °C to evaluate muscle fiber diameter. After being washed in PBS, the slides were incubated with appropriate secondary antibodies for 60 min at RT. After washing rigorously in PBS, the sections were embedded in PermaFluor. The slides were stored in the dark at 4 °C until further analyses were carried out.

2.7. Survival and motor function analyses After the administration, we carefully observed the rats and evaluated their BW and motor functions (grip strength and landing foot splay test) at days 0, 2, 4, 6, 8, 10, and 12. The endpoint was defined as the day when the BW decreased by 20% of the maximum according to an agreement established with the Animal Committee of the Tohoku University Graduate School of Medicine. Functional assessments were performed by an examiner blinded to treatment assignment in the ASO administration study, while the noggin administration study was conducted in an open-label manner, but was carried out in triplicate (n = 10, 10, and 5 per treatment group) because each quantity of noggin protein was needed to be adjusted by BW of each rat.

2.8. Grip strength 2.11. Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDSPAGE) and immunoblotting

The fore- and hindlimb grip strength were examined using a grip strength meter (MK-380CM/R, Muromachi Kikai, Tokyo, Japan), as described previously (Meyer et al., 1979). For forelimb grip strength, the rat was held from the back and was placed into a trough with the forepaws inside a grasping wire mesh. Using one hand, the rat was grasped by the tail and steadily pulled away from the wire mesh until the grip was broken. For the hindlimb grip strength, the rat continued to be pulled along the trough until the hindlimbs grasped the wire mesh. The trial was completed when the hindlimb grip was broken. We recorded the average of five trials for each rat.

SDS-PAGE and immunoblotting were performed as described previously (Shijo et al., 2018). The deeply anesthetized rats (n = 49 for correcting lumbar cords and n = 10 for correcting TA muscles) were transcardially perfused with ice-cold heparinized saline. The lumbar cord and the left TA muscle were dissected out, and the tissues were collected using 5-mm Derma Punch (Maruho, Osaka, Japan). The tissue specimens were quickly frozen and stored at −80 °C. To extract proteins, the tissues were homogenized in a lysis buffer (T-PER, Thermo Fisher Scientific) containing a protease inhibitor cocktail (cOmplete Mini, Roche Diagnostics, Basel, Switzerland) and a phosphatase inhibitor cocktail (PhosSTOP, Roche Diagnostics) with a disperser (ULTRA-TURRAX® T25 basic, IKA, Staufen, Germany) at 9500 rpm for 45 s on ice. After centrifugation, the protein concentration of the supernatant was determined using a BCA protein assay kit (Thermo Fisher Scientific). The protein samples were boiled and subjected to SDSpolyacrylamide gel electrophoresis through a 10–20% gradient gel (Atto, Tokyo, Japan). Lysate samples containing approximately 20 μg of protein were electrophoresed for 75 min with an electrophoresis

2.9. Landing foot splay test The landing foot splay test was performed as described previously (Aoki et al., 2011). The rat was held from the back with its body parallel to the floor and dropped from a height of approximately 30 cm over a landing point. Landing foot splay distance was defined as the distance between the hind feet measured between the heels. We again recorded the average of five trials for each rat. 166

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

2.13.2. Skeletal muscle immunofluorescence Four to six horizontal sections of TA muscle were used for image analysis. All images were acquired using a confocal laser scanning microscope (A1, Nikon) equipped with an Ar (488 nm) laser unit. At 200× magnification, we randomly captured 10 square images of the regions of interest (1024 × 1024 pixels for 512 × 512 μm) per animal using image acquisition software (NIS ver. 4.30, Nikon). These images were analyzed using the ImageJ 64. The short diameter of each muscle fiber was calculated automatically, and the average value was then determined.

apparatus (PageRunAce, Atto). The samples were electro-blotted onto polyvinylidene difluoride membranes (Immobilon-P, Merck Millipore, MA) using the semi-dry method. The membranes were blocked for 60 min with an appropriate blocking buffer and incubated with a primary antibody (Supplementary Table S3) overnight at 4 °C. After washing in 0.05% Tween-20 in Tris-buffered saline, the membranes were incubated with appropriate horseradish peroxidase-conjugated secondary antibodies for 60 min at RT. Bands were detected using a kit (ECL, GE Healthcare Life Sciences, Little Chalfont, UK) and a luminescent image analyzer (Omega Lum G Imaging System, Gel company, California, USA).

2.13.3. Immunoblotting The all-immunoreactive bands were quantified using the ImageJ 64. We calculated the ratios of target to reference from the same blot.

2.12. Quantitative reverse transcription polymerase chain reaction (qRTPCR)

2.14. Statistical analysis

RNA extraction and qRT-PCR were performed as previously described (Shijo et al., 2018). The deeply anesthetized rats (n = 49) were transcardially perfused with ice-cold heparinized saline. The entire spinal cord was quickly removed, and the lumbar spinal cord tissue was collected using a 5-mm Derma Punch (Maruho). The tissue specimens were immersed in a protection buffer (Allprotect Tissue Reagent, QIAGEN Inc., California, USA) and incubated overnight at 4 °C. The buffer was removed, and the tissue specimens were stored at −80 °C. For RNA extraction, a kit (RNeasy Tissue Mini, QIAGEN) was used according to the manufacturer's instructions. One μg of RNA was reversetranscribed using a kit (QuantiTect Reverse Transcription, QIAGEN). Real-time quantitative RT-PCR (qRT-PCR) was performed using a CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories, California, USA) with iQ SYBR Green Supermix (Bio-Rad) according to the manufacturer's instructions. The primers used in this study are summarized in Supplementary Table S4. We used melting curve analyses to assess each reaction. We evaluated the rat lumbar spinal cord mRNA expressions of 12 reference genes using the Rat Housekeeping Gene Primer Set (Takara, Kyoto, Japan). We further examined their stability using two major software programs, geNorm, and Normfinder (Vandesompele et al., 2002) (Andersen et al., 2004). Consequently, TATA-box binding protein (Tbp) and tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein, zeta (Ywhaz) were chosen as the most appropriate reference genes.

All statistical analyses were performed using GraphPad Prism 6 (GraphPad Software, California, USA). All data were determined to be normally distributed using the D'Agostino-Pearson omnibus test. We performed a one-way analysis of variance (ANOVA), followed by the Tukey-Kramer post hoc test or unpaired t-test. Pearson's correlation coefficient was used to evaluate the relationship between BMP4 or noggin protein levels in human CSF and disease duration. We then used the Mantel-Cox survival analysis test and two-way ANOVA, followed by the Tukey-Kramer post hoc test for the motor function assessment. All results are presented as means ± standard deviations (SD) and P < .05 was considered statistically significant. 3. Results 3.1. BMP4 is up-regulated in the spinal cords and in CSF in symptomatic ALS Tg rats We first examined the expression levels of BMPs (BMP2, 4, 6, and 7), noggin (a major inhibitor of BMP signaling), and BMP receptors in spinal cords of ALS Tg rats. Compared to non-Tg rats, the BMP4 protein levels were significantly increased in the lumbar spinal cords of late symptomatic ALS Tg rats (Fig. 1A1 and B1), although the levels of BMP2, 6, and 7 proteins did not significantly differ among the groups (data not shown). In contrast to BMP4, noggin protein levels significantly decreased in late symptomatic ALS Tg rats (Fig. 1A2 and B2). BMP receptor 1A (BMPR1A) also decreased in symptomatic ALS Tg rats (Fig. 1A3 and B3), although there was no significant between-groups differences in the protein levels of BMP receptor 1B (BMPR1B) and 2 (data not shown). In conjunction with increases in BMP4, mothers against decapentaplegic homolog (Smad) 1/5/8, a major downstream signaling molecule of BMP, was significantly activated in symptomatic ALS Tg rats, even from the early phases of the disease (Fig. 1A4, B4, and B5). As for non-Smad pathways of BMP signaling, p38 mitogen-activated protein kinase was also activated in symptomatic ALS Tg rats, as reported previously (data not shown) (Tortarolo et al., 2003). As expected, we observed a significant increase of Bmp4 expression in the lumbar spinal cords of symptomatic ALS Tg rats by qRT-PCR (Fig. 1C1). On the other hand, the levels of Noggin and Bmpr1a expression did not change (Fig. 1C2 and C3). When we evaluated CSF levels of BMP4 and noggin, both of which are secreted glycoproteins, the BMP4 protein was significantly increased in the CSF of late symptomatic ALS Tg rats, compared with presymptomatic ALS Tg rats (Fig. 1D). However, there were no significant between-group differences in noggin protein levels (Supplementary Fig. S1A). To confirm similar changes in the CSF levels of BMP4, we examined six samples from patients with sALS and found a significant increase of BMP4, in accordance with the disease duration (Fig. 1E2). There was no significant difference in BMP4 protein levels between the sALS and the disease controls (Fig. 1E1). As for noggin protein levels in CSF, we found no significant between-group

2.13. Image analyses and quantification 2.13.1. Spinal cord immunofluorescence Image analysis and quantification were performed as described previously (Shijo et al., 2018). Four to six transverse sections, separated by at least 36 μm, were used for the analysis. All images were acquired with a confocal laser scanning microscope (A1; Nikon, Tokyo, Japan) equipped with Ar (488 nm), HeNe-green (561 nm), and HeNe-red (640 nm) laser units. At 200× magnification, square regions of interest (1024 × 1024 pixels for 256 × 256 μm) were captured in the bilateral ventral horns, dorsal horns, and central gray matter using image acquisition software (NIS ver. 4.30, Nikon). Using each antibody, all images were captured with identical settings for normalization. The digital images were stored as TIFF files using ImageJ 64 software (Wayne Rasbad, NIH). We set a constant threshold for positive, and all images were binarized. We calculated the number of positive pixels to determine their areas. Immmunoreactivity was evaluated independently in each immunostaining, and semi-quantified in a relative manner. We employed cell-selective markers as follows: neuronal nuclear antigen (NeuN)/Fox3 for neurons, polysialic acid -neural cell adhesion molecule (PSA-NCAM) with Hu C/D for neural precursors, GFAP for astrocytes, glutathione S-transferase-pi (GST-pi) for oligodendrocytes, neuron-glial antigen 2 (NG2) for glial precursors, ionized calcium-binding adaptor molecule 1 (Iba-1) for microglia, and CD68 (clone ED1) for activated microglia. The average values calculated from four to six sections, obtained from each individual rat were statistically analyzed. All images were evaluated by a blinded examiner. 167

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

Fig. 1. BMP4 but not noggin is up-regulated in spinal cords of ALS Tg rats and CSF of ALS Tg rats and sporadic ALS patients. A: Immunoblotting of BMP4 (A1), noggin (A2), BMP receptor 1A (BMPR1A, A3), and pSmad1/5 (A4) in the lumbar cords of age-matched littermates (non-Tg) and ALS transgenic (Tg) rats. GAPDH: glyceraldehyde 3-phosphate dehydrogenase. B: Quantitative immunoblotting analyses of A. The graph charts show the relative optical density values of the bands. The ratios of BMP4 (B1), noggin (B2), BMPR1A (B3), pSmad1/5 (B4), and Smad5 (B5) to those of the internal control protein are shown. n = 4 per group, mean ± SD, one-way analysis of variance (ANOVA) followed by Tukey-Kramer post hoc test. C: Quantitative reverse transcription polymerase chain reaction (RT-PCR) analyses of Bmp4 (C1), Noggin (C2) and Bmpr1a (C3) in the rat lumbar cords. The ratios of target gene to those of the internal control (tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein, zeta, Ywhaz) gene are shown. The same results are obtained with using TATA-box binding protein (Tbp) as the internal control (data not shown). n = 4 per group, mean ± SD, one-way ANOVA followed by Tukey-Kramer post hoc test. D: Enzyme-linked immunosorbent assay (ELISA) analysis of BMP4 protein in CSF of ALS Tg rats. n = 4 per group, Mean ± SD, one-way ANOVA followed by Tukey-Kramer post hoc test. E: ELISA analysis of BMP4 protein in CSF of sporadic ALS (sALS) patients. The BMP4 levels between sALS and control (E1, P = .6045, mean ± SD, n = 6 per group) or the correlation of BMP4 and disease durations of sALS patients were evaluated (E2, R2 = 0.9135, P = .0029, n = 6). Unpaired t-test (E1) or Pearson's correlation coefficient (E2). *P < .05, **P < .01, ***P < .001. ALS: amyotrophic lateral sclerosis, BMP: bone morphogenetic protein, CSF: cerebrospinal fluid, SD: standard deviation. Pre: presymptomatic, ES: early symptomatic, LS: late symptomatic. NS: not significant.

BMP4 expression in CD68+ reactive microglia in ALS Tg rats (Supplementary Fig. S2A4–6). In contrast to the BMP4, noggin expression was mainly observed in neurons of both ALS Tg and non-Tg rats (Fig. 2B1–6). In ALS Tg rats, the expression of noggin decreased significantly in the ventral horn neurons, in accordance with disease progression (Fig. 2F1–3). Those differences were not detected in the dorsal horns (data not shown). We could not detect noggin expression in CD68+ reactive microglia in ALS Tg rats (Supplementary Fig. S2B4–6). BMPR1A was also expressed mainly in neurons of both ALS Tg and non-Tg rats (Fig. 2C1–6). BMPR1A expression, particularly in the ventral horn neurons, significantly decreased according to neuronal degeneration in symptomatic ALS Tg rats (Fig. 2G1 and G2). In contrast, GFAP+ astrocytes showed progressive expression of the BMPR1A even from the presymptomatic phase (Fig. 2G3). The reactive microglia, which express BMPR1A, were rarely detected (Supplementary Fig. S2C4–6). Phosphorylated Smad1/5/8 (pSmad1/5/8) was expressed mainly in astrocytes and neurons of both ALS Tg and non-Tg rats

differences and no correlation with disease duration (Supplementary Fig. S1B1 and B2). 3.2. BMP4 selectively increases in astrocytes of ALS Tg rats after disease onset Next, we examined the cellular source of up-regulated BMP4 and the down-regulation of its inhibitor noggin in the spinal ventral horns of ALS Tg rats, where the pathological changes were remarkable. Multiple immunofluorescence staining revealed that BMP4 is predominantly expressed in GFAP-positive (GFAP+) astrocytes of symptomatic ALS Tg rats (Fig. 2A5 and A6), in contrast to the expression of BMP4 in NeuN+ neurons of presymptomatic ALS Tg rats (Fig. 2A4) and non-Tg rats (Fig. 2A1–3). Total BMP4 expression (Fig. 2E1), and its astrocytic expression in ALS Tg rats (Fig. 2E3), significantly increased in accordance with progressive astrocytosis (Fig. 2E2). Those differences were not detected in the dorsal horns (data not shown). We could not detect 168

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

(caption on next page)

169

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

Fig. 2. BMP4, BMP receptor 1A (BMPR1A) and pSmad1/5/8 are increased in astrocytes of ALS Tg rats and conversely noggin decreased in neurons in accordance with disease progression. Immunofluorescence analyses of the lumbar spinal ventral horns of age-matched littermates (non-Tg) and ALS transgenic (Tg) rats. A: Triple immunofluorescence staining of the expression of BMP4 (green), neuronal nuclear antigen (NeuN, red), and glial fibrillary acidic protein (GFAP, cyan as a pseudo color or blue in merged images) in non-Tg (A1–A3) and ALS Tg (A4–A6) rats. The arrowheads in enlarged pictures indicate neurons expressing BMP4, and the arrows indicate astrocytes expressing BMP4. B: Triple immunofluorescence staining of the expression of noggin (green), NeuN (red), and GFAP (cyan as a pseudo color or blue in merged images) in non-Tg (B1–B3) and ALS Tg (B4–B6) rats. The arrowheads in enlarged pictures indicate neurons expressing noggin. C: Triple immunofluorescence staining of the expression of BMPR1A (green), NeuN (red), and GFAP (cyan as a pseudo color or blue in merged images) in non-Tg (C1–C3) and ALS Tg (C4–C6) rats. The arrowheads in enlarged pictures indicate neurons expressing BMPR1A and the arrows indicate astrocytes expressing BMPR1A. D: Triple immunofluorescence staining of the expression of pSmad1/5/8 (green), neuronal nuclear antigen (NeuN, red), and glial fibrillary acidic protein (GFAP, cyan as a pseudo color or blue in merged images) in non-Tg (D1–D3) and ALS Tg (D4–D6) rats. The arrows in enlarged pictures indicate astrocytes expressing pSmad1/5/8. Scale bars = 100 μm (in merged pictures) or 50 μm (in enlarged pictures), V: ventral. E: Semi-quantification of A. BMP4+ (E1), GFAP+ (E2), and BMP4+GFAP+ (E3) areas are evaluated. F: Semiquantification of B. noggin+ (F1), NeuN+ (F2), and noggin+NeuN+ (F3) areas are evaluated. G: Semi-quantification of C. BMPR1A+ (G1), BMPR1A+NeuN+ (G2), and BMPR1A+GFAP+ (G3) areas are evaluated. H: Semi-quantification of D. pSmad1/5/8+ (H1) and pSmad1/5/8+GFAP+ (H2) areas are evaluated. E–H: n = 4 per group, Mean ± SD, one-way analysis of variance (ANOVA) followed by Tukey-Kramer post hoc test. Mean ± SD. *P < .05, **P < .01, ***P < .001. ALS: amyotrophic lateral sclerosis, BMP: bone morphogenetic protein, SD: standard deviation, Pre: presymptomatic, ES: early symptomatic, LS: late symptomatic. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

and vimentin/GFAP double-positive (vimentin+/GFAP+) activated astrocytes were significantly decreased in the lumbar spinal ventral horns of ALS Tg rats (Fig. 4A1, A2, B1, and B2). Immunoblotting and qRT-PCR using samples extracted from whole lumbar spinal cords strengthened the findings of down-regulation of GFAP and vimentin by noggin administration (Fig. 4C1, C2, D1, D2, E1, and E2). Moreover, the number of BrdU+ astrocytes, which proliferated or were newly generated in the first half of the noggin supplementation, decreased significantly (Fig. 4F and G). Additionally, ependymal cells containing a potential neural stem/progenitor population around the central canal of the lumbar spinal cord were more abundant in the noggin-treated group (Supplementary Fig. S4A1, A2, and B). Regardless of the treatment group, we could not detect BrdU+ neuronal or neuroblastic (BrdU+NeuN+ or BrdU+PSA-NCAM+Hu+) cells in the ventral horns (data not shown). The cell types of BrdU+ are summarized in Supplementary Fig. S4C.

(Fig. 2D1–6). Total pSmad1/5/8 expression (Fig. 2H1), and its astrocytic expression in ALS Tg rats (Fig. 2H2), significantly increased in accordance with progressive astrocytosis. 3.3. Exogenous noggin supplementation attenuates disease progression and motor dysfunction in ALS Tg rats Based on these observations, we hypothesized that the BMP4-expressing reactive astrocytes might cause progressive astrocytosis, resulting in sustained neuroinflammation at the site of neurodegeneration in ALS Tg rats. Therefore, antagonizing BMPs might reduce astrocytosis and possibly attenuate motor neuron disease. To inhibit the BMP signaling in the spinal cords of ALS Tg rats, we administered human recombinant noggin protein intrathecally using osmotic pumps just after disease onset, as shown in Fig. 3A. Before administration, we confirmed that there were no significant differences between the vehicle- and noggin-treated rat groups with regard to age (Vehicle versus Noggin: 178.1 ± 3.278 versus 180.6 ± 3.156 days, mean ± SD, P = .5895) and BW (271.6 ± 4.277 versus 276.7 ± 3.853 g, mean ± SD, P = .3853) at day 0. Compared with the vehicle-treated group, the noggin-treated ALS Tg rats showed significant retention of BW (Fig. 3B) over days 8–12. Moreover, exogenous noggin supplementation improved measures of motor dysfunction (Fig. 3C1, C2 and D), and eventually extended the duration of disease by 7 days (Fig. 3E, P = .0014, median survival: 10.00 days in vehicle- and 17.00 days in noggin-treated group). In addition, the noggin administration attenuated the progressive neurogenic atrophy of TA muscle fibers in ALS Tg rats, likely illustrating the abovementioned phenotypical improvement (Fig. 3F1, F2 and G). We further confirmed that both the phosphorylated neurofilament heavy chain (pNF–H) with ubiquitin (Ub)+ aggregates and GFAP+ astrocytes containing Ub+ aggregates (the hallmarks of neurodegeneration in this model of ALS) significantly decreased in the noggin-administered group (Fig. 3H1, H2, I, and J), although there was no significant difference in the remaining lumbar spinal ventral horn neurons between the noggin- and vehicle-treated groups (Supplementary Fig. S3A1, A2, and S3B). We confirmed that endogenous Bmp4 expression increased and Noggin expression decreased in the lumbar spinal cords following exogenous noggin administration (Supplementary Fig. S3C and S3D), although no significant differences were detected in the protein levels within the lumbar spinal cords (Supplementary Fig. S3E, S3F, S3G, and S3H) and left TA muscles (data not shown).

3.5. Noggin also attenuates microgliosis and neuroinflammation in ALS Tg rats Microglial activation was also suppressed by noggin administration in ALS Tg rats. CD68+ activated microglia were significantly decreased in the lumbar spinal ventral horns of ALS Tg rats after 14 days of noggin supplementation (Fig. 5A1, A2, and B1). Immunoblotting revealed a significant decrease in CD68 in the lumbar spinal cords of noggintreated ALS Tg rats, supporting a reduction in microglial activation (Fig. 5C and D). In contrast, the immunoreactivity of Iba-1+ pan-microglia did not significantly differ between the vehicle- or noggintreated groups (Fig. 5A1, A2, and B2). Among neuro-inflammatory cytokines, interleukin-1 beta (IL-1β) expression significantly decreased, particularly in astrocytes, in the noggin-administered ALS Tg rats (Fig. 5E1, E2, F1, and F2). Additionally, inducible nitric oxide synthase (iNOS) expression also significantly decreased, both in astrocytes and microglia, by noggin administration (Fig. 5G1, G2, and H1–3). Moreover, tumor necrosis factor alpha (TNFα) decreased significantly, both in astrocytes and activated microglia (Fig. 5I1, I2, and J1–3), as well as iNOS. 3.6. Smad and non-Smad BMP pathways in astrocytes are inhibited by noggin We found that the Smad1/5/8 pathway was significantly downregulated by noggin (Fig. 6A, C1, and C2). The pSmad 1/5/8 levels of noggin-treated ALS Tg rats decreased to those of non-Tg rats (Supplementary Fig. S5A and C). Moreover, GFAP+ astrocytes expressed less pSmad1/5/8 in the lumbar spinal ventral horns of noggin-treated ALS Tg rats (Fig. 6E1, E2, F1, and F2). However, we could not detect the pSmad1/5/8+ activated microglia in any group (Fig. 6E1 and E2, enlarged pictures). As with the Smad pathway, p-p38MAPK significantly

3.4. Noggin suppresses astrocytosis in the spinal ventral horns of ALS Tg rats Continuous intrathecal noggin administration was found to inhibit reactive astrocytosis of ALS Tg rats. At day 14, both GFAP+ astrocytes 170

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

3.7. Rat Bmp4-targeted ASO also suppresses astrocytosis and microgliosis in the spinal ventral horn of ALS Tg rats

decreased by noggin administration (Fig. 6B, D1, and D2). However, the p-p38MAPK levels of noggin-treated ALS Tg rats were still higher than those of non-Tg rats (Supplementary Fig. S5B and D). In the noggin-treated ALS Tg rats, both the astrocytes and the activated microglia expressed less p-p38MAPK in the lumbar spinal ventral horns (Fig. 6G1, G2, and H1–3).

In addition to noggin, we administered Bmp4-targeted VivoMorpholino® ASO (Bmp4-ASO) intrathecally to determine if selective down-regulation of BMP4 led to decreased gliosis in the spinal ventral horns of ALS Tg rats. This was thought to be because noggin

(caption on next page) 171

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

Fig. 3. Noggin attenuates muscle fiber atrophy, neuronal degeneration, motor dysfunction, and extends survival of ALS Tg rats. A: Experimental paradigm. Recombinant human noggin or phosphate buffered saline (PBS) was intrathecally injected to ALS Tg rats at day 0, the day after disease onset (day −1). 5-Bromo-2′-deoxyuridine (BrdU) was administrated subcutaneously for first half 7 days. Body weights (BW) and motor functions were evaluated every other day and rats were then sacrificed at day 14 for histological analyses or kept alive for survival analysis. B: BW compared with maximum weight of each rat. Mean ± SD, n = 10 per group, two-way analysis of variance (ANOVA) followed by Tukey-Kramer post hoc test. C, D: The motor function evaluations. Grip strength of fore- (C1) and hindlimb (C2) and the scores of landing foot splay (D) are evaluated. Mean ± SD, n = 10 per group, two-way ANOVA followed by Tukey-Kramer post hoc test. E: The Mantel-Cox sumulative survival plot. n = 10 per group. P = .0014. F: Reprensative microphotographs of immunofluoresence of ALS Tg rat tibial anterior (TA) muscles (F1: vehicle- and F2: noggin-treated). Scale bar = 100 μm. G: Quantification of F. P = .0074. Mean ± SD, n = 5 per group, unpaired t-test. H: Triple immunofluorescence staining of the expression of phosphorylated neurofilament heavy chain (pNF–H, green), ubiquitin (Ub, red), and glial fibrillary acidic protein (GFAP, cyan as a pseudo color or blue in merged images) of ALS Tg rat lumbar spinal ventral horns (H1: vehicle- and H2: noggin-treated). The arrowheads indicate pNF-H accumulations which contain Ub+ aggregates. The arrows indicate GFAP+ astrocytes which contain Ub+ aggregates. Scale bar = 50 μm. I, J: Semiquantification of H. pNF+Ub+ (I, P = .0008) and GFAP+Ub+ (J, P = .0009) area are evaluated. Mean ± SD, n = 5 per group, unpaired t-test. *P < .05, **P < .01, ***P < .001. ALS: amyotrophic lateral sclerosis, SD: standard deviation. Pre: presymptomatic, ES: early symptomatic, LS: late symptomatic. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 4. Noggin inhibits both astrocytic hypertrophy and astrocytogenesis in ALS Tg rats. A: Double immunofluorescence staining of the expression of vimentin (green) and glial fibrillary acidic protein (GFAP, magenta) of ALS Tg rat lumbar spinal ventral horns (A1: vehicle- and A2: noggin-treated). Scale bar = 100 μm, V: ventral. B: Semi-quantification of A. GFAP+ (B1, P = .0014) and vimentin+GFAP+ (B2, P = .0155) areas are evaluated. Mean ± SD, n = 5 per group, unpaired t-test. C: Immunoblotting of GFAP (C1) and vimentin (C2) in the lumbar cords of ALS Tg rats. GAPDH: glyceraldehyde 3-phosphate dehydrogenase. D: Quantitative immunoblotting analyses of C. The graph charts show the relative optical density values of the bands. The ratios of GFAP (D1, P = .0039) and vimentin (D2, P = .0004) to those of GAPDH are shown. Mean ± SD, n = 5 per group, unpaired t-test. E: Quantitative reverse transcription polymerase chain reaction (RT-PCR) analyses of Gfap (E1, P = .0035) and Vimentin (E2, P = .0008) in the ALS Tg rat lumbar cords. The ratios of target gene to those of the internal control (tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein, zeta, Ywhaz) gene are shown. The same results are obtained with using TATA-box binding protein (Tbp) as the internal control (data not shown). Mean ± SD, n = 5 per group, unpaired t-test. F: Triple immunofluorescence staining of the expression of 5-Bromo-2′-deoxyuridine (BrdU, green), vimentin (red), and GFAP (cyan as a pseudo color or blue in merged image) of a vehicle-treated ALS Tg rat lumbar spinal ventral horn. The arrowhead indicates a newly generated astrocyte. Scale bar = 50 μm. G: Quantification of BrdU+ newly generated astrocytes. P < .0001. Mean ± SD, n = 5 per group, unpaired t-test. *P < .05, **P < .01, ***P < .001. ALS: amyotrophic lateral sclerosis, SD: standard deviation. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

172

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

extracted from rat lumbar spinal cords, we confirmed that Bmp4 mRNA decreased significantly in the group that received Bmp4-ASO (Fig. 7F), but not Bmp2, 6, or 7 mRNA levels (Supplementary Fig. S6A1–A3). The BMP4 protein and its downstream signaling Smad1/5/8 phosphorylation also decreased in the Bmp4-ASO-administered group (Fig. 7G1, G2, H1, and H2), although the total Smad5 protein level did not change (Fig. 7H3). Next, we examined the cellular source of BMP4 down-regulation and gliosis in the lumbar spinal ventral horns of ALS Tg rats with immunofluorescence staining. We confirmed that BMP4 decreased significantly (Fig. 7I1–3 and K1), particularly in vimentin-positive reactive astrocytes (Fig. 7I1–3 and K4). Selective down-regulation of Bmp4 inactivated both astrocytes (Fig. 7I1–3, K2 and K3) and microglia (Fig. 7J1–3 and K5), as was shown by the noggin supplementation. The Iba-1+ pan-microglia (Fig. 7J1–3 and K6) and NeuN+ remaining neurons (Supplementary Fig. S6B1, B2, and C) were not influenced by Bmp4 suppression.

antagonizes, not only BMP4, but other BMPs such as BMP2, 6, and 7. The Bmp4-ASO was administered as indicated in Fig. 7A. Before administration, we confirmed that there were no significant betweengroup differences in average age (Vehicle versus negative control [NC] -ASO versus Bmp4-ASO: 188.3 ± 5.382 versus 186.8 ± 5.382 versus 190.5 ± 4.983 days, mean ± SD, P = .7551) and BW (267.7 ± 13.20 versus 278.8 ± 13.20 versus 280.5 ± 12.22 g, mean ± SD, P = .3853) at day 0. As with our comparison to the NC-ASO-treated group, the Bmp4-ASO-treated ALS Tg rats showed significant BW retention (Fig. 7B) at day 12. In addition, the selective suppression of endogenous Bmp4 expression improved hindlimb motor dysfunction (Fig. 7C2 and D), although forelimb grip strength was not changed (Fig. 7C1). Moreover, the Bmp4-ASO-administered rats showed an extended survival compared with the vehicle- and NC-ASO-treated group (Fig. 7E, P = .0232, median survival: 10.00 days in vehicle, 8.00 days in NC-ASO, and undefined in Bmp4-ASO-treated group). Using samples

(caption on next page) 173

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

Fig. 5. Noggin suppresses neuroinflammation in spinal ventral horns of ALS Tg rats. A: Double immunofluorescence staining of the expression of CD68 (green) and ionized calcium-binding adaptor molecule 1 (Iba-1, magenta) of ALS Tg rat lumbar spinal ventral horns (A1: vehicle- and A2: noggin-treated). Scale bar = 100 μm, V: ventral. B: Semi-quantification of A. CD68+ (B1, P = .0064) and Iba-1+ (B2, P = .9668) area are evaluated. Mean ± SD, n = 5 per group, unpaired t-test. C: Immunoblotting of CD68 in the lumbar cords of ALS Tg rats. GAPDH: glyceraldehyde 3-phosphate dehydrogenase. D: Quantitative immunoblotting analysis of C. The graph chart shows the relative optical density values of the bands. The ratios of CD68 to those of GAPDH are shown. P = .0114. Mean ± SD, n = 5 per group, unpaired t-test. E: Triple immunofluorescence staining of the expression of interleukin 1 beta (IL-1β, green), CD68 (red), and glial fibrillary acidic protein (GFAP, cyan as a pseudo color or blue in merged pictures) of ALS Tg rat lumbar spinal ventral horns (E1: vehicle- and E2: noggin-treated). The arrows in enlarged pictures indicate GFAP+ astrocytes which express IL-1β. Note that the activated microglia which express IL1β can not be detected (see enlarged pictures). Scale bar = 100 μm (in a merged picture) or 50 μm (in an enlarged picture), V: ventral. F: Semi-quantification of E. IL1β+ (F1, P = .0015) and IL-1β+GFAP+ (F2, P = .0002) areas are evaluated. Mean ± SD, n = 5 per group, unpaired t-test. G: Triple immunofluorescence staining of the expression of inducible nitric oxide synthase (iNOS, green), Iba-1 (red), and GFAP (cyan as a pseudo color or blue in merged pictures) of ALS Tg rat lumbar spinal ventral horns (G1: vehicle- and G2: noggin-treated). The arrows in enlarged pictures indicate GFAP+ astrocytes which express iNOS and the arrowheads indicate Iba1+ microglia which express iNOS. Note that the astrocytes predominantly express iNOS compared with activated microglia (see enlarged pictures). Scale bar = 100 μm (in a merged picture) or 50 μm (in an enlarged picture), V: ventral. H: Semi-quantification of G. iNOS+ (H1, P = .0082), iNOS+GFAP+ (H2, P = .0057), and iNOS+Iba-1+ (H3, P = .0021) areas are evaluated. Mean ± SD, n = 5 per group, unpaired t-test. I: Triple immunofluorescence staining of the expression of tumor necrosis factor alpha (TNFα, green), CD68 (red), and GFAP (cyan as a pseudo color or blue in merged pictures) of ALS Tg rat lumbar spinal ventral horns (I1: vehicle- and I2: noggin-treated). The arrows in enlarged pictures indicate GFAP+ astrocytes which express TNFα and the arrowheads indicate CD68+ activated microglia which express TNFα. Note that the astrocytes predominantly express TNFα compared with activated microglia (see enlarged pictures). Scale bar = 100 μm (in a merged picture) or 50 μm (in an enlarged picture), V: ventral. J: Semi-quantification of I. TNFα+ (J1, P = .0003), TNFα+GFAP+ (J2, P = .0001), and TNFα+CD68+ (J3, P = .0008) areas are evaluated. Significantly decrease by noggin administration. Mean ± SD, n = 5 per group, unpaired t-test. *P < .05, **P < .01, ***P < .001. ALS: amyotrophic lateral sclerosis, SD: standard deviation, NS: not significant. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

4. Discussion

insufficient against chronic and progressive increases of BMP4 from reactive astrocytes. Although acute injuries such as stroke reportedly increase noggin in response to BMP4 up-regulation (Lei et al., 2012), previous reports of neurodegenerative disease models support the idea that noggin is down-regulated, in contrast to BMP4 up-regulation (Tang et al., 2009) (Meyers et al., 2016). In our study, antagonizing BMPs, particularly BMP4, held a considerable role in the development of progressive and sustained astrocytosis in this ALS model. Continuous intrathecal administration of human recombinant noggin successfully inhibited the Smad1/5/8 pathway, and reduced GFAP and vimentin expressions in spinal astrocytes. Moreover, the exogenous noggin suppressed generation of mature astrocytes in the spinal ventral horn of ALS model rats, which reportedly constituted approximately 7% of proliferating cells in ALS model mice (Lepore et al., 2008). Together, our results suggest that antagonism of the up-regulated BMP4 affects both resident and newly generated astrocytes in the microenvironment surrounding spinal motor neurons. Additionally, noggin promoted proliferation of neural stem/progenitor-like cells around the central canal and NG2+ glial progenitors in the spinal ventral horn. The activation of adult neural stem/progenitor cells around the central canal in the ALS model mice was previously associated with absent neurogenesis, as a response to motor neuron degeneration (Ohta et al., 2006) (Guan et al., 2007). Overexpression of noggin in mice expands neural progenitor cells in the adult neurogenic niche (Meyers et al., 2016) (Morell et al., 2015). Oligodendrocyte progenitors also reportedly express BMP receptors (Kondo and Raff, 2004), and are proliferated and differentiated by noggin-overexpression in hypoxia-ischemia model mice (Dizon et al., 2011). Therefore, BMPs may be directly involved in proliferation and differentiation of both neural stem/progenitor and glial progenitor cells in the adult spinal cord. Consequently, BMPs may serve as possible targets for regulating endogenous regenerative responses under ALSlike neurodegenerative conditions. By inhibiting BMP signaling mainly within the BMPR1A-expressing astrocytes themselves, exogenous human recombinant noggin supplementation, after disease onset, effectively reduced reactive astrocytosis, indirectly but convincingly attenuated progressive motor dysfunction, preserved muscle fiber volume, and eventually extended the survival of ALS Tg rats. Behind the improvement (intriguingly without significant change of NeuN+ neuronal loss), attenuation of aberrant perinuclear pNF-H accumulation and axonal/astrocytic ubiquitinated protein deposition, which are pathological features in the ALS model (Bruijn et al., 1997) (Stieber et al., 2000), were confirmed in noggin-treated rat spinal ventral horn neurons. Although BMPs and their receptors are

We found that BMP4, a secreted glycoprotein, was predominantly up-regulated in the reactive astrocytes, as well as in the CSF of symptomatic Tg rats with ALS-linked mutant SOD1. The induction of BMP4 paralleled remarkable astrocytosis and progressive neuronal loss, where the activated astrocytes expressed the receptor BMPR1A. This suggests a vicious cycle underlying the sustained neuroinflammation at the site of neurodegeneration. Moreover, inhibiting excessive BMP4 by both exogenous noggin and Bmp4-targeted ASO, beginning in the early symptomatic stage, significantly attenuated the disease progression by suppressing BMP-Smad signaling, astrocytosis, and microgliosis in the spinal cords of ALS Tg rats. Although this study employed a candidate molecule approach, rather than a comprehensive one, these novel findings illustrate the significance of extracellular microenvironments surrounding the motor neurons, where excessive BMP4-linked progressive astrocytosis provokes persistent neuroinflammation as an aggravating neurotoxic factor in the mutant SOD1-induced ALS. Among the BMP family expressed in the adult rat spinal cord, BMP4 was selectively elevated in the spinal ventral horn astrocytes of ALS Tg rats, beginning in the early symptomatic stage. Even from the presymptomatic phase, BMPR1A, a membrane receptor for BMP4, was upregulated in ventral horn astrocytes. Although total BMPR1A levels in the spinal cords of symptomatic ALS Tg rats decreased, this may reflect BMPR1A-expressing neuronal degeneration. Moreover, Smad1/5/8, the downstream molecules activated by BMP4-BMPR1A binding, were increasingly phosphorylated in the spinal cords of symptomatic ALS Tg rats, with astrocytes as the major source. The Smad1/5/8 pathway promotes Gfap expression by cooperating with signal transducers and activators of the transcription 3 (STAT3) pathway (Nakashima et al., 1999), which is also commonly activated in mutant SOD1-Tg mice and in patients with sALS (Shibata et al., 2010). The activated astrocytes mainly produce BMP4, possibly in an autocrine manner, and partially contribute to the sustained and progressive astrocytosis in this model. In contrast to BMP4, endogenous noggin decreased in late symptomatic ALS Tg rats, although Noggin mRNA levels did not change. Noggin is a secreted glycoprotein, as well as BMP. As one of the over 20 physiological antagonists of BMP, it can bind directly to BMP2/4 with high affinity, inhibiting Smad1/5/8 activation (Krause et al., 2011). However, its physiological functions in the adult mammalian CNS are poorly understood. In the adult spinal ventral horns, neurons mainly express noggin, which decreased in accordance with neurodegeneration in this ALS model. Therefore, the loss of noggin-producing neurons results in consumption of endogenous noggin, likely rendering the supply 174

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

Fig. 6. Noggin inactivates Smad1/5/8 pathway in astrocytes and p38MAPK pathway in astrocytes and activated microglia in ALS Tg rats. A, B: Immunoblotting of pSmad1/5 (A) and p-p38MAPK (B) in the lumbar cords of ALS Tg rats. GAPDH: glyceraldehyde 3-phosphate dehydrogenase. C, D: Quantitative immunoblotting analyses of A and B. The graph charts show the relative optical density values of the bands. The ratios of pSmad1/5 (C1, P = .0002), Smad5 (C2, P = .1752), p-p38MAPK (D1, P = .0009), and p38MAPK (D2, P = .1406) to those of the internal control protein are shown. Mean ± SD, n = 5 per group, unpaired t-test. E: Triple immunofluorescence staining of the expression of pSmad1/5/8 (green), CD68 (red), and glial fibrillary acidic protein (GFAP, cyan as a pseudo color or blue in merged pictures) of ALS Tg rat lumbar spinal ventral horns (E1: vehicle- and E2: noggin-treated). The arrows in enlarged pictures indicate GFAP+ astrocytes which express pSmad1/5/8. Scale bar = 100 μm (in a merged picture) or 50 μm (in an enlarged picture), V: ventral. F: Semi-quantification of E. pSmad1/5/8+ (F1, P = .0022) and but pSmad1/5/8+GFAP+ (F2, P = .0039) areas are evaluated. Mean ± SD, n = 5 per group, unpaired t-test. G: Triple immunofluorescence staining of the expression of p-p38MAPK (green), CD68 (red), and GFAP (cyan as a pseudo color or blue in merged pictures) of ALS Tg rat lumbar spinal ventral horns (G1: vehicle- and G2: noggin-treated). The arrows in enlarged pictures indicate GFAP+ astrocytes which express p-p38MAPK and the arrowheads indicate CD68+ activated microglia which express p-p38MAPK. Scale bar = 100 μm (in a merged picture) or 50 μm (in an enlarged picture), V: ventral. H: Semiquantification of G. p-p38MAPK+ (H1, P = .0011), p-p38MAPK+GFAP+ (H2, P = .0001), and p-p38MAPK+CD68+ (H3, P = .0023) areas are evaluated. Mean ± SD, n = 5 per group, unpaired t-test. *P < .05, **P < .01, ***P < .001. ALS: amyotrophic lateral sclerosis, MAPK: mitogen- activated protein kinase, SD: standard deviation. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

and neuroinflammation in a mouse model of ALS reportedly preserved the NMJ without affecting the remaining motor neurons (Ringer et al., 2017). Additionally, noggin administration enhanced axonal growth in a rat spinal cord injury model (Matsuura et al., 2008), also supporting this idea. However, the precise mechanisms underlying improvements in motor neuron function by BMP antagonism are unclear. The effects of BMP inhibition by intrathecally administered noggin were not restricted to the reduction of progressive astrocytosis. We also

expressed in neurons (Miyagi et al., 2012), the role of BMPs on mature neurons remains unclear. Our findings strongly suggest non-cell autonomous neurotoxicity by the activated astrocytes of the animals with the ALS-linked mutant SOD1 (Nagai et al., 2007) (Yamanaka et al., 2008) and from the patients with sporadic ALS (Haidet-Phillips et al., 2011). This might influence intrinsic neuronal functions such as intracellular protein metabolism, axonal transport machinery, and re-innervation at the neuromuscular junction (NMJ). Suppression of gliosis 175

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

(caption on next page) 176

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

Fig. 7. Bmp4 targeted ASO also attenuates disease progression and extends survival of ALS Tg rats via suppressing gliosis. A: Experimental paradigm. Bmp4 targeted ASO (Bmp4-ASO), negative control ASO (NC-ASO) or physiological saline (vehicle) was intrathecally injected to ALS Tg rats at day 0, the day after disease onset (day −1). Body weights (BW) and motor functions were evaluated every other day and rats were then sacrificed at day 14 for further analyses. B: BW compared with maximum weight of each rat. Mean ± SD, n = 6 (Vehicle) or 8 (ASO), two-way analysis of variance (ANOVA) followed by Tukey-Kramer post hoc test. C, D: The motor function evaluations. Grip strength of fore- (C1) and hindlimb (C2) and the scores of landing foot splay (D) are evaluated. Mean ± SD, n = 6 (Vehicle) or 8 (ASO), two-way ANOVA followed by Tukey-Kramer post hoc test. C, D: * indicates a significant difference between Vehicle and Bmp4-ASO treated groups, and # indicates a significant difference between N. C. and Bmp4-ASO treated groups. * or # P < .05, ** or ## P < .01.E: The Mantel-Cox sumulative survival plot. n = 6 (Vehicle) or 8 (ASO). P = .0232. F: Quantitative reverse transcription polymerase chain reaction (RT-PCR) analysis of Bmp4 in the ALS Tg rat lumbar cords. The ratios of target gene to those of the internal control (tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein, zeta, Ywhaz) gene are shown. Vehicle versus Bmp4: P = .0007, NC versus Bmp4: P = .0008, and Vehicle versus NC: P = .7077. Mean ± SD, n = 3 (Vehicle) or 5 (ASO), one-way ANOVA followed by Tukey-Kramer post hoc test. G: Immunoblotting of BMP4 (G1) and pSmad1/5 (G2) in the lumbar cords of ALS Tg rats. H: Quantitative immunoblotting analyses of G. The graph charts show the relative optical density values of the bands. The ratios of BMP4 (H1, Vehicle versus Bmp4: P = .0086, NC versus Bmp4: P < .0001, and Vehicle versus NC: P = .0583), pSmad1/5 (H2, Vehicle versus Bmp4: P = .0002, NC versus Bmp4: P = .0015, and Vehicle versus NC: P = .1422) and Smad5 (D1, Vehicle versus Bmp4: P = .8651, NC versus Bmp4: P = .4891, and Vehicle versus NC: P = .3142) to those of the internal control protein are shown. Mean ± SD, n = 3 (Vehicle) or 5 (ASO) per group, one-way ANOVA followed by Tukey-Kramer post hoc test. I: Triple immunofluorescence staining of the expression of BMP4 (green), vimentin (red) and GFAP (cyan as a pseudo color or blue in merged images) of ALS Tg rat lumbar spinal ventral horns (I1: vehicle, I2: NC-ASO, and I3: Bmp4-ASO treated). Scale bar = 100 μm, V: ventral. J: Double immunofluorescence staining of the expression of CD68 (green) and ionized calcium-binding adaptor molecule 1 (Iba-1, magenta) of ALS Tg rat lumbar spinal ventral horns (J1: vehicle, J2: NC-ASO, and J3: Bmp4-ASO treated). Scale bar = 100 μm, V: ventral. K: Semi-quantifications of I and J. BMP4+ (K1, Vehicle versus Bmp4: P = .0230, NC versus Bmp4: P = .0253, and Vehicle versus NC: P = .9961), GFAP+ (K2, Vehicle versus Bmp4: P = .0245, NC versus Bmp4: P = .0031, and Vehicle versus NC: P = .1926), vimentin+GFAP+ (K3, Vehicle versus Bmp4: P = .0230, NC versus Bmp4: P = .0253, and Vehicle versus NC: P = .9961), BMP4+ vimentin+GFAP+ (K4, Vehicle versus Bmp4: P = .0493, NC versus Bmp4: P = .0143, and Vehicle versus NC: P = .6098), CD68+ (K5, Vehicle versus Bmp4: P = .0112, NC versus Bmp4: P = .0226, and Vehicle versus NC: P = .8077), and Iba-1+ (K6, Vehicle versus Bmp4: P = .7668, NC versus Bmp4: P = .1384, and Vehicle versus NC: P = .0556) areas are evaluated. Mean ± SD, n = 3 per group, one-way ANOVA followed by Tukey-Kramer post hoc test. *P < .05, **P < .01, ***P < .001. ALS: amyotrophic lateral sclerosis, ASO: antisense oligonucleotide, SD: standard deviation, NS: not significant. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

microenvironment. Therefore, the inhibitory effects of Bmp4-targeted ASO might be more moderate than those of noggin. Given that BMP4 was exclusively up-regulated at both the mRNA and protein levels in the spinal cords of ALS Tg rats, we are convinced that BMP4 plays the most crucial role in astrocytic activation among the BMPs. BMP4 increased in the CSF and spinal cord tissues from ALS-linked mutant SOD1-Tg rats during the late symptomatic stage. This supports an idea that extracellular levels of BMP4 are likely elevated by upregulation in the spinal cord parenchyma, followed by leakage across the lenient CSF-spinal cord barrier. Although no significant differences were detected between ALS patients and disease controls, the BMP4 levels in the CSF samples from sALS cases also increased according to disease progression, as did the BMP4 levels in ALS Tg rats. Three of six patients we examined as disease controls had neurodegenerative disease accompanied by robust astrocytosis. The BMP4 in the CSF could be a novel surrogate biomarker which reflects the degree of astrocytosis in patients with neurological diseases. A prospective study involving a large number of age-matched controls and pathologically diagnosed cases with neurodegenerative disorders, or a longitudinal investigation in the same patients with ALS will be needed to confirm this theory. In recent years, oligonucleotide therapeutics were used to treat an increasing number of hereditary neurological diseases. Nusinersen, an ASO drug targeted survival motor neuron (SMN) 2, has already been on the market for use in patients with spinal muscular atrophy. There are at least two ongoing clinical trials for mutant SOD1-linked familial ALS using SOD1- targeted ASO and for Huntington's disease employing HTTtargeted ASO. In the case of Bmp4, a much broader range of clinical application to neurodegenerative diseases may be warranted if further studies uncover the detailed molecular mechanisms regulating BMP4associated neuroinflammation that is provoked by heterogeneous population of glial cells. Caution should be exercised because the whole pool of astrocytes is not responsible for motor neuron degeneration in ALS. Genetically knock down of intermediate filament genes (Gfap and Vimentin) in reactive astrocytes, or simply ablating proliferating astrocytes, did not extend survival nor protect motor neurons within ALS model mice (Ben Haim et al., 2015). More recent studies have suggested a heterogeneity of reactive astrocytes. They consist of neurotoxic and neuroprotective astrocytes in various neurodegenerative diseases, including ALS (Liddelow et al., 2017). Noggin and BMP inhibition may convert the cellular phenotype of reactive astrocytes. In this study, however, we could not distinguish the neuroprotective effects from the

confirmed suppression of microgliosis and decreased levels of neuroinflammatory cytokines in the spinal cords of noggin-treated rats. These results strengthen the possibility that reactive astrocytes can activate microglia, although the direct effects of noggin on microglia cannot be fully excluded (Shin et al., 2014). Recently, reactive astrocytes were reportedly up-regulated TGFβ1, then interfered with neuroprotective functions of microglia, eventually exacerbating neuroinflammation in SOD1-linked ALS model mice (Endo et al., 2015). Astrocyte-selective ablation of mutant SOD1 in the ALS mice led remarkable decrease of activated microglia (Yamanaka et al., 2008). On the contrary, activated microglia can also induce “A1,” neurotoxic astrocytes by interleukin1α, TNFα and C1q (Liddelow et al., 2017). Thus, astrocytes and microglia can probably activate each other by humoral signaling molecules under neurodegenerative conditions. In this study, the p38MAPK pathway and its downstream cytokine expressions were predominantly activated in astrocytes, and less so in the microglia of ALS Tg rats, which were suppressed by noggin administration. The p38MAPK pathway is activated in neurons, astrocytes (Tortarolo et al., 2003), and activated microglia (Veglianese et al., 2006) of ALS model mice and patients with ALS, and has a deteriorating effect on motor neurons by promoting expressions of IL-1β in astrocytes (Ralay Ranaivo and Wainwright, 2010), iNOS and TNFα in astrocytes, and microglia (Pyo et al., 1998) (Xu et al., 2006). Based on increasing evidence indicating that chronic and sustained neuroinflammation accelerates ALS disease progression, several anti-inflammatory drugs have recently been tested in clinical trials (Liu and Wang, 2017). Our findings also suggest that p38MAPK-related neuroinflammation may serve as a potential diseasemodifying target, especially in mutant SOD1-linked ALS. Among BMPs, BMP4 has the most important part in astrocytosis, along with disease progression in this ALS model. The endogenous BMP-antagonist noggin is known to bind, not only to BMP4, but to BMP2, 6, and 7 (Krause et al., 2011). By selective knockdown of Bmp4targeted ASO in vivo, we tried to prove that the BMP4 has a major impact on promoting the mutant SOD1-induced astrocytosis. In contrast to noggin, intrathecal infusion of Bmp4-targeted ASO provided a delayed and limited effect in BW, hindlimb (but not forelimb) function, and survival. Noggin directly binds to BMP4, and then immediately inhibits its function from the extracellular space (Krause et al., 2011). On the other hand, ASO exerts its action beginning from intracellular uptake by endocytosis, followed by binding to the targeted mRNA. This cannot inhibit the existing BMP4 within the extracellular 177

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

disease-aggravating reactive astrocytes in vivo. In conclusion, this study revealed that BMP4 and its downstream signaling play a key role in enhancing astrocytosis, microgliosis, and the neuroinflammation that exacerbates the disease progression in the ALS model. Given that the dysregulated neuroinflammation occurs in a variety of neurodegenerative diseases, targeting the astrocytes and antagonism to the BMP signaling in an adult CNS may help develop new disease-modifying therapies.

familial and sporadic ALS patients are toxic to motor neurons. Nat. Biotechnol. 29, 824–828. doi. https://doi.org/10.1038/nbt.1957. Horner, P.J., Power, A.E., Kempermann, G., Kuhn, H.G., Palmer, T.D., Winkler, J., Thal, L.J., Gage, F.H., 2000. Proliferation and differentiation of progenitor cells throughout the intact adult rat spinal cord. J. Neurosci. 20, 2218–2228. Ishigaki, A., Aoki, M., Nagai, M., Warita, H., Kato, S., Kato, M., Nakamura, T., Funakoshi, H., Itoyama, Y., 2007. Intrathecal delivery of hepatocyte growth factor from amyotrophic lateral sclerosis onset suppresses disease progression in rat amyotrophic lateral sclerosis model. J. Neuropathol. Exp. Neurol. 66, 1037–1044. doi. https://doi. org/10.1097/nen.1030b1013e318159886b. Katagiri, T., Watabe, T., 2016. Bone morphogenetic proteins. Cold Spring Harb. Perspect. Biol. 8, 1–25. Kondo, T., Raff, M.C., 2004. A role for noggin in the development of oligodendrocyte precursor cells. Dev Biol. 267, 242–251. doi. https://doi.org/10.1016/j.ydbio.2003. 1011.1013. Krause, C., Guzman, A., Knaus, P., 2011. Noggin. Int J Biochem Cell Biol. 43, 478–481. doi. https://doi.org/10.1016/j.biocel.2011.1001.1007 Epub 2011 Jan 1021. Lepore, A.C., Dejea, C., Carmen, J., Rauck, B., Kerr, D.A., Sofroniew, M.V., Maragakis, N.J., 2008. Selective ablation of proliferating astrocytes does not affect disease outcome in either acute or chronic models of motor neuron degeneration. Exp. Neurol. 211, 423–432. doi. https://doi.org/10.1016/j.expneurol.2008.1002.1020 Epub 2008 Mar 1017. Lei, Z.N., Liu, F., Zhang, L.M., Huang, Y.L., Sun, F.Y., 2012. Bcl-2 increases stroke-induced striatal neurogenesis in adult brains by inhibiting BMP-4 function via activation of βcatenin signaling. Neurochem. Int. 61, 34–42. Liddelow, S.A., Guttenplan, K.A., Clarke, L.E., Bennett, F.C., Bohlen, C.J., Schirmer, L., Bennett, M.L., Munch, A.E., Chung, W.S., Peterson, T.C., Wilton, D.K., Frouin, A., Napier, B.A., Panicker, N., Kumar, M., Buckwalter, M.S., Rowitch, D.H., Dawson, V.L., Dawson, T.M., Stevens, B., Barres, B.A., 2017. Neurotoxic reactive astrocytes are induced by activated microglia. Nature 541, 481–487. doi. https://doi.org/10.1038/ nature21029 Epub 22017 Jan 21018. Liu, J., Wang, F., 2017. Role of neuroinflammation in amyotrophic lateral sclerosis: cellular mechanisms and therapeutic implications. Front. Immunol. 8. Maragakis, N.J., Rothstein, J.D., 2006. Mechanisms of disease: astrocytes in neurodegenerative disease. Nat Clin Pract Neurol. 2, 679–689. doi. https://doi.org/10.1038/ ncpneuro0355. Matsuura, I., Taniguchi, J., Hata, K., Saeki, N., Yamashita, T., 2008. BMP inhibition enhances axonal growth and functional recovery after spinal cord injury. J. Neurochem. 105, 1471–1479. Meyer, O.A., Tilson, H.A., Byrd, W.C., Riley, M.T., 1979. A method for the routine assessment of fore- and hindlimb grip strength of rats and mice. Neurobehav. Toxicol. 1, 233–236. Meyers, E.A., Gobeske, K.T., Bond, A.M., Jarrett, J.C., Peng, C.Y., Kessler, J.A., 2016. Increased bone morphogenetic protein signaling contributes to age-related declines in neurogenesis and cognition. Neurobiol. Aging 38, 164–175. Miyagi, M., Mikawa, S., Sato, T., Hasegawa, T., Kobayashi, S., Matsuyama, Y., Sato, K., 2012. BMP2, BMP4, noggin, BMPRIA, BMPRIB, and BMPRII are differentially expressed in the adult rat spinal cord. Neurosci. 203, 12–26. Morell, M., Tsan, Y.C., O'Shea, K.S., 2015. Inducible expression of noggin selectively expands neural progenitors in the adult SVZ. Stem Cell Res. 14, 79–94. doi. https:// doi.org/10.1016/j.scr.2014.1011.1001 Epub 2014 Nov 1022. Nagai, M., Aoki, M., Miyoshi, I., Kato, M., Pasinelli, P., Kasai, N., Brown Jr., R.H., Itoyama, Y., 2001. Rats expressing human cytosolic copper-zinc superoxide dismutase transgenes with amyotrophic lateral sclerosis: associated mutations develop motor neuron disease. J. Neurosci. 21, 9246–9254. Nagai, M., Re, D.B., Nagata, T., Chalazonitis, A., Jessell, T.M., Wichterle, H., Przedborski, S., 2007. Astrocytes expressing ALS-linked mutated SOD1 release factors selectively toxic to motor neurons. Nat. Neurosci. 10, 615–622. doi. https://doi.org/10.1038/ nn1876 Epub 2007 Apr 1015. Nakashima, K., Yanagisawa, M., Arakawa, H., Kimura, N., Hisatsune, T., Kawabata, M., Miyazono, K., Taga, T., 1999. Synergistic signaling in fetal brain by STAT3-Smad1 complex bridged by p300. Science 284, 479–482. Ohta, Y., Nagai, M., Nagata, T., Murakami, T., Nagano, I., Narai, H., Kurata, T., Shiote, M., Shoji, M., Abe, K., 2006. Intrathecal injection of epidermal growth factor and fibroblast growth factor 2 promotes proliferation of neural precursor cells in the spinal cords of mice with mutant human SOD1 gene. J. Neurosci. Res. 84, 980–992. doi. https://doi.org/10.1002/jnr.21017. Pyo, H., Jou, I., Jung, S., Hong, S., Joe, E.H., 1998. Mitogen-activated protein kinases activated by lipopolysaccharide and beta-amyloid in cultured rat microglia. Neuroreport 9, 871–874. Ralay Ranaivo, H., Wainwright, M.S., 2010. Albumin activates astrocytes and microglia through mitogen-activated protein kinase pathways. Brain Res. 1313, 222–231. Ringer, C., Tune, S., Bertoune, M.A., Schwarzbach, H., Tsujikawa, K., Weihe, E., Schutz, B., 2017. Disruption of calcitonin gene-related peptide signaling accelerates muscle denervation and dampens cytotoxic neuroinflammation in SOD1 mutant mice. Cell Mol. Life Sci. 74, 339–358. doi. https://doi.org/10.1007/s00018-00016-0233700014 Epub 02016 Aug 00023. Setoguchi, T., Nakashima, K., Takizawa, T., Yanagisawa, M., Ochiai, W., Okabe, M., Yone, K., Komiya, S., Taga, T., 2004. Treatment of spinal cord injury by transplantation of fetal neural precursor cells engineered to express BMP inhibitor. Exp. Neurol. 189, 33–44. doi. https://doi.org/10.1016/j.expneurol.2003.1012.1007. Shibata, N., Yamamoto, T., Hiroi, A., Omi, Y., Kato, Y., Kobayashi, M., 2010. Activation of STAT3 and inhibitory effects of pioglitazone on STAT3 activity in a mouse model of SOD1-mutated amyotrophic lateral sclerosis. Neuropathology 30, 353–360. Shijo, T., Warita, H., Suzuki, N., Kitajima, Y., Ikeda, K., Akiyama, T., Ono, H., Mitsuzawa, S., Nishiyama, A., Izumi, R., Aoki, M., 2018. Aberrant astrocytic expression of

Declaration of interest None. Supplementary data to this article can be found online at https:// doi.org/10.1016/j.expneurol.2018.06.009. Acknowledgements The authors are grateful to Dr. Takafumi Hasegawa, Dr. Naoto Sugeno, Prof. Hiroshi Funakoshi, and Prof. Yasuto Itoyama for fruitful discussions and generous support. We also deeply thank Akiko Machii, Naoko Shimakura, Seiichi Otake, Yayoi Aita, and Kiminobu Shishido for the excellent technical assistance. Funding This work was partly supported by the Grants-in-Aid for Scientific Research [26461288, 16H05318, 16K15474, and 17K09773] from the Japanese Ministry of Education, Culture, Sports, Science and Technology; and a Grant-in-Aid for Research on Rare and Intractable Diseases, the Research Committee on Establishment of Novel Treatments for Amyotrophic Lateral Sclerosis from the Japan Agency for Medical Research and Development (16kk0105002h0001), AMED. References Andersen, C.L., Jensen, J.L., Orntoft, T.F., 2004. Normalization of real-time quantitative reverse transcription-PCR data: a model-based variance estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets. Cancer Res. 64, 5245–5250. doi. https://doi.org/10.1158/0008-5472.CAN-52040496. Aoki, M., Warita, H., Mizuno, H., Suzuki, N., Yuki, S., Itoyama, Y., 2011. Feasibility study for functional test battery of SOD transgenic rat (H46R) and evaluation of edaravone, a free radical scavenger. Brain Res. 1382, 321–325. Ben Haim, L., Carrillo-De Sauvage, M.A., Ceyzeriat, K., Escartin, C., 2015. Elusive roles for reactive astrocytes in neurodegenerative diseases. Front. Cell. Neurosci. 9, 278–304. Brooks, B.R., Miller, R.G., Swash, M., Munsat, T.L., 2000. El Escorial revisited: revised criteria for the diagnosis of amyotrophic lateral sclerosis. Amyotroph Lateral Scler. Other Motor Neuron Disord. 1, 293–299. Bruijn, L.I., Becher, M.W., Lee, M.K., Anderson, K.L., Jenkins, N.A., Copeland, N.G., Sisodia, S.S., Rothstein, J.D., Borchelt, D.R., Price, D.L., Cleveland, D.W., 1997. ALSlinked SOD1 mutant G85R mediates damage to astrocytes and promotes rapidly progressive disease with SOD1-containing inclusions. Neuron 18, 327–338. Burda, J.E., Sofroniew, M.V., 2014. Reactive gliosis and the multicellular response to CNS damage and disease. Neuron 81, 229–248. doi. https://doi.org/10.1016/j.neuron. 2013.1012.1034. Cui, Z.S., Zhao, P., Jia, C.X., Liu, H.J., Qi, R., Cui, J.W., Cui, J.H., Peng, Q., Lin, B., Rao, Y.J., 2015. Local expression and role of BMP-2/4 in injured spinal cord. Genet. Mol Res. 14, 9109–9117. doi. https://doi.org/10.4238/2015.August.9107.9120. Dizon, M.L., Maa, T., Kessler, J.A., 2011. The bone morphogenetic protein antagonist noggin protects white matter after perinatal hypoxia-ischemia. Neurobiol Dis. 42, 318–326. doi. https://doi.org/10.1016/j.nbd.2011.1001.1023 Epub 2011 Feb 1017. Endo, F., Komine, O., Fujimori-Tonou, N., Katsuno, M., Jin, S., Watanabe, S., Sobue, G., Dezawa, M., Wyss-Coray, T., Yamanaka, K., 2015. Astrocyte-derived TGF-beta1 accelerates disease progression in ALS mice by interfering with the neuroprotective functions of microglia and T cells. Cell Rep. 11, 592–604. doi. https://doi.org/10. 1016/j.celrep.2015.1003.1053 Epub 2015 Apr 1016. Guan, Y.J., Wang, X., Wang, H.Y., Kawagishi, K., Ryu, H., Huo, C.F., Shimony, E.M., Kristal, B.S., Kuhn, H.G., Friedlander, R.M., 2007. Increased stem cell proliferation in the spinal cord of adult amyotrophic lateral sclerosis transgenic mice. J. Neurochem. 102, 1125–1138. doi. https://doi.org/10.1111/j.1471-4159.2007.04610.x Epub 02007 Apr 04630. Haidet-Phillips, A.M., Hester, M.E., Miranda, C.J., Meyer, K., Braun, L., Frakes, A., Song, S., Likhite, S., Murtha, M.J., Foust, K.D., Rao, M., Eagle, A., Kammesheidt, A., Christensen, A., Mendell, J.R., Burghes, A.H., Kaspar, B.K., 2011. Astrocytes from

178

Experimental Neurology 307 (2018) 164–179

T. Shijo et al.

progression. Mol. Cell. Neurosci. 23, 180–192. Vandesompele, J., De Preter, K., Pattyn, F., Poppe, B., Van Roy, N., De Paepe, A., Speleman, F., 2002. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. 3 RESEARCH0034. Epub 2002 Jun 0018. Veglianese, P., Lo Coco, D., Bao Cutrona, M., Magnoni, R., Pennacchini, D., Pozzi, B., Gowing, G., Julien, J.P., Tortarolo, M., Bendotti, C., 2006. Activation of the p38MAPK cascade is associated with upregulation of TNF alpha receptors in the spinal motor neurons of mouse models of familial ALS. Mol. Cell Neurosci. 31, 218–231. doi. https://doi.org/10.1016/j.mcn.2005.1009.1009 Epub 2005 Oct 1010. Xu, Z., Wang, B.R., Wang, X., Kuang, F., Duan, X.L., Jiao, X.Y., Ju, G., 2006. ERK1/2 and p38 mitogen-activated protein kinase mediate iNOS-induced spinal neuron degeneration after acute traumatic spinal cord injury. Life Sci. 79, 1895–1905. doi. https:// doi.org/10.1016/j.lfs.2006.1806.1023 Epub 2006 Jun 1821. Yamanaka, K., Komine, O., 2018. The multi-dimensional roles of astrocytes in ALS. Neurosci. Res. 126, 31–38. Yamanaka, K., Chun, S.J., Boillee, S., Fujimori-Tonou, N., Yamashita, H., Gutmann, D.H., Takahashi, R., Misawa, H., Cleveland, D.W., 2008. Astrocytes as determinants of disease progression in inherited amyotrophic lateral sclerosis. Nat. Neurosci. 11, 251–253. doi. https://doi.org/10.1038/nn2047 Epub 2008 Feb 1033.

chondroitin sulfate proteoglycan receptors in a rat model of amyotrophic lateral sclerosis. J. Neurosci. Res. 96, 222–233. doi. https://doi.org/10.1002/jnr.24127 Epub 22017 Jul 24128. Shin, J.A., Lim, S.M., Jeong, S.I., Kang, J.L., Park, E.M., 2014. Noggin improves ischemic brain tissue repair and promotes alternative activation of microglia in mice. Brain Behav. Immun. 40, 143–154. Stieber, A., Gonatas, J.O., Gonatas, N.K., 2000. Aggregates of mutant protein appear progressively in dendrites, in periaxonal processes of oligodendrocytes, and in neuronal and astrocytic perikarya of mice expressing the SOD1 (G93A) mutation of familial amyotrophic lateral sclerosis. J. Neurol. Sci. 177, 114–123. Takahashi, M., Arai, Y., Kurosawa, H., Sueyoshi, N., Shirai, S., 2003. Ependymal cell reactions in spinal cord segments after compression injury in adult rat. J. Neuropathol. Exp. Neurol. 62, 185–194. Tandan, R., Bradley, W.G., 1985. Amyotrophic lateral sclerosis: part 1. Clinical features, pathology, and ethical issues in management. Ann. Neurol. 18, 271–280. Tang, J., Song, M., Wang, Y., Fan, X., Xu, H., Bai, Y., 2009. Noggin and BMP4 co-modulate adult hippocampal neurogenesis in the APP(swe)/PS1(DeltaE9) transgenic mouse model of Alzheimer's disease. Biochem. Biophys. Res. Commun. 385, 341–345. Tortarolo, M., Veglianese, P., Calvaresi, N., Botturi, A., Rossi, C., Giorgini, A., Migheli, A., Bendotti, C., 2003. Persistent activation of p38 mitogen-activated protein kinase in a mouse model of familial amyotrophic lateral sclerosis correlates with disease

179